Commission, February 27, 2013, No M.6663

THE EUROPEAN COMMISSION,

Having regard to the Treaty on the Functioning of the European Union, Having regard to the Agreement on the European Economic Area, and in particular Article 57 thereof, Having regard to Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of concentrations between undertakings (1), and in particular Article 8(3) thereof, Having regard to the Commission's decision of 29 August 2012 to initiate proceedings in this case, Having given the undertakings concerned the opportunity to make known their views on the objections raised by the Commission, Having regard to the opinion of the Advisory Committee on Concentrations (2), Having regard to the final report of the Hearing Officer in this case (3), Whereas:

1. THE NOTIFICATION AND THE PROCEDURE

(1) On 24 July 2012, the European Commission received a notification of a proposed concentration pursuant to Article 4 of the Merger Regulation by which the undertaking Ryanair Holdings plc ("Ryanair", Ireland) acquires within the meaning of Article 3(1)(b) of the Merger Regulation control of the whole of Aer Lingus Group plc ("Aer Lingus", Ireland) by way of public bid announced on 19 June 2012 ("the Transaction").The Transaction would be implemented through a wholly-owned subsidiary of Ryanair, Coinside Limited.

(2) After examination of the notification, the Commission concluded that the Transaction fell within the scope of the Merger Regulation and raised serious doubts as to its compatibility with the internal market and with the EEA Agreement. The Commission therefore initiated proceedings in accordance with Article 6(1)(c) of the Merger Regulation (hereafter the "decision opening the proceedings") on 29 August 2012.

(3) Following a request by Ryanair of 31 August 2012, a non-confidential version of certain key submissions of third parties collected during the first phase investigation was provided to Ryanair on 6 September 2012.

(4) On 6 September 2012, Ryanair requested an extension of the time period for the second phase investigation by 15 working days pursuant to the second subparagraph of Article 10(3) of the Merger Regulation.

(5) Ryanair submitted its written comments on the decision opening the proceedings on 13 September 2012.

(6) Aer Lingus submitted its written comments on the decision opening the proceedings on 28 September 2012.

(7) Ryanair submitted the first package of formal commitments on 17 October 2012 pursuant to Article 8(2) of the Merger Regulation (the "Commitments of 17 October 2012"). These Commitments were not market tested.

(8) On 13 November 2012, a Statement of Objections ("SO") was sent to Ryanair by the Commission pursuant to Article 18 of the Merger Regulation.

(9) Ryanair submitted its reply to the SO on 28 November 2012. On 14 December 2012, the Commission sent Ryanair a Letter of Facts. Ryanair replied to the Letter of Facts on 20 December 2012.

(10) In order to address competition concerns identified in the SO, Ryanair submitted a second set of commitments on 7 December 2012 ("Commitments of 7 December 2012"). The Commission launched a market test to gather the views of relevant market participants on the effectiveness of the proposed commitments ("the first market test").

(11) On 15 January 2013, Ryanair submitted a third and new version of the commitments (the "Commitments of 15 January 2013"). On 21 January 2013, the Commission launched a market test of Commitments of 15 January 2013 ("the second market test").

(12) The Commitments of 15 January 2013 were subsequently modified on 1 February 2013 (the "final Commitments"). On 4 February 2013, the Commission launched a market test of these modified commitments ("the final market test").

(13) The Advisory Committee was convened on 18 February 2013.

2. THE PARTIES

(14) Ryanair is a low-fares carrier operating point-to-point scheduled air services essentially in Europe. The company has a fleet of 305 aircraft (4) and 51 bases across Europe, the most important bases being London Stansted (114 routes, 39 aircraft), Brussels Charleroi (78 routes, 15 aircraft), Milan Bergamo (78 routes, 15 aircraft) and Dublin (18 aircraft in summer 2012). In the summer season 2012, Ryanair operated in particular 62 short-haul routes ex-Dublin (5).

(15) Ryanair is not a member of any airline alliance and does not have interlining operations with any other carrier.

(16) Aer Lingus is a publicly-listed carrier based in Ireland. It offers essentially point-topoint scheduled air transport services. (6) In the International Air Transport Association ("IATA") summer season of 2012, it operated on 108 routes, across Ireland, the United Kingdom, Continental Europe, and the United States of America, with 45 aircraft. Aer Lingus explains that it carried 10.4 million passengers in 2011 (7). Aer Lingus is based principally at Dublin Airport from where it operates a substantial portion of its scheduled flights. In the 2012 summer season, Aer Lingus (including Aer Arann) operated 66 short-haul routes ex-Dublin (8). In the Union, Aer Lingus also has bases in Cork, Shannon, Belfast, and Gatwick. Aer Lingus is not a member of any airline alliance (it left oneworld in April 2007). It develops a concept of "open network architecture" whereby its neutrality allows it to partner across alliances and offer connectivity through major hubs to worldwide destinations in addition to carrying point-to-point traffic. (9)

(17) Ryanair and Aer Lingus are referred to together as "the Parties" in this Decision.

(18) Ryanair's minority shareholding in Aer Lingus represents 29.82% of Aer Lingus' total issued share capital and makes Ryanair the largest shareholder in Aer Lingus. (10) The Irish Government (Minister for Finance) is the next largest shareholder with a stake of around 25.1%. (11)

3. THE TRANSACTION

(19) On 19 June 2012, Ryanair announced its intention to acquire Aer Lingus, by means of an all-cash offer for all of Aer Lingus' outstanding shares (70.18%) not already owned by Ryanair. The offer document was sent to Aer Lingus shareholders on 17 July 2012 with a deadline for acceptance of 15 September 2012. If successful, Ryanair would gain the majority of Aer Lingus's shares and thus sole control over Aer Lingus.

(20) The Transaction constitutes therefore a concentration within the meaning of Article 3(1)(b) of the Merger Regulation.

(21) The Transaction is the third attempt by Ryanair to acquire sole control over Aer Lingus. In a Commission Decision of 27 June 2007 declaring a concentration to be incompatible with the common market and the EEA Agreement in Case No COMP/M.4439 - Ryanair/Aer Lingus (the "2007 Decision") (12), the Commission conducted an in-depth investigation of Ryanair's first bid to takeover Aer Lingus, and adopted the 2007 Decision under Article 8(3) of the Merger Regulation ("the 2007 Decision") whereby it declared the notified concentration incompatible with the internal market. The 2007 Decision was upheld by the General Court of the European Union in Case T-342/07 Ryanair Holdings plc v European Commission [2010] ECR II-03457 ("the 2010 Judgment").

(22) In 2008-2009, Ryanair initiated a second takeover bid for Aer Lingus (13), but eventually withdrew its offer.

4. UNION DIMENSION

(23) The undertakings concerned have a combined aggregate world-wide turnover of more than EUR 5 000 million (Ryanair: EUR 4 390 million; (14) Aer Lingus: EUR 1 288 million) (15). Both Ryanair and Aer Lingus have a Union-wide turnover in excess of EUR 250 million. Ryanair did not achieve more than two-thirds of its turnover within one and the same Member State.

(24) Consistent with the approach adopted in the 2007 Decision, turnover has been calculated using the "50/50" methodology and the "point of departure" methodology. (16) The turnover thresholds set out in of Article 1(2) of the Merger Regulation are met under both methodologies.

(25) The concentration therefore has a Union dimension within the meaning of Article 1(2) of the Merger Regulation.

5. INVESTIGATION OF THE CASE

(26) The assessment of the case has proven highly complex. The Commission has sought to make use of all available means of investigation pursuant to Article 11 of the Merger Regulation. It not only analysed questionnaires which were sent inter alia to competing scheduled carriers, charter carriers, travel agents, corporate customers, consumers associations, and airports, but also undertook other written and oral contacts with these market participants and other third parties such as slot coordinators, civil aviation authorities, and transport authorities. In addition, the Commission analysed a substantial amount of internal documents originating from the potential upfront buyers ("UFB") identified by Ryanair.

(27) Against this background, it is important to stress that the assessment of the competitive impact of the Transaction involves a complex legal and economic analysis, the results of which are based on the totality of the available evidence (17).

(28) In particular as regards the questionnaires, it is important to note that the market investigation is by no means an opinion poll. For instance, the fact that the majority of third parties provide a similar opinion in reply to a specific question, can only be an indication for the Commission's own investigation, not a foregone conclusion. Likewise, it would not be appropriate to assume that the answers to the questionnaires can always be considered to be fully informed and objective. The specific level of knowledge of respondents might vary, the questions might have been misunderstood, the replies might be more or less representative, and the opinion provided might be biased to influence the Commission's decision-making process in a certain way (18).

(29) The Commission has also collected and used quantitative data. Once verified for accuracy and relevance, quantitative data obtained were used in the decision-making process for all relevant issues, complementing the qualitative assessment of those issues.

(30) Ryanair claims in its response to the SO that the "Commission has failed to make use of all available means of investigation pursuant to Article 11 of the EU Merger Control regulation. In particular, given the complexities of this case, it should have sent questionnaires to, and should have sought ample contact with, customers, as it did in the context of the 2007 Ryanair/Aer Lingus proceeding". (19) Furthermore, Ryanair suggests that "the SO contains no quantitative analysis (such as survey evidence) using the SSNIP test for any of the airports it deems substitutable. The SO analysis proceeds entirely in the abstract, leading to conclusions that are inconsistent with basic facts and common sense". (20)

(31) In this regard, the Commission considers that a survey similar to the one performed in 2007 would not have been useful in the case at hand for the purposes (in particular the market definition) alleged by Ryanair. The principal goal of the survey conducted in 2007 was to test "Ryanair's claim that, from the perspective of the customer, Aer Lingus and Ryanair do not compete with each other". (21) In particular, the survey examined that issue through questions of whether consumers would have considered a given competitor in different types of routes ("same airport", "different airport" and "oligopoly" routes). However, for the present Transaction, there is ample evidence demonstrating that Ryanair and Aer Lingus compete. Therefore, such a survey would have provided no material added value. In its assessment of the Transaction, the Commission has gathered cogent alternative qualitative and quantitative (like the correlation analysis) evidence supporting its conclusions on market definition (22).

Furthermore, for the vast majority of airport pairs, the issue of market definition was also analysed in detail in the 2007 Decision, and the Commission assessed whether changes in market circumstances had occurred since 2007.

(32) The Commission also remarks in this context that the views of the customers were gathered in the market investigation through the usual proxy used by the Commission in airline cases, sending questionnaires to corporate customers, and travel agents (being in principle in close touch with customers, even if, in the case at hand, the bulk of sales is achieved through the internet).

(33) As in any other merger investigation, the Commission has therefore carefully analysed, interpreted and weighed all views expressed during the market investigation.

6. GENERAL CONTEXT FOR THE ASSESSMENT OF THE CASE

(34) The 2007 Decision and the 2010 Judgment constitute important background elements for the assessment of the Transaction. However the Transaction is assessed considering present circumstances, while changes in market conditions since 2007 provide very relevant insights for the understanding and analysis of the present circumstances.

(35) Ryanair points to "material market changes since Ryanair's unsuccessful 2006 bid for Aer Lingus" (23).

(36) Ryanair highlights first, that Europe's flag carriers are inexorably consolidating into five large scheduled airlines groups led by Air France, British Airways, easyJet, Lufthansa and Ryanair. It considers that the most significant of these consolidations is the recent takeover by IAG of bmi. Contrary to these trends, Ryanair claims that Aer Lingus has, since 2006, failed to find a consolidation partner or make itself attractive to one of Europe's three "mega-carrier" groups (Air France-KLM, IAG, Lufthansa) and that Aer Lingus is a subscale peripheral carrier that is not consistently profitable and cannot grow or compete with the much larger carriers in Europe. According to Ryanair, Aer Lingus has no sustainable growth plan, with annual traffic having fallen by almost one million passengers (or 9%) since 2009, and has significant accumulated losses since its 2006 initial public offering (IPO). In this context, the Transaction would be "the only way to secure Aer Lingus' future as part of a well financed and highly competitive European carrier that continues to grow and prosper in the single market." (24) Ryanair claims that the Transaction would therefore benefit consumers.

(37) The Commission acknowledges that some consolidation is indeed taking place in the air transport sector in the Union (25). However, such consolidation cannot take place at the expense of consumers. The specific effects on competition of the Transaction must therefore be assessed. The Commission also notes that Aer Lingus has been profitable since 2010 and presents a rather strong balance sheet, at a time of difficulties both in Ireland and in the transport sector. (26)

(38) Secondly, Ryanair claims that Dublin Airport is now operating at just over 50% capacity. When Ryanair first bid for Aer Lingus in late 2006, Dublin Airport was suffering from congestion. The Commission will examine this specific issue in Section 8.5.2.

7. RELEVANT MARKETS

7.1. Introduction

(39) Ryanair and Aer Lingus both provide scheduled passenger air transport services within the EEA. There are no other markets affected by the Transaction other than those relating to passenger air transport services.

7.1.1. Operating models of carriers in the airline industry

(40) Carriers which currently offer scheduled air transport services within Europe form a heterogeneous group, with significant differences between each single carrier. Such differences between carriers relate mainly to the operating model of the respective company (in principle hub & spoke or "network" carriers as opposed to point-to-point models) and to the level of service that is offered to passengers (full service as opposed to low-frills or no-frills model). Intermediate levels of services can also be provided by some carriers.

(41) Some carriers operate a hub-and-spoke network model providing connecting services throughout their system by using a centrally located airport. Within a hub-and-spoke network, flights are concentrated in time and space to an airport (hub). These network carriers serve various different passenger segments, such as passengers flying short-haul or long-haul, time sensitive and non-time sensitive passengers and passengers who are connecting or fly point-to-point, with a series of complex operations at their hub airports. The hub-and-spoke network provides increased flexibility to network airlines with regard to operations from their hub airport. On the other hand, it is less likely that network carriers would build up significant operations on a stand-alone basis, at airports not connected to their hub airport.

(42) Other carriers concentrate on local traffic and offer point-to-point services on many different city pairs. They focus on fewer passenger segments and have a simplified operating model without incurring the additional cost and complexity of creating and operating a hub. Point-to-point carriers decide on the business case for a new or existing route based on the operational economics of individual routes. This provides them with more flexibility to enter individual routes if sufficient demand exists. For increased operational efficiency, point-to-point carriers often also concentrate their activities on particular airports (bases). A "base" airport is an airport where carriers base a certain number of aircraft and on which they concentrate their operations and from which they operate several routes. A large majority of the respondents to the market investigation agree with this definition for the purpose of the assessment of this case (27). At a base airport, carriers overnight aircraft, entertain additional facilities and services, such as maintenance services, customer care services, ground-handling services or stand-by aircraft. A base airport requires the commitment of infrastructure, personnel and equipment (with the concomitant capital expenditure) at an airport, which are used to operate several routes from that airport. Furthermore, as crew are attached to it, a base airport leads to fewer costs (as there are no hotel overnight expenditures for crew) and it tends to allow airlines to employ a lower number of reserve crew while still ensuring adequate reserves are on hand to ensure replacement of personnel (due to for example illness or delays) in a timely manner (28).

(43) The difference between network carriers and point-to-point carriers often coincides with the distinction between full-service carriers and low-frills or no frills carriers. Network carriers operating a hub-and-spoke system generally provide an integrated passenger product with additional service features such as complimentary luggage transport, on board meals, seat allocation, access to airport lounges, frequent flyer programmes, ticket flexibility. Point-to-point carriers, on the other hand, usually operate as low frills or no frills carriers. They either charge extra for so-called "frill" services like catering, checked-in luggage, hand luggage, lounges, seat allocation, priority boarding and extra comfort or, alternatively, they choose not to offer these services at all.

(44) The differences between the operating models which carriers use within the EEA cannot, however, be grouped clearly into distinct categories. There are many variations between the extremes of strictly no frills point-to-point services of certain carriers and the traditional full service network product of others. As part of their product differentiation, carriers may create their individual combinations of service elements without clearly falling into any category.

7.1.2. Services provided by Ryanair and Aer Lingus

(45) Ryanair and Aer Lingus compete on the provision of scheduled air passenger transport services on routes to and from Ireland and in particular to and from Dublin, Cork, Shannon and Knock.

(46) Aer Lingus operated as a traditional full service carrier until 2001, when it introduced a low fares model to compete with low cost carriers. Following the global economic and financial crisis in 2008, Aer Lingus initiated a review of its business model and since 2009, Aer Lingus has emphasised its positioning as a value carrier between the low cost carriers and full service carriers, with a strong core product and offering additional paid options. As such, Aer Lingus targets both leisure / non-time sensitive and business / time sensitive passengers (29).

(47) Ryanair on the other hand describes itself as an ultra "low-fares" or "low cost" carrier and it operates a point-to-point no-frills service.

(48) For the purpose of the assessment of the Transaction, therefore, it is not appropriate to define separate markets according to the distinctions between the different operating models of carriers. These differences between operating models will be taken into account when assessing the competitive impact of the Transaction. Each competitor's incentives and ability to compete effectively with the merged entity based on its business model will exert different levels of competitive constraints as shown in particular in Section 8.9.

7.2. Origin and destination approach (O&D)

7.2.1. Demand side considerations

(49) The Commission's Notice on the definition of relevant market for the purposes of Union competition law (the "Relevant Market Notice") (30) provides that the main purpose of market definition is to identify in a systematic way the competitive constraints faced by the undertakings involved. In so doing, the Commission takes into account a range of evidence which would enable it to assess the extent to which substitution would take place. The Commission considers the characteristics and specificity of the industry and products or services analysed (31).

(50) In its decisional practice, the Commission has traditionally defined the relevant market for scheduled passenger air transport services on the basis of the "point of origin/point of destination" ("O&D") city-pair approach (32). Such a market definition reflects the demand-side perspective whereby passengers consider all possible alternatives of travelling from a city of origin to a city of destination, which they do not consider substitutable for a different city-pair. As a result, every combination of a point of origin and a point of destination is considered a separate market.

(51) Ryanair does not object to the O&D approach (33).

(52) The majority of the respondents to the market investigation confirm that the traditional O&D approach is the most appropriate tool for defining markets in this case (34).

(53) It follows from the O&D approach, as the Commission has traditionally considered, that connecting passengers (35) are not part of the same market as O&D passengers. In particular, this was the conclusion reached in the 2007 Decision, when considering the specific situation of Ryanair and Aer Lingus (36). Most recently, in the Commission's decision of 30 March 2012 in Case No COMP/M.6447 - IAG/bmi (37), the Commission reached a similar conclusion. Therefore, the Commission considers for the purposes of this Transaction that connecting passengers are not part of the same market as O&D passengers.

(54) Therefore, the effects of this Transaction are assessed on the basis of an O&D citypair approach (38).

7.2.2. Supply side and other considerations

(55) In the Commission's Decision of 11 February 2004 in Case No COMP/M.3280 - Air France/KLM (39), the Commission has also taken into consideration supply-side elements such as network competition between airlines based on the hub-and-spoke structure of traditional carriers. However the Commission considered that the degree of supply-side substitutability between different O&Ds was limited.

(56) In the 2007 Decision, the Commission examined whether supply-side substitution could warrant a departure from the established O&D approach (that is to say a socalled market for short-haul flights from and to Ireland) (40). The question was raised whether the bundle of routes out of Dublin forms one single relevant market, since suppliers operating from Ireland could switch between the different routes. The Commission concluded in the 2007 Decision that non-Irish airlines face significant barriers to entry and more difficulties to win sufficient customers for their Irish routes than the two well-established Irish airlines Aer Lingus and Ryanair. The Commission did not identify effects of supply-side substitution strong enough to be equivalent to those of demand side substitution in terms of effectiveness and immediacy. It thus upheld the use of the traditional O&D approach.

(57) Ryanair argues (41) that with the growth of point-to-point airlines, supply-side substitution is an increasingly important aspect of market definition. However, as explained in detail in Section 8.5.2, it remains the case that there are significant barriers to entry on the Irish market which makes it difficult for airlines to switch their services to Irish routes.

(58) In any event, as explained in detail in Section 7.2.1, the Commission considers that the O&D approach is the most appropriate tool for defining markets in this case. The evidence collected through the market investigation indicates the same (42). Furthermore, in line with the Commission's Relevant Market Notice (43) and with its previous decisional practice, the competitive constraint arising from supply side substitutability is normally only considered in the market definition when it has an immediate and effective impact on the relevant product market. As this is not the case here (44), the Commission gives precedence to demand-side substitution (45).

(59) Some respondents to the market investigation also pointed out the specificity of "leisure travel". In their view, such passengers are not fixed on a given O&D pair, but have the flexibility to switch between a number of O&D pairs depending on various criteria, amongst which are prices and convenience (46). Another respondent argued that "[...] some consumers have a much wider set of criteria for their destination choice, for example they may trade-off different holiday destinations. The product market here may therefore be 'city destinations', 'access to ski stations' or 'beach destinations" (47).

(60) As already acknowledged in the 2007 Decision (48), the Commission considers that some customers might consider flying to different city or holiday airports without having a clear preference for one destination ("destination insensitive customers"). However, it also remains the case that for the vast majority of passengers, a flight from Ireland to one destination is not simply substitutable with a flight to another destination. On the contrary, from a demand-side perspective, every combination of a point of origin and a point of destination forms a separate market from a customers' viewpoint as customers require transportation from one specific point to another.

(61) Therefore, for the assessment of the Transaction, the Commission considers that competition still takes place on an O&D basis.

7.3. Analysis of the relevant routes (airport pairs vs. city pairs)

7.3.1. Introduction: analytical framework

(62) In establishing whether an O&D pair forms a relevant market, the Commission assesses the possibilities which are offered to consumers to travel between a given point of origin and destination. Many cities are connected to two or more airports and therefore when defining the relevant O&D markets for passenger air transport services, the Commission examines whether flights from or to airports which have sufficiently overlapping catchment areas can be considered as substitutes in the eyes of passengers (49). To this extent, both the point of origin and the point of destination should include all airports that are substitutable in the eyes of passengers as passengers beginning or ending their journey in the catchment area of two or more airports can choose between flights offered to any of those airports (50).

(63) From the demand side, a number of factors influence passengers' choice of airport. Passengers take into account the convenience and cost of getting to the airport, the services offered from an airport such as the flight times, schedules and frequencies. Certain passengers also take into account the availability of their preferred airline or extra services like lounges, which they are entitled to use by reason of their membership of loyalty programmes. Airport substitution increases competition if the choice between airports also implies a wider choice between different airlines, to the extent that they are independent from each other.

(64) According to the Relevant Market Notice "differences in product characteristics are not in themselves sufficient to exclude demand substitutability, since this will depend to a large extent on how customers value different characteristics" (51).

(65) In order to correctly capture the competitive constraint that flights from and to two (or more) different airports exert on each other, a detailed analysis is necessary by taking into consideration the specific characteristics of the case at hand. As already identified and explained in the 2007 Decision (52), passengers take into account a number of elements like travel time, travel costs, flight times/schedules/frequencies and the quality of service when it comes to choosing between air transport services to and from different airports. The passenger's choice for one or the other airline service will ultimately be driven by a combination of these elements. These elements continue to be relevant for the purpose of the assessment of the Transaction. Indeed as Ryanair itself acknowledges, it competes with other airlines "primarily with respect to fare levels, frequency and dependability of service, name recognition, passenger amenities (such as access to frequent flyer programs), and the availability and convenience of other passenger services." (53)

7.3.2. Parameters considered

(66) The 2010 Judgment approved the Commission's approach to market definition as set out in the 2007 Decision, that is to analyse the question of airport substitutability from the perspective of customers using the technique of bundling evidence (54).

(67) This approach involved defining, for the purposes of examining a concentration between two carriers, the markets for passenger air transport on the basis of air routes between two cities or bundles of air routes to the extent that there is substitutability between them (according to the specific features of the transaction) and by using different parameters to characterise the substitutability of these routes. These parameters included inter alia a comparison of distances and travelling times to the indicative benchmark of 100km/1 hour driving time, the responses of competitors, airports, other civil aviation authorities to the market investigation or Ryanair's own marketing practices.

(68) In its assessment of airport substitutability from the perspective of customers the Commission will follow a similar approach and will use its findings in the 2007 Decision and the following pieces of evidence, as described in detail in the Sections that follow: a "first proxy" catchment area analysis, the price monitoring behaviour of Aer Lingus and Ryanair, the outcome of the market investigation (views of the airports, the competitors, the civil aviation authorities and other market participants), Ryanair's marketing practices, the price correlation analysis, and any other relevant element.

7.3.2.1. The 2007 Decision and other Commission precedents

(69) The Commission has in its previous decisional practice investigated the question of airport substitutability as regards some of the airport pairs which are relevant for the assessment of the current Transaction. The question of airport substitutability is relevant for specific routes in particular for assessing whether passenger air transport services between one point of origin and one airport at a point of destination is substitutable for passenger air transport services between the same point of origin and a different suitable airport for the same point of destination.

(70) In the 2007 Decision, the Commission assessed airport substitutability for certain routes out of Ireland (Dublin, Cork, or Shannon in particular) operated by Ryanair and Aer Lingus and which continue to be relevant for the current Transaction.

(71) The General Court held in its Judgment in Case T-158/00 ARD v Commission [2003] ECR II-3825 that "a comparison with other merger cases can be relevant only if it is established that they raise the same competition problems and concern markets with the same charaterstics and where conditions have not changed." (55) (emphasis added)

(72) For the current Transaction, a comparison with the 2007 Decision is particularly relevant. First, for market definition purposes, the current Transaction raises the same competition problems as in 2007; both Ryanair and Aer Lingus target essentially the same type of passengers as in 2007 and their service offerings have not materially changed (56). Secondly, the current Transaction concerns for most of the routes exactly the same markets (O&D routes) as in 2007 and lastly, the market conditions have not changed to such a degree as to justify a different approach.

(73) The Commission conducted an in-depth investigation into the matter of airport substitutability and it considered both qualitative and quantitative evidence in its 2007 Decision. The Commission's methodology and conclusions have been upheld by the 2010 Judgment (57). The Commission hence treats its findings in the 2007 Decision as important factual elements which it takes into account in its assessment of the question of airport substitutability. Indeed, the Commission sought to ascertain during its market investigation into the Transaction, whether there have been any significant changes in the market circumstances such as to warrant a different conclusion than the one reached in the 2007 Decision as regards the relevant airport pairs for flights to and from Dublin (or Cork or Shannon as relevant).

(74) As will be shown below in Section 7.3.3, the Commission nevertheless closely examined the question of airport substitutability and scrutinised various parameters and it did not exclusively rely on its findings in the 2007 Decision.

(75) As regards other Commission precedents which dealt with airport substitutability issues for some of the airport pairs which are relevant for the current Transaction, these concerned routes to and from other countries than Ireland. Furthermore, the other precedents generally did not concern carriers or destinations which have the same characteristics as the destinations served or the passengers carried by Ryanair and Aer Lingus. Therefore, such precedents are of much more limited relevance than the 2007 Decision.

7.3.2.2. Airport catchment area of 100km or 1 hour's drive time

(76) In the 2007 Decision, the Commission established a proxy to estimate an airport's typical minimum catchment area on the basis of responses from airports (58). Accordingly, the Commission considered that airports are likely to have a catchment area serving the same city if they are within 100km or 1 hour from the city centre (the "100km/1 hour benchmark").

(77) However the Commission explained that the 100km or 1 hour's drive time is a conservative estimate of an airport's typical catchment area and that within such travelling distances or times to the airport, most passengers would not consider that flying from one airport or the other to the same destination is manifestly inconvenient. As a result, most passengers would openly consider flying from one airport or the other, to the effect that competing air transport services between a point in Ireland, on the one hand and each of these airports on the other, may exercise a competitive constraint on each other (59).

(78) The 100km/1 hour benchmark was used in the 2007 Decision merely as a "first proxy" (60) and the General Court confirmed that "the Commission cannot be criticised for having used the 100 km or one hour driving time benchmark in defining the catchment areas of the airports" (61).

(79) In the present Transaction, Ryanair argues that the Commission's decisional practice creates a strong rebuttable presumption that, where airports are outside the 100km/1 hour benchmark, they are unlikely to be considered substitutes for antitrust purposes (62).

(80) Furthermore, Ryanair argues that the 100km/1 hour benchmark is less instructive when two airports primarily serve different major cities or regions and are not secondary airports but are the "main" airports for a certain city (63). For these airports, Ryanair claims that the Commission's following two presumptions should be much weaker (or the validity of the presumption should be less clear) (64) : (i) if an airport falls within 100km or 1 hour's drive time of a city centre, a large pool of the airport's customers will reside between the airport and the city centre and (ii) when more than one airport is within the catchment area of a city then customer pools overlap, that is to say there will be an appreciable group of customers who consider the two or more airports substitutable. Ryanair also argues that the 100km/1 hour benchmark is less instructive for leisure routes (for example the Dublin-Bilbao/Santander route). Lastly, Ryanair argues that, besides distance, the availability of transport connections will also be a key factor in identifying the catchment areas of each airport (65).

(81) On the contrary, Aer Lingus considers that the 100km/1 hour benchmark is an overly conservative estimate of an airport's typical catchment area and that the rule of thumb should be extended to 200km or 2 hours for leisure non-time sensitive passengers (66).

(82) The Commission considers that it continues to be appropriate to use a first proxy based on a catchment area of 100km/1 hour drive time for determining whether airports appear prima facie as substitutable (67). This does not mean that, if airports do not fall within this benchmark, such airports would not be substitutable. It also does not mean that once airports fall within this benchmark, they are to be automatically treated as substitutable.

a) Appropriateness of using an "airport catchment area" filter

(83) As explained by the Civil Aviation Authority of the United Kingdom (the "United Kingdom CAA") in its study Catchment area analysis (the "Catchment area analysis"), a catchment area is "a way of estimating the geographic area from which a large proportion of an airport's outbound passengers originate, or inbound passengers travel to, and their geographic distribution within this area. The size of catchment areas and overlaps between catchment areas of neighbouring airports could provide useful evidence of the potential for, and the strength of competition between these airports" (68).

(84) The Commission considers that catchment areas provide useful evidence for determining whether flights between two airports and the same destination could be part of the same relevant market for competition law purposes. However such evidence would need to be complemented with other elements which are relevant for determining whether in the eyes of passengers, for airports situated in the same catchment area (or outside such catchment area), particular scheduled point-to-point flights from one airport to a given destination would be substitutable for flights from another airport to the same destination. This approach was followed by the Commission in its 2007 Decision (69). Therefore, contrary to Ryanair's contentions, the Commission's decisional practice cannot create any presumption that where airports are outside the 100km/1 hour benchmark, they are unlikely to be considered substitutes for competition purposes, let alone a strong rebuttable presumption.

(85) Indeed, as explained by the United Kingdom CAA in its Catchment area analysis, catchment areas by themselves should not be taken as establishing which airports are within the same geographic market, for two main reasons (70) : (i) while on the one hand, catchment areas illustrate overlaps between all airports in the regions being considered, there could be fewer airports in the same relevant market than indicated by the catchment area analysis because the services offered at these airports may be differentiated and (ii) on the other hand, catchment areas might in some circumstances be smaller than the relevant geographic market and even if two neighbouring airports with similar services and prices might have relatively small catchment areas with little or no overlaps, an increase in price or reduction in service quality might result in a significant shift of passengers from one airport to another, suggesting they might be in the same (geographic) market.

b) Appropriateness of using a 100km/1hour benchmark

(86) The United Kingdom CAA explains in its Catchment area study that surface access drive time seems to be an important consideration in passengers' choice of airport and that an analysis of surface access travel time can provide an indication of the potential usage of an airport and the potential for passengers to make choices between different airports (71). The United Kingdom CAA has previously used for indicative purposes, 120 minutes to construct the drive time isochrones for United Kingdom short-haul leisure passengers and 60 minutes for United Kingdom international short-haul business passengers (72) and the United Kingdom Office of Fair Trading (the "OFT") also previously used drive time isochrones of 60, 90 and 120 minutes (73).

(87) The Commission considers that the 100km/1hour benchmark, while being conservative for the purpose of defining the appropriate catchment area in the Transaction, continues to be a relevant first proxy. In the majority of cases, as appears from its' own commercial behaviour, Ryanair seeks and manages to attract passengers which are bound for major cities (74). Indeed Ryanair generally markets itself toward price-sensitive leisure travellers who are largely indifferent to using secondary airports and are generally less time-sensitive (75). It therefore appears that at least Ryanair's passengers generally accept a longer total journey time in exchange for a competitive airline fare. These passengers seem to be willing to undertake longer journey times in surface transportation to and from other airports if this would lower the total transport costs between the point of departure and the final point of destination.

(88) In its airport-by-airport analysis, the Commission ascertains whether a given airport falls within the 100km/1 hour benchmark in the following manner:

(89) The distance in kilometres, as provided by Ryanair in the Form CO and as complemented by the Commission where necessary, was calculated as a distance from each individual airport served by Ryanair to the city centre of the closest city to Aer Lingus' served airport (76).

(90) As regards the duration, in its assessment of whether a certain airport falls within the one hour drive time benchmark, the Commission scrutinised the time needed to travel by car as well as the time needed to travel via public transport such as bus (city bus, shuttle bus, coach) and train and metro (if available) on the basis of the information provided by Ryanair and complemented by the Commission where necessary. Therefore the Commission captured (insofar as provided by Ryanair) all possible means of accessing the city centre from a given airport.

(91) As regards the use of a benchmark based on distance and drive time to the city centre, the Commission explained in its 2007 Decision that, by using this approach, if the centre of a city falls within the benchmark, a substantial part of the city and any suburbs or other urban areas located in between is included in this catchment area (77). Furthermore, where the catchment areas overlap over densely-populated areas, the number of inhabitants, potential air passengers who would consider flying from either airport, is substantial. If this number is high enough, carriers serving one of the two or more overlapping airports, will take this into account when setting the level of their own fares (78). This approach continues to be relevant for the purposes of the Transaction.

(92) Most of the airports considered in this Decision are located close to a large city that is likely to be the destination of most passengers, with the secondary airport typically being further away from that city than the primary airport. This is the case for example for Barcelona El Prat/Girona/Reus, Brussels Airport/Charleroi, Glasgow/Prestwick, Frankfurt/Hahn, the London airports, Milan Linate/Malpensa/Bergamo, Paris Roissy-Charles de Gaulle/Beauvais/Orly, Rome Fiumicino/Ciampino, Stockholm/Skavsta, Venice/Treviso or Warsaw/Modlin.

(93) However the 100km/1 hour benchmark is also relevant for those airports which Ryanair claims primarily serve different cities, for example the airports serving Bilbao/Santander, Vienna/Bratislava, Birmingham/East Midlands, Manchester/Liverpool, Toulouse/Carcassonne or Munich/Memmingen. The Commission considers that, even for these pairs, it is likely that a substantial part of the first city is captured by the catchment area of the other airport (or vice-versa) and that an airline which serves one of the two airports will take the other into account when setting the level of its own fares. Indeed, for the routes between Dublin, Cork, Shannon or Knock (as applicable) and the airports which Ryanair claims are primary airports for major cities, at least Aer Lingus (or Aer Arann) monitors Ryanair's fares to the other airport (79). Furthermore, it seems unlikely for passengers not to consider alternative flights from the same origin to other airports in the area merely because such alternative flights would be operated to another airport serving a different main city. This is particularly true when the distance is small (as is the case for Birmingham/East Midlands or Manchester/Liverpool). In addition customers in the airline industry face great transparency on the market and they can easily compare fares even from alternative airports (80). Therefore, it becomes increasingly easy for the customers to see alternative routes to the same city or region.

(94) In addition, the Commission found in its 2007 Decision that even for some holiday destinations (like Bilbao/Santander, Toulouse/Carcassonne or even Vienna/Bratislava) the city centre criterion is an indicative benchmark showing the relative distance of the airports from the local centre and thus also the difference in their ability to serve the tourist resorts in the vicinity. The Commission considers that this approach is still relevant for the purpose of thes Transaction as explained in more depth in the airport-by-airport assessment in Section 7.3.3.

(95) The Commission also points out the fact that the actual duration of a flight or the frequency of flights on a certain route are not crucial elements in establishing the catchment area of an airport for the purposes of the Transaction. The Commission acknowledges that passengers would be more likely to travel for considerably longer in order to reach an airport serving a long-haul destination and that this is less true for short-haul destinations (81). However, as explained above, considering the passenger profile in the Transaction, the 100km/1 hour benchmark remains suitable as explained above. Furthermore, the question of frequencies is relevant when assessing the strength of the competitive constraints and less so in determining whether flights to and from one airport could be substitutable with flights to and from another airport serving the same destination (82).

(96) The ultimate purpose of using the benchmark is to provide a starting point for the Commission to ascertain, on the basis of all information available, whether air passengers would consider air transport services to/from neighbouring airports as an alternative for a given route. Therefore, even if airports would fall outside the 100km/1 hour benchmark (which is only the case for a few of the airport pairs in the Transaction (83)), the Commission considers that on the basis of the bundle of evidence it can reach its conclusions. Therefore, the benchmark is merely one parameter in the bundle of evidence used by the Commission.

(97) To conclude, establishing a 100km/1 hour benchmark is an indicative first step and does not eliminate the need for detailed case-by-case analysis concerning the individual airport pairs. The Commission needs to consider all relevant elements as will be shown in the airport-by-airport descriptions in Section 7.3.3.

7.3.2.3. Ryanair's and Aer Lingus' Price Monitoring

(98) The Commission further investigated whether Ryanair and Aer Lingus monitor each other's fares on the same routes (to nearby airports) or if they take into account these fares in their normal course of business, for example when setting their own fares.

(99) [Ryanair's monitoring practices]*, it has continuously used "QL2" price comparison software [Ryanair's monitoring practices]*. The QL2 software gathers pricing across a range of dates at regular intervals. [Ryanair's monitoring practices]* (84). Ryanair's monitoring activity will be described in detail for each airport pair individually in Section 7.3.3.

(100) Aer Lingus [...]* monitors Ryanair's prices on all short-haul routes operated by Ryanair where these overlap with a route operated by Aer Lingus from Dublin, Cork, Knock or Shannon to the same or a nearby airport. For this it uses several systems (85) : it uses a Sabre revenue management system which allows it to monitor competitor pricing in real time and to amend their own pricing as required. In addition, Aer Lingus uses [...]*, QL2, the "screen scraping" software which harvests fares from competitor websites. Aer Lingus runs reports from QL2 and benchmarks Aer Lingus prices against Ryanair's. Furthermore, Aer Lingus monitors the Ryanair website manually to have a full understanding of the product attributes offered by Ryanair and to monitor promotions and special events. Aer Lingus explained that it needs to determine the product offered by Ryanair in the base fare and relevant ancillary charges, in order to ensure that Aer Lingus offers a comparable product at a competitive price and to adjust its prices accordingly in order to remain competitive.

(101) Aer Lingus itself does not monitor Ryanair prices on routes served by Aer Arann under the Aer Lingus Regional brand. Revenue management on these routes is the responsibility of Aer Arann and Aer Arann uses the Aer Lingus revenue management system to do so. Aer Arann monitors Ryanair's fares on all routes on which Aer Arann operates in competition with Ryanair. It conducts the research manually on Ryanair's website on a daily or weekly basis (86).

(102) [Ryanair and Aer Lingus monitoring practices]* Furthermore, such evidence shows that Ryanair and Aer Lingus (including Aer Arann) largely compete for the same customers.

(103) Furthermore, it is sufficient that only one of the two carriers (Ryanair or Aer Lingus/Aer Arann) monitors the other on a given route to confer significant relevance to the monitoring criterion for the purpose of airport substitutability. In this sense, the fact that only one of the two Parties monitors the other and considers the fares of the other in setting its own fares on a given route is still relevant evidence showing that it is being constrained in its competitive conduct. Therefore, there is a high likelihood that air passenger transport services on the same city pair (at different airports) would be part of the same relevant market for the purpose of the competitive analysis of the Transaction, and even more so when other evidence also points to the same direction.

7.3.2.4. Ryanair's marketing practices

(104) The Commission further investigated Ryanair's practices in marketing its services from and to certain airports. These marketing practices indicate whether Ryanair itself considers that a certain airport is appropriate for passengers destined for a certain region or city. Furthermore, Ryanair uses these practices to better inform its passengers and to make it easier for customers who might be interested in booking a flight with Ryanair to identify the flight destinations available and to dispel any uncertainties as regards the means of reaching a certain city or region from the airport which is served by Ryanair (87).

(105) In this sense, the Commission considers that it is most relevant to look at Ryanair's and not Aer Lingus' marketing practices because Ryanair typically serves the secondary airports and needs to promote flights to these airports as suitable alternatives to flights operated by Aer Lingus to a different airport which is frequently more conveniently located.

(106) The scrutiny of such practices was conducted on the basis of the information provided by Ryanair on its website on the dedicated page entitled "Destinations". The Commission analysed the description of the city, the airport information, the main attractions and the "Top 5 things to do".

(107) The Commission would like to point out that, for a number of airports, Ryanair flies to both airports which might be considered as substitutable (either from Dublin or from other Irish airports). This is the case in particular for the United Kingdom cities Manchester, Liverpool, Leeds Bradford, Birmingham, East Midlands but also for some cities in Spain like Alicante, Murcia, Bilbao and Santander. Therefore as regards these cities (and in particular the United Kingdom cities), Ryanair markets these cities as destinations in their own right. However this does not mean that from the perspective of Ryanair's passengers, these airports would not be substitutable, particularly if considering other qualitative and quantitative evidence as described in Section 7.3.3.

7.3.2.5. Outcome of the market investigation

(108) The Commission contacted relevant stakeholders in order to gather their views as regards the question of airport substitutability during its first phase and second phase investigations.

(109) The Commission paid particular attention to the views expressed by airlines and airports. As regards airlines, the Commission analysed the response of the other airlines active on the specific route. When there were no such airlines, the Commission took into account the responses of other airlines with a presence at the airports involved (for example by having a hub or a base at the airport in question) or airlines which operated on the route in question in the past. As regards airports, the Commission considered the information received from the airports involved and, where available, from the civil aviation authorities.

(110) Lastly, the Commission also took into account the overall outcome of its market investigation and in particular the responses from other carriers, airports, travel agents, civil aviation authorities, consumer associations, trade associations and corporate customers.

7.3.2.6. Empirical analysis (price correlation) and Ryanair's Map-Based analysis

(111) The Commission also conducted a price correlation analysis which provides quantitative support to the conclusions reached by the Commission on market definition and in particular concerning the substitutability of neighbouring airports.

(112) The results of this analysis are included in Section 7.3.3, while explanations regarding the methodology and more detailed description of the results are provided in Annex I.

(113) Ryanair also submitted a map-based analysis based on illustrative price rises ("IPR") which according to Ryanair relates to the substitutability of airports within O&D pairs, as defined by the Commission in the decision opening the proceedings. The analysis was performed on five city pairs (Dublin to Paris, Frankfurt, Munich, Stockholm and Vienna) where Aer Lingus and Ryanair operate to different destination airports, and uses customer location data from Ryanair's booking system. The first goal of the analysis is to determine to what extent the catchment areas of Ryanair and Aer Lingus overlap in these city pairs on the basis of this booking data.

(114) The Commission demonstrates in Annex II that Ryanair's Map-Based analysis relies on several strong and arbitrary assumptions. First, the Commission does not consider that the assumptions made by Ryanair (notably the 100km benchmark and the assumptions on routes with competitors) provide a sufficiently meaningful comparator for measuring actual diversion ratios for the purposes of market definition and effects based analysis. Secondly, the results of the analysis are not robust to changes in the assumptions of the computation of the critical diversion ratio. Thirdly, the Commission cannot exclude that the sample provided is not a representative sample. Therefore, the Commission does not consider that the results of this analysis provide any evidence that the relevant airports belong to different markets; on the contrary, even under the strong assumptions made by Ryanair in several robustness checks they indicate the opposite.

7.3.2.7. No customer survey was necessary in this case

(115) As explained in Section 5 and contrary to what Ryanair alleges, the Commission considers that a survey similar to the one performed in 2007 would not have been useful in the case at hand for the purpose of market definition.

7.3.2.8. Argument raised by Ryanair: common sense and practical realities facing consumers

(116) Ryanair claims in its Response to the Statement of Objections that, as regards airport substitutability, the SO's conclusions are in several cases contrary to common sense and the practical realities facing consumers (88).

(117) Ryanair argues that passengers who purchase flights on Ryanair would not switch to Aer Lingus in response to a 5-10% price increase by Ryanair because the schedule and destination would not change and Ryanair will generally still be by a wide margin the cheapest available option. Ryanair argues that it is only Aer Lingus' passengers who are likely to decide whether to switch between Aer Lingus' service and a cheaper (but potentially less convenient) Ryanair alternative.

(118) Therefore, in Ryanair's view, Aer Lingus' and Ryanair's services belong to the same market if in the event of a 5-10% price increase by Aer Lingus, a sufficient number of Aer Lingus' passengers would switch to Ryanair such as to make the price increase unprofitable. Ryanair uses the following example to support its point: Aer Lingus' average fare from Dublin to Frankfurt Main Airport is EUR 89.76. A 10% price increase in Aer Lingus' fare would amount to around EUR 9. Ryanair then concludes that it is not credible that individual travellers would switch from Aer Lingus to Ryanair, considering the increase in the total travel duration, the increased transportation costs and the significant differences in service levels between Ryanair and Aer Lingus.

(119) Aside from the fact that Ryanair's proposition is inconsistent with its own approach elsewhere (89) and the unsubstantiated claim that its own passengers would not switch to Aer Lingus (90), even if one would only assess the likelihood that Aer Lingus'passengers would switch to Ryanair, the example and explanations provided by Ryanair are not convincing.

(120) Generally, in cases of differentiated products, and notably in this case as Ryanair and Aer Lingus sometimes fly from different airports, differences in fares alone do not provide information on whether a product exerts a competitive constraint on the other. In their decision as to whether or not to switch, passengers would look at all cost elements as well as the fact that they would get a differentiated service. Considering that before an increase in Aer Lingus' fares, certain passengers are willing to pay a higher price to Aer Lingus for certain additional services (which they would not obtain on a Ryanair flight), there is no convincing evidence that Aer Lingus' passengers would value these services more than a price increase of EUR 9. In the Frankfurt example which Ryanair used to prove its assertions, it appears reasonable that a sufficient number of passengers would consider that an additional EUR 9 (10%) price difference is not justified for the services and convenience offered by Aer Lingus and would therefore choose to fly Ryanair instead. As explained in Section 7.3.3.6 the Commission has cogent evidence pointing to a single market for these two airports.

(121) Lastly, customers in the airline industry face great transparency on the market and they can easily compare fares from alternative airports (91). Furthermore, there is a wealth of information available on the internet as to the means and costs for reaching a city from an alternative airport.

7.3.2.9. Conclusion

(122) The Commission's analysis of airport substitutability has been carried out by taking into account the parameters and factors described in Section 7.3.2. The Commission has carefully analysed, interpreted and weighted all views expressed by the Parties and all other stakeholders during the market investigation.

(123) The Commission's investigation into the Transaction revealed information and evidence which sometimes pointed to different directions.

(124) However, the Commission considered all evidence which was available and it is by bundling the evidence that the Commission reaches its conclusions as regards airport substitutability.

7.3.3. Airport-by-airport assessment

(125) In the present case, the question of substitutability of scheduled air transport services from different airports is relevant for three reasons: first for determining to what extent the activities of the Parties overlap; secondly for the assessment of competitive constraints from airlines operating at other airports and thirdly for the assessment of entry prospects at other relevant airports.

(126) The activities of Ryanair currently overlap with Aer Lingus on 16 routes, listed in Table 1, on which Aer Lingus and Ryanair fly between the same two airports and on which there is no issue of airport substitutability because there are no other identified relevant airports.

Table 1: Overlap routes for which there is no airport substitutability issue

(1) Dublin-Berlin (92) / (7) Dublin-Lanzarote / (13) Cork-Malaga

(2) Dublin-Budapest / (8) Dublin-Madrid / (14) Cork-Faro

(3) Dublin-Faro / (9) Dublin-Malaga / (15) Cork-Lanzarote

(4) Dublin-Fuertaventura / (10) Dublin-Marseille / (16) Cork-Palma

(5) Dublin-Gran Canaria / (11) Dublin-Nice

(6) Dublin-Ibiza / (12) Dublin-Palma

(127) On 11 additional routes, although there is more than one airport in the origin or destination city, Ryanair and Aer Lingus fly to the same airport ("airport pairs"). For some routes, the other airport is equally served by one of the Parties or by another airline. For these routes, airport substitutability would be relevant in particular for the assessment of entry projects of Parties' potential competitors.

(128) As the Commission has not identified any entry or expansion plans, which would lead to timely, likely and sufficient entry or expansion to dispel the competition concerns identified for the relevant routes, it is not necessary to reach a definitive conclusion as regards the substitutability of these airports for routes from Dublin, Cork and Shannon as relevant.

(129) For one additional route, Dublin-Bristol/Cardiff/Exeter, the question of airport substitutability is relevant only for ascertaining whether the Parties would face competition constraints from flights from Exeter to Dublin.

(130) The routes for which airport substitutability is relevant for entry or for competition are listed in Table 2.

Table 2: Airport pair overlap routes (substitutability relevant for entry or competition)

Route / Airport to which both Ryanair and Aer Lingus fly / Name of the Party or third party who flies to an additional airport and name of relevant airport

(1) Dublin - Alicante / Murcia / Alicante / Ryanair to Murcia

(2) Dublin - Barcelona El Prat/ Girona / Reus / Barcelona El Prat / Ryanair to Girona and Reus

(3) Dublin - Birmingham/East Midlands / Birmingham / Ryanair to East Midlands

(4) Dublin - Bristol/Cardiff/Exeter / Bristol / Aer Lingus to Cardiff (Aer Lingus Regional) Flybe to Exeter

(5) Dublin - Edinburgh / Glasgow / Edinburgh / Aer Lingus to Glasgow (Aer Lingus Regional)

(6) Dublin - London Heathrow / Gatwick / Luton / Stansted / City / Southend / London Gatwick / Ryanair to Stansted and Luton, Aer Lingus to Heathrow and Southend (Aer Lingus Regional), British Airways to Heathrow Cityjet to London City

(7) Dublin - Manchester / Liverpool / Leeds Bradford / Manchester / Ryanair to Liverpool and Leeds

(8) Dublin - Tenerife South / Tenerife North / Tenerife / South None

(9) Cork - Alicante / Murcia / Alicante / None

(10) Cork - London Heathrow/Gatwick/Stansted / London Gatwick / Aer Lingus to Heathrow, Ryanair to Stansted

(11) Cork - Tenerife South / Tenerife North / Tenerife South / None

(131) On the remaining 19 routes ("city pairs"), Ryanair and Aer Lingus fly from Ireland (Dublin, Shannon, Cork or Knock as relevant) to different destination airports. For these routes, in the absence of airport substitutability the Parties would operate in different relevant markets. Airport substitutability is therefore relevant for establishing whether there is an overlap on these routes.

(132) The Commission will therefore assess for these airport pairs, listed in Table 3, whether they are substitutable for the routes from Dublin, Cork, Shannon, and Knock as relevant:

Table 3: City pair overlap routes (substitutability relevant for overlap) Route Ryanair flies to Aer Lingus flies to

(1) Dublin - Bilbao / Santander Santander Bilbao

(2) Dublin - Brussels / Charleroi Charleroi Brussels

(3) Dublin - Frankfurt / Hahn Hahn Frankfurt

(4) Dublin - Glasgow / Prestwick Prestwick Glasgow International

(5) Dublin - Milan Malpensa / Milan Linate / Bergamo Bergamo Malpensa and Linate

(6) Dublin - Munich / Memmingen Memmingen Munich

(7) Dublin - Paris CDG/ Paris Beauvais/Orly93 Beauvais Paris CDG

(80) Dublin - Rome Ciampino / Rome Fiumicino Ciampino Fiumicino

(9) Dublin - Stockholm Arlanda / Skavsta Skavsta Arlanda

(10) Dublin - Toulouse / Carcassonne Carcassonne Toulouse

(11) Dublin - Venice / Treviso Treviso Venice

(12) Dublin - Vienna / Bratislava Bratislava Vienna

(13) Dublin - Warsaw / Warsaw Modlin Modlin Warsaw

(14) Cork - Barcelona El Prat/Girona/Reus Girona and Reus El Prat

(15) Cork - Manchester / Liverpool Liverpool Manchester

(16) Knock - Birmingham / East Midlands East Midlands Birmingham

(17) Knock - London Heathrow / Gatwick / Luton / Stansted / City / Southend94 Luton and Stansted Gatwick

(18) Shannon - London Heathrow / Gatwick / Luton / Stansted / City / Southend95 Gatwick and Stansted Heathrow96

(19) Shannon - Manchester / Liverpool Liverpool Manchester 93 In this context, it is noted that substitutability of Orly would be relevant for potential new entry. 94 In this context, it is noted that substitutability of other London airports would be relevant for potential new entry. 95 In this context, it is noted that substitutability of other London airports would be relevant for potential new entry. 96 Aer Lingus also flew to London Gatwick in the IATA summer season 2011; however it did not fly to this airport in the IATA winter season 2011/12 and previously (except for a limited number of passengers to London Gatwick in IATA winter season 2010/11. EN 39 EN

(133) The detailed description of each individual airport pair is assessed in Sections 7.3.3.1 to 7.3.3.18. 7.3.3.1. Barcelona-El Prat Airport / Girona-Costa Brava Airport / Reus Airport

(134) Ryanair operates a service from Dublin to Girona-Costa Brava Airport (GRO airport), Reus Airport (REU airport) and Barcelona-El Prat Airport (BCN airport) while Aer Lingus operates a service from Dublin to BCN airport.

(135) Ryanair contends that, since GRO airport and REU airport are a substantial distance from Barcelona city centre, they are not substitutable for services at BCN airport97.

(136) The Commission concluded in its 2007 Decision that scheduled point-to-point passenger air transport services between Dublin on the one hand and BCN airport, GRO airport or REU airport, on the other, belong to the same market98. The Commission drew this conclusion by taking account of a number of elements including the market investigation, Ryanair's marketing practices and the price correlation analysis.

(137) In its Decision in case M.5747 Iberia/ British Airways99, the Commission left open the question whether BCN airport, GRO airport and REU airport belong to the same market for flights out of London while at the same time indicating that "several respondents consider that non-time sensitive passengers who focus on low airfares will find these three airports substitutable to each other"100. In its Decision in case M.5364 Iberia/Vueling/Clickair101, the Commission found that there was a significant number of passengers of the parties in that transaction for which BCN airport, GRO airport and REU airport were not substitutable for a number of domestic and European routes102.

(138) The question of airport substitutability as regards BCN, GRO and REU airport is relevant for the route ex-Cork Airport, as Ryanair and Aer Lingus operate to/from different airports. For the route ex-Dublin Airport, Ryanair and Aer Lingus both fly to BCN airport (and in addition Ryanair flies from GRO airport and REU airport).

(139) Table 4 summarises the following: (i) the distance from BCN airport, GRO airport and REU airport to Barcelona city centre (in km and time), (ii) the destination airport and frequency103 of both Ryanair and Aer Lingus on this route and (iii) the total travel time from city centre to city centre. 97 Annex 7.7.a to the Form CO. 98 See recitals 226-233 of the 2007 Decision. 99 Case M.5747 Iberia/ British Airways, recital 33. 100 Case M.5747 Iberia/ British Airways, recital 32. 101 Case M.5364 Iberia/Vueling/Clickair, recitals 56-63. 102 Not involving routes to or from Dublin though. 103 Frequency per week, one way. Applies to all tables in this Section. EN 40 EN Table 4: Barcelona-El Prat Airport / Girona-Costa Brava Airport / Reus Airport Dublin / Cork - Barcelona and Girona and Reus Relevant airport(s) Barcelona (BCN) Girona (GRO) Reus (REU) Distance and travel times to Barcelona city centre BCN - 13 km GRO - 100 km REU - 80 km Car: BCN - 25 mins; REU - 70 mins; GRO - 60 mins Bus: BCN - 25 min; REU - 80 mins; GRO - 70 mins Train: BCN - 20 mins; REU - none; GRO - none. Airlines active on route from Dublin (and airport served) Ryanair: 7 per week ("pw") from BCN (and 4 pw from GRO, 7 pw from REU) Aer Lingus: 14 pw from BCN Total Travel Time: Dublin City Centre to Barcelona City Centre Ryanair: via BCN 315 mins Ryanair: via GRO 330 mins Ryanair: via REU 375 mins Aer Lingus: via BCN 290 mins Airlines active on route from Cork (and airport served) Ryanair: 2 pw GRO (and 4 pw from REU) Aer Lingus: 3 pw BCN Total Travel Time Cork City Centre to Barcelona City Centre Ryanair: 335 mins to GRO Ryanair: 330 mins to REU Aer Lingus: 269 mins to BCN Source: Form CO, Annex 7.3.a

(140) On the basis of the 100km / 1 hour benchmark, BCN airport, REU airport and GRO airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services ex-Dublin or Cork in terms of distance and if the travel duration by car is considered. If transportation by bus is considered, the duration would be outside the Commission's benchmark (for REU airport also if travel time by car is considered).

(141) [...]*104. As explained above, Aer Lingus monitors the Ryanair prices on all shorthaul routes where Ryanair operates to the same or a nearby airport, including Ryanair's fares on Dublin-GRO airport, Dublin-REU airport, Cork-GRO airport and Cork-REU airport.

(142) Ryanair's marketing activities also indicate that Ryanair considers that GRO Airport and REU airport would be appropriate for passengers destined for Barcelona. Indeed, in its list of destinations on its website, under the destinations "Barcelona Girona" 104 Annex 7.7 to the Form CO. EN 41 EN and "Barcelona Reus", Ryanair describes the city of Barcelona and its main attractions105. In addition, in its airport information section, Ryanair provides information on the various public transport options for travelling to Barcelona106.

(143) The evidence collected through the market investigation also indicates that the Commission's conclusion in the 2007 Decision is still appropriate for the current Transaction, namely that scheduled point-to-point flights between Dublin on the one hand and BCN airport, GRO airport or REU airport, on the other, belong to the same market107 and that this conclusion is also relevant for flights to or from Cork. Indeed the operator of BCN airport, GRO airport and REU airport explained that "it can be said that Girona and Barcelona on one hand, and Barcelona and Reus on the other, share part of their catchment areas. In this sense, they share part of the market in the overlap of the catchment area. Nevertheless there are also areas in Girona, Barcelona and Reus that do not fall under the catchment area of the other two airports."108 Furthermore, IAG 109 and Vueling which has its main hub at BCN airport both confirm this view and Vueling further stated that "In our specific experience in the Catalonian region (BCN, GRO and REU), we believe that all these three airports work as a focal inbound and outbound point of entry for the city of Barcelona"110. The majority of respondents to the market investigation also share this view.

(144) Lastly, the market investigation has not revealed any indication that there have been significant changes in the market circumstances such as to warrant a different conclusion than the one reached in the 2007 Decision as regards these airports for flights ex Dublin111. 105 http://www.ryanair.com/en/flights-to-barcelona-reus/ and http://www.ryanair.com/en/flights-tobarcelona- girona/ 106 "Girona Airport is located just 18 Km from Girona city. Buses, operated by Sagales", connect the airport to Girona. The bus journey takes approximately 20-25 minutes and costs €2.50 one way. The bus stop is located outside arrivals. Alternatively, there is a taxi rank outside arrivals. A taxi to Girona takes approximately 15-20 minutes and costs €20-25." and "Reus Airport is located 5 km from the town of Reus. Reus Transport bus no.50 links the airport to the town centre; tickets cost €2.20 and the journey takes about twenty minutes. Hispano Igualadina operates a connection to Barcelona airport; the journey time is 1 hour and 30 minutes and tickets cost €12.00. There is also a bus from the airport to the coastal resorts of La Pineda, Salou and Cambrils; it takes about 30 minutes and costs €5.60. Buses leave from outside the Arrivals Terminal. For more information, visit www.reustransport.cat , ryanair.hispanoigualadina.com and www.planabus.com" http://www.ryanair.com/en/flights-tobarcelona- reus/ and http://www.ryanair.com/en/flights-to-barcelona-girona/ 107 See recitals 226-233 of the 2007 Decision. 108 Aena Aeropuertos, response to question 31.4.1 of questionnaire Q4 - Airports. At the same time Aena Aeropuertos clarified that the issue of substitutability requires a detailed analysis, as the answer not only depends on the catchment areas of the airports, but also on the passenger profile that potentially could use any of the mentioned airports. 109 IAG is the holding company of both British Airways plc ("British Airways") and Iberia Líneas Aéreas de España, S.A. ("Iberia"), (hereafter "IAG"), response to question 11.4 of questionnaire Q1 - Competitors. Iberia has large operations in Spain. 110 Vueling, response to question 11.4 of questionnaire Q1 - Competitors. 111 Responses to question 11.4.1 of questionnaire Q1 - Competitors, question 18.4.1 of questionnaire Q2 - Travel Agents, question 31.4.1 of Q4 - Airports and question 26.4.1 of questionnaire R1 - Corporate Customers. EN 42 EN

(145) The Commission's empirical analysis (price correlation) also provides evidence that air transport services between Dublin or Cork on the one hand and BCN airport, GRO airport or REU airport on the other hand belong to the same market (see further Annex I).

(146) Therefore, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin or Cork and BCN airport, GRO airport and REU airport, belong to the same market. 7.3.3.2. Bilbao Airport and Santander Airport112

(147) Ryanair operates a service from Dublin to Santander Airport (SDR airport) while Aer Lingus operates a service from Dublin to Bilbao Airport (BIO airport).

(148) Ryanair contends in its response to the decision opening the proceedings that there is no substitutability between BIO airport and SDR airport because of the following: (i) SDR airport is on the outer margin of the Commission's 100km/1 hour benchmark,

(ii) this benchmark is less instructive in light of the leisure nature of routes to SDR airport and BIO airport, (iii) SDR airport is not a "secondary" Bilbao airport but is rather the primary airport of Santander and the resort area West of Santander, (iv) the closest large city to BIO airport is Bilbao and multiple transport connections are in place between BIO airport and Bilbao, therefore it would be reasonable to assume that an appreciable amount of BIO airport's customers are travelling ex Bilbao, (v) the closest large city to SDR airport is Santander and multiple transport connections are in place between SDR airport and Santander, therefore it would be reasonable to assume that an appreciable amount of SDR airport's customers are travelling to or from Santander and (vi) other qualitative elements like Ryanair's marketing practices, Aer Lingus' marketing practices or the names of the airports.

(149) Table 5 summarises the following: (i) the distance from Bilbao Airport (BIO airport) and Santander airport (SDR airport) to Bilbao city centre (in km and time), (ii) the destination airport and frequency of each of Ryanair and Aer Lingus on this route and (iii) the total travel time from city centre to city centre. 112 Recitals 242-247 of the 2007 Decision concluded that there was substitutability between Bilbao Airport and Vitoria Airport. However for the purpose of the assessment of the current Transaction, the question of substitutability between Bilbao Airport and Vitoria Airport is not relevant as there are no operations between Dublin and Vitoria airport and there are no entry plans by competitors from Vitoria airport on this route. EN 43 EN Table 5: Bilbao Airport /Santander Airport Dublin - Bilbao and Santander Relevant airport(s) Bilbao (BIO) Santander (SDR) Distance and travel times to Bilbao city centre BIO - 13km SDR - 96km Car: BIO- 17 mins; SDR - 56 mins Bus: BIO - 20 mins; SDR - 65 mins (fastest). Train: None Airlines active on route (and airport served) Aer Lingus: 3 pw from BIO Ryanair: 2 pw from SDR Total Travel Time Dublin to Bilbao Aer Lingus from BIO - 270 mins Ryanair from SDR - 315 mins Source: Form CO, Annex 7.3.a

(150) On the basis of the 100 km / 1 hour benchmark, BIO airport and SDR airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin in terms of distance and if the travel duration by car is considered. If transportation by bus is considered, the travel time would be outside the Commission's benchmark.

(151) The fact that SDR airport is closer to the benchmark does not per se point to the absence of substitutability of the two airports for this Transaction, as the 100 km / 1 hour benchmark is less instructive when, as argued by Ryanair, two airports primarily serve a leisure destination. In fact, as mentioned above, even for these destinations the city centre criterion is an indicative benchmark showing the relative distance of the airports from the local centre and thus also the difference in their ability to serve the tourist resorts in the vicinity. For the Dublin-Bilbao/Santander route, the Cantabria region which lies along the coastal strip of Northern Spain stretches almost as far as Bilbao and at least three well-known tourist centres lie on the coastline between Bilbao and Santander, as presented by Ryanair on its website. These are Castro Urdiales, Santona and Laredo113.

(152) In the same manner, the fact that SDR airport is not a "secondary" airport to BIO airport and the fact that the names of the airports do not have a reference to the cities, in general cannot preclude a finding that SDR airport and BIO airport could be substitutable for scheduled point-to-point flights from Dublin, as explained in Section 7.3.2.2. The Commission needs to take account of all elements which are 113 "Other well-know tourist centres lie on the coastline: Castro Urdiales and Santona, famous for the anchovies; Laredo with it's 6 kilometer long beach; the fishing village of San Vicente de La Barquera; Comillas with the "Capricho", built by the world-known architect Gaudi; etc." [sic] http://www.ryanair.com/en/flights-to-santander/ EN 44 EN relevant for determining whether scheduled point-to-point flights between Dublin and BIO airport are substitutable for flights between Dublin and SDR airport.

(153) [...]*114. [...]*, as explained above, Aer Lingus does monitor Ryanair's prices on all short-haul routes where Ryanair operates to the same or a nearby airport, including Ryanair's fares on Dublin-SDR airport.

(154) Ryanair's marketing activities regarding its services to SDR airport are more targeted to passengers destined for Santander than for Bilbao. In its list of destinations on its website, under the destination "Santander", Ryanair describes the city of Santander and its main attractions.115 However, in its airport information section, Ryanair describes the various public transport options for travelling from SDR airport to Bilbao city centre.116

(155) The evidence collected through the market investigation indicates that these two airports would be substitutable. Indeed, the airport operator of both SDR airport and BIO airport explained that "it can be said that [...] Santander and Bilbao [...] share part of their catchment areas. In this sense, they share part of the market in the overlap of the catchment area. Nevertheless there are also areas in Bilbao, [...] and Santander that do not fall under the catchment area of the other two airports. "117. Furthermore, IAG (which has large operations in Spain and which, according to Ryanair, has a base at BIO airport) also considered that scheduled flights between Dublin and BIO Airport are substitutable with scheduled flights between Dublin and SDR airport for point-to-point passengers, further stating that "Bilbao is not substitutable with Victoria but is substitutable with Santander."118 Lastly, the majority of respondents in the market investigation also share this view.

(156) The Commission's empirical analysis (price correlation) also provides evidence that air transport services between Dublin on the one hand and BIO airport and SDR airport on the other belong to the same market (see further Annex I Price correlation analysis).

(157) In light of all the above, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin and BIO airport and SDR airport belong to the same market. 114 Ryanair's response to Commission's request of September 20, 2012 and Annex 7.7 to the Form CO. 115 http://www.ryanair.com/en/flights-to-santander/ 116 "Santander Airport is situated 6.5 Km outside of Santander city. There is a bus connection to the city centre, operated by Alsa; the journey time is 15 min; cost is €1.50. There is a connection to Bilbao from the airport, operated by Alsa. Journey time is 1 hour and 30 minutes. Both buses leave from outside the Arrivals Terminal." http://www.ryanair.com/en/flights-to-santander/ 117 Aena Aeropuertos, response to question 31.1.1 of questionnaire Q4 - Airports. At the same time Aena Aeropuertos clarified that the issue of substitutability requires a detailed analysis, as the answer not only depends on the catchment areas of the airports, but also on the passenger profile that potentially could use any of the mentioned airports. 118 IAG, response to question 13.1.1 of questionnaire Q1 - Competitors. EN 45 EN 7.3.3.3. Birmingham Airport and East Midlands Airport: Nottingham, Leicester & Derby

(158) Both Ryanair and Aer Lingus operate a service from Dublin to Birmingham Airport

(BHX airport) and in addition Ryanair operates a service to East Midlands Airport: Nottingham, Leicester & Derby (EMA airport). Furthermore, Ryanair operates a service from Knock to EMA airport while Aer Lingus operates a service from Knock to BHX airport.

(159) The question of airport substitutability as regards BHX airport and EMA airport is relevant for the route ex Knock Airport, as Ryanair and Aer Lingus operate ex different airports. For the route ex Dublin Airport, Ryanair and Aer Lingus both fly to BHX airport (and in addition Ryanair flies to EMA airport).

(160) Ryanair contends in its response to the decision opening the proceedings that there is no substitutability between BHX airport and EMA airport because (i) EMA airport falls close to the Commission's 1 hour travel time benchmark, (ii) the Commission's benchmark of 100km/1 hour is less instructive when two airports primarily serve different major cities as is the case with BHX airport and EMA airport, (iii) EMA airport is not a "secondary" Birmingham airport but rather the primary airport for three nearby cities (iv) the closest large city to BHX airport is Birmingham and multiple transport connections are in place between BMX airport and Birmingham,

(v) EMA airport is located centrally between three significant population centres Nottingham, Leicester and Derby and multiple transport connections are in place between EMA airport and these three cities while there is no public transport connections in place between EMA airport and Birmingham and (vi) other qualitative elements like Ryanair's marketing practices, Aer Lingus' marketing practices or the names of the airports.

(161) The Commission concluded in its 2007 Decision that scheduled point-to-point passenger air transport services between Dublin on the one hand and BHX airport or EMA airport on the other, belong to the same market119. The Commission drew this conclusion by taking account of a number of elements such as the market investigation, Ryanair's internal documents and the price correlation analysis on the routes Dublin-BHX airport and Dublin-EMA airport.

(162) In its Decision in Case No COMP/M.5403 - Lufthansa/British Midland120, the Commission concluded that BHX airport and EMA airport are not substitutable for flights ex Brussels. However, as the Commission recalled at recital 49 of that Decision, "airport substitutability cannot be assessed in the abstract but can only be determined taking into account the characteristics of the passengers travelling on the routes at stake." Indeed, in that Decision, the Commission thoroughly described the type of passengers travelling between Brussels and BHX airport and between Brussels and EMA airport (business) and their needs (particularly fast connectivity to the place of meetings). Therefore, the findings in that Decision were adapted to the specific 119 Recitals 136-142 of the 2007 Decision. 120 Commission's Decision of 14 May 2009 in Case No COMP/M.5403 - Lufthansa/British Midland, recitals 49 to 57. EN 46 EN circumstances of that case, to the routes in question (out of Brussels) and to a type of passenger that is not typical for Ryanair and Aer Lingus121.

(163) In the same manner, the OFT found in its decision on reference regarding the acquisition by Flybe Group Limited of the BA Connect business122 that for routes to Edinburgh and Glasgow, "services from EMA [airport] would provide at most a weak competitive constraint on Flybe's services from BHX [airport] post-merger." The OFT reached this conclusion on the basis of passenger information which showed that, on these routes, passengers living within one hour of BHX airport by and large used that airport for travel and that business passengers suggested that EMA airport was not a substitute for BHX airport. Lastly, the OFT also argued that, as the airline operating on the route operated services to both Edinburgh and Glasgow from both EMA airport and BHX airport, this suggested that the airline does not consider these airports to be close enough substitutes for a sufficient number of passengers. As with the Commission's Decision of 14 May 2009 in Case No COMP/M.5403 - Lufthansa/British Midland, the case investigated by the OFT concerned different markets, different carrier types and different customer groups than the ones relevant for this Transaction. The arguments used by the OFT are specific to United Kingdom internal routes between Edinburgh or Glasgow and BHX airport and EMA airport. Furthermore, these arguments by and large take into account passenger characteristics which are different to those of the typical passengers travelling with Ryanair or Aer Lingus on the route from Dublin or Knock123.

(164) Table 6 summarises the following: (i) the distance from BHX airport and EMA airport to Birmingham city centre (in km and time), (ii) the destination airport and frequency of each of Ryanair and Aer Lingus on the route to Dublin and Knock and

(iii) the total travel time from city centre to city centre for both Dublin and Knock. 121 Ryanair indicated that the Dublin-BHX/EMA airport route is a mixed route business/leisure, however the Knock-BHX/EMA route is mainly leisure according to Annex 7.3.a to the Form CO. Ryanair's recent offering on the Dublin-Brussels route targeting more business passengers does not sustain a similarity because it remains the case that the majority of Ryanair's passengers on Knock-BHX/EMA are leisure, hence non-time sensitive passengers. 122 OFT decision on reference under section 33(1) of the Enterprise Act 2002 given on 7 February 2007, recital 30. Full text of decision published 15 February 2007, http://www.oft.gov.United Kingdom/shared_oft/mergers_ea02/361227/Flybe.pdf. 123 Ryanair indicated that the Dublin-BHX/EMA airport route is a mixed route business/leisure, however the Knock-BHX/EMA route is mainly leisure according to Annex 7.3.a to the Form CO. EN 47 EN Table 6: Birmingham Airport / East Midlands Airport Dublin / Knock - Birmingham and East Midlands Relevant airport(s) Birmingham International (BHX) East Midlands (EMA) Distance and travel times to Birmingham city centre BHX - 13km EMA - 64km Car: BHX - 15 mins; EMA - 52 mins Bus: BHX - 25 mins; EMA - no service Train: BHX - 17 mins; EMA - 74 mins Airlines active on route from Dublin (and airport served) Aer Lingus: 27 pw from BHX Ryanair: 20 pw from BHX (and 14 pw from EMA) Total Travel Time from Dublin city centre to Birmingham city centre Ryanair DUB-BHX 202 mins Aer Lingus DUB-BHX 202 mins Ryanair DUB-EMA 269 mins Airlines active on route from Knock (and airport served) Aer Lingus: 7 pw from BHX (operated by Aer Arann) Ryanair: 5 pw from EMA Total Travel Time Knock to Birmingham City Centre Aer Lingus (via BHX) 217 mins Ryanair (via EMA) 232 mins Source: Form CO, Annex 7.3.a

(165) On the basis of the 100 km / 1 hour benchmark, BHX airport and EMA airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services ex Dublin and ex Knock.

(166) The fact that the benchmark is less instructive when two airports primarily serve different major cities, as argued by Ryanair, does not per se point to the absence of substitutability of the two airports for the Transaction. In the same manner, the fact that EMA airport is not a "secondary" airport to BHX airport but a primary airport for three nearby cities and the fact that the names of the airports do not have a reference to the cities, in general cannot preclude a finding that BHX airport and EMA airport could be substitutable for scheduled point-to-point flights from Dublin or Knock, as explained in Section 7.3.2.2. The Commission needs to take account of all elements which are relevant for determining whether scheduled point-to-point flights between Dublin or Knock and BHX airport are substitutable for flights between Dublin or Knock and EMA airport. EN 48 EN

(167) [...]*124. [...]*, Aer Arann, operating under the Aer Lingus regional brand on the Knock-BHX airport route does monitor Ryanair's fares on the Knock-EMA airport route125. As explained above, Aer Lingus monitors the prices of Ryanair on all shorthaul routes where Ryanair operates to the same or a nearby airport.

(168) As explained in Section 7.3.2.4, Ryanair's marketing practices are less relevant for its flights to United Kingdom airports such as BHX airport or EMA airport because Ryanair flies to both airports and by definition would try to advertise each airport individually. The Commission therefore needs to take into account other more relevant information.

(169) The evidence collected through the market investigation points to the fact that the Commission's conclusion in the 2007 Decision remains appropriate for the current Transaction, namely that scheduled point-to-point flights between Dublin on the one hand and BHX airport or EMA airport, on the other, belong to the same market126. The Commission has not identified any indications that the situation might be different for flights out of Knock. While the operator of BHX airport considered that Commission's conclusion in the 2007 Decision is no longer appropriate for the current Transaction, it only explained that "East Midlands and Birmingham International for this short-haul route (Birmingham Dublin) [and Birmingham Knock] serve separate market places."127 Flybe also disagrees with the Commission's view, however it nuances its response as regards routes ex Dublin and states that "[...] non- time sensitive (leisure) travellers may consider a switch (say from Birmingham to East Midlands) [...] In summary, and taking all of the above into consideration, we doubt that the 5-10% trigger for a change in fares would have a significant switching impact on passenger volumes."128 On the other hand, for the routes to/from Knock from BHX airport and EMA airport, Flybe explains that "Whilst there are [sic] much more non-time (leisure) sensitive travellers travelling in the Knock market, and one might consider the impact of fare changes to have a greater effect, we still feel that the price trigger of 5-10% would have a limited impact (given the low average yield levels) and that other considerations, such as flight timing and schedule frequency, will have as much influence as fares."129 Flybe further clarified in its response to the Commission's Statement of Objections130 that a 5-10% fare increase is unlikely to have a significant switching effect whether the airport at the other end is Dublin or Knock. Flybe further suggested that the question of market substitutability between Knock and these two airports can be left open. As explained above, the question of airport substitutability is particularly relevant for flights out of Knock because a finding of non-substitutability would lead to no overlap between the Parties' activities from Knock. 124 Ryanair's response to Commission request of September 20, 2012 and Annex 7.7 to the Form CO. 125 Aer Arann, response to question 22.1 of Q1 - Competitors and agreed minutes of conference call of 11 October 2012 with Aer Arann. 126 See recitals 136-142 of the 2007 Decision. 127 Birmingham Airport Limited, response to question 31.2 and 31.2.3 of Q4 - Airports. 128 Flybe, response to question 11.2 of Q1 - Competitors. 129 Flybe, response to question 11.2.3.1 of Q1 - Competitors. 130 Flybe, response to the Commission's Statement of Objections. EN 49 EN

(170) On the other hand, the operator of EMA airport confirmed that it is still appropriate to consider that BHX airport and EMA airport are substitutable for scheduled pointto- point passengers ex Dublin and Knock and explained that "The markets for [...], East Midlands/Birmingham [...] all significantly overlap."131

(171) Furthermore, the United Kingdom CAA also explained that "although the airports are imperfect substitutes, a significant proportion of passengers are likely to see the two airports as substitutes if an equivalent airline service was offered from both of them. Routes between Dublin, Cork and Knock, on the one hand, and Birmingham/East Midlands are short-haul flights, which means that passengers are not likely to be willing to travel particularly far to reach a substitute airport. However, given the high degree of catchment area overlap between Birmingham and East Midlands airports, a large proportion of passengers would have similar drive times to both airports and could therefore consider them to be substitutes. As a result, if the catchment areas of these two airports overlap for routes to Dublin, then they are likely to also overlap for routes to Cork and Knock."132 The majority of respondents in the market investigation also share this view as regards flights from Dublin or Knock.

(172) Lastly, the market investigation has not revealed any other indication that there have been significant changes in the market circumstances such as to warrant a different conclusion than the one reached in the 2007 Decision as regards these airports for flights ex Dublin133.

(173) The Commission's empirical analysis (price correlation) also provides evidence that air transport services between Dublin on the one hand and BHX airport or EMA airport on the other belong to the same market (see further Annex I Price correlation analysis). The Commission considers that such evidence is also relevant for air transport services from Knock, as the market investigation has not revealed substantial differences between Dublin and Knock.

(174) In light of all the above, the Commission takes the view that scheduled point-to-point passenger air transport services between (i) Knock and BHX airport and (ii) Knock and EMA airport belong to the same market while the question can be left open for air transport services between Dublin and BHX airport or EMA airport because the competitive assessment would not change irrespective of the precise market definition for this route (see Section 8.9.4.1). 7.3.3.4. Brussels Airport and Brussels South Charleroi Airport

(175) Ryanair operates a service from Dublin to Brussels South Charleroi Airport (CRL airport) while Aer Lingus operates a service from Dublin to Brussels Airport (BRU airport). 131 Manchester Airports Group, response to question 8.1 and 31.2 of Q4 - Airports. 132 United Kingdom Civil Aviation Authority, response to question 9.2.1 of questionnaire Q8 - Civil Aviation Authorities. 133 Responses to question 11.2.1 of questionnaire Q1 - Competitors, question 18.2.1 of questionnaire Q2 - Travel Agents and question 31.2.1 of questionnaire Q4 - Airports and question 26.2.1 of questionnaire R1 - Corporate Customers. EN 50 EN

(176) Ryanair submits that although it does not believe there is full substitutability between BRU airport and CRL airport, it accepts that they form part of a single market for the purposes of the competition assessment in the present case.

(177) The Commission concluded in its 2007 Decision that scheduled point-to-point passenger air transport services between Dublin on the one hand and BRU airport or CRL airport, on the other, belong to the same market134. The Commission drew this conclusion by taking account of a number of elements including the market investigation, Ryanair's marketing practices, its internal documents and the customer survey.

(178) In its Decision in case M.5335 Lufthansa/SN135, the Commission left open the question whether BRU airport and CRL airport would be substitutable for flights to and from various destinations in Germany and Switzerland. In its Decision in case M.5440 Lufthansa/Austrian136, the Commission found Ryanair's activities on route between Bratislava Airport and CRL airport did not constrain the parties' activities between Vienna Airport and BRU airport to any meaningful degree and that the market investigation in that transaction had confirmed that flights from Vienna airport to BRU airport and Bratislava airport to CRL airport were not in the same market for the purposes of that transaction. The issue at stake in that Decision was the question of the substitutability of flights to and from two primary airports

(Vienna Airport and BRU airport) with flights to and from two secondary airports

(Bratislava Airport and CRL airport). As the Commission explained in that Decision "the operations between two secondary airports represent a far less immediate constraint than operations between two primary airports."137 However, this issue is not at stake in the current Transaction as the question of airport substitutability is only relevant at one end of the route (Brussels) and not both.

(179) Table 7 summarises the following: (i) the distance between each of BRU airport and CRL airport to Brussels city centre (in km and time), (ii) the destination airport and frequency of each of Ryanair and Aer Lingus on this route and (iii) the total travel time from city centre to city centre. 134 Recitals 184-196 of the 2007 Decision. 135 Recitals 51-57. 136 Recital 259. 137 Recital 259. EN 51 EN Table 7: Brussels Airport / Brussels South Charleroi Airport Dublin - Brussels and Charleroi Relevant airport(s) Brussels Airport (BRU airport) Brussels South Charleroi Airport (CRL airport) Distance and travel times to Brussels city centre BRU airport- 12km CRL airport- 46km Car: BRU airport- 16 mins, CRL airport- 45 mins Bus: BRU airport- 18 mins; CRL airport- 45 mins Train: BRU airport- 17 mins; CRL airport- 50 mins

(via Charleroi Sud) Airlines active on route (and airport served) Aer Lingus: 18 pw from BRU airport Ryanair: 7 pw from CRL airport Total Travel Time Dublin City Centre to Brussels City Centre Aer Lingus via BRU airport: 237 mins Ryanair via CRL airport: 265 mins Source: Form CO, Annex 7.3.a

(180) On the basis of the 100km or 1 hour drive time benchmark, BRU airport and CRL airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services ex Dublin.

(181) [...]*138. As explained above, Aer Lingus also monitors the Ryanair prices on all short-haul routes where Ryanair operates to the same or a nearby airport, including Ryanair's fares on Dublin-CRL airport.

(182) Ryanair's marketing activities also indicate that Ryanair considers that CRL airport would be appropriate for passengers destined for Brussels. Indeed, in its list of destinations on its website, under the destination "Brussels Charleroi", Ryanair describes the city of Brussels and its main attractions139. In addition, in its airport information section, Ryanair mentions the distance to Brussels city and refers to the website of CRL airport140 which includes information on various transport options to Brussels city (so-called Brussels City Shuttle)141.

(183) The evidence collected through the market investigation also indicates that the Commission's conclusion in the 2007 Decision is still appropriate for the current 138 Annex 7.7 to the Form CO. 139 http://www.ryanair.com/en/flights-to-brussels-charleroi/ 140 "Brussels South Charleroi Airport is located approximately 8 km from the centre of Charleroi and 46 Km from the centre of Brussels. A bus connects the airport to the train station. There are rail connections to the rest of Belgium. Please check www.charleroi-airport.com for further information." http://www.ryanair.com/en/flights-to-brussels-charleroi/ 141 http://www.charleroi-airport.com/en/passengers/acces-and-parking/brussels-city-shuttle/index.html EN 52 EN Transaction, namely that scheduled point-to-point flights between Dublin on the one hand and BRU airport or CRL airport, on the other, belong to the same market142. Indeed the operator of BRU airport143, and Brussels Airlines144 which has a base at BRU airport, all confirm this view. The fact that Ryanair will increase its frequencies on the Dublin-CRL airport route during the Irish Presidency of the European Union is also an indication that BRU airport and CRL airport are substitutable145. In addition, the majority of respondents to the market investigation also agree with this view.

(184) Lastly, the market investigation has not revealed any indication that there have been significant changes in the market circumstances so as to warrant a different conclusion than the one reached in the 2007 Decision as regards these airports for flights ex Dublin146.

(185) The Commission's empirical analysis (price correlation) shows inconclusive results for this airport pair as explained in Annex I.

(186) Therefore, the Commission takes the view that scheduled point-to-point passenger air transport services between (i) Dublin and BRU airport and (ii) Dublin and CRL airport belong to the same market. 7.3.3.5. Glasgow Airport and Glasgow Prestwick Airport

(187) Ryanair operates a service from Dublin to Glasgow Prestwick Airport (PIK airport) while Aer Lingus operates a service from Dublin to Glasgow Airport (GLA airport).

(188) Ryanair considers that GLA airport and PIK airport are substitutable and part of the same market for the purpose of competitive assessment in this case147.

(189) The Commission concluded in its 2007 Decision that scheduled point-to-point passenger air transport services between Dublin on the one hand and GLA airport or PIK airport on the other, belong to the same market148. The Commission drew this conclusion by taking account of a number of elements like Ryanair's marketing practices, the market investigation and the customer survey.

(190) Table 8 summarises the following: (i) the distance from each of GLA airport and PIK airport to Glasgow city centre (in kilometres and time), (ii) the destination airport and frequency of both Ryanair and Aer Lingus on this route and (iii) the total travel time from city centre to city centre. 142 Recitals 184-196 of the 2007 Decision. 143 The Brussels Airport Company, response to question 31.5 of Q4 - Airports. 144 Brussels Airlines, response to question 11.5 of Q1 - Competitors. 145 Ryanair confirmed that the Irish Government accepted its proposal for low fares for civil servants and diplomats and the agreement between Ryanair and the Irish Government became effective as of 10 December 2012 (Ryanair's response to Commission's request for information of 9 January 2012). 146 Responses to question 11.5.1 of questionnaire Q1 - Competitors, question 18.5.1 of questionnaire Q2 - Travel Agents, question 31.5.1 of questionnaire Q4 - Airports and question 26.5.1 of questionnaire R1 - Corporate Customers. 147 Annex 7.3.a to the Form CO. 148 Recitals 149-156 of the 2007 Decision. EN 53 EN Table 8: Glasgow Airport / Glasgow Prestwick Airport Dublin - Glasgow and Prestwick Relevant airport(s) Glasgow Airport (GLA airport) Glasgow Prestwick Airport (PIK airport) Distance and travel times to Glasgow city centre GLA airport- 17 km PIK airport- 54 km Bus: GLA airport- 15 mins PIK airport - 55 mins Train: GLA airport none; PIK airport- 45 mins Airlines active on route (and airport served) Aer Lingus: 27 pw from GLA airport, operated by Aer Arann Ryanair: 11 pw from PIK airport Total Travel Time Dublin City Centre to Glasgow City Centre Aer Lingus: 215 mins Ryanair: 215 mins Source: Form CO, Annex 7.3.a and www.ryanair.com

(191) On the basis of the 100 km/1 hour benchmark, GLA airport and PIK airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services ex Dublin.

(192) [...]*149. Aer Arann, operating under the Aer Lingus regional brand on the route [...]*monitors Ryanair's fares150.

(193) Furthermore, Ryanair's marketing activities also indicate that Ryanair considers that PIK airport would be appropriate for passengers destined for Glasgow. Indeed, in its list of destinations on its website, under the destination "Glasgow Prestwick", Ryanair describes the city of Glasgow and its main attractions151. In addition, in its airport information section, Ryanair includes information on the various public transport options for travelling from PIK airport to Glasgow city152.

(194) The evidence collected through the market investigation also indicates that the Commission's conclusion in the 2007 Decision is still appropriate for the current Transaction, namely that scheduled point-to-point flights between Dublin on the one hand and GLA airport or PIK airport, on the other, belong to the same market153. Indeed, except for the operator of PIK airport which considers that "there is little 149 Annex 7.7 to the Form CO. 150 Aer Arann, response to question 22.1 of questionnaire Q1 - Competitors. 151 http://www.ryanair.com/en/flights-to-glasgow-prestwick/ 152 "Glasgow Prestwick International Airport is located 51 Km from Glasgow. There are rail, bus and taxi connections to Glasgow. Busses and trains leave every 30 minutes. The train takes 45 minutes and, with a 50% discount, costs approximately £2.70 one way. Two "Stagecoach" busses go to Glasgow city: the X77 (£4.15 one way, journey time is approximately 55 minutes) and the No.4 (£5.10 one way, journey time is approximately 90 minutes)." http://www.ryanair.com/en/flights-to-glasgow-prestwick/ 153 Recitals 149-156 of the 2007 Decision. EN 54 EN evidence to support this being the same market"154, the operator of GLA airport155 and Aer Arann156, the operator on this route, the majority of the other respondents to the Commission's market investigation157 confirm that scheduled point-to-point flights between GLA airport and PIK airport belong to the same market.

(195) Lastly, the market investigation has not revealed any indication that there have been significant changes in the market circumstances such as to warrant a different conclusion than the one reached in the 2007 Decision as regards these airports for flights to/from Dublin158.

(196) The Commission's empirical analysis (price correlation) shows that a conclusion can only be reached as regards the question of whether air transport services between Dublin on the one hand and GLA airport and PIK airport on the other belong to the same market after taking into account other qualitative evidence (see further Annex I Price correlation analysis).

(197) Therefore, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin and GLA airport and Dublin and PIK airport belong to the same market. 7.3.3.6. Frankfurt Airport and Frankfurt Hahn Airport

(198) Ryanair operates a service from Dublin to Frankfurt Hahn Airport (HHN airport) while Aer Lingus operates a service from Dublin to Frankfurt Airport (FRA airport).

(199) Ryanair contends in its response to the decision opening the proceedings that there is no substitutability between FRA airport and HHN airport mainly because HHN airport falls outside the Commission's 100km/1 hour benchmark and Ryanair's Map Based analysis indicates that FRA airport and HHN airport are not substitutable from a consumer perspective.

(200) The Commission concluded in its 2007 Decision that, for the purposes of that case, scheduled point-to-point passenger air transport services between Dublin, on the one hand, and FRA airport or HHN airport, on the other, belong to the same market159. The Commission drew this conclusion by taking account of a number of elements like Ryanair's marketing practices, Ryanair's competitive monitoring, a customer survey and the market investigation. 154 Infratil Airports Europe Ltd, response to question 33.3.1 of questionnaire Q4 - Airports. On the other hand, Infratil Airports Europe Ltd considered that there would be one way substitutability between Edinburgh airport and GLA airport for flights to and from Dublin (see response to question 33.2.2 of questionnaire Q4 - Airports. However, as explained in Section 7.3.3.18 below, it is not relevant to conclude on the question whether there is airport substitutability between Edinburgh airport and Glasgow Airport. 155 British Airport Authority, response to question 33.3 of questionnaire Q4 - Airports. 156 Aer Arann, response to question 13.3 of questionnaire Q1 - Competitors. 157 Flybe does not agree with the Commission's conclusion however, in its explanation it refers to timesensitive passengers only. Flybe, response to question 13.3.1 of questionnaire Q1 - Competitors. 158 Responses to question 13.3.1 of questionnaire Q1 - Competitors, question 20.3.1 of questionnaire Q2 - Travel Agents and question 33.3.1 of questionnaire Q4 - Airports and question 27.3.1 of questionnaire R1 - Corporate Customers. 159 Recitals 204-211 of the 2007 Decision. EN 55 EN

(201) In its Decision in case M.5335 Lufthansa/SN160, the Commission left open the question of whether FRA airport and HHN airport would be substitutable for flights to/from BRU airport and CRL airport respectively. In its Decision in case M.5440 Lufthansa/Austrian161, the Commission found that Ryanair's services between HHN airport and Bratislava airport did not constrain services between FRA airport and Vienna airport to any meaningful degree given that the total travel time would have been much longer than a trip involving Vienna airport and FRA airport, that the respondents to the market investigation in that case did not consider Bratislava airport as a substitute for Vienna airport and HHN airport as a substitute for FRA airport and lastly because of lack of data to prove the contrary. In that case, the Commission assessed the question of airport substitutability at both ends of the route, that is to say both at the Vienna end where it assessed the substitutability of Vienna airport with Bratislava airport and at the Frankfurt end where it assessed the substitutability of FRA airport with HHN airport and concluded that, for that specific case, FRA airport and HHN airport would not be substitutable. However, this issue is not at stake in the current Transaction as the question of airport substitutability is only relevant at one end of the route (Frankfurt). Furthermore, that Decision concerned two airlines with a different profile and targeting different types of customers than is the case in the Transaction.

(202) Table 9 summarises the following: (i) the distance from FRA airport and HHN airport to Frankfurt city centre (in km and time), (ii) the destination airport and frequency of Ryanair, Aer Lingus and other airlines flights' on this route and (iii) the total travel time from city centre to city centre. Table 9: Frankfurt Airport / Frankfurt Hahn Airport Dublin - Frankfurt and Hahn Relevant airport(s) Frankfurt Main (FRA) Frankfurt Hahn (HHN) Distance and travel times to Frankfurt city centre FRA - 9km HHN - 124km Car: FRA- 20 mins; HHN - 85 mins Bus: FRA - 15 mins; HHN - 105 mins Train: FRA - 12 mins; HHN - None. Airlines active on route

(and airport served) Ryanair: 4 pw (from HHN) Aer Lingus: 14 pw (from FRA) Lufthansa: 21 pw (from FRA) Total Travel Time Dublin to Frankfurt City Centre Ryanair: (via HHN) 345 mins Aer Lingus: (via FRA) 257 mins Lufthansa: (via FRA) 257 mins Source: Form CO, Annex 7.3.a 160 Recital 116. 161 Recital 155. EN 56 EN

(203) On the basis of the 100km or 1 hour drive time benchmark, FRA airport and HHN airport do not appear prima facie to be substitutable from the demand side for pointto- point scheduled passenger air transport services to/from Dublin.

(204) The fact that HHN airport falls slightly outside the benchmark does not per se point to absence of substitutability with FRA airport for scheduled point-to-point flights to and from Dublin as explained in Section 7.3.2.2. The Commission needs to take account of all elements which are relevant for determining whether scheduled pointto- point flights between Dublin and FRA airport are substitutable for flights between Dublin and HHN airport.

(205) [...]*162. Furthermore, Aer Lingus also monitors Ryanair's fares on the Dublin-HHN airport route163.

(206) Ryanair's marketing activities indicate that Ryanair considers that HHN airport would be appropriate for passengers destined for Frankfurt. Indeed, in its list of destinations on its website, under the destination "Frankfurt Hahn", Ryanair describes the city of Frankfurt and its main attractions164. In its airport information section, Ryanair refers to the shuttle bus services connecting HHN airport with Frankfurt city centre165.

(207) The evidence collected through the market investigation also indicates that the Commission's conclusion in the 2007 Decision is still appropriate for the current Transaction, namely that scheduled point-to-point flights between Dublin on the one hand and FRA airport or HHN airport, on the other, belong to the same market166. Indeed, the operator of FRA airport167, of HHN airport168 and Lufthansa, the other operator on this route169, all confirm this view in line with the majority of the other respondents to the Commission's market investigation.

(208) Lastly, the market investigation has not revealed any indication that there have been significant changes in the market circumstances such as to warrant a different 162 Ryanair's response to Commission request of September 20, 2012 and Annex 7.7 to the Form CO. 163 Aer Lingus' response to question 3 of the Commission's request for information of 20 September 2012. 164 http://www.ryanair.com/en/flights-to-frankfurt-hahn/ 165 "Frankfurt-Hahn is located approximately 124 km from the city of Frankfurt in the low mountain range area Hunsrück and the federal state Rheinland-Pfalz (Rhineland-Palatinate). There is no rail service to the airport. There is a coach service connecting Hahn Airport to Neu-Isenburg, Offenbach and Hanau for €8. Other shuttle bus services connect the airport with Frankfurt-city centre, Trier, Koblenz, Mainz, Worms and other towns. There are connections the gemstone-city Idar-Oberstein and the health resort Bad Kreuznach in the Nahe Valley. Bingen and Koblenz in the Middle Rhine Valley have shuttle bus connections as well as Bullay and Zell in the Moselle valley. Cars can be hired in the terminal building and taxis are available outside. The airport is reachable by car over the highway A 61, exit Rheinböllen, from where you take the trunk road B 50 to Hahn airport." http://www.ryanair.com/en/flights-to-frankfurt-hahn/ 166 Recitals 204-211 of the 2007 Decision. 167 Fraport AG, response to question 31.6 of questionnaire Q4 - Airports. 168 Flughafen Frankfurt-Hahn GmbH, response to question 31.6 of questionnaire Q4 - Airports. 169 Deutsche Lufthansa AG ("Lufthansa"), response to question 11.6 of questionnaire Q1 - Competitors. EN 57 EN conclusion than the one reached in its 2007 Decision as regards these airports for flights to/from Dublin170.

(209) The Commission's empirical analysis (price correlation) shows that a conclusion can only be reached as regards the question of whether air transport services between Dublin on the one hand and FRA airport and HHN airport on the other belong to the same market after taking into account other qualitative evidence (see further Annex I).

(210) As regards Ryanair's Map-Based analysis, it relies on several strong and arbitrary assumptions and the Commission does not consider that the results of this analysis provide any evidence that the relevant airports belong to different markets as explained in detail in Annex II171.

(211) Therefore, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin and FRA airport and Dublin and HHN airport belong to the same market. 7.3.3.7. London airports

(212) Ryanair operates from Dublin to London Gatwick (LGW airport), to Stansted (STN airport) and to Luton (LTN airport) and Aer Lingus from Dublin to London Heathrow (LHR airport) and Gatwick (LGW airport). Aer Arann operates from Dublin to London Southend (SEN airport) under the Aer Lingus Regional brand.

(213) Ryanair operates from Cork to LGW airport and STN airport. Aer Lingus operates from Cork to LHR airport and LGW airport.

(214) On the route to Shannon, Ryanair operates from LGW airport and STN airport while Aer Lingus operates from LHR airport.

(215) Lastly, Ryanair operates from Knock to LGW airport and Aer Lingus operates from Knock to LTN airport and STN airport.

(216) Ryanair submits that although it does not believe there is full substitutability between the various London airports, it accepts that they form part of a single market for the purposes of competition assessment in the present case172.

(217) The Commission concluded in its 2007 Decision that scheduled point-to-point passenger air transport services between Dublin, Cork or Shannon, and London Heathrow (LHR airport), Gatwick (LGW airport), Stansted (STN airport), Luton

(LTN airport) and City (LCY airport) airports (the "London Airports"), belong to the same market173. The Commission drew this conclusion by taking account of a 170 Responses to question 11.6.1 of questionnaire Q1 - Competitors, question 18.6.1 of questionnaire Q2 - Travel Agents and question 31.6.1 of questionnaire Q4 - Airports and question 27.3.1 of questionnaire R1 - Corporate Customers. 171 See Section 7.3.2.8 for the dismissal of Ryanair's argument that it is simply not credible that passengers would switch from Aer Lingus to Ryanair for a 10% price increase. 172 Annex 7.7.a to the Form CO. 173 Recitals 109-125 of the 2007 Decision. EN 58 EN number of elements such as the market investigation, the customer survey and other studies.

(218) In other previous Decisions the Commission concluded that the London Airports might be substitutable174 or left the question of substitutability open for some airports175. For example in its Decision in case M.5747 - Iberia/British Airways176 the Commission concluded that LHR airport, LGW airport and LCY airport are to be considered substitutable177 and it left open the question whether LTN airport and STN airport were also substitutable for LHR airport, LGW airport and LCY airport while also pointing out that the market investigation had provided indications that the London Airports might be substitutable due to the existing transport infrastructure and the availability of cheaper airfares at some airports which are less conveniently located. In its Decision in case M.6447 IAG/bmi, the Commission left open the question whether all London Airports (including SEN airport)178 are substitutable.

(219) Table 10 summarises the following: (i) the distance from the London airports and SEN airport to London city centre (in km and time), (ii) the destination airport and frequency of each of Ryanair and Aer Lingus on the route to Dublin, Cork, Shannon and Knock and (iii) the total travel time from city centre to city centre for Dublin, Cork, Shannon and Knock. Table 10: London airports Dublin / Cork / Shannon / Knock - London airports Relevant airport(s) Heathrow (LHR) Gatwick (LGW) Luton (LTN) Stansted (STN) London City (LCY) London Southend (SEN) Distance and travel times to London city centre LHR - 28 km LGW - 46 km STN - 59 km LTN - 51 km LCY - 14 km SEN - 68 km Car: LHR - 65 mins; LGW - 85 mins; STN - 85 mins; LTN - 44 mins; LCY - 20 mins; SEN - 64 mins 174 See also Commission's Decision of 10 December 2003 in Case No COMP/D2/38479 British Airways/Iberia/GB Airways, recital 21 and following of the summary of the Commission's assessment. 175 Commission's Decision of 22 September 1997 in Case No COMP/M.967 KLM/Air United Kingdom, recital 24. 176 Commission's Decision of 14 July 2010 in Case No COMP/M.5747 - Iberia/British Airways, recital 30. 177 In Commission Decision of 26 January 2011 in Case No COMP/M.5830 - Olympic/Aegean Airlines, recital 1644, it was concluded that LHR airport and LGW airport are substitutes for all passengers on the London-Athens route. 178 Commission's Decision of 30 March 2012 in case M.6447 IAG/bmi, recital 58. EN 59 EN Bus: LHR - 65 mins; LGW - 95 mins; STN - 90 mins; LTN - 80 mins; Train: LHR - 15 mins; LGW - 30 mins; STN - 46 mins; LTN - 26 mins plus 10 minute shuttle bus connection ; LCY - 22 mins; SEN - 53 mins Airlines active on Dublin route (and airport served) Aer Lingus: 89 pw from LHR, 37 pw from LGW, 21 pw from SEN

(Operated by Aer Arann) BA/Bmi: 37 pw from LHR Air France: 31 pw from LCY (operated by CityJet) Ryanair: 31 pw from LGW, 47 pw from STN, 21 pw from LTN Total Travel Time City Centre to City Centre from Dublin Aer Lingus: from LHR 215 mins, from LGW 215 mins, from SEN 283 mins BA/bmi: from LHR 220 mins Air France: from LCY 237 mins Ryanair: from LGW 230 mins, from STN 246 mins, from LTN 231 mins Airlines active on Cork route (and airport served) Aer Lingus: 28 pw from LHR, 6 pw from LGW Ryanair: 7 pw from LGW, 16 pw from STN Total Travel Time City Centre to City Centre from Cork Cork to London (via LHR) 205 mins Cork to London (via LGW) 220 mins Cork to London (via STN) 231 mins Airlines active on Shannon route Aer Lingus: 21 pw from LHR Ryanair: 7 pw from LGW, 14 pw from STN Total Travel Time City Centre to City Centre from Shannon Shannon to London (via LHR) 220 mins Shannon to London (via LGW) 235 mins Shannon to London (via STN) 246 mins Airlines active on Knock route (and airport served) Aer Lingus: 7 pw from LGW Ryanair: 5 pw from LTN, 7 pw from STN Total Travel Time Knock to London Aer Lingus via LGW 230 mins Ryanair via LTN 225 mins Ryanair via STN 235 mins Source: Form CO, Annex 7.3

(220) On the basis of the 100 km / 1 hour benchmark, the London Airports and SEN airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services ex Dublin, ex Cork, ex Shannon and ex Knock in terms of distance and travel time by train.

(221) [...]*179. [...]*Aer Lingus monitors Ryanair's operations to all the Ryanair London airports for its Dublin-LHR airport route as well as for its Dublin-LGW airport 179 Ryanair's response to Commission request of September 20, 2012 and Annex 7.7 to the Form CO. EN 60 EN route.180 [...]*, Aer Arann, operating under the Aer Lingus regional brand on the Dublin-SEN airport route does monitor Ryanair's fares181.

(222) Ryanair's marketing activities also indicate that Ryanair considers that the airports to which it operates (LGW airport, STN airport and LTN airport, the "Ryanair London airports") would be appropriate for passengers destined for London. Indeed, in its list of destinations on its website, under the destination "London Gatwick", "London Luton" and "London Stansted", Ryanair describes the city of London and its main attractions182. In addition, in its airport information section, Ryanair indicates the distance to the city of London and includes information on the various public transport options for travelling to London city.

(223) The evidence collected through the market investigation indicates that the Commission's conclusion in the 2007 Decision is still appropriate for the current Transaction, namely that scheduled point-to-point flights between Dublin, Cork or Shannon on the one hand and the London Airports, on the other, belong to the same market. The Commission has not identified any indications that the situation might be different for flights out of Knock. The United Kingdom CAA indicated that they had analysed the interactions and potential for competitive constraints between the four major London airports (excluding LCY airport) in two studies183. Overall, the United Kingdom CAA found that these airports have varying degrees of catchment area overlaps and substitutability. However, the CAA considers on the whole that, although the London airports are imperfect substitutes for scheduled point-to-point passenger air transport services to Dublin, the Commission's 2007 conclusion still appears to be appropriate for current Transaction and this should also cover flights from/to Knock184. British Airways185 and Air France/CityJet186 also agree with this conclusion (including for flights out of Knock). Furthermore the majority of the other respondents to the Commission's market investigation also confirm this view.

(224) Lastly, the market investigation has not revealed any indication that there have been significant changes in the market circumstances such as to warrant a different conclusion than the one reached in the 2007 Decision as regards these airports for flights to/from Dublin, Cork or Shannon187.

(225) The Commission acknowledges that the degree of substitutability between the London Airports might vary depending on which airports are precisely taken into consideration as regards flights to/from Dublin, Cork, Shannon or Knock. However, 180 Aer Lingus' response to question 3 of the Commission's request for information of 20 September 2012. 181 Aer Arann, response to question 22.1 of questionnaire Q1. 182 http://www.ryanair.com/en/flights-to-london-stansted/, http://www.ryanair.com/en/flights-to-londongatwick/ and http://www.ryanair.com/en/flights-to-london-luton/ 183 Catchment area analysis http://www.caa.co.United Kingdom/docs/5/Catchment%20area%20analysis%20working%20paper%20-%20FINAL.pdf and Passengers' airport preferences.http://www.caa.co.United Kingdom/docs/5/Passenger%20survey%20results%20-%20FINAL.pdf 184 UK CAA, response to question 11.2 of questionnaire Q8 - Civil Aviation Authorities. 185 IAG, response to question 13.2 of questionnaire Q1 - Competitors. 186 Air France, response to question 13.2 of questionnaire Q1 - Competitors. 187 Responses to questions 11.1.1 and 11.1.2 of questionnaire Q1 - Competitors, questions 18.1.1 and 18.1.2 of questionnaire Q2 - Travel Agents, questions 31.1.1 and 31.1.2 of questionnaire Q4 - Airports and question 26.1.1 and 26.1.2 of questionnaire R1 - Corporate Customers. EN 61 EN the Commission notes that in any event for flights from Dublin and Cork Ryanair and Aer Lingus both operate from LGW airport.

(226) The Commission's empirical analysis (price correlation) leads to inconclusive results as regards the question of whether air transport services between Dublin, Cork, Shannon or Knock on the one hand and all London Airports on the other belong to the same market (see further Annex I).

(227) As regards the substitutability of SEN airport with the other London Airports, the outcome of the market investigation was more nuanced. While a considerable number of respondents considered that SEN airport is substitutable with all London Airports, other respondents indicated some or the other London airports as substitutable. However, for the purpose of this investigation, the Commission can leave open the question whether SEN airport is part of the same market as the other London Airports as it would not change the outcome of the competitive assessment.

(228) Therefore, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin, Cork, Shannon or Knock and the London Airports belong to the same market, while the question of whether SEN airport should also be treated as substitutable with the other London Airports can be left open. 7.3.3.8. Manchester Airport and Liverpool John Lennon Airport

(229) Ryanair and Aer Lingus both operate a service from Dublin to Manchester Airport

(MAN airport) and in addition, Ryanair operates services to Liverpool John Lennon Airport (LPL airport) and Leeds airport (LBA airport). Furthermore, Ryanair operates a service from Cork and Shannon to LPL airport while Aer Lingus operates a service from Cork and Shannon to MAN airport.

(230) Ryanair contends in its response to the decision opening the proceedings that there is no substitutability between MAN airport and LPL airport because of the following:

(i) the Commission's benchmark of 100km/1 hour is less instructive when two airports primarily serve different major cities as is the case with MAN airport and LPL airport, (ii) LPL airport is not a "secondary" Manchester airport but rather the primary airport for Liverpool as well as nearby cities such as Warrington (iii) the closest large city to MAN airport is Manchester and multiple transport connections are in place between MAN airport and Manchester and (iv) other qualitative elements like Ryanair's marketing practices, the fact that Ryanair serves both MAN airport and LPL airport, Aer Lingus' marketing practices or the names of the airports.

(231) The Commission concluded in its 2007 Decision that scheduled point-to-point passenger air transport services between Dublin, Cork or Shannon, on the one hand, and MAN airport, LPL airport or LBA airport on the other, belong to the same market188. The Commission drew this conclusion by taking account of a number of 188 Recitals 126-135 of the 2007 Decision. For the current Transaction, the substitutability of Leeds airport is not relevant for the overlap analysis. For flights out of Dublin, only Ryanair also flies to Leeds airport, whereas there are no flights to Leeds airport from Cork or Shannon. EN 62 EN elements such as the market investigation, Ryanair's internal documents and the price correlation analysis on the Dublin-MAN airport and Dublin-LPL airport routes189.

(232) The question of airport substitutability as regards MAN airport and LPL airport is relevant for the routes to/from Cork Airport and to/from Shannon Airport, as Ryanair and Aer Lingus operate to/from different airports190. For the routes to/from Dublin Airport, Ryanair and Aer Lingus both fly to MAN airport (and in addition, Ryanair flies to LPL airport and LBA airport).

(233) Table 11 summarises the following: (i) the distance from MAN airport and LPL airport to Manchester city centre (in km and time), (ii) the destination airport and frequency of each of Ryanair and Aer Lingus on the route to Dublin, Cork and Shannon and (iii) the total travel time from city centre to city centre for Dublin, Cork and Shannon. Table 11 : Manchester Airport and Liverpool John Lennon Airport Dublin / Cork / Shannon - Manchester and Liverpool Relevant airport(s) Manchester (MAN) Liverpool (LPL) Distance and travel times to Manchester city centre MAN - 17km LPL - 52 km Car: MAN - 15 mins; LPL - 41 mins; Bus: MAN - 30 mins; LPL - 50 mins (from MAN) & 35 mins from LPL; Train: MAN - 18 mins; Liverpool City Centre - 67 mins

(from MAN); Airlines active on route from Dublin (and airport served) Aer Lingus: 24 pw from MAN Ryanair: 24 pw from MAN (and 24 pw from LPL) Total Travel Time from Dublin city centre to Manchester city centre Aer Lingus DUB-MAN 200 mins Ryanair DUB-MAN 195 mins Ryanair DUB-LPL 220 mins Airlines active on route from Cork

(and airport served) Aer Lingus: 16 pw from MAN, operated by Aer Arann Ryanair: 5 pw (from LPL) 189 The 2007 Decision clearly explains at recitals 130-131 why the conclusion reached by the Commission is not put into question by arguments which were discussed in the OFT decision on reference under section 33(1) of the Enterprise Act 2002, given on 7 February 2007 regarding the acquisition by Flybe Group Limited of the BA Connect business. Full text of decision published 15 February 2007, http://www.oft.gov.United Kingdom/shared_oft/mergers_ea02/361227/Flybe.pdf . These arguments are still valid for the current Transaction. In its Decision of 14 May 2009 in Case No COMP/M.5403 - Lufthansa/British Midland, the Commission left open the question of substitutability between Manchester Airport and Leeds Airport. 190 As there are currently no flights offered between Cork or Shannon and LBA airport and no likely, timely and sufficient entry has been identified on these routes, it is not relevant to assess the substitutability of LBA airport for these two routes while for the Dublin route the question can be left open as both Ryanair and Aer Lingus operate to MAN airport. EN 63 EN Dublin / Cork / Shannon - Manchester and Liverpool Total Travel Time from Cork city centre to Manchester city centre Cork to Manchester (via MAN) 218 mins Cork to Manchester (via LPL) 235 mins Airlines active on route from Shannon (and airport served) Aer Lingus: 13 pw from MAN (operated by Aer Arann) Ryanair: 3 pw (from LPL) Total Travel Time from Shannon city centre to Manchester city centre Shannon to Manchester (via MAN) 218 mins Shannon to Manchester (via LPL) 245 mins Source: Form CO, Annex 7.3.a as corrected (in underline)

(234) On the basis of the 100 km or 1 hour drive time benchmark, MAN airport and LPL airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin, to/from Cork and to/from Shannon.

(235) The fact that the benchmark is less instructive when two airports primarily serve different major cities, as argued by Ryanair, does not per se point to the absence of substitutability of the two airports for the purposes of the Transaction. In the same manner, the fact that LPL airport is not a "secondary" airport to MAN airport but a primary airport for Liverpool and the fact that the names of the airports do not have a reference to the cities, in general cannot preclude a finding that MAN airport and LPL airport could be substitutable for scheduled point-to-point flights from Dublin, Cork or Shannon as explained in Section 7.3.2.2. The Commission needs to take account of all elements which are relevant for determining whether scheduled pointto- point flights between Dublin, Cork or Shannon and MAN airport are substitutable for flights between Dublin, Cork or Shannon and LPL airport.

(236) [...]*191. [...]*, Aer Arann, operating under the Aer Lingus regional brand on the Cork-MAN airport route and on the Shannon-MAN airport route does monitor Ryanair's fares192. Furthermore, as explained above, Aer Lingus monitors Ryanair prices on all short-haul routes where Ryanair operates to the same or a nearby airport, including Ryanair's fares on Dublin-MAN airport, Dublin-LPL airport and Dublin-LBA airport.

(237) As explained in Section 7.3.2.4 Ryanair's marketing practices are less relevant for its flights to United Kingdom airports like MAN airport or LPL airport because Ryanair flies to both airports and by definition would try to advertise each airport individually. The Commission therefore needs to take into account other more relevant information.

(238) The evidence collected through the market investigation points to the fact that the Commission's conclusion in the 2007 Decision is still appropriate for the current Transaction, namely that scheduled point-to-point flights between Dublin, Cork or Shannon on the one hand and MAN airport or LPL airport, on the other, belong to 191 Ryanair's response to Commission request of September 20, 2012 and Annex 7.7 to the Form CO. 192 Aer Arann, response to question 22.1 of Q1 - Competitors. EN 64 EN the same market193. The operator of LPL airport considers that: "LPL-DUB, MANDUB and LBA-DUB services do not serve the same market. Similarly, we believe that LPL-SNN or LPL-ORK and LBA-SNN or LBA - ORK services serve separate markets etc."194 Flybe expressed a more nuanced view that "A combination of the additional transport costs and additional logistics adding to the total journey times are unlikely to be outweighed by the 5-10% increase in airline tariffs, particularly for time sensitive (ie business) travellers, on the Liverpool-Manchester-Leeds catchment."195 On the other hand, the operator of MAN airport confirmed that it is still appropriate to consider that MAN airport and LPL airport are substitutable for scheduled point-to-point passengers to/from Dublin, Cork or Shannon and explained that "The markets for Manchester/Liverpool, [...] all significantly overlap."196 Furthermore, Wizz Air, in its internal analysis of the Dublin-MAN airport route, considers as "competition" the Ryanair services both between Dublin-MAN airport and Dublin-LPL airport197 . The United Kingdom CAA also confirmed that: "The catchment area overlap is most significant between Manchester and Liverpool airports, which would suggest that these airports may be seen as substitutable by passengers in their respective catchment areas, were a suitable flight available."198 Furthermore, the majority of the other respondents to the Commission's market investigation also confirm this view.

(239) Lastly, the market investigation has not revealed any other indication that there have been significant changes in the market circumstances such as to warrant a different conclusion than the one reached in the 2007 Decision as regards these airports for flights to/from Dublin, Cork or Shannon199.

(240) The Commission's empirical analysis (price correlation) also provides evidence that air transport services between Dublin on the one hand and MAN airport and LPL airport on the other belong to the same market (see further Annex I)200. The Commission considers that such evidence is also relevant for air transport services from Cork and Shannon, as the market investigation has not revealed substantial differences between Dublin, Cork and Shannon.

(241) Therefore, the Commission takes the view that scheduled point-to-point passenger air transport services between Cork or Shannon and MAN airport and LPL airport belong to the same market. The question can be left open as regards substitutability of air transport services between Dublin and MAN airport, LPL airport or LBA airport as the Parties' activities overlap on the Dublin-MAN airport route. 193 Recitals 126-135 of the 2007 Decision. 194 Peel Airports Ltd, response to questions 31.1, 31.1.1 and 31.1.2 of questionnaire Q4 - Airports. 195 Flybe, response to questions 11.1, 11.1.1 and 11.1.2 of questionnaire Q1 - Competitors. 196 Manchester Airports Group, response to question 8.1 and 31.1 of Q4 - Airports. 197 Wizz Air's response to Commission's RFI of 27 November 2012. 198 United Kingdom Civil Aviation Authority, response to question 9.1.1 of questionnaire Q8 - Civil Aviation Authorities. 199 Responses to questions 11.1.1 and 11.1.2 of questionnaire Q1 - Competitors1, questions 18.1.1 and 18.1.2 of questionnaire Q2 - Travel Agents, questions 31.1.1 and 31.1.2 of questionnaire Q4 - Airports and question 26.1.1 and 26.1.2 of questionnaire R1 - Corporate Customers. 200 For Leeds, the results are inconclusive. EN 65 EN 7.3.3.9. Milano Linate and Milano Malpensa and Orio al Serio International Airport Il Caravaggio (Bergamo)

(242) Ryanair operates a service from Dublin to Orio al Serio International Airport Il Caravaggio (BGY airport) while Aer Lingus operates a service from Dublin to Milano Linate (LIN airport) and Milano Malpensa (MXP airport).

(243) Ryanair contends that LIN airport, MXP airport and BGY airport are not substitutable and do not overlap for competition purposes because there is a substantial distance between these destination airports201.

(244) The Commission concluded in its 2007 Decision that scheduled point-to-point passenger air transport services between Dublin, on the one hand, and LIN airport, MXP airport or BGY airport on the other, belong to the same market202. The Commission drew this conclusion by taking account of a number of elements such as the market investigation and Ryanair's marketing. In its Decision in Case M.3280 Air France /KLM203, the Commission concluded that LIN airport and MXP airport are substitutable and that "Bergamo is generally considered an option only for non-time sensitive passengers." In its Decision in Case M.5335 Lufthansa/SN204, the Commission acknowledged that Milan is served by three airports (LIN airport, MXP airport and BGY airport), however ultimately left the question open whether these airports were substitutable.

(245) Table 12 summarises the following: (i) the distance from LIN airport, MXP airport and BGY airport to Milan city centre (in km and time), (ii) the destination airport and frequency of each of Ryanair and Aer Lingus on this route and (iii) the total travel time from city centre to city centre. Table 12: Milano Linate / Milano Malpensa / Orio al Serio International Airport Il Caravaggio (Bergamo) Dublin - Milan Linate and Malpensa and Bergamo Relevant airport(s) Milano Linate (LIN) Milano Malpensa (MXP) Orio al Serio International Airport Il Caravaggio (BGY) Distance and travel times to Milan city centre LIN - 7km MXP - 45km BGY - 45km Car: LIN - 45 mins; MXP -56 mins; BGY - 50 mins Bus: LIN - 20 mins; MXP -50 mins; BGY - 60 mins Train: MXP - 40 mins Airlines active on route

(and airport served) Aer Lingus: 7 pw from MXP, 7 pw from LIN Ryanair: 7 pw from BGY Total Travel Time Dublin to Milan City Centre Aer Lingus: via MXP 325 mins via LIN 305 mins Ryanair: via BGY 345 mins Source: Form CO, Annex 7.3.a 201 Annex 7.7.a and Annex 6.2.a to the Form CO. 202 Recitals 262-267 of the 2007 Decision. 203 Commission Decision of 11 February 2004 in case M.3280 Air France/KLM, recitals 31 and 33. 204 Commission Decision of 22 June 2009 in case M.5335 Lufthansa/SN, recitals 343-344. EN 66 EN

(246) On the basis of the 100 km or 1 hour drive time benchmark, LIN airport, MXP airport and BGY airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin.

(247) [...]*. Aer Lingus monitors the prices of Ryanair on all short-haul routes where Ryanair operates to the same or a nearby airport, including Ryanair's fares on Dublin-BGY airport.

(248) Ryanair's marketing activities regarding its services to BGY airport are focused on Bergamo rather than Milan. In its list of destinations on its website, under the destination "Milan Bergamo", Ryanair describes the city of Bergamo and its main attractions205. However, there are other elements in Ryanair's marketing practices which point to the fact that Ryanair continues to consider that BGY airport would be appropriate for passengers destined for Milan. Ryanair markets its services to BGY airport as "Milan Bergamo" and in its airport information section, Ryanair provides information on the various public transport options for travel to Milan city centre and includes a link to the website on which tickets to Milan can be purchased206. Furthermore, under the "Top things to do" in its destination city "Milan Bergamo", Ryanair lists the top attractions in Milan (but not Bergamo).

(249) The evidence collected through the market investigation also points to the fact that the Commission's conclusion in the 2007 Decision is still appropriate for the current Transaction, namely that scheduled point-to-point flights between Dublin on the one hand and LIN airport, MXP airport or BGY airport, on the other, belong to the same market207. The operator of BGY airport considers that the Commission's conclusion of 2007 is not appropriate because "each airport has also an own local market"208 On the other hand, the operator of LIN airport and MXP airport considers that the three airports "are potentially partially overlapping"209 and Alitalia210 confirms the Commission's view in line with the majority of the other respondents to the Commission's market investigation.

(250) Lastly, the market investigation has not revealed any other indication that there have been significant changes in the market circumstances so as to warrant a different conclusion than the one reached in the 2007 Decision as regards these airports for flights to/from Dublin (211).

(251) The Commission's empirical analysis (price correlation) also provides evidence that air transport services between Dublin on the one hand and MXP airport, LIN airport or BGY airport on the other belong to the same market (see further Annex I Price correlation analysis).

(252) Therefore, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin and MXP airport, LIN airport and BGY airport belong to the same market.

7.3.3.10. Munich Airport and Allgäu Airport Memmingen

(253) Ryanair operates a service from Dublin to Allgäu Airport Memmingen (FMM airport) while Aer Lingus operates a service from Dublin to Munich Airport (MUC airport).

(254) Ryanair contends in its response to the decision opening the proceedings that there is no substitutability between MUC airport and FMM airport for flights to and from Dublin for the following reasons: (i) FMM airport falls well outside the Commission's 100km / 1 hour benchmark, (ii) FMM airport is substantially closer to and within the 100km distance benchmark to other regional cities like Augsburg, Friedrichshafen, Ulm, Kempten or Bregenz and FMM should be more appropriately characterised as a regional airport serving a multitude of smaller cities in its vicinity rather than as a secondary airport serving mainly Munich, (iii) Ryanair's marketing is attempting to capture as many customers as possible including holiday makers with varied itineraries that may end in/or otherwise include Munich, (iv) [30-40]*% of the passengers on the DUB-FMM route originate in Ireland and passengers travelling from Germany who live in Munich are unlikely to find an airport located over 100km from their city a substitute for Munich Airport and lastly (v) Ryanair's Map Bases analysis.

(255) In its previous Decision in case M.5335 Lufthansa/SN (212), the Commission considered for the purpose of the investigation in that case, that there was only one airport serving Munich, namely MUC airport. However, the Commission in its Decision also referred to FMM airport as being "sometimes considered as a potential secondary airport for Munich." (213) However, the Commission found in that case that, considering that FMM airport was only used by low cost/charter companies for a limited number of domestic flights and leisure destinations it was unlikely to be substitutable for MUC airport for flights to/from Brussels as there was no indication that an airline could serve BRU from FMM in the near future (214):

(256) Table 13 summarises the following: (i) the distance from MUC airport and FMM airport to Munich city centre (in km and time), (ii) the destination airport and frequency of Ryanair, Aer Lingus and other airlines flights' on this route and (iii) the total travel time from city centre to city centre.

Table 13: Munich Airport /Allgäu Airport Memmingen

Dublin - Munich and Memmingen

Relevant airport(s) / Munich (MUC) / Memmingen Airport (FMM)

Distance and travel times to Munich city centre

MUC - 40 km / FMM - 110 km / Car: MUC 30 mins; FMM 80 mins / Bus: MUC 40 mins; FMM 85 mins / Train: MUC 45 mins; FMM none

Airlines active on route (and airport served)

Aer Lingus: 11 pw from MUC / Ryanair: 4 pw (from FMM) / Lufthansa: 1 pw from MUC

Total Travel Time Dublin to Munich City Centre / Aer Lingus (via MUC): 300 mins / Ryanair (via FMM): 345 mins / Lufthansa (via MUC): 300 mins

Source: Form CO, Annex 7.3.a

(257) On the basis of the 100km or 1 hour drive time benchmark, MUC airport and FMM airport do not appear prima facie to be substitutable from the demand side for point to-point scheduled passenger air transport services to/from Dublin.

(258) However, the fact that FMM airport falls outside the benchmark does not per se indicate an absence of substitutability between scheduled point-to-point flights from Dublin to FMM airport and scheduled point-to-point flights from Dublin to MUC airport. Furthermore, it is undisputed that one airport could be suitable for reaching various cities or regions, therefore the fact that an airport like FMM airport would be closer to other regional cities (than Munich) cannot change the conclusion as to the substitutability between flights from Dublin to FMM airport or to MUC airport. The Commission needs to take account of all elements which are relevant for determining whether scheduled point-to-point flights between Dublin and MUC airport are substitutable for flights between Dublin and FMM airport.

(259) [...]*. (215) Aer Lingus [...]* monitors Ryanair on the Dublin-FMM route and considers Ryanair to be the primary competitor on this route due to the lack of real alternatives in terms of frequency with Lufthansa only operating on a Sunday. Aer Lingus also considers that the leisure segment is highly competitive between the two airlines on this route (216).

(260) Ryanair's marketing activities indicate that Ryanair considers that FMM airport would be appropriate for passengers destined for Munich. Indeed, in its list of destinations on its website, under the destination "Munich Memmingen", Ryanair describes the city of Munich and its main attractions (217). In its airport information section, Ryanair states that FMM airport "not only serves Memmingen and the Allgäu area but also provides a low-cost alternative to Munich Airport" and includes the link to the Allgäu Express which inter alia offers direct bus connections to Munich city centre. (218) While Ryanair is attempting to capture passengers with varied itineraries, its marketing practices clearly show that it also targets passengers destined for Munich and it considers that FMM airport would be "the low cost alternative" to MUC airport.

(261) The evidence collected through the market investigation also points to the substitutability of these two airports for flights to or from Dublin. Indeed, Lufthansa, the other airline operating on this route, considers that "for some customers (nontime- sensitive and cost sensitive) flights between Dublin and Memmingen Airport might be an alternative to flights between the Dublin and Munich Airport." (219) Lufthansa estimates that around 0 to 10% of passengers would switch to scheduled point-to-point flights between Dublin and FMM airport in case of a price increase of scheduled point-to-point flights between Dublin and MUC airport (220) and that these would be "Only non-time-sensitive and very cost sensitive passengers" because "the additional transportation cost to/from Munich would probably be higher than the price increase." (221) The operator of MUC airport also considers that these two airports are substitutable and that "recent developments have shown that the opening of new routes at FMM [airport] have automatically resulted in a reduction of passengers at MUC [airport] to the same destinations" (222). MUC airport therefore estimates that between 10% to 30% of passengers would switch from scheduled point-to-point fligts between MUC airport and Dublin to scheduled point-to-point flights between FMM airport and Dublin and about 0 to10% would switch the other way around "as average fares at FMM are generally very low, those passengers that need to fly will still accept the higher fares, whereas those passengers that do not necessarily need to fly would rather refrain from flying at all" (223). FMM airport also considers that, generally, scheduled point-to-point flights from Dublin to FMM airport or MUC airport "are substitutable to each other, because of the relative small distance betweeen Munich and Memmingen. But it also depends on the capacity utilisation of the flights" (224). While some of the respondents to the Commission's market investigation have diverging views, the majority considered that for the purposes of this Transaction, scheduled point-to-point flights between Dublin and MUC airport and between Dublin and FMM airport belong to the same market.

(262) The Commission's empirical analysis (price correlation) is inconclusive as regards Munich airport and Memmingen airport (see further Annex I).

(263) As indicated by Ryanair, [30-40]*% of its passengers on the DUB-FMM route originate in Ireland. [...]*. However, this does not mean that all Ryanair's passengers originate in Munich and therefore, they would not consider flying from FMM airport because they have the closer alternative MUC airport.

(264) As regards Ryanair's Map-Based analysis, it relies on several strong and arbitrary assumptions and the Commission does not consider that the results of this analysis provide any evidence that the relevant airports belong to different markets as explained in detail in Annex II.

(265) In light of all the above, the Commission takes the view that scheduled point-to-point passenger air transport services between (i) Dublin and MUC airport and (ii) Dublin and FMM airport belong to the same market.

7.3.3.11. Roissy-Charles de Gaulle Airport, Paris Beauvais Airport and Orly Airport

(266) Ryanair operates a service from Dublin to Paris Beauvais Airport (BVA airport) while Aer Lingus operates a service from Dublin to Paris Roissy-Charles de Gaulle Airport (CDG airport).

(267) Ryanair contends that CDG airport and BVA airport are not substitutable or do not overlap for competition purposes since there is a substantial distance between these two destination airports (225).

(268) The Commission concluded in its 2007 Decision that scheduled point-to-point passenger air transport services between Dublin, on the one hand, and CDG airport or BVA airport on the other, belong to the same market (226). The Commission drew this conclusion by taking account of a number of elements like Ryanair's marketing practices, the market investigation, the price correlation analysis and the customer survey.

(269) In its Decision in case M.3280 Air France /KLM (227), the Commission concluded that CDG airport and Orly airport (ORY airport) are substitutable from the demand side for point-to-point traffic (including for intra-European flights). However the Commission did not further investigate the substitutability with BVA airport as the parties to that transaction had not argued for substitutability (228). Similarly, in its Decision in case in COMP/M.5830 Olympic / Aegean Airlines the Commission concluded that CDG airport and ORY airport are substitutable for all passengers travelling to and from Athens (229).

(270) Table 14 summarises the following: (i) the distance from CDG airport, BVA airport and ORY airport to Paris city centre (in km and time), (ii) the destination airport and frequency of each of Ryanair and Aer Lingus on this route and (iii) the total travel time from city centre to city centre (230).

Table 14: Roissy-Charles de Gaulle Airport, Paris Beauvais Airport and Orly Airport

Dublin - Paris Charles de Gaulle and Beauvais and Orly

Relevant airport(s) / Paris Charles de Gaulle (CDG) / Paris Beauvais (BVA) / Paris Orly (ORY)

Distance and travel times to Paris city centre

CDG - 23km / BVA- 80km / ORY - 20 km / Car: CDG - 31 mins; BVA - 60 mins; ORY - 30 mins / Bus: CDG - 55 mins; BVA - 75 mins; / Train: CDG - 35 mins; ORY - 30 mins

Airlines active on route (and airport served) / Aer Lingus: 21 pw from CDG / Air France: 41 pw from CDG / Ryanair: 12 pw from BVA

Total Travel Time Dublin City Centre to Paris City Centre / Aer Lingus: (via CDG) 265 mins / Air France: (via CDG) 270 mins / Ryanair: (via BVA) 285 mins

Source: Form CO, Annex 7.3.a, www.viamichelin.com and www.aeroportsdeparis.fr (underlined)

(271) On the basis of the 100km/1 hour benchmark, CDG airport, BVA airport and ORY airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin in terms of distance and driving time by car (231).

(272) [...]* (232). Furthermore, Aer Lingus compares its fare against Ryanair's fare on DUBBVA airport primarily, and against Air France on DUB-CDG airport secondarily (233).

(273) Ryanair's marketing activities indicate that Ryanair considers that BVA airport would be appropriate for passengers destined for Paris. Indeed, in its list of destinations on its website, under the destination "Paris Beauvais", Ryanair describes the city of Paris and its main attractions (234). In its airport information section, Ryanair provides information on the various public transport options for travel to Paris city centre and includes a link to the website of the Paris Beauvais Airport Shuttle for which tickets can be purchased online (235).

(274) The evidence collected through the market investigation indicates that the Commission's conclusion in its 2007 Decision is still appropriate for the current Transaction, namely that scheduled point-to-point flights between Dublin on the one hand and CDG airport or BVA airport, on the other, belong to the same market (236). Air France, the other operator on this route explained that "BVA [airport is] a clear viable option for Paris travellers, whether time sensitive or non-time sensitive passengers for the following reasons: BVA is part of the Paris Airport System: it appears in the Global Distribution Systems (GDS) under the Paris (PAR) designator code. Although Beauvais is further from central Paris than are Paris-Orly and Paris- CDG airports, a substantial proportion of Ryanair's passengers is travelling to or from Paris" (237). Furthermore, the operator of CDG airport also confirmed the conclusion and indicated that nothing has changed since the 2007 Decision (238). On the other hand, the operator of BVA airport stated that it does not regard itself in competition with CDG airport and explained: "La situation de concurrence entre la plateforme de Beauvais Tillé et la plateforme de l'aéroport de Roissy-Charles de Gaulle doit être fortement relativisée. En effet, l'aéroport de Roissy-Charles de Gaulle est beaucoup plus accessible que celui de Beauvais Tillé (en termes de distance, de temps de trajet, d'infrastructures de transport telles que le réseau express régional) pour les passagers en provenance de Paris et de la banlieue parisienne" (239). However, as explained in detail by Air France in its response, "the very low airport transit time at BVA compensates for the longer travelling time between Paris and the airport: in fact, total voyage time from central Paris to the aircraft at BVA can be comparable to ORY and CDG as each ground shuttle leaves 2 hours and 30 minutes before the planned take-off of the aircraft.[...]. The majority of the other respondents to the Commission's market investigation also confirm the findings in the 2007 Decision.

(275) Lastly, the market investigation has not revealed any other indication that there have been significant changes in the market circumstances such as to warrant a different conclusion to the one reached in the 2007 Decision as regards these airports for flights to/from Dublin (240).

(276) As regards the substitutability between CDG airport, BVA airport and ORY airport, the market investigation broadly confirmed that ORY airport would be substitutable with CDG airport and BVA airport respectively for flights to/from Dublin.

(277) The Commission's empirical analysis (price correlation) also points to the fact that air transport services between Dublin on the one hand and CDG airport and BVA airport on the other belong to the same market (see further Annex I).

(278) As regards Ryanair's Map-Based analysis, it relies on several strong and arbitrary assumptions and the Commission does not consider that the results of this analysis provide any evidence that the relevant airports belong to different markets as explained in detail in Annex II.

(279) Therefore, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin and CDG airport, BVA airport and ORY airport belong to the same market.

7.3.3.12. Rome: Leonardo da Vinci International Airport (Fiumicino) and G.B Pastine International Airport (Ciampino)

(280) Ryanair operates a service from Dublin to G.B Pastine International Airport (Ciampino) (CIA airport) while Aer Lingus operates a service from Dublin to Leonardo da Vinci International Airport (Fiumicino) (FCO airport).

(281) Ryanair contends that, since there is a substantial distance between these two destination airports they are not substitutable or do not overlap for competition purposes (241).

(282) The Commission concluded in its 2007 Decision, that, for the purposes of that case, scheduled point-to-point passenger air transport services between Dublin, on the one hand, and FCO airport or CIA airport, on the other, belong to the same market (242). The Commission drew this conclusion by taking account of a number of elements such as the market investigation, the customer survey and the price correlation analysis.

(283) Table 15 summarises the following: (i) the distance from FCO airport and CIA airport to Rome city centre (in km and time), (ii) the destination airport and frequency of each of Ryanair and Aer Lingus on this route and (iii) the total travel time from city centre to city centre.

Table 15: Leonardo da Vinci International Airport (Fiumicino) / G.B Pastine International Airport (Ciampino)

Dublin - Rome Fiumicino and Ciampino

Relevant airport(s) / Rome Fiumicino (FCO) / Rome Ciampino (CIA)

Distance and travel times to city centre / FCO - 26km / CIA - 18km / Car: FCO - 27 mins; CIA - 26 mins / Bus: FCO - 40 mins / Train: FCO - 30 mins ; CIA - 30 mins

Airlines active on route (and airport served) / Ryanair: 7 pw from CIA / Aer Lingus: 11 pw from FCO

Total Travel Time City Centre to City Centre / Ryanair: (from CIA) 345 mins / Aer Lingus: (from FCO) 345 mins

Source: Form CO, Annex 7.3.a

(284) On the basis of the 100 km or 1 hour drive time benchmark, FCO airport and CIA airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin.

(285) [...]* (243). As explained above, Aer Lingus monitors Ryanair prices on all short-haul routes where Ryanair operates to the same or a nearby airport, including Ryanair's fares on Dublin-CIA airport.

(286) The evidence collected through the market investigation also indicates that the Commission's conclusion of 2007 is still appropriate for the current Transaction, namely that scheduled point-to-point flights between Dublin on the one hand and FCO airport or CIA airport, on the other, belong to the same market. The operator of FCO airport and of CIA airport confirmed this (244) as did Alitalia (245), which has a hub at FCO airport and used to operate on this route, and the majority of the other respondents to the Commission's market investigation.

(287) Lastly, the market investigation has not provided any indication that there have been significant changes in the market circumstances such as to warrant a different conclusion than the one reached in the 2007 Decision as regards these airports for flights to/from Dublin (246).

(288) The Commission's empirical analysis (price correlation) also provides evidence that air transport services between Dublin on the one hand and FCO airport and CIA airport on the other, belong to the same market (see further Annex I).

(289) Therefore, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin and FCO airport and Dublin and CIA airport belong to the same market.

7.3.3.13. Stockholm Arlanda Airport and Stockholm Skavsta Airport (247)

(290) Ryanair operates a service from Dublin to Stockholm Skavsta Airport (NYO airport) while Aer Lingus operates a service from Dublin to Stockholm Arlanda Airport (ARN airport).

(291) Ryanair contends in its response to the decision opening the proceedings that there is no substitutability between ARN airport and NYO airport mainly for the following reasons: (i) NYO airport falls outside the Commission's 100km/1 hour benchmark and (ii) Ryanair's data supports the conclusion that ARN airport and NYO airport are not substitutable.

(292) The Commission's market investigation in case M.3770 Lufthansa/Swiss, regarding the substitutability of ARN airport, NYO airport and Bromma Stockholm Airport for flights to and from Zurich has found that: "in principle airport substitutability might be justified from a demand-side view for all three airports [ARN airport, NYO airport and Bromma Stockholm airport] for non-time sensitive passengers and between [ARN airport] and Bromma [Stockholm airport] for non-time sensitive passengers" (248).

(293) Table 16 summarises the following: (i) the distance from ARN airport and NYO airport to Stockholm city centre (in km and time), (ii) the destination airport and frequency of Ryanair, Aer Lingus and other carriers flights' on this route and (iii) the total travel time from city centre to city centre.

Table 16: Stockholm Arlanda Airport /Stockholm Skavsta Airport

Dublin - Arlanda and Skavsta

Relevant airport(s) / Stockholm Arlanda Airport (ARN airport) / Stockholm Skavsta Airport (NYO airport)

Distance and travel times to Stockholm city centre / 45 km (ARN airport) / 106 km (NYO airport) / Car: (ARN airport) 35 mins, (NYO airport) 80 mins / Bus: (ARN airport) 45 mins, (NYO airport) 80 mins / Train: (ARN airport) 20 mins; (NYO airport) none

Airlines active on route (and airport served) / Aer Lingus: 4 pw from ARN airport / Ryanair: 4 pw from NYO airport / SAS: 7 pw from ARN airport

Total Travel Time City Centre to City Centre / Aer Lingus Dublin to Stockholm (via ARN airport) 295 mins / SAS Dublin to Stockholm (via ARN airport) 300 mins / Ryanair Dublin to Stockholm (via NYO airport) 360 mins

Source: Form CO, Annex 7.3.a

(294) On the basis of the 100 km / 1 hour benchmark, ARN airport and NYO airport do not appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin.

(295) However, the fact that NYO airport falls outside the benchmark does not per se point to absence of substitutability with ARN airport for scheduled point-to-point flights to/from Dublin. The Commission needs to take account of all elements which are relevant for determining whether scheduled point-to-point flights between Dublin and ARN airport are substitutable for flights between Dublin and NYO airport.

(296) [...]* (249). Furthermore, Aer Lingus also monitors Ryanair and currently references Ryanair as the primary competitor on this new route (250).

(297) Ryanair's marketing activities indicate that Ryanair considers that NYO airport would be appropriate for passengers destined for Stockholm. Indeed, in its list of destinations on its website, under the destination "Stockholm Skavsta", Ryanair describes the city of Stockholm and its main attractions (251). In its airport information section, Ryanair refers to the airport coaches that operate directly to and from Stockholm (252).

(298) The evidence collected through the market investigation also points to the substitutability of these two airports. Indeed, the other airline active on this route SAS considers that these airports are substitutable for flights to/from Dublin and that a "5-10 % price increase [on Dublin-Arlanda or on Dublin-Skavsta] would probably render some deviation in the price sensitive segment. Probably not more than 10% deviation as passengers are willing to pay a premium for proximity to Stockholm, reliability, convenience and service" (253). The operator of ARN airport also agrees that there is a certain degree of substitutability between these two airports (254). The manager of NYO airport also stated that NYO airport is a "very viable alternative for a price sensitive market" (255). Indeed, NYO airport advertises and describes itself as "Stockholm's second airport" (256). Furthermore, Norwegian Airlines also took the view that "Arlanda and Skavsta belong clearly to the same catchment area, even if demographic characteristics are different between the two airports. Norwegian considers Ryanair, which flies to Skavsta, as a direct competitor on its routes to Arlanda (for example from London Gatwick)" (257). The majority of other market participants which issued an opinion also consider that ARN airport and NYO airport are substitutable.

(299) As regards Ryanair's Map-Based analysis, it relies on several strong and arbitrary assumptions and the Commission does not consider that the results of this analysis provide any evidence that the relevant airports belong to different markets as explained in detail in Annex II. On the other hand, the Commission cannot draw any meaningful conclusions from its empirical analysis (price correlation) as regards air transport services between Dublin on the one hand and ARN airport or NYO airport on the other due to the small number of observations (see further Annex I).

(300) Therefore, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin and ARN airport and Dublin and NYO airport belong to the same market.

7.3.3.14. Toulouse Blagnac Airport and Carcassonne Airport

(301) Ryanair operates a service from Dublin to Carcassonne Airport (CCF airport) while Aer Lingus operates a service from Dublin to Toulouse Blagnac Airport (TLS airport).

(302) Ryanair contends in its response to the decision opening the proceedings that there is no substitutability between TLS airport and CCF airport mainly for the following reasons: (i) CCF airport is exactly at the Commission's 1 hour drive time benchmark and just below the Commission's 100km distance benchmark, (ii) this benchmark is less instructive when two airports primarily serve different major cities as is the case with TLS airport and CCF airport, (iii) CCF airport is not a "secondary" Toulouse airport but is rather a gateway to the Languedoc-Roussillon region, including the town of Carcassonne itself and (iv) other qualitative elements like Ryanair's marketing practices, Aer Lingus' marketing practices or the names of the airports.

(303) The Commission concluded in its 2007 Decision, that, for the purposes of that case, scheduled point-to-point passenger air transport services between Dublin, on the one hand, and TLS airport or CCF airport, on the other, belong to the same market (258). The Commission drew this conclusion by taking account of a number of elements such as the market investigation and the price correlation analysis.

(304) Table 17 summarises the following: (i) the distance from TLS airport and CCF airport to Toulouse city centre (in km and time), (ii) the destination airport and frequency of Ryanair and Aer Lingus flights on this route and (iii) the total travel time from city centre to city centre.

Table 17: Toulouse Blagnac Airport / Carcassonne Airport

Dublin - Toulouse and Carcassonne

Relevant airport(s) / Toulouse (TLS) / Carcassonne (CCF)

Distance and travel times to Toulouse city centre / TLS - 8km / CCF - 90km / Car: TLS - 17 mins; CCF - 60 mins / Bus: TLS - 20 mins CCF - none / Train: TLS-none; CCF - none

Airlines active on route (and airport served) / Aer Lingus: 3 pw from TLS / Ryanair: 3 pw from CCF

Total Travel Time Dublin City Centre to Toulouse City Centre / Dublin to Toulouse (via TLS) 270 mins / Dublin to Toulouse (via CCF) 310 mins

Source: Form CO, Annex 7.3.a

(305) On the basis of the 100km/1 hour benchmark, TLS airport and CCF airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin.

(306) The fact that CCF airport is closer to the benchmark and that the benchmark is less instructive when two airports primarily serve different major cities, as argued by Ryanair, does not per se indicate the absence of substitutability of the two airports for this Transaction. In the same manner, the fact that CCF airport is not a secondary airport to Toulouse but a gateway to the Languedoc-Roussillon region and the fact that the names of the airports do not have a reference to the cities, in general cannot preclude a finding that TLS airport and CCF airport could be substitutable for scheduled point-to-point flights from Dublin as explained in Section 7.3.2.2. In fact, even for these destinations the city centre criterion is an indicative benchmark showing the relative distance of the airports from the local centre and thus also the difference in their ability to serve the tourist resorts in the vicinity. For the leisure route Dublin-Toulouse/Carcassonne, while the two airports are located in different administrative regions (Midi-Pyrénées and Languedoc-Roussillon), they are suitable for passengers destined for one or the other region. For example, Toulouse airport advertises on its website the accessibility to the Mediterranean coast (259) while the Carcassonne airport mentions the attractiveness of the Canal du Midi which reaches Toulouse (260). In any event, the Commission needs to take account of all elements which are relevant for determining whether scheduled point-to-point flights between Dublin and TLS airport are substitutable for flights between Dublin and CCF airport.

(307) [...]* (261). [...]*. As explained above, Aer Lingus monitors Ryanair prices on all shorthaul routes where Ryanair operates to the same or a nearby airport, including Ryanair's fares on the Dublin-CCF airport route.

(308) Ryanair's marketing activities regarding its services to CCF airport are more targeted to passengers destined for Carcassonne than for Toulouse. In its list of destinations on its website, under the destination "Carcassonne", Ryanair describes the city of Carcassonne and its main attractions (262) and in its airport information section it describes the various ways to travel to Carcassonne city centre (263). However, in its "Top things to do" section, Ryanair lists the Canal du Midi (264) which reaches Toulouse (265). While there are no direct bus or train connections between CCF airport and Toulouse city centre, there is a bus connecting CCF airport to the city of Carcassonne (266) and there are direct trains from there to the city of Toulouse (267). The same situation prevailed at the time of the Commission's 2007 Decision.

(309) The evidence collected through the market investigation points rather to the fact that the Commission's conclusion of 2007 is still appropriate for the current Transaction, namely that scheduled point-to-point flights between Dublin on the one hand and TLS airport or CCF airport, on the other, belong to the same market (268). While the operator of CCF airport claims that these two airports are not substitutable for each other for scheduled point-to-point flights to/from Dublin (269), the operator of TLS airport takes the opposite view (270). Furthermore, Air France, which according to Ryanair has a base at TLS airport, also confirms (271), as does the majority of the respondents which issued an opinion, that the Commission's conclusion of 2007 is still appropriate.

(310) Lastly, the market investigation has not revealed any other indication that there have been significant changes in the market circumstances such as to warrant a different conclusion to the one reached in the 2007 Decision as regards these airports for flights to/from Dublin (272).

(311) The Commission's empirical analysis (price correlation) also provides evidence that air transport services between Dublin on the one hand and TLS airport and CCF airport on the other, belong to the same market (see further Annex I).

(312) Therefore, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin and TLS airport and Dublin and CCF airport belong to the same market.

7.3.3.15. Venice Airport and Treviso Airport

(313) Ryanair operates a service from Dublin to Treviso Airport (TSF airport) while Aer Lingus operates a service from Dublin to Venice Airport (VCE airport).

(314) Ryanair accepts that VCE airport and TSF airport form part of a single market for the purposes of competition assessment (273).

(315) The Commission concluded in its 2007 Decision, that, for the purposes of that case, scheduled point-to-point passenger air transport services between Dublin, on the one hand, and VCE airport or TSF airport, on the other, belong to the same market (274). The Commission drew this conclusion by taking account of a number of elements such as the market investigation and Rynanair's marketing practices.

(316) Table 18 summarises the following: (i) the distance between each of VCE airport and TSF airport to Venice city centre (in km and time), (ii) the destination airport and frequency of Ryanair and Aer Lingus flights on this route and (iii) the total travel time from city centre to city centre.

Table 18: Venice Airport / Treviso Airport

Dublin - Venice and Treviso

Relevant airport(s) / Venice Marco Polo (VCE) / Venice Treviso (TSF)

Distance and travel times to Venice city centre / VCE - 12km / TSF - 31km / Car: VCE - 20 mins; TSF - 35 mins / Bus: VCE - 20 mins; TSF - 45 mins

Airlines active on route (and airport served) / Aer Lingus: 4 pw from VCE / Ryanair: 3 pw from TSF

Total Travel Time City Centre to City Centre / Dublin to Venice (via VCE) 300 mins / Dublin to Venice (via TSF) 315 mins

Source: Form CO, Annex 7.3.a

(317) On the basis of the 100 km or 1 hour drive time benchmark, VCE airport and TSF airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin.

(318) [...]* (275). As explained above, Aer Lingus monitors the prices of Ryanair on all shorthaul routes where Ryanair operates to the same or a nearby airport, including Ryanair's fares on Dublin-TSF airport.

(319) Furthermore, Ryanair's marketing activities also indicate that Ryanair considers that TSF airport would be appropriate for passengers destined for Venice. Indeed, in its list of destinations on its website, under the destination "Venice Treviso", Ryanair describes the city of Venice and its main attractions (276). In addition, in its airport information section, Ryanair includes explanations about the transport options for travelling from TSF airport to Venice city and includes the link to the website on which tickets can be purchased (277).

(320) The evidence collected through the market investigation also indicates that the Commission's conclusion in the 2007 Decision is still appropriate for the current Transaction, namely that scheduled point-to-point flights between Dublin on the one hand and VCE airport or TSF airport, on the other, belong to the same market (278). Indeed, the operator of VCE airport and of TSF airport (279) and Alitalia (280) confirm this view. In addition, the majority of respondents to the market investigation also agree with this view.

(321) Lastly, the market investigation has not revealed any indication that there have been significant changes in the market circumstances such as to warrant a different conclusion to the one reached in the 2007 Decision as regards these airports for flights to/from Dublin (281).

(322) The Commission's empirical analysis (price correlation) also provides evidence that air transport services between Dublin on the one hand and VCE airport and TSF airport on the other, belong to the same market (see further Annex I).

(323) Therefore, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin and VCE airport and Dublin and TSF airport belong to the same market.

7.3.3.16. Vienna Airport and Bratislava Airport

(324) Ryanair operates a service from Dublin to Bratislava Airport (BTS airport) while Aer Lingus operates a service from Dublin to Vienna Airport (VIE airport).

(325) Ryanair contends in its response to the decision opening the proceedings that there is no substitutability between VIE airport and BTS airport mainly for the following reasons: (i) BTS airport is closer to the outer limit of the 100km/1 hour benchmark; (ii) this benchmark is less instructive when two airports primarily serve different major cities as is the case with VIE airport and BTS airport; (iii) BTS airport is not a "secondary" Vienna airport but is rather a primary airport for Bratislava and Slovakia in general; (iv) a large portion of Ryanair's BTS airport originating passengers in fact come from locations east of BTS airport and not from Vienna; (v) other qualitative elements like Ryanair's marketing practices, Aer Lingus' marketing practices or the names of the airports; and (vi) Ryanair's credit card data.

(326) The Commission concluded in its 2007 Decision that, for the purposes of that case, scheduled point-to-point passenger air transport services between Dublin, on the one hand, and VIE airport or BTS airport, on the other, belong to the same market (282). The Commission drew this conclusion by taking account of a number of elements such as Ryanair's marketing practices, the market investigation and the customer survey.

(327) In its Decision in case M.5440 Lufthansa/Austrian (283), the Commission found that Ryanair's services between HHN airport and BTS airport did not constrain services between FRA airport and VIE airport to any meaningful degree given that the total travel time would have been much longer than a trip involving VIE airport and FRA airport, that the respondents to the market investigation in that case did not consider both BTS airport as a substitute for VIE airport and HHN airport as a substitute for FRA airport and lastly because of lack of data to prove the contrary. As is evident, in that case, the Commission assessed the question of airport substitutability at both ends of the route, that is to say both at the Vienna end where it assessed the substitutability of VIE airport with BTS airport and at the Frankfurt end where it assessed the substitutability of FRA airport with HHN airport. The Commission concluded that for that specific route, the parties' contention that BTS airport would be substitutable for VIE airport had not been confirmed by the market investigation (284). However, while in that Decision the question of airport substitutability was relevant for both ends of the route, this issue is not relevant to the current Transaction. Furthermore, that Decision concerned two airlines with a different profile and targeting different type of customers than is the case in the Transaction. In its Decision in case M.6447 IAG/bmi, the Commission ultimately left open the question of the substitutability between VIE airport and BTS airport for flights to/from London (285).

(328) Table 19 summarises the following: (i) the distance from VIE airport and BTS airport to Vienna city centre (in km and time), (ii) the destination airport and frequency of Ryanair and Aer Lingus flights on this route and (iii) the total travel time from city centre to city centre.

Table 19: Vienna Airport / Bratislava

Airport Dublin - Vienna and Bratislava

Relevant airport(s) / Vienna (VIE) / Bratislava (BTS)

Distance and travel times to Vienna city centre / VIE - 16km / BTS - 87km / Car: VIE - 21 mins; BTS - 55 mins / Bus: VIE - 20 mins; BTS - 100 mins / Train: VIE - 24 mins; BTS - none.

Airlines active on route (and airport served) / Aer Lingus: 7 pw from VIE / Ryanair: 6 pw from BTS

Total Travel Time Dublin City Centre to Vienna City Centre / Dublin to Vienna (via VIE) 309 mins / Dublin to Vienna (via BTS) 380 mins

Source: Form CO, Annex 7.3.a

(329) On the basis of the 100km/1 hour benchmark, VIE airport and BTS airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin in terms of distance and if the travel duration by car is considered. If transportation by bus is considered, the duration would be outside the Commission's benchmark.

(330) The fact that BTS airport is closer to benchmark and that the benchmark is less instructive when two airports primarily serve different major cities, as argued by Ryanair, does not per se indicate the absence of substitutability of the two airports for the purpose of the Transaction. In the same manner, the fact that BTS airport is not a "secondary" airport to VIE airport but a primary airport for Bratislava and Slovakia and the fact that the names of the airports do not have a reference to the cities (286), in general cannot preclude a finding that VIE airport and BTS airport could be substitutable for scheduled point-to-point flights from Dublin as explained in Section 7.3.2.2. The Commission needs to take account of all elements which are relevant for determining whether scheduled point-to-point flights between Dublin and VIE airport are substitutable for flights between Dublin and BTS airport.

(331) [...]* (287). Aer Lingus monitors the prices of Ryanair on all short-haul routes where Ryanair operates to the same or a nearby airport, including Ryanair's fares on Dublin-BTS airport.

(332) Ryanair's marketing activities regarding its services to BTS airport are focussed on Bratislava rather than on Vienna. In its list of destinations on its website, under the destination "Bratislava", Ryanair describes the city of Bratislava and its main attractions (288) and in its airport information section, it describes the means to travel to Bratislava city centre. Furthermore, Ryanair groups its destinations by countries so that Bratislava appears under destinations in Slovakia (289). However, there are other elements in Ryanair's marketing practices which seem to point to the fact that Ryanair continues to consider that BTS airport would also be appropriate for passengers destined for Vienna. While Ryanair no longer markets its services to Bratislava as "Bratislava (Vienna)" on its website, (290) it still lists under the "Top things to do" in its destination city "Bratislava", the top attractions in Vienna (and not so much Bratislava) (291). Furthermore, Ryanair also markets transportation services between BTS airport and Vienna by including the link to its shuttle direct transfer bus through which transfer tickets from BTS airport to Vienna city centre can be purchased online (292).

(333) The evidence collected through the market investigation indicates that the Commission's conclusion in the 2007 Decision is still appropriate for the current Transaction, namely that scheduled point-to-point flights between Dublin on the one hand and VIE airport or BTS airport, on the other, belong to the same market (293). The operators of VIE airport (294), BTS airport (295), Austrian Airlines (296) and the majority of respondents confirm this view. Furthermore, BTS airport also seems to be marketing itself as an advantageous alternative and a smart and flexible option to the more expensive VIE airport (297).

(334) Lastly, the market investigation has not provided any indication that there have been significant changes in the market circumstances such as to warrant a different conclusion to the one reached in the 2007 Decision as regards these airports for flights to/from Dublin (298).

(335) The Commission's empirical analysis (price correlation) also provides evidence that air transport services between Dublin on the one hand and VIE airport and BTS airport on the other, belong to the same market (see further Annex I).

(336) As regards Ryanair's Map-Based analysis, it relies on several strong and arbitrary assumptions and the Commission does not consider that the results of this analysis provide any evidence that the relevant airports belong to different markets as explained in detail in Annex II.

(337) In light of all the above, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin and VIE airport and Dublin and BTS airport belong to the same market.

7.3.3.17. Warsaw Chopin Airport and Warsaw - Modlin Mazovia Airport

(338) Ryanair operates a service from Dublin to Warsaw - Modlin Mazovia Airport (WMI airport) while Aer Lingus operates a service from Dublin to Warsaw Chopin Airport (WAW airport).

(339) Ryanair accepts that WAW airport and WMI airport form part of a single market for the purpose of the competitive assessment in the present case (299).

(340) Table 20 summarises the following: (i) the distance from WAW airport and WMI airport to Warsaw city centre (in km and time), (ii) the destination airport and frequency of Ryanair and Aer Lingus flights on this route and (iii) the total travel time from city centre to city centre.

Table 20: Warsaw Chopin Airport / Warsaw - Modlin Mazovia Airport

Dublin - Warsaw and Modlin

Relevant airport(s) / Warsaw Chopin Airport (WAW airport) / Warsaw - Modlin Mazovia Airport (WMI airport)

Distance and travel times to Warsaw city centre / WAW airport - 8 km / WMI airport - 42 km

Car / WAW airport 15 mins / WMI airport 56 mins

Bus / WAW airport 20 mins / WMI airport 50 mins

Train / WAW airport 25 mins / WMI airport none

Airlines active on route (and airport served) / Aer Lingus: 7 pw from WAW airport / Ryanair plans to commence 5 pw from WMI airport

Total Travel Time City Centre to City Centre / Dublin to Warsaw (via WAW airport) 320 mins / Dublin to Warsaw (via WMI airport) 350 mins

Source: Form CO, Annex 7.3.a

(341) On the basis of the 100 km or 1 hour drive time benchmark, WAW airport and WMI airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin.

(342) [...]* (300). As explained above, Aer Lingus monitors the prices of Ryanair on all shorthaul routes where Ryanair operates to the same or a nearby airport, including Ryanair's fares on Dublin-WMI airport.

(343) Ryanair's marketing activities also indicate that Ryanair considers that WMI airport would be appropriate for passengers destined for Warsaw. Indeed, in its list of destinations on its website, under the destination "Warsaw Modlin", Ryanair describes the city of Warsaw and its main attractions (301). In addition, under the airport information section, Ryanair includes explanations about the various transport options for travel from WMI airport to Modlin and to Warsaw city (302).

(344) The evidence collected through the market investigation indicates that these two airports would be substitutable. Indeed, the airport operators of WAW airport (303) and of WMI airport (304) consider that these two airports are substitutable. The Polish operator LOT (305) and Wizz Air (306) who have a base at WMI airport also responded that flights between Dublin and WAW airport and flights between Dublin and WMI airport are substitutable, in line with the majority of respondents which issued an opinion in the Commission's market investigation.

(345) The Commission cannot draw any meaningful conclusions from its empirical analysis (price correlation) as regards air transport services between Dublin on the one hand and WAW airport or WMI airport on the other due to the small number of observations (see further Annex I).

(346) In light of all the above, the Commission takes the view that scheduled point-to-point passenger air transport services between Dublin and WAW airport and Dublin and WMI airport belong to the same market.

7.3.3.18. Other airports for which airport substitutability is relevant for entry or competition from other carriers

(347) The question of airport substitutability is further relevant for a number of airport pairs for the purpose of assessing any entry plans by competitors at the other airport or for ascertaining whether flights operated by other carriers would be part of the same market as flights operated by the Parties. These airport pairs are Alicante and Murcia (for flights to/from Dublin and Cork), Edinburgh and Glasgow (for flights to/from Dublin), Tenerife North and Tenerife South (for flights from Dublin and Cork) and Bristol/Cardiff and Exeter (for flights from Dublin).

(348) There are no indications that carriers, whether full service, low cost or charter carriers, are making plans which post-Transaction would lead to timely, likely and sufficient entry or expansion on routes between Ireland and the airports listed in sub- Sections (a) to (c) below. Furthermore, competition from Dublin-Exeter seems to be limited. Therefore, as the question of airport substitutability can be left open as regards these airport pairs, the Commission has not conducted an in-depth assessment.

a) Alicante and Murcia

(349) For the routes to/from Dublin Airport or Cork Airport to Alicante Airport (ALC airport) or Murcia San Javier Airport (MJV airport), Ryanair and Aer Lingus both fly to ALC airport (and in addition Ryanair flies to MJC airport from Dublin).

(350) Ryanair contends in its response to the decision opening the proceedings that there is no substitutability between ALC airport and MJV airport because of the following:

(1) MJV airport falls at the outer margin of the Commission's 100km /1 hour benchmark, (2) this benchmark is less instructive when two airports primarily serve different major cities or regions as is the case with ALC airport and MJV airport, ALC airport is the natural gateway to major beach resorts located in the north of Alicante, (3) MJV airport is not a "secondary" airport for Alicante, but the natural gateway to a number of popular tourist destinations, (4) the Commission's benchmark is less instructive in light of the leisure nature of routes to ALC airport and MJV airport and (5) other qualitative factors like Ryanair's marketing practices, Aer Lingus' marketing practices or the names of the airports.

(351) Table 21 summarises the following: (i) the distance from ALC airport and MJV airport to Alicante city centre (in km and time), (ii) the destination airport and frequency of Ryanair and Aer Lingus flights on the route from Dublin and Cork and (iii) the total travel time from city centre to city centre.

Table 21: Alicante Airport / Murcia San Javier Airport

Dublin / Cork -Alicante and Murcia

Relevant airport(s) / Alicante (ALC) / Murcia San Javier (MJV)

Distance and travel times to Alicante city centre / ALC - 9 km: / MJV - 71 km / Car: ALC -19 mins; MJV - 52 mins / Bus: ALC - 40 mins; MJV - None direct to Alicante. / Train: None

Airlines active on route from Dublin (and airport served) / Aer Lingus: 5 pw (from ALC) / Ryanair: 7 pw from ALC (and 7 pw from MJV)

Total Travel Time City Centre to City Centre (Dublin to Alicante) / Aer Lingus (from ALC) 345 mins / Ryanair: (from ALC) 345 mins / Ryanair (from MJV) 362 mins

Airlines active on route from Cork (and airport served) / Ryanair: 3 pw (from ALC) / Aer Lingus: 2 pw (from ALC)

Total Travel Time City Centre to City Centre / Ryanair (from ALC): 307 mins / Aer Lingus (from ALC): 307 mins

Source: Form CO, Annex 7.3.a

(352) On the basis of the 100 km / 1 hour benchmark, ALC airport and MJV airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin or Cork.

(353) The majority of respondents to the Commission's market investigation confirmed the conclusion reached in the 2007 Decision (307) that flights between ALC airport and Dublin are substitutable for flights between MJV airport and Dublin. As none of the Parties or other third parties operates flights between Cork and MJV airport and no sufficient, likely and timely entry has been identified, it is not relevant to address the issue of airport substitutability between ALC and MJV airport for flights out of Cork.

(354) Therefore, the Commission takes the view that the question whether scheduled pointto- point passenger air transport services between Dublin or Cork and ALC airport or MJV airport belong to the same market can be left open as it does not have any consequences on the competitive assessment in this Transaction.

b) Edinburgh/Glasgow

(355) For the routes to/from Dublin Airport to Edinburgh Airport (EDI airport) or Glasgow Airport (GLA airport), Ryanair and Aer Lingus both fly to EDI airport (and in addition Aer Lingus flies to GLA airport from Dublin).

(356) Ryanair contends in its response to the decision opening the proceedings that there is no substitutability between EDI airport and GLA airport, because: (i) GLA airport is on the outer margin of the Commission's 100km/1 hour benchmark, (ii) the Commission's benchmark is less instructive (particularly in a relatively small country such as Scotland) in a situation when two airports are neighbouring and they primarily serve different major cities, (iii) GLA airport is not a secondary Edinburgh airport but rather the primary airport for Glasgow, (iv) the lack means of transportation from GLA airport to Edinburgh, (v) other qualitative elements like Ryanair's marketing practices, Aer Lingus' marketing services or the name of the airport.

(357) Table 22 summarises the following: (i) the distance from EDI airport and GLA airport to Edinburgh city centre (in km and time), (ii) the destination airport and frequency of Ryanair and Aer Lingus flights on this route and (iii) the total travel time from city centre to city centre.

Table 22: Edinburgh Airport /Glasgow Airport

Dublin - Edinburgh and Glasgow

Relevant airport(s) / Edinburgh (EDI Glasgow International (GLA)

Distance and travel times to Edinburgh city centre / EDI - 11 km / GLA - 89 km / Car: EDI- 15 mins; GLA - 90 mins / Bus: EDI - 25 mins; GLA - no direct bus. / Train: None

Airlines active on route (and airport served) / Aer Lingus: 26 pw from EDI, operated by Aer Lingus and Aer Arann and 27 pw from GLA operated by Aer Arann. / Ryanair: 18 pw from EDI

Total Travel Time City Centre to City Centre / Aer Lingus (via EDI): 210 mins / Ryanair (via EDI): 205 mins / Aer Lingus (via GLA): 268 mins

Source: Form CO, Annex 7.3.a, www.viamichelin.com and www.glasgowairport.com

(358) On the basis of the 100 km / 1 hour benchmark, EDI airport and GLA airport do not appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin, if looking at the travel time.

(359) About half of the respondents to the market investigation which issued an opinion considered that these airports are substitutable while a significant number also considered that these airports are not substitutable. On the other hand, BAA Airports Limited, the operator of GLA airport and the operator of EDI airport responded that these two airports would be substitutable (308).

(360) Therefore, the Commission takes the view that the question of whether scheduled point-to-point passenger air transport services between Dublin and EDI airport or GLA airport belong to the same market can be left open as it does not have any consequences for the competitive assessment in this Transaction.

c) Tenerife South / North

(361) As none of the Parties or third parties operate flights between Dublin and Tenerife North airport and no sufficient, likely and timely entry has been identified on this route from Tenerife North, it is not relevant to address the issue of airport substitutability between Tenerife South and Tenerife North airport for flights out of Dublin or Cork in this Transaction (309).

d) Bristol/Cardiff/Exeter

(362) Ryanair and Aer Lingus/Aer Arann (310) both operate from Dublin to Bristol airport (BRS airport). In addition, Aer Lingus/Aer Arann operates from Dublin to Cardiff airport (CWL airport) and Flybe operates from Dublin to Exeter airport (EXT airport).

(363) Ryanair contends in its response to the decision opening the proceedings that there is no substitutability between BRS airport, CWL airport and EXT airport because of the following: (i) the Commission's benchmark of 100km/1 hour is less instructive when two airports primarily serve different major cities as is the case with BRS airport, CWL airport and EXT airport, (ii) the closest largest city to CWL airport is Cardiff itself, (iii) CWL airport is not a "secondary" BRS airport but rather the primary airport for Cardiff, (iv) EXT airport falls outside the 100km/1hour benchmark, (v) CWL airport is not a "secondary" Bristol airport but rather the primary airport for Exeter, (vi) other qualitative elements like Ryanair's marketing practices, Aer Arann's marketing practices or the name of the airports.

(364) Table 23 summarises the following: (i) the distance from of BRS Airport, CWL Airport and EXT airport to Bristol city centre (in km and time), (ii) the destination airport and frequency of Ryanair, Aer Lingus and their competitors flights on this route and (iii) the total travel time from city centre to city centre.

Table 23: Bristol Airport /Cardiff Airport / Exeter Airport

Dublin - Bristol and Cardiff and Exeter

Relevant airport(s) / Bristol Airport (BRS) / Cardiff Airport (CWL) / Exeter Airport(EXT)

Distance and travel times to Bristol city centre / BRS - 13km / CWL - 95km / EXT - 124 km / Car: BRS - 15 mins; CWL - 56 mins; EXT - 95 mins / Bus: BRS - 30 mins; CWL - none, EXT - none / Train: No rail link at BRS

Airlines active on route (and airport served) / Aer Lingus: 14 pw from BRS, operated by Aer Arann (and 13 pw from CWL, operated by Aer Arann) / Ryanair: 19 pw from BRS / Flybe: 14 pw from EXT

Total Travel Time City Centre to City Centre (Dublin to Bristol) / Aer Lingus: BRS 235 mins; / Aer Lingus CWL 251 mins / Ryanair: BRS 210 mins

Source: Form CO, Annex 7.3.a, market investigation and www.viamichelin.com

(365) On the basis of the 100 km / 1 hour drive time benchmark, BRS airport and CWL airport appear prima facie to be substitutable from the demand side for point-to-point scheduled passenger air transport services to/from Dublin. On the contrary, prima facie flights between BRS airport and EXT airport do not appear to be substitutable for point-to-point scheduled passenger air transport services to/from Dublin.

(366) As regards the substitutability between BRS airport and CWL airport and between BRS airport and EXT airport, the respondents to the Commission's market investigation have diverging views. However the majority considered that for the purposes of this Transaction, BRS airport and CWL airport and BTS airport and EXT airport are substitutable for flights out of Dublin.

(367) Therefore, the Commission takes the view that the question whether scheduled point-topoint passenger air transport services between Dublin and BRS airport or CWL airport and between Dublin and BRS airport or EXT airport belong to the same market can be left open as it does not have any consequences for the competitive assessment in this Transaction.

7.3.4. Supply- side considerations of airport substitution

(368) From the supply-side, that is to say from the airlines' point of view, the substitutability of airports depends on the needs of the airlines' passengers, on their operating model and the services they wish to provide from the airport.

(369) From the network airlines' point of view secondary airports are in principle not viable substitutes for primary airports due to the needs of their passengers and their operating model. Conversely, point-to-point, low-cost airlines may regard secondary airports and primary airports as providing substitutable services, subject to the costs of the services provided to carriers (which would be higher in primary airports) (311). The Commission's market investigation also confirmed these findings (312).

(370) In any event, in line with the Commission's Notice on market definition (313), the Commission will focus its analysis on demand-side substitutability.

7.3.5. Conclusion

(371) Therefore, the Commission concludes that the airports listed in Table 24 are substitutable for a substantial proportion of Aer Lingus' and Ryanair's passengers who travel from/to these airports to/from Dublin, Cork, Shannon or Knock as relevant.

Table 24: Summary: airports substitutable

Barcelona El Prat, Girona and Reus / London Airports (Heathrow, Gatwick, Stansted, Luton and City) / Stockholm Arlanda and Skavsta

Bilbao and Santander / Manchester and Liverpool (for Cork and Shannon) / Toulouse and Carcassonne

Birmingham and East Midlands (for Knock) / Milan Linate, Malpensa and Bergamo / Venice and Treviso

Brussels and Charleroi / Munich and Memmingen / Vienna and Bratislava

Glasgow and Prestwick / Paris CDG, Beauvais and Orly / Warsaw and Modlin

Frankfurt and Hahn / Rome Fiumicino and Ciampino

(372) The Commission takes the view that the question whether scheduled point-to-point passenger air transport services between Dublin or Cork (as relevant) and the following airport pairs belong to the same market, can be left open as it does not have any consequences for the competitive assessment of the Transaction:

Table 25: Summary: airport substitutability left open

London Southend and the London Airports / Birmingham and East Midlands (for Dublin) / Alicante and Murcia

Edinburgh and Glasgow / Bristol and Cardiff ; Bristol and Exeter / Tenerife North/South

Manchester, Liverpool, Leeds Bradford (for Dublin)

7.4. Markets for direct flights and indirect flights

(373) On a given O&D pair, passengers can travel either by way of a direct flight between the point of origin and the point of destination (for example Dublin-Budapest) or by way of an "indirect" flight on the same O&D pair but via an intermediate destination

(for example Dublin-London-Budapest).

(374) The level of substitutability of indirect flights for direct flights largely depends on the duration of the flight. As a general rule, the longer the flight, the higher the likelihood that indirect flights exert a competitive constraint on direct flights (314).

(375) When defining the relevant O&D markets for air transport services, the Commission has considered in previous decisions (315) that with respect to short-haul routes

(generally below 6 hours flight duration) indirect/one-stop flights do not generally provide a competitive constraint to direct/non-stop flights absent exceptional circumstances (for example the share of indirect flights in the overall market is significant) (316).

(376) In the 2007 Decision, the Commission concluded that, for the assessment of that transaction, indirect flights should not be considered as part of the relevant market (317).

(377) Ryanair submits that, in principle, indirect flights may also constrain direct services on flights of a duration of less than three hours (318). According to Ryanair, for the overlap routes to and from Ireland, potentially relevant indirect flights would largely be flights via the hubs of Air France/KLM/Alitalia, British Airways/Iberia/bmi or Lufthansa/Austrian/Swiss (319).

(378) Ryanair argues that the following routes would be constrained by indirect flights: Dublin-Budapest, Dublin-Fuerteventura, Dublin-Gran Canaria, Dublin-Lanzarote, Dublin-Malaga, Dublin-Palma, Dublin-Rome, Dublin-Tenerife and Dublin- Warsaw/Modlin, Cork-Lanzarote and Cork-Tenerife (320). Ryanair further claims that this list identifies only services offered by the same airline on each leg and if alliance partner or codeshare services were included, the list of available indirect services would increase significantly (321). Ryanair did not however identify the list of routes on which indirect flights would exist and constrain the merged entity post-Transaction.

(379) Aer Lingus contends that all overlap routes are short-haul or medium-haul routes and are well-served by direct flights and that there are essentially no circumstances under which an indirect service would be a substitute for a direct service on these routes. Furthermore, Aer Lingus argues that a direct flight is superior to an indirect flight from the customer's perspective on a number of dimensions: not only on travel time

(in Aer Lingus' view, indirect routes involve a significant increase to travel time) but also convenience and vulnerability to disruption. The magnitude of these differences is such that there are no practical circumstances under which an indirect flight would be viewed as a substitute for a direct flight by a customer. Lastly, Aer Lingus also stated that it does not monitor the prices of indirect flights on the overlap routes, it never monitors the fares of indirect routings using QL2 and it never inputs such fares into Sabre Inflight for use in yield management. Generally speaking, Aer Lingus believes that indirect flights on the overlap routes would be significantly more expensive than the direct services offered from Dublin (322).

(380) Other carriers which responded to the Commission's market investigation have different views on the question of whether for the purposes of the Transaction indirect flights would be substitutable for direct flights on the overlap routes. While on the one hand most carriers consider that for their passengers indirect flights are substitute services for direct flights when the extension of the total flights duration is limited and the total travel time is still comparable with the non-stop service, others consider that indirect flights would be substitute services under other circumstances for example depending on the fare level or the frequency and scheduling (323).

(381) The Commission considers that, because the overlap routes in the present Transaction are short-haul or medium-haul flights, indirect flights are unlikely to exercise a competitive constraint on direct flights (324). However, the question of whether indirect flights would belong to the same market can be left open, as it would not ultimately change the outcome of the competitive assessment in this Transaction. This is a very cautious and conservative approach.

7.5. Distinction between groups of passengers

(382) The Commission has in its decisional practice (mostly concerning network carriers) considered distinguishing, for a given O&D route, between: a relevant market for time sensitive (TS) passengers who tend to travel for business purposes, require significant flexibility for their tickets and are willing to pay higher prices for this flexibility and a relevant market for non-time sensitive (NTS) passengers who travel predominantly for leisure purposes, do not require flexibility with their booking and are more price-sensitive than the first category. Moreover, they tend to buy their tickets longer before the flight date than TS passengers. The Commission considered that NTS passengers typically tend to purchase restricted tickets while TS tend to purchase unrestricted tickets. If this approach was followed, it would mean that on a given O&D pair, competition conditions may be different on the TS segment and on the NTS segment, which should be assessed separately.

(383) However, in the 2007 Decision (325), while acknowledging the existence of different passenger categories and although an overall majority of Ryanair and Aer Lingus' customers on the overlap routes were rather non-time-sensitive and were flying for private purposes, the Commission found that it was not appropriate to define separate markets for different categories of passengers, whether according to the distinction between TS and NTS passengers or other distinctions such as the distinction between business and leisure passengers, or the "time between booking and departure" approach. The Commission's conclusion was based in particular on the finding that both airlines did not discriminate between different types of passengers by offering unrestricted alongside restricted tickets and offered one-way tickets only. (326).

(384) Ryanair submits that this approach remains applicable for the assessment of the Transaction. Ryanair, which offers only one class of non-refundable tickets (that is to say only restricted tickets) does not think that market circumstances have changed in a way to warrant an analysis different from that undertaken in the 2007 Decision (327).

(385) As mentioned above, since 2009, Aer Lingus has emphasised its positioning as a value carrier between the low cost and full service carriers, with a strong core product and offering additional paid options. As such, Aer Lingus targets leisure NTS and business TS passengers. In 2011, Aer Lingus introduced "flexi" tickets on short-haul routes (328). Aer Lingus argues that they do not believe that for purposes of the present case it is appropriate to segment point-to-point scheduled passengers into different categories. Aer Lingus also claims that, whilst various European Commission Decisions have previously concluded that it may be appropriate to distinguish between different types of passengers, this is not appropriate in relation to the market for travel ex-Ireland as it remains the case that the vast majority of Aer Lingus' short-haul fares are one-way restricted tickets. Aer Lingus' flexible fares represent a small portion of their overall sales and they do not capture all TS or business passengers (329).

(386) Overall, a large majority of respondents to the market investigation in the present case support the Commission's approach in the 2007 Decision, considering that no significant changes in the market conditions have taken place since 2007 which would warrant a different conclusion as regards passenger differentiation in the Transaction (330). They generally confirm that there is no need to define separate markets for the different categories of passengers for the purpose of analysing the Transaction.

(387) Therefore the Commission concludes that, for the purposes of the Transaction, it is not appropriate to define separate markets for different categories of passengers, whether according to the distinction between TS and NTS passengers, the distinction between business and leisure passenger or other approaches such as the "time between booking and departure" approach.

7.6. Substitutability between charter and scheduled services

7.6.1. General issues regarding charter services

(388) Charter flights, as opposed to scheduled flights, are usually defined as air transport services that take place outside normal schedules, normally through a hiring arrangement with a particular customer (in particular a tour operator). Charter companies often fly to destinations where no scheduled airline is active and usually operate on a seasonal basis with a relatively low frequency of flights, in response to the requirements of tour operators (for example, once a week on Saturday, only during the summer or only during the ski season).

(389) Indeed, charter companies did not traditionally sell tickets directly to passengers. They sold seats on their aircraft to tour operators which included the flight in a holiday package. As such, the flight (air transport) is part of a package holiday, the price of which includes flights, accommodation and other services. However, charter companies now sell to some extent so-called "dry seats" directly to passengers (that is to say seats only without other services), in addition to sales of seats to tour operators for inclusion in package holidays.

(390) Ryanair have significantly expanded their operations to traditional charter destinations and it argues in its response to the decision opening the proceedings that the decline in the charter business ex-Ireland is a direct consequence of Ryanair's expansion into several routes that were traditionally served by charter companies. In Ryanair's view this is clear evidence of the substitutability between charter services and Ryanair's services. Furthermore, Ryanair contends that charter companies have the proven ability to expand capacity on routes they wish to serve at short notice and can do so in response to surges in market demand.

(391) In past Decisions regarding the airline sector and in particular in the 2007 Decision (331), the Commission concluded that most of the services offered by charter companies are not in the same market as scheduled air transport services (package holiday sales (332) and seat sales to tour operators). However, the Commission has so far left open the question of whether or not dry seat sales of charter companies are part of the same relevant market since this would not have changed the competitive assessment in the cases where this was considered.

(392) Some respondents to the market investigation in the Transaction indicated that there would be competition between the services provided by scheduled carriers and those provided by charter carriers, "where the two providers are operating to a similar destination market. For example, [...] for people seeking a holiday to a Greek island (or indeed a 'beach' destination) sufficient numbers of people will compare options across charter and scheduled operations to put these in the same markets" (333). Furthermore, other respondents explained that "both charter and scheduled services are providing simply a means of transport between the point of departure and destination of the passengers who use those services. [...] As such, [...] that there is no difference between the customers who book charter airlines and those who book scheduled airlines. It is more likely that distinctions may be drawn by destinations [...]" (334).

(393) In line with the Commission's decisional practice, the Commission will distinguish different types of activities by charter companies in its analysis: sales of package holidays to end customers, sales of seats to tour operators, and sales of dry-seats to end customers.

7.6.1.1. Market for package holidays

(394) The Commission has assessed to what extent Ryanair and Aer Lingus might face competition from providers of package holidays.

(395) Evidence from the market investigation pointed to the fluid character of the market as regards competition between the various services. One respondent indicated that "the disappearance of traditional demarcation lines in the sale of travel products (package vs. independent, charter vs. scheduled vs. low-cost carriers, travel agent v tour operator) has also created a fluid market in which distinct business roles and models have broken down and competition has increased even more" (335). This has therefore provided consumers with increased choice of individual travel components, packages, suppliers and independent, impartial information that can easily be researched and compared to capture the best deal. The increasing tendency of consumers to "self-package" flights and accommodation (and other travel products) has given rise to demand-side substitution between package holidays and independently arranged holidays, placing the charter company element of the package under indirect competitive pressure from low cost carriers and scheduled airlines that may otherwise be used in an independently arranged holiday. As regards sales made to tour operators, charter companies face a similar degree of demand-side substitutability with low-cost carriers and scheduled airlines (336).

(396) Furthermore, another respondent considered that "it is possible that some customers, on routes where charter services overlap with those of Ryanair or another low cost carrier, may consider a point-to-point seat-only ticket which they would buy as part of assembling their dynamic package as (imperfectly) substitutable with a tour package offered by a charter airline. Indeed, Ryanair and other low cost carriers typically have links on their respective websites to hotel bookings, car hire and other ancillary services, which are potential components of a "dynamic" package" (337).

(397) Indeed, Ryanair and Aer Lingus both provide the possibility to book a hotel and to rent a car on their websites. However, Ryanair does not sell and invoice accommodation or rental cars, but receives a fee if such products are booked via Ryanair's website (338). In that respect, Ryanair's service would rather be similar to that of a travel agency or retailer. Similarly Aer Lingus does not sell and invoice accommodation or rental cars, but receives a fee if such products are booked via its website. Aer Lingus confirmed that it "sources ancillary partners who can provide its passengers an added value service that fits with their travel booking e.g. car hire, hotels, travel insurance etc. Aer Lingus partners with the leading companies in each relevant area but the service is completely provided by each partner and the passenger contracts directly with the relevant partner with Aer Lingus merely acting as an agent. Aer Lingus receives a commission for each booking purchased via its website, and therefore does not directly sell car hire or accommodation which are sold via partners" (339). In that respect, Aer Lingus' service would also rather be similar to that of a travel agency or retailer.

(398) Therefore, while Ryanair and Aer Lingus could arguably be deemed to be, at least to a certain extent, active in the market for package holidays and therefore exercise pressure on tour operators and charter carriers, this is not the case the other way around (that is to say that tour operators or charter carriers do not exercise pressure on Ryanair and Aer Lingus) for routes to/from Ireland. [Ryanair and Aer Lingus monitoring practices of charter companies]* (340). The fact that tour operators and charter carriers are negatively affected by Ryanair's and Aer Lingus' services does not mean that the tour operators and charter carriers exert any competitive pressure on the Parties.

(399) Furthermore, significant differences in the characteristics of the two services can be clearly identified. The providers of package holidays by definition offer a number of additional services and request their customers to acquire a number of complementary products such as accommodation, transportation at destination, meals, travel guides, animation (341). These services go well beyond the product that can be booked on Ryanair's and Aer Lingus' website (flight plus hotel or car). Furthermore, the Parties' customers assemble the package themselves and customise which services they require, having therefore a lot of flexibility.

(400) In addition, tour operators offer fixed packages with fixed flight dates (typically with flights only at weekends) and frequencies, while low-cost airlines such as Ryanair and Aer Lingus offer scheduled and thus flexible frequencies. This means for instance that a customer willing to stay 3 o 5 nights would not be able to use charter services, which would operate once a week only, imposing a stay of 7 or 14 nights at a destination (342).

(401) The Commission therefore considers that if services provided by charter companies and tour operators were to exercise a constraint on the Parties this would be only regarding those customers who want to book a flight and accommodation and who would have considered booking a package holiday from a tour operator or charter company (so-called "care-free" service).

(402) Indeed, evidence collected through market investigation indicates that customers of tour operators are looking for "the complete package including the flights, accommodation and transfers" (343) and they would choose a tour operator because these will provide them "with a convenient way of delivering their travel requirements within a group of like-minded travellers" (344).

(403) Ryanair and Aer Lingus on the other hand do not target these passengers. Indeed, many passengers flying with Ryanair or Aer Lingus would not need such a package (for example passengers flying to holiday homes, passengers flying to visit friends and family, etc). Even those customers who book a hotel via Ryanair and Aer Lingus, or independently, are likely not to consider package holidays as an alternative, since they would not want to pay for or are not interested in the extra services provided by a tour operator such as hotel transfer, travel guides and animation. Indeed, Ryanair's business objective is to offer low fares that generate increased passenger traffic (345).

(404) As a result, a test based on a small but significant non-transitory increase ("SSNIP") of 5 % or 10 % in the price of scheduled flights is unlikely to lead a sufficient number of passengers to purchase a package holiday so as to render the SSNIP unprofitable for a hypothetical airline with a monopoly on all scheduled flights.

(405) In its 2007 Decision, the Commission concluded that sales of package holidays offered by charter companies and tour operators are not in the same market as the scheduled air transport services provided by Ryanair and Aer Lingus on the then affected routes (346). No material changes in circumstances have occurred in this regard since 2007. Indeed, the majority of the respondents to the market investigation confirm the Commission's conclusion in the 2007 Decision that most of the services offered by charter companies (package holiday sales and seat sales to tour operators) are not in the same market as scheduled point-to-point air transport services (347).

(406) Therefore, seats included in package holidays could not be considered as substitutable for seats on scheduled flights as most passengers relevant for the purpose of the investigation of the Transaction purchase seats only and not package holidays.

(407) As a conclusion, the Commission considers that sales of package holidays offered by charter companies and tour operators are not in the same market as the scheduled air transport services provided by Ryanair and Aer Lingus on the overlap routes.

7.6.1.2. Sales of seats to tour operators

(408) Charter companies sell seats (or entire flights) to holiday companies, who then integrate the flight into a package holiday (348). In previous decisions, the Commission regarded the wholesale supply of airline seats to tour operators as a distinct market from the supply of scheduled air transport services to end customers (349).

(409) Indeed, from a demand-side perspective, conditions on the market for sales of seats to tour operators differ significantly from those on the market for sales to individuals. The market for sales of seat packages to tour operators is one which is upstream to the market for seat sales to individuals. Accordingly, the competitive conditions in this market are manifestly different, since tour operators have different requirements from those of individual customers (for example, buying of large seat packages, negotiation of rebates, taking into account of customers' needs in terms of flight times).

(410) It can thus be concluded that wholesale sales of seat packages to tour operators are not in the same market as scheduled air transport services for end-customers.

7.6.1.3. Market for dry seat sales to end-customers

(411) Ryanair has identified charter companies offering seat only services on the relevant city-pairs. Ryanair submits also that, where a charter company currently operates on the route, that airline is clearly capable of adding additional flights or additional routes to and from Dublin at any time and in any season (350).

(412) Indeed, on the same O&D route, the dry seat sales of charter companies are similar to the sale of scheduled air transport passenger transport services. While there are some differences between these services, from the customer's perspective, under certain circumstances, these could be seen as substitutes.

(413) The market investigation indicated that since the Commission's 2007 Decision, Ryanair have significantly expanded their operations to traditional charter destinations and one respondent highlighted the fact that "there has been some convergence between the charter flight operations and scheduled services and as a result in many cases it is now possible for passengers to independently book seats on charter services" (351).

(414) However, regarding the sale of dry seats, some respondents to the market investigation pointed out that charter flights are of inferior quality compared to scheduled flights not least because of the limited capacity of dry seats and the limited visibility to consumers (352) in addition to the frequencies and fare price (353). Indeed respondents consider that dry seats should not be part of the same market as scheduled services because "The typical travel customer would not be aware of the existence of dry seats as a) customers are not generally aware that charter airlines offer flight-only seats b) these would usually only be available via a tour operator, who do not generally specialise in last-minute travel and c) it is difficult to envision how a charter flight would effectively sell last minute dry seats due to their business model generally favouring the packaged approach and not being set-up for such sales" (354).

(415) Ryanair submits that given the specificities of this case, in terms of market definition, charter flights are relevant for the following routes:

Table 26: Overlap routes where charter companies are present

Relevant route / Charter competitor

1 / Dublin-Barcelona / Thomson, Orbest (earlier Iberworld)

2 / Dublin-Faro / Thomson, Germania, Orbest (earlier Iberworld)

3 / Dublin-Gran Canaria / Thomson, Orbest (earlier Iberworld)

4 / Dublin-Ibiza / Thomson, Orbest (earlier Iberworld)

5 / Dublin-Lanzarote / Thomson, Orbest (earlier Iberworld)

6 / Dublin-Malaga / Thomson, Germania

7 / Dublin-Palma / Thomson, Air Contractors, Orbest (earlier Iberworld)

8 / Cork-Barcelona / No specific charter company mentioned

9 / Cork-Lanzarote / Monarch

10 / Cork-Malaga / Monarch

11 / Cork-Palma / Monarch

Source: Form CO, Annex 7.3.(a)(i)

(416) In any event, the Commission does not need to reach a definitive conclusion on the question whether dry seats are part of the same market as scheduled air transport services in the Transaction because it would not change the outcome of the competitive assessment.

(417) Therefore, the relevance or otherwise of dry seats will be considered by the Commission in the competitive assessment in the route- by-route analysis (section 8.9).

7.6.2. Conclusion on competition from charter companies

(418) The Commission therefore concludes for the purposes of the Transaction that most of the services offered by charter companies (package holiday sales, seat sales to tour operators) are not in the same market as scheduled point-to-point air transport services.

(419) Furthermore, it can be left open whether or not dry seat sales are part of the same relevant market as scheduled point-to-point air transport services, because this would not change the conclusion of the competitive assessment of the Transaction.

7.7. Intermodal substitution

(420) When determining the scope of the relevant markets on various city-pairs, the Commission analysed in earlier Decisions the level of substitutability of air travel with other available transport modes (355). Such intermodal substitutability was in particular considered in cases where relevant alternative modes of transport were deemed comparable options for prospective passengers in terms of price, aggregate travel time, schedule or quality.

(421) No intermodal substitutability issue appears relevant for the assessment of the Transaction, nor has any been put forward by Ryanair.

(422) The Commission therefore concludes, for the purposes of this Transaction, that intermodal substitution is not relevant for any of the overlap routes.

8. COMPETITIVE ASSESSMENT

8.1. General framework for the assessment

(423) As already described in Section 6, Ryanair argues that the air transport market in Europe has materially changed since 2007 and there are compelling reasons why the Commission should approve the Transaction. (356)

(424) Ryanair refers to the fact that Europe's flag carriers are consolidating and Aer Lingus has failed to find a consolidation partner. Aer Lingus is said to be a subscale peripheral carrier that is not consistently profitable and cannot grow or compete with larger carriers in Europe. Ryanair also points out that Dublin Airport is currently operating at just over 50% capacity. The earlier congestion problems no longer exist due to the drop in annual traffic and the expansion carried out in the meantime, in particular, with the opening of Terminal 2.

(425) Changes in market circumstances since 2007 have been analysed by the Commission. In broad overview, for the purposes of the competition analysis, the main changes appear to be the following:

- The financial and economic crisis, which started in 2008, affecting Member States, including Ireland. The crisis has also hit the air transport sector in Ireland and in other Member States relevant for the assessment of the present Transaction, such as Spain and the United Kingdom.

- Some further consolidation is taking place in the air transport sector.

- The number of airlines operating at Dublin airport has decreased significantly compared to 2007.

- Aer Lingus and Ryanair compete on a greater number of routes compared to the 2007 Decision. The number of routes of overlap has increased significantly, from 35 in 2007 to 46 in 2012. Furthermore, on almost 85% of these routes (63% in 2007) there are no other scheduled airlines active.

- With regard to short-haul point-to-point traffic out of Dublin airport, concentration has further increased. The combined market share of Ryanair and Aer Lingus of flights out of Dublin (in terms of passenger volumes) in IATA summer season of 2012 reaches 85%, an increase from the 73% in IATA summer season of 2007.

- The operation of charter companies out of Dublin appears to have decreased by 70% between 2007 and 2011.

- Finally although Dublin Airport Authority ("DAA") has added capacity at Dublin airport since 2007, especially with the opening of Terminal 2 in late 2010, Dublin airport is still a level 3 coordinated airport (357).

(426) Ryanair notes that "a relatively weak domestic economy does not only represent an opportunity for future growth, but it also ensures that airlines have an incentive to lower prices in order to increase consumer demand." (358)

(427) However, Ryanair's stance is somewhat contradicted by other statements, for example, where Ryanair claimed that "We expect market conditions in Europe to remain tough as recession, austerity, high fuel costs, and excessive Government taxes dampen air travel demand." (359)

(428) As mentioned by DAA, there is considerable evidence that traffic volumes at Dublin Airport are affected by economic growth (see Figure 1 below) (360). For instance, Boeing has reported that "Worldwide economic activity is the most powerful driver of growth in commercial air transport." (361) There is also economic literature suggesting that, while the demand for passenger services arises from the complex interaction of a large number of factors which affect the different market segments differentially, as far as the impact of economic growth is concerned, there appears to be a very strong correlation between the annual rates of growth in the world's GDP and the growth rates in air travel, measured in revenue passenger-kilometers. Broadly speaking, there would be a two to one relationship between demand for air travel and world GDP (362).

Figure 1: DAA Airports Passenger Performance v Irish GDP growth rates 2000-2012 (363)

<Graphic 1>

(429) There has been a significant reduction in economic output in Ireland with real GDP falling by 8.2% between 2007 and 2010 according to most recent estimates. Although real GDP growth of 1.4% was recorded in 2011, this was driven by net exports, with domestic demand (the sum of private consumption, government consumption, domestic investment and changes in stocks) continuing to contract for the year, and down by a cumulative 21% between 2007 and 2011 in real terms (see Figure 2).

Figure 2: Irish domestic demand and air passenger movements

<Graphic 2>

(430) Turning to the economic outlook for the medium term, the most recent Commission forecast projects a continuation of modest real GDP growth over the 2012-2015 period. However, domestic demand is set to make a negative contribution to growth (of about 1 percentage point) over the period as the adjustment in the domestic economy continues (364).

8.2. Treatment of Aer Arann

(431) Aer Arann is a scheduled regional carrier operating between Ireland and the United Kingdom under a franchise arrangement with Aer Lingus.

(432) The company aims at providing "regional connectivity to major regions in the United Kingdom from Ireland's 3 main airports by providing frequency of service at the right times under Aer Lingus' respected brand and product". It also aims at improving connectivity for these regions by timing flights to connect with transatlantic flights from Dublin (365).

(433) Aer Lingus does not have a shareholding in Aer Arann (366).

(434) Aer Arann operates on routes between Ireland and the United Kingdom. It has 12 aircraft, eight of which are based in Dublin, two in Cork, one in Shannon and one in London Southend (367).

(435) Aer Arann carried 1 million passengers in 2011 and in the last financial year, Aer Arann generated a turnover of nearly EUR 100 million. Aer Arann has not been profitable since its restructuring in 2010, but expects to generate profits in 2013 (368).

(436) Aer Arann operates on 7 of the routes subject to this Decision, namely Dublin - Bristol/Exeter/Cardiff, Dublin-Edinburgh/Glasgow, Dublin - Glasgow/Prestwick, Dublin - London, Cork - Manchester/Liverpool, Knock - Birmingham/East Midlands, Shannon - Manchester/Liverpool (369).

(437) The Dublin - Edinburgh/Glasgow (370) and Dublin - London (371) routes are operated by both Aer Arann and Aer Lingus (372).

8.2.1. The Franchise agreement

(438) Aer Lingus and Aer Arann entered into a franchise agreement in 2010. The agreement was expanded in 2012 to include all services operated by Aer Arann. As a result, from April 2012 all Aer Arann flights are operated under the franchise agreement with Aer Lingus and branded "Aer Lingus Regional".

(439) A 10 year extension of the franchise agreement was agreed in late 2012 (373).

(440) Aer Lingus describes the key objectives of the franchise agreement as the following: (i) to expand the Aer Lingus brand reach at United Kingdom regional airports; (ii) to improve connectivity with Aer Lingus' North Atlantic network; (iii) to provide increased travel options for Aer Lingus customers and retain loyalty; (iv) to positively differentiate Aer Lingus' product from competition based on improved access and schedules; and (v) to deliver new revenue streams from franchise fees and commissions (374).

(441) The key terms of the franchise agreement are as follows:

- Under the franchise arrangement, Aer Arann operates its routes under the "Aer Lingus Regional" brand (375).

- Aer Arann assumes full operation and commercial responsibility for the services covered by the franchise agreement (376).

- Aer Arann employs the crew and engineers and has its own aircraft (377).

- Franchise services are sold exclusively through Aer Lingus direct and indirect sales channels (378). Aer Arann does not have a booking engine or a website (tickets on Aer Arann flights are bought directly from the Aer Lingus' website) (379).

- While Aer Arann has principal responsibility for the marketing and promotion of the franchise services, Aer Lingus incorporates routes operated as franchise services in relevant information and promotional activity, including, but not limited to, network based advertising, route group promotions, and web and print route and timetable information (380).

- Aer Lingus provides customer relations management for the franchise services (381).

- Aer Arann pays a franchise fee to Aer Lingus (382).

8.2.2. Aer Arann is not an independent competitor of Aer Lingus

(442) Ryanair claims that Aer Arann should be treated as an independent competitor. It argues that Aer Arann is a separately owned and managed airline that is legally and operationally independent of Aer Lingus. It further argues that when a route is operated by Aer Arann (and not by Aer Lingus), Aer Lingus has no control over Aer Arann's prices, revenues, costs, schedules, or its route network, which are all determined independently by Aer Arann. According to Ryanair, Aer Lingus would merely receive a fee per passenger for the use of its brand. It claims that Aer Arann views itself as an independent competitor and that Aer Arann nowhere "suggest[s] that it was not a competitor of its own, operating independently from Aer Lingus" (383). Ryanair submits that its claim that Aer Arann must be treated as an independent competitor is further supported by the fact that the Stobart Group has taken control over Aer Arann (384).

(443) Aer Arann stated that it does not compete directly with Aer Lingus (on fares or any other commercial aspect) (385). It further considers that its schedule complements the Aer Lingus schedule and product offering by providing frequency benefits with smaller aircraft. Aer Arann fears that its future at all Irish airports could be seriously in doubt if the Transaction goes ahead. It explains that its ability to operate following its financial restructuring in 2010 is due to the fact that it now has access to a brand that enables a small airline like Aer Arann to enter competitive routes and provide the customer with frequency, connectivity and choice and it considers it unlikely that there would be such cooperation with a new entity as it is unclear what their business model would be. It considers that it is not really possible for Aer Arann to create a brand as strong as the Aer Lingus brand. It further indicates that it is unlikely that Aer Arann operating under Aer Lingus Regional brand would be able to enter any new routes that Ryanair is already operating should the Transaction take place (386).

(444) Aer Lingus does not consider Aer Arann to be a competitor of Aer Lingus while Aer Lingus considers that Aer Arann is a competitor of Ryanair (387).

(445) Aer Lingus stresses that, under the terms of the franchise agreement, all of Aer Arann's operations are operated under the Aer Lingus brand umbrella and all tickets are sold exclusively via Aer Lingus distribution channels (principally aerlingus.com). In the absence of the franchise agreement, Aer Arann would not have the resources or systems necessary to carry out the required investment in distribution, brand development, revenue accounting, and marketing. According to Aer Lingus, it is therefore incorrect to assert that Aer Lingus has no control over Aer Arann's prices, revenues, costs, schedules, or route network. On the contrary, Aer Arann no longer exists as a separate brand and does not have any material commercial operations outside of its franchise agreement with Aer Lingus (388).

(446) Aer Lingus believes that without the franchise agreement, Aer Arann would not be able to operate routes in competition with Ryanair as it currently does. It indicates that at a brand level the competition between Aer Lingus and Ryanair includes those routes operated by Aer Arann under the Aer Lingus Regional brand and is broader than that indicated by the routes on which Aer Lingus and Ryanair directly operate. As such, it feels that the competitive interaction between Aer Arann and Ryanair is also relevant to the analysis of the Transaction (389).

(447) In the OFT's recent decision under reference on Ryanair's minority interest in Aer Lingus (390), whilst acknowledging the commercial relationship between Aer Lingus and Aer Arann, the OFT did not find it necessary for the purposes of its decision to decide on the extent to which Aer Arann overlaps should be included or excluded from the analysis.

(448) The Commission has already been confronted with the existence of a franchise agreement between carriers in the past. In the Iberia/Clickair/Vueling case (391), the Commission had to assess the relationship between Iberia and Air Nostrum. Despite the fact that Air Nostrum was an independent regional airline, in which Iberia was not a shareholder, the Commission decided to add Air Nostrum's shares of the market to those of Iberia because of the existence of an exclusive franchise agreement between them.

(449) The fact that Aer Arann is a separately owned and managed airline does not mean that it is an independent competitor to Aer Lingus. To the contrary, the Commission is of the view that Aer Arann is actually closely linked to and dependent on Aer Lingus through the franchise agreement.

(450) Under the franchise agreement, Aer Arann undertakes to operate the agreed scheduled flights exclusively under the Aer Lingus Regional brand and using Aer Lingus flight numbers. All of Aer Arann's operations are governed by the franchise agreement. Aer Arann therefore always appears in the market as "Aer Lingus Regional" as opposed to under its own trading name.

(451) A main feature of the franchise cooperation between Aer Lingus and Aer Arann is to provide regions in the United Kingdom with connectivity for example through Aer Lingus transatlantic flights. Accordingly, the schedules of the two carriers are complementary. The fact that Aer Arann can decide on frequencies and schedule of the flights and that it can set prices does not imply that they are to be considered an independent competitor of Aer Lingus. The fact that on two of the routes, Aer Arann and Aer Lingus operate in parallel does not contradict this. (392) Moreover, the franchise agreement requires Aer Lingus' approval for the use of the Aer Lingus Regional brand on any new routes (393).

(452) Aer Lingus integrates the routes operated by Aer Arann under the franchise agreement in the relevant information and promotional activity, including but not limited to network-based advertising, route group promotions, route map, and timetable information (394).

(453) Tickets for Aer Arann's flights are sold exclusively through Aer Lingus' sales channels and Aer Arann also depends on Aer Lingus for a number of other supporting services (such as customer relations management).

(454) In addition, Aer Arann is not a very well-known brand in Ireland. Market research carried out in June 2012 and provided by DAA, demonstrates that Aer Arann, despite being present on the Irish market since 1970, is significantly behind Aer Lingus and Ryanair. The respondents were asked to name Irish travel/transport companies. Two measures of spontaneous and total awareness were taken into account. Three airlines were mentioned with the following scores: Aer Lingus (52%/98%); Ryanair (50%/97%), Aer Arann (25%/88%) (395).

(455) Aer Lingus does not consider that it competes with Aer Arann, and Aer Arann does not consider that it competes with Aer Lingus (on fares or any other commercial aspect) (396). On the other hand, Aer Arann monitors Ryanair's fares on all routes on which Aer Arann operates in competition with Ryanair (397).

(456) The results of the market investigation show that the majority of those competitors that expressed a view on the matter consider that Aer Arann is a competitor of Ryanair but not of Aer Lingus (398).

(457) Moreover, it is noted that Ryanair itself has included the Aer Arann frequencies and market shares in the data provided in the Form CO and its annexes and it has described the relationship between Aer Lingus and Aer Arann as a "partnership" (399).

(458) The fact that the Stobart Group has taken control over Aer Arann (400) does not change the Commission's assessment as the Stobart Group has confirmed that the current relationship between Aer Arann and Aer Lingus will continue post-Transaction and that, in particular, the franchise agreement will be maintained (401). This was confirmed by Aer Arann that indicated that "the proposed restructuring and new business plan focuses on further developing the success of our franchise agreement with Aer Lingus. The plan will focus on extending the scale of the franchise agreement in tandem with a fleet replacement programme at Aer Arann. The proposed agreement is franchise agreement-centric. [.../...] The proposed investment is dependent on the franchise agreement as it contains the vital economical attributes to allow for the Aer Arann business to be sustainable and grow. It is therefore vital that the franchise agreement remains instact including all lines of flying as contained today, along with proposed additional routes." (402)

(459) It appears therefore that Aer Arann cannot be considered as an independent competitor from Aer Lingus, while it is a competitor of Ryanair.

(460) It is thus appropriate to add Aer Arann's shares of the market to those of Aer Lingus for the purposes of the competitive assessment.

8.2.3. Conclusion

(461) In the light of the above, the Commission considers that Aer Arann is a competitor of Ryanair, but not of Aer Lingus. The market shares of Aer Arann (operating under the Aer Lingus Regional brand) have to be attributed to Aer Lingus for the purposes of the competitive assessment.

8.3. Market shares and concentration levels

(462) Ryanair is of the view that, in previous airline merger decisions, the Commission has repeatedly found that high post-Transaction market shares do not necessarily give rise to competitive concerns, particularly where one or more competitors are active on the relevant route (403). This would apply in particular on 7 routes ex Dublin: Frankfurt, London, Madrid, Munich, Paris, Stockholm, and Bristol/Cardiff/Exeter.

(463) According to the "Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings" (404) (" the Horizontal Merger Guidelines"), market shares and concentration levels provide useful first indications of the market structure and of the competitive importance of both the merging parties and their competitors.

(464) The Horizontal Merger Guidelines also state that according to well established caselaw, very large market shares (50% or more) may in themselves be evidence of the existence of a dominant market position (405).

(465) The General Court confirmed that although the importance of market shares may vary from one market to another, very large shares are in themselves, and save in exceptional circumstances, evidence of the existence of a dominant position. (406) In another case concerning the airline sector, the General Court also stated that "market shares held by the parties to the merger led the Commission to conclude that commitments should be offered on the markets affected and on which those parties enjoyed a market share of almost 50%, thereby respecting the presumption of dominance as laid down by the case-law". (407)

(466) Finally, in the 2010 Judgment, the General Court also established with respect to the 2007 Decision that "the Commission was rightly able to find that the acquisition of very high market shares as a result of both the implementation of the concentration and the concentration level associated with it were relevant indicators of the market power which would have been acquired by Ryanair-Aer Lingus combined". (408)

(467) Ryanair's and Aer Lingus' activities overlap only in relation to scheduled short-haul passenger services, from or to airports located in Dublin, Shannon, Cork and Knock.

(468) This decision deals with 46 direct/direct overlap routes: 33 from Dublin, 9 from Cork, 2 from Knock, and 2 from Shannon (409).

(469) Of the 46 routes (410), 26 were also assessed in the 2007 Decision. Compared to the situation in 2007, the merged entity would have larger market shares on most of these 26 routes, in particular due to the exit of competitors on some routes. In addition, there are 20 new routes, on which Ryanair and Aer Lingus now compete and on which they did not in 2007.

(470) Several route overlaps are seasonal: Ryanair and Aer Lingus overlap only in IATA summer season on 9 routes (411).

(471) In addition, on 27 of the overlap routes there are direct airport-to-airport overlaps as both Parties fly from and to the same airports at both ends of the O&D routes (412). On the remaining 19 routes there is a city-to-city overlap with Ryanair and Aer Lingus using different but substitutable airports (413).

(472) As detailed in Section 8.9.4, the Transaction would create a monopoly on 28 routes compared to 16 in the 2007 Decision.

(473) Furthermore, there are 18 overlap routes on which the merged entity would operate alongside other carriers. On 11 of these 18 routes, the only other operating carrier(s) is a charter company. As described in Section 8.9.6, the Parties' combined market shares on each of these routes exceeds 80%.

(474) On 6 of the 18 routes on which the Parties operate alongside other airlines, the competitors of the Parties are scheduled carriers, whose business model is different from those of the Parties. On these routes, as described in Section 8.9.5, the combined market shares of Ryanair and Aer Lingus would be above 50% according to the most recent available data. On the last remaining route, Dublin- Bristol/Cardiff/Exeter, if flights between Dublin and Exeter are included in the relevant market, as described in Section 8.9.5.1, the Parties' competitor is Flybe, a scheduled regional carrier with a marginal market share between [5-10%] (414).

(475) Upon the Commission's request, Ryanair computed HHIs for all routes using passenger data (415). The HHIs are high in absolute terms, and often nearly double post- Transaction, going up to 10 000 where the Transaction leads to a monopoly. According to Ryanair, the lowest HHIs post-merger were 5009 in IATA winter 2011/12 and 5035 in IATA summer 2011 for the Knock-Birmingham route (416). The Commission disagrees with Ryanair's computation and, in fact, considers that a monopoly on city basis would be created post-Transaction on the Knock- Birmingham route. Despite already indicating a concentrated market, the HHI computation by Ryanair on the Dublin-Stockholm route does not take into account the significant market share increase by Aer Lingus on that route in summer 2012.

(476) Therefore, the Commission concludes that the Parties would have very high market shares on all routes on which their activities overlap. For 28 routes the Transaction would create a monopoly, depriving customers of the only available alternative. On the other routes, given the very high market shares of the Parties, convincing evidence is necessary to conclude that the Transaction would not result in the creation or strengthening of a dominant position.

(477) The Commission has carefully analysed whether there are circumstances which might exclude a significant impediment to effective competition as a result of the creation of a dominant position despite the high market shares. In order to take due account of the particularities of each route, the Commission has also analysed the competitive situation and in particular the main factors that could counteract the effects of the Transaction: closeness of competition (Section 8.4), and entry or barriers to entry (Sections 8.5, 8.6 and 8.7). Furthermore, additional analysis is conducted where relevant in the route-by route-assessment (Section 8.9).

8.4. Closeness of competition

(478) The Commission considers that Ryanair is in competition with Aer Lingus. In this regard, the market investigation has not provided material indications that market circumstances have changed since 2007. If anything, competition may have increased between the Parties.

(479) It is worth noting that in response to the decision opening the proceedings Ryanair has stated that it "is relatively little influenced by the activities of competitors" (417). Yet, Ryanair "does not claim that it belongs to "a market on its own" or that it does not in any circumstance compete against Aer Lingus" (418). Ryanair maintains a view that, at least on the seven routes, where the merged entity would face competing airlines, these competitors would exert a sufficient competitive constraint on the merged entity post-Transaction.

(480) Based on the evidence collected in the investigation, the Commission considers that it is clear that Ryanair and Aer Lingus (including Aer Arann) are in competition with each other. Both Aer Lingus and Ryanair monitor prices of the other when setting their own prices on a given route (419) and they react to each other's promotions and advertising campaigns. Aer Lingus has confirmed that it "monitors in particular the pricing of its main competitor, Ryanair, on routes served by both airlines at both the airport and city pair level [...] Ryanair's pricing has a direct effect on Aer Lingus pricing. Aer Lingus sees a reciprocal effect in how Ryanair responds to Aer Lingus' price changes" (420). [...]* (421).

(481) Therefore, it is apparent that Aer Lingus and Ryanair compete on the routes covered by this Decisison.

(482) The concept of "closeness of competition" may play an important role in better understanding the competitive constraint exerted by different competitors on each other in differentiated markets such as airline markets (422). This concept is however of limited added-value on markets in which from the outset only two competitors are active (such as on the 28 duopoly markets affected by the Transaction), since both competitors in a duopoly are by definition each other's closest competitors (423).

(483) For two airlines to be considered as offering services which are close substitutes to each other it is not necessary that the two services are identical. Some services may be closer substitutes than others. What matters from a substantive competition perspective is the high degree of substitutability between the services of Aer Lingus and Ryanair on the overlap city pairs.

(484) In view of this, it is considered that the analysis of closeness of competition will focus on those competitors who actually compete with Ryanair and Aer Lingus on the affected markets (424). Therefore, it is relevant to compare the Parties, but also the remaining competitors in the affected markets (425).

(485) In this context, there is a strong presumption that the Parties are each other's closest competitors in those markets where each of the Parties are by far the largest competitors in terms of market shares (426). This is obviously the case for the 28 merger to monopoly routes, given that the Parties are by definition the closest competitors on these routes. Even on those routes on which Ryanair and Aer Lingus face one or more competitors, the distance in market share between the competitor(s) and each of the Parties is in most cases significant. This is true for 11 routes where the Parties face competition from charter companies. This approach is also in line with the 2010 judgement which notes "for example, that, on the 22 routes [i.e., 16 routes leading to monopoly and 6 other routes where the Parties face charter airlines] on which Ryanair and Aer Lingus are the only airlines in operation, at present there are no other airlines in a position to offer scheduled air transport services. On those markets [...] Aer Lingus thus remains Ryanair's closest competitor" (427). Furthermore, on 4 out of the 7 remaining routes where the Parties are facing other scheduled operators (namely, Dublin-Bristol/Cardif/Exeter, Dublin-London, Dublin-Madrid and Dublin-Munich), the distance in market share between the competitor(s) and each of the Parties is also significant.

(486) For the remaining 3 out of 7 routes, namely the Dublin-Stockholm (428), Dublin- Frankfurt, and Dublin-Paris routes, there are only three carriers (Ryanair, Aer Lingus and the other incumbent carrier) with significant market shares, that is in the range of [20-30]%, [30-40]% and/or [40-50]% depending on the route and IATA season. However, even on those routes the Parties appear to be very close (if not the closest) competitors of each other based on their business models, and generally the closest competitors in terms of brand recognition, their well-established positions in Irish airports and bases in Ireland.

(487) In response to the Decision opening the proceedings, Ryanair reiterated its claim that product offerings of both airlines are significantly different, in particular in terms of fares, services and airports, and insisted that the question of closeness of competition should be addressed on a route-by-route basis, in particular given the differences of competitive conditions on many overlap routes.

(488) Individual route-by-route assessment of closeness of competition will be set out in Sections 8.9.5-8.9.6 for those routes, where the merged entity will face competition from other carriers.

(489) The following Sections 8.4.1-8.4.3 will address general arguments in relation to closeness of competition, concerning the Parties' strong market position on routes ex- Ireland and their business model and brand and their significant bases in Ireland. In the final Section concerning the closeness of base operations of the Parties. There, the Commission will first describe the advantages arising from base operations, such as increased flexibility in reacting to demand side (or supply-side) shocks, and cost savings arising from economies of scale and scope, and will then analyse the closeness of base competition between airlines with bases in Ireland and those with bases at the destination airports. Lastly Section 8.4.4 will outline the Commission's price regression analysis.

8.4.1. Ryanair and Aer Lingus have the strongest market positions on routes ex-Ireland

(490) Based on the evidence collected in the market investigation it is concluded that Ryanair and Aer Lingus have a particularly strong position in Ireland. The Transaction would lead to a monopoly on the 28 out of 46 routes concerned by this Decision. Even on routes where both airlines face competition, the position of their competitors is, in most cases, significantly weaker than the Parties' combined position.

(491) Ryanair has not contested the Commission's conclusion that Ryanair and Aer Lingus have a particularly strong position in Ireland in its response to the Decision opening the proceedings (429). However, in its response to the Statement of Objections, Ryanair has considered that the Commission's attempt to show that there is high degree of substitutability between Ryanair and Aer Lingus by reference to their market position in Ireland was unsubstantiated within the meaning of paragraphs 28-30 of the Horizontal Merger Guidelines (430), albeit without giving further explanations.

(492) As acknowledged in the 2007 Decision, the Commission considers that the fact that both carriers have equally strong positions in Ireland and face few competitors out of Dublin is a relevant consideration, among a number of elements, leading to the conclusion that Ryanair and Aer Lingus are each other closest competitors (431). The Commission considers that there are no material indications that market circumstances have changed since 2007 to an extent, that would warrant a different approach in the present Decision.

(493) Ryanair and Aer Lingus serve the single largest number of destinations ex-Dublin; the operations of Ryanair and Aer Lingus currently overlap on 48 routes out of Ireland. Aer Lingus and Ryanair also accounted for the largest share of passenger traffic at Dublin airport compared to the next largest carriers. Both Ryanair and Aer Lingus accounted for approximately [30-40]*% of passengers ex-Dublin in 2011, whereas all other carriers combined (for example, IAG, Lufthansa, Air France) accounted for less than [20-30]%*. Table 27 shows that, compared to the situation in 2007, the combined share of Ryanair and Aer Lingus of European departing seats at Dublin airport has increased from [80-90]*% (2007) to [80-90]*% (2012). Therefore, if anything, concentration on European short-haul flights ex-Dublin has steadily increased, and the combined position of the Parties has strengthened since 2007.

Table 27: Ryanair and Aer Lingus share of European departing seats as % of European departing seats at Dublin (432)

Year / Ryanair ('000) / Aer Lingus ('000) / Total ('000) / % Ryanair / % Aer Lingus / Combined Ryanair/Aer Lingus Share

2007 / [...]* / [...]* / [...]* / [40-50]*% / [30-40]*% / [80-90]*%

2008 / [...]* / [...]* / [...]* / [40-50]*% / [30-40]*% / [80-90]*%

2009 / [...]* / [...]* / [...]* / [40-50]*% / [30-40]*% / [80-90]*%

2010 / [...]* / [...]* / [...]* / [40-50]*% / [40-50]*% / [80-90]*%

2011 / [...]* / [...]* / [...]* / [40-50]*% / [40-50]*% / [80-90]*%

2012 / [...]* / [...]* / [...]* / [40-50]*% / [40-50]*% / [80-90]*%

(494) The figures in Table 27 also show that in number of seats, Aer Lingus has remained rather stable over the period, [...]*.

(495) The information provided by the DAA (433) also demonstrates the strengthening of the market positions of Aer Lingus and Ryanair at Dublin airport over the last five years. The combined share of European point-to-point passenger volumes of Ryanair and Aer Lingus increased from 73% in 2007 to 85% in 2011 (434). Overall slot capacity allocated at Dublin airport to the Parties increased from 60% (2008) to 75% (2012) and the allocation of peak slot capacity from 80% (2007) to 90% (2012) (435). Currently, Ryanair and Aer Lingus are the two largest customers of DAA accounting for 82% of total traffic (in the first half of 2012), while the remaining 18% are divided among over 60 other airlines (436).

(496) The DAA data (437) demonstrates furthermore that the short-haul passenger services market has become more concentrated over the last 5 years. The number of overlap routes has increased from 35 to 48 (438) over the last five years. Although in 2011 the Parties overlapped on 48 out of 139 routes from the airports managed by the DAA, these 48 routes represented around 78% of the total short-haul capacity for both airlines across Dublin, Cork and Shannon airports, compared to 65% for the 35 routes that overlapped in 2007. These 48 overlap routes (including the 44 out of 46 routes concerned by this Decision (439)) comprise 8 of the 10 most important routes to or from Dublin (on airport-to-airport basis). (440) The number of duopoly overlap routes has increased from 16 in 2007 to 28 in 2012, and the number of routes where the parties face compeitition only from the sale of dry seats by charter operators has increased from 6 in 2007 to 11 in 2012. In 2012 the Parties operated around 90 noncompeting routes out of Dublin, Cork and Shannon airports, while in 2007 they operated 122 non-competing routes. Other evidence collected by the Commission in the market investigation has also confirmed that the competitive relationship between Ryanair and Aer Lingus has at least persisted, if not increased, since 2007 (441).

(497) Therefore, the Commission considers that the Parties have by far the strongest market positions in Ireland (442). Therefore, the Commission concludes that Aer Lingus and Ryanair are each other's closest competitors in terms of their overall market positions in Ireland.

8.4.2. Business models: operations and brand

(498) As pointed out in Section 7.1, carriers are highly differentiated and do not often fulfil all characteristics of a particular category (full service or low frills model, or pure network carriers and point-to-point carriers).

(499) Ryanair claims that it has a unique business model and that "it would be uncontroversial to any traveller who has flown both airlines that its product offering is significantly different from that of Aer Lingus in terms of services, fares and airports" (443), and that "in pricing and product terms other low fare airlines such as easyJet, Norwegian, and Wizz Air are more similar to Ryanair than Aer Lingus is to Ryanair" (444).

8.4.2.1. Services offering

(500) The comparison of service offering is a relevant factor for assessing the closeness of competition between carriers (445).

(501) As regards the services offering, Ryanair submits, inter alia, that other low fares airlines are more similar to Ryanair than Aer Lingus. Unlike Ryanair, Aer Lingus targets business customers (through flexi fares) and provides services on a connecting basis through Dublin and under various partnership agreements (including code share agreements with various airlines and its franchise agreement with Aer Arran) (446). According to Ryanair, the Commission's admission that Ryanair now offers seat reservations and priority boarding products is only further evidence that Commission's no frills or low frills analysis is incorrect (447).

(502) First it is to be noted that easyJet, Norwegian and Wizz Air are not competitors currently flying on the routes subject to this Decision. The analysis of closeness of competition must focus on those competitors who actually compete with Ryanair and Aer Lingus on the affected markets.

(503) Moreover, there are a number of elements that differ in the services offered between Aer Lingus and Ryanair. As set out in Section 7.1.2, Ryanair describes itself as an ultra low-fares or low cost carrier and it operates point-to-point no-frills services. Although Ryanair operates mostly to secondary or regional airports, its business model does not prevent it from operating in central airports, where Aer Lingus is present (448) (on 27 of the city pairs (449) concerned by this Decision, Ryanair and Aer Lingus both fly to the same airport). According to Ryanair, it targets only pricesensitive leisure passengers, and all Ryanair passengers are asked to check-in online (or pay a charge at the airport) and pay for food and drinks and checked-in luggage.

(504) Aer Lingus operated as a traditional full service carrier until 2001, when it introduced a low-fares model to compete with low-cost carriers. Following the global economic and financial crisis in 2008, Aer Lingus initiated a review of its business model and since 2009 emphasised it's positioning as a value carrier between the low cost and full service carriers, with a strong core product and offering additional paid options. Aer Lingus offers business lounges, frequent flyer programme and flies into central airports, however at a more competitive prices than most of the legacy carriers (450). As such, Aer Lingus targets both leisure or NTS and business or TS passengers. As a result of business model changes, Aer Lingus has introduced a few differentiated and discretionary product options (flexi-fare targeting business passengers and plus-fare tickets (451)). With respect to connecting traffic, the most significant change was in expansion of the partnerships with airlines on its own longhaul business (with United Airlines and JetBlue), and the franchise agreement with Aer Arran. According to Aer Lingus, the partnerships with airlines for its' short-haul business within Europe has already been a consistent element of the Aer Lingus model before the introduced changes in 2009 (452).

(505) This view was also confirmed by the responses to the market investigation, showing there was some differentiation in terms of brand and service offering between Aer Lingus and Ryanair (453).

(506) Indeed, in 2009-2010, Aer Lingus revised its strategic direction and positioned itself as a value carrier. It considered that the pure low cost and low fares model was not sustainable while a full service model would not be viable in serving its key markets. The market positioning adopted by Aer Lingus was shown as follows:

Figure 3: Aer Lingus' comparative view of its service offering

<Graphic 3>

Source: Aer Lingus (454)

(507) In the analysis of closeness of competition, for two airlines to be considered as offering services which are close substitutes for each other, it is not necessary that the two services are identical. Although Aer Lingus' services can be regarded as midfrills when compared to Ryanair, whose services are no-frills, the Commission considers that the services included in the Ryanair fare base are closer to those included in the Aer Lingus base fare than to services of traditional full-service airlines, which are in stark contrast to Ryanair.

(508) The other scheduled carriers that operate on the affected routes are mainly traditional flag network carriers that provide full service and for whom feeder-oriented traffic and business customers are important (for example Air France, through its subsidiary City Jet, IAG - British Airways, Iberia, Lufthansa or SAS) (455). All these flag carriers also belong to worldwide alliances. Therefore they offer services that are not offered at all by Ryanair, and are not offered to any significant extent by Aer Lingus.

(509) The Commission considers that operations of network carriers on a particular route are essentially driven by the number of transfer passengers they can feed into their network. (456) Both Ryanair's and Aer Lingus' primarily "point-to-point" operating model differs from "feeder-oriented" services of network carriers. Although Aer Lingus provides incremental services on a connecting basis through its base/hub in Dublin and under various partnership agreements (457), Aer Lingus' primary operations remain on a scheduled point-to-point basis. (458) On its corporate website Aer Lingus explains that the full service model is not competitive for its operations, because Dublin hub is in a disadvantageous geographical position for short-haul connecting flows" (459). This is further evidenced by the submission of Aer Lingus, that it remains the case that "the majority of Aer Lingus traffic is point-to-point" (460). In fact on most of the routes of concern, where full service operators are active, these latter have carried a higher ratio of connecting passengers than Aer Lingus (461).

(510) There is only one route, the Dublin-Bristol/Cardiff/Exeter route, where a competitor - Flybe - with a mid-frills model is active. Although Flybe serves both business and leisure market segments, (462) Flybe has confirmed that there is an important product differentiation not only between Flybe and Ryanair, but also between Flybe and Aer Lingus (463). However, in any event, Flybe is a distant competitor of both Ryanair and Aer Lingus on the Dublin-Bristol/Cardiff/Exeter route, in particular because its market share on this route is very low (464).

(511) As regards the "frills" spectrum, Ryanair also acknowledged that "it is clear that among scheduled carriers Ryanair sits at one end of the spectrum, full service network carriers such as British Airways, Air France and Lufthansa sit at the other end of spectrum, and Aer Lingus falls in between" (465). However, Ryanair claims that although "in the context of "frills" spectrum, Aer Lingus service offering will generally be somewhere between Ryanair and a network carrier, it does not follow that Aer Lingus and Ryanair are, in every case, each other's closest competitor" and that on at least 7 routes "Aer Lingus' service offering is considerably closer to that of major network carriers than to Ryanair's" (466). According to Ryanair, the differences between Ryanair and Aer Lingus service offering are qualitatively substantial, while the differences between Aer Lingus and full service offering are minor. In particular, unlike Aer Lingus and full service carriers, Ryanair claims that these key differences include the fact that Rynair does not operate to primary hub airports, does not offer multiple check-in methods, flexible fares, business lounges, allocated seating, fast track check-in and does not carry children unaccompanied (467).

(512) However, the Commission considers that the services offering of Ryanair on the short-haul flights ex Ireland is closer to Aer Lingus' than to the service model mix offered by network carriers. As regards the service offering of Aer Lingus on shorthaul routes out of Ireland, the Commission considers that it is also very close (if not the closest) to Ryanair's.

(513) First, full service/network carriers operate into main airports (their "hubs"), offer wider connectivity options and mostly provide full-service.

(514) Secondly, both Ryanair and Aer Lingus apply true one-way pricing models (that is to say that a passenger pays the same price for a ticket regardless of whether it is purchased as part of the return journey) (468). Lufthansa, (469) IAG (British Airways (470) and Iberia (471)), and Air France (on the Dublin-Paris route) (472) do not offer true one-way pricing on the routes of concern. However SAS on the Dublin-Stockholm route and Air France/CityJet on the Dublin-London route sell true one-way tickets.

(515) Thirdly, as for Ryanair, the vast majority of Aer Lingus short-haul fares are one-way, one-class and restricted non-refundable tickets. Aer Lingus confirmed that, compared to 2007, it remains the case that the vast majority of Aer Lingus' short-haul fares are one-way restricted tickets (non-refundability and changes being subject to additional fees and fare difference). Aer Lingus allocates the same fare revenue for both Low fare and Plus Fare tickets; the only difference is accounted in bundling of ancillary products, such as free checked-in luggage and seat selection. Aer Lingus' Flexi Fare tickets offer refundability and free changes plus other ancillary elements (473); however Flexi Fares (and Plus Fares) represent only a very small portion of Aer Lingus sales (474). On the contrary, Lufthansa (475), British Airways (476), Iberia (477), Air France (478) and SAS (479) offer a larger variety of fare classes on the routes of concern (with varying degrees of flexibility, often using "fare fencing" and return fares).

(516) As in 2007, it also remains the case that both companies operate a single class economy cabin on their European short-haul flights, while this is not the case for full service operators on the routes of concern (480). Several full service providers strongly target business and TS customers, and in particular Lufthansa on Dublin-Frankfurt and Dublin-Munich routes (where TS passengers were estimated at [80-90]% in summer 2012 season) (481) and Air France/CityJet on Dublin-Paris and Dublin-London routes (where TS passengers were estimated at [60-80]% in the last three IATA seasons) (482). IAG (British Airways and Iberia) and SAS also operate business cabin and estimate that [10-20]% (483), [5-10]% (484) and, respectively, [10-20]% (485), of their passengers are TS passengers on the routes of concern. On the contrary, Ryanair argues that it does not specifically target business customers. Aer Lingus also confirms that its sales of Flexi-fare and Plus fare tickets represents a very small portion of overall sales and that the sales of these ticket in any event would "not accurately reflect the level of business / time sensitive passengers carried by Aer Lingus" (486). [...]* (487).

(517) In addition, the Commission considers that the fact that Aer Lingus offers limited access to frequent flyer programme and business lounges does not amount to a substantial difference in the overall service offering between Ryanair and Aer Lingus. Only Aer Lingus' Flexi fare passengers and the passengers who have accrued a sufficient number of points to qualify for Aer Lingus frequent flyer programme (The Gold Circle Club (488)) or the passengers who opt for "pay-as-you-go" access at a small number of airports (for example, Dublin and Heathrow), can enjoy access to business lounges. However, as set out above flexi fares still represents a very small portion of Aer Lingus sales. Furthermore, the frequent flyer program of Aer Lingus differs from programmes offered by full service carriers in that Aer Lingus rewards members for the number of flights booked, while the programmes of most full service carriers (including BA and Lufthansa) base rewards by reference to the level of fares, (489) that is, in order to reward less price sensitive passengers.

(518) Furthermore, it is important to note that Ryanair's service offering has also evolved over time to include various ancillary paid options proposing new services to passengers. According to Ryanair's half year results "Ancillary revenues increased by 20% to €583.8m, faster than the 7% increase in passenger volume, due to a combination of an improved product mix, the network roll out of reserved seating, and higher internet related revenues." (490) (emphasis added) Contrary to 2007, Ryanair now offers both priority boarding and reserved seating at an additional charge. Aer Lingus' reserved seating is free only for those passengers who can avail themselves of Flexi-fare and Plus fare tickets which account for a very small percentage of sales. (491) Both Ryannair and Aer Lingus impose ancillary charges for baggage, food and drinks (492), facts which further differentiate them from full service network carriers that do not charge for checked-in baggage (or allow at least one piece) and offer some level of complimentary food or beverage onboard. In this context it is to be noted that Ryanair's response to the SO omits to consider these further aspects of service offering that are very similar between Ryanair and Aer Lingus and differentiate them from full service carriers (493).

(519) Further elements also differentiate full service carriers from Ryanair and Aer Lingus, including greater seat pitch in economy class (494), higher ratio of crew to passengers (495), and options for children carried unaccompanied under the age of 12 years (496) offered by full service carriers.

(520) Other evidence collected in the market investigation also supports the view that Aer Lingus and Ryanair indeed are very close ( if not the closest) competitors in terms of quality and sophistication of the product on the routes of concern: for instance, according to mid-frills carrier Flybe, "Aer Lingus's short haul model utilises Airbus aircraft and has in recent years moved more towards the Ryanair model in its ambition to operate at high load factors ( 75% plus), with similar yield management policies, high density seating, and a broad range of ancillary revenue charges for the customer. Although there are certain facilities to attract the business passenger, such as a frequent flyer programme and business lounges and assigned seating, in Flybe's view the proposition to the majority of their customers is low ticket prices, similar to Ryanair, which is its biggest short haul competitor. Aer Lingus's regional operations are conducted through Aer Arann as a franchise of Aer Lingus. These are delivered by relatively old ATR turboprop aircraft on routes that cannot justify Airbus operations due to their small scale. Again the main proposition to the customer is price" (497).

(521) In addition, both Ryanair and Aer Lingus realise most of their bookings from the internet rather than via travel agents. Ryanair sells almost all its tickets through its own website with [...]* sold through call centres and airport ticket desks. In 2010- 2011 Aer Lingus sold [80-90] % of its tickets through direct sales channels (that is, online or via call center) as opposed to only [10-20] % sold indirectly on the routes to and from Ireland (498). In the Irish market Aer Lingus distributes its short haul flights exclusively through direct channels (particularly, aerlingus.com) (499). This is in sharp contrast to, for example, IAG (British Airways) that realised only [30-40] % tickets directly and over [60-70] % through indirect channels, such as travel agents (500). These results of the market investigation are in line with the view that full service or network carriers more widely distribute fares through Computer Reservation Systems ("CRSs") and have much lower level of direct or online sales compared to Ryanair and Aer Lingus (501).

(522) Ryanair further argues that on the routes, where Aer Lingus flies to primary hub airports along with another network carrier, while Ryanair flies to a secondary airport, it is simply not credible to claim that the Parties are the closest competitors. From Ryanair's viewpoint, the market shares in this context would significantly overstate the position of the merged entity.

(523) In this respect, the Commission notes that it remains the case that on the majority (namely, 27 (502) out of 46) of the city pairs concerned by this Decision, Ryanair and Aer Lingus both fly to the same airport. On a further 14 routes, Ryanair and Aer Lingus fly to different airports; however, all these routes are merger to monopoly routes, where Ryanair and Aer Lingus are by definition the closest competitors.

(524) On the other hand, for the remaining routes (in particular Dublin-Frankfurt, Dublin- Munich, Dublin-Paris, and Dublin-Stockholm), the fact that the Parties fly to different airports needs to be taken into account on an individual basis. However, as analysed in the individual route-by-route assessment, there are other important differences between the competing carriers (that operate to the same airports as Aer Lingus) and the Parties on these routes. (503) On the routes, where the Parties fly to different airports, it is clear that Ryanair's closest competitor is Aer Lingus. It is also apparent that Aer Lingus is a very close (if not the closest) competitor to Ryanair in comparison to other full service carriers. Aer Lingus and Ryanair appear as very close (if not the closest) in terms of services offering and business models. Furthermore, Ryanair and Aer Lingus are the closest competitors with respect to Irish originating passengers, due to their well-established positions in Irish airports, brand recognition and base presence in Ireland. The competitive constraints exercised by individual competitors will be discussed in detail in the individual route-by-route analyses.

(525) Regarding charter companies, the services offering is generally different from that of scheduled carriers, like Ryanair and Aer Lingus. The sales of seat only flight tickets is of secondary importance for charters companies, unlike for Ryanair or Aer Lingus.

(526) As regards the sale of dry seats, as set out in Section 7.6.1.3, some respondents to the market investigation pointed out that charter flights are of inferior quality when compared to scheduled flights not least because of the limited capacity of dry seats and the limited visibility to consumers (in addition to the frequencies and fare price) (504). "The typical travel customer would not be aware of the existence of dry seats as a) customers are not generally aware that charter airlines offer flight-only seats b) these would usually only be available via a tour operator, who do not generally specialise in last-minute travel and c) it is difficult to envision how a charter flight would effectively sell last minute dry seats due to their business model generally favouring the packaged approach and not being set-up for such sales." (505) In addition, the responses of competitors, travel agents, tour operators and corporate customer confirm the view that charter companies are not the closest competitors to either of the Parties.

(527) In any event, further comparison of closeness in the level of charter services for dry seats vis-à-vis the Parties (for example, in terms of fare types) would have no material impact on the current competitive assessment, in particular given the negligible activities of charter companies on the routes concerned subject to this Decision.

(528) Therefore, it is concluded that, even if not identical, Ryanair's services offering is the closest to that of Aer Lingus in comparison with other carriers on the routes subject to this Decision. Furthermore, it is considered that Aer Lingus' level of services is also generally very close (if not the closest) to that of Ryanair in comparison with other carriers on the routes subject to this Decision (506).

8.4.2.2. Brand recognition

(529) Ryanair claims that brand awareness is not an important competitive factor. Given that competition in the Irish market is predominantly driven by price and consumers are price-sensitive and are purchasing tickets predominantly online, Ryanair considers that band awareness is a much less relevant factor, while value for money is the key criterion.

(530) The Commission is of the view that the claims of Ryanair cannot be accepted on several grounds. First, Ryanair does not provide evidence that the role of brand awareness is insignificant. By contrast, the majority of respondents to the Commission's market investigation have considered that brand awareness still plays an important role in the competitive process on the short-haul routes to and from Ireland. (507) Furthermore, the scheduled air transport services markets are not a homogeneous product market. On the contrary, there is a high degree of product differentiation with airline business models ranging from a high degree of sophistication to a no-frills business models. It is also relevant to note that the role of brand awareness in these markets has also been acknowledged by the 2010 judgement (508).

(531) As regards brand recognition, there are certain differences between Ryanair and Aer Lingus. Ryanair was generally associated with low fares, secondary airports, basic services, a price-orientated marketing, and an extensive European network, making it a low-cost carrier with a strong identity. (509)

(532) The responses to the market investigation were less clear in relation to Aer Lingus' perceived brand positioning and operating model. Respondents generally considered it to be a solid brand (in particular in Ireland but to a lesser extent elsewhere in Europe) with a long history, still perceived by many as the former Irish national carrier serving city centre airports. (510) However, it remains the case, as in 2007, that no other competing airline enjoys the same high level of brand recognition across a large number of routes to and from Ireland, which is a very important competition factor (as confirmed by majority of respondents to the market investigation (511)) and taking into account the high level of online sales in the Irish market (512). In this context, the Commission considers that brand awareness is important in delivering high level of direct sales. In other words, customers are less likely to book flights on the websites they are less familiar with. (513)

(533) The evidence collected in the market investigation shows that other competing scheduled operators have strong brands but which are distinct from those of the Parties. The evidence suggests that IAG (British Airways and Iberia) have strong brand recognition (514), but they have only very limited activities on the routes ex Ireland (that is, the Dublin-London route for British Airways and the Dublin-Madrid route for Iberia). Air France's CityJet brand is mostly a business focused brand on the London City network (515). While Air France is also a strong brand in France (for the Dublin-Paris route), Air France considers that Ryanair and Aer Lingus both dominate the Irish market (516). Similarly, SAS considered it has a strong brand, but rather on the Swedish side of the Dublin-Stockholm route (517). Flybe has been established only for some time on the Dublin-Exeter route and therefore has much lower brand recognition in Ireland compared to the Parties (518). Lufthansa, although it has a strong brand across Europe, considered that when it comes to traffic ex-Ireland it is weak compared to Ryanair and Aer Lingus (519).

(534) Furthermore, the DAA has provided recent market research showing that both Ryanair and Aer Lingus rank very highly in terms of spontaneous brand awareness in Ireland. The DAA runs a brand tracker on a quarterly basis, involving phone-based interviews. For the fieldwork conducted from 9 to 11 June 2012, the respondents were asked to name Irish travel or transport companies. Two measures of spontaneous and total awareness were taken into account. Three airlines were mentioned with the following scores: Aer Lingus (52%/98%); Ryanair (50%/97%), Aer Arann (25%/88%) (520). These results demonstrate that Aer Arann, despite being present on the Irish market since 1970, significantly lags behind the Parties in terms of spontaneous brand recognition. This also indicates that non-Irish based carriers would have only limited abilities to replicate the strength of brand awareness that the Parties enjoy.

(535) According to respondents, at least two years or more would be required to create a strong brand or to upgrade their own brand on routes ex Ireland, notably by means of increased marketing expenditure (521).

(536) Ryanair has also argued that brand awareness is not an important competitive factor (and not a barrier to entry), as demonstrated by recent development of brand presence in Ireland by carriers, such as Etihad, Emirates or Qatar Airways. (522) These arguments will be further analysed in the Section 8.5.2.5 on barriers to entry. In this context the Commission notes that the examples given by Ryanair concern Etihad, Emirates and Qatar which are all global brands that can afford intensive marketing expenditure (523). These examples therefore tend to confirm the Commission's finding that marketing and promotion costs required to upgrade a brand would be significant.

(537) Therefore, the Commission concludes that the Parties are closest competitors in terms of brand recognition in Ireland. (524)

8.4.2.3. The Parties' average fare levels remain very close (if not the closest to each other)

(538) As regards fare levels, it is considered that the Parties are also very close (if not the closest) competitors in terms of their fare levels. Although Aer Lingus' fares are usually higher, Aer Lingus' customers pay a certain premium for additional services (525).

(539) Table 28 indicates the evolution of "short-haul average fares since 2006" as computed by Ryanair (526). It also shows the evolution of the price difference between Ryanair's short-haul average fares and Aer Lingus' average short-haul fares since 2006.

Table 28: Evolution of short-haul average fares since 2006

(In Euro)

Aer Lingus / 2006 : 91 / 2007 : 94 / 2008 : 88 / 2009 : 77 / 2010 : 86 / 2011 : 91

Ryanair / 2006 : 44 / 2007 : 44 / 2008 : 40 / 2009 : 35 / 2010 : 39 / 2011 : 45

Difference / 2006 : 47 / 2007 : 50 / 2008 : 48 / 2009 : 42 / 2010 : 47 / 2011 : 46

Difference as % of Ryanair fare / 2006 : 107% / 2007 : 114% / 2008 : 120% / 2009 : 120% / 2010 : 102% / 2011 : 102%

Source: Ryanair response to RFI of 6 August 2012, table in paragraph 5.1.

(540) If anything, Table 28 shows that in 2011 Ryanair and Aer Lingus had short-haul average fares which were very similar to those in 2007. Also, as Ryanair itself acknowledges, "the gap between Ryanair and Aer Lingus pricing has remained largely unchanged" (527). Furthermore, according to the data provided by Ryanair, the difference between average fares of Ryanair (or Aer Lingus) and other scheduled competitors is even more significant compared to a difference between the fares of Ryanair and Aer Lingus, as shown in Table 29:

Table 29: Comparison of average fares (March 2011-March 2012)

Carrier / Average fare (in euro, 2006) / Average fare (in euro, 2011/2012) / Comparison to Ryanair's fare (2011/2012)

Ryanair / 41 / 45 / -

Aer Lingus / 91 / 91 / +102%

Iberia / na / 169 / +276%

Lufthansa / 225 / 244 / +442%

British Airways / 268 / 248 / +451%

Air France / 216 / 254 / +464%

Source: Form CO, paragraph 7.69 (528)

(541) The Commission has taken into account that the reliability of Table 29 provided by Ryanair could be contested on several grounds. For example, the average fares of competitors include both short-haul and long-haul activities, while for Ryanair and Aer Lingus, Ryanair has calculated the average fares for short-haul operations only. Secondly, the figures for Aer Lingus and Ryanair concern a different portfolio of routes. While Aer Lingus primarily operates from or to Irish airports, the same is not true for Ryanair operations, with over 50 aircraft bases across Europe.

(542) It is worth noting that although Aer Lingus average fares are generally higher than those charged by Ryanair, this is not always the case. When booking a ticket, it is possible that a customer can buy a flight at a cheaper price from Aer Lingus than from Ryanair. Furthermore the difference in prices charged by Ryanair and those charged by Aer Lingus also reflects some quality advantages associated with Aer Lingus's service, such as the fact that it serves primary airports, offers business lounges and higher service orientation. For a customer who takes into account the price and time for transfer to the secondary airport when purchasing a ticket, the effective price difference may therefore be lower than the mere difference in bare flight ticket price. In view of the above, the exact fare comparison is made more complicated.

(543) In this regard, the price correlation analysis conducted by the Commission provides additional evidence that, at least with respect to the vast majority of the overlap routes, Aer Lingus and Ryanair prices have also systematically moved together over time. As the price difference reflects perceived differences in quality, a price change by one carrier would affect demand for services provided by the other on the same route. Likewise at given prices, increases in frequencies or improved flight schedules by either of the Parties can be expected to divert passengers to the other (529).

(544) No material change in market circumstances has been evidenced regarding fare levels in comparison with the situation prevailing in 2006-2007.

(545) Therefore, it is concluded that the Parties remain very close (if not the closest competitors) in terms of their average fare levels on the routes subject to this Decision. For the Dublin-Bristol/Cardiff/Exeter route, the Commission considers that it is not necessary to conclude whether Flybe is a close competitor to either Ryanair or Aer Lingus in terms of the average fare levels, in particular because its market share on this route is very low (530).

8.4.2.4. Charter companies are distant competitors to the Parties

(546) Charter companies often operate on a seasonal basis with relatively low flight frequencies set in response to the requirements of tour operators.

(547) As set out in Section 7.6, the majority of services offered by charter companies (in particular package holiday sales and seat sales to tour operators) do not belong to the same market as scheduled air transport services and the number of dry seat sales is limited (531). Moreover, sales of dry seats are very limited in volumes (see Section 8.9.1).

(548) The Parties are each other's closest competitors also considering the charter routes where each of the Parties are by far the largest competitors in terms of market shares (see Section 8.9.6).

(549) [...]* (532), [...]*.

(550) In conclusion, even if not identical, Aer Lingus' business model is much closer to that of Ryanair than to the business models of charter companies active on the routes subject to this Decision. The same is true for Ryanair's business model, which is much closer to that of Aer Lingus than to charter companies' business models.

8.4.3. Both airlines operate with a large base at Dublin but also significant bases in Cork and Shannon

8.4.3.1. General advantages of operating from a base

(551) The term "base" is used to characterise the advantages of concentrating significant operations at one single airport (533). A base airport is an airport where carriers base a certain number of aircraft and on which they concentrate their operations and from which they operate several routes. At a base airport airlines do not simply overnight aircraft but entertain additional facilities and services.

(552) At its base airport an airline can carry out various activities: overnight aircraft, entertain additional facilities and services (maintenance, customer care, groundhandling or stand-by planes). A base requires the commitment of infrastructure at an airport, personnel and equipment (with the concomitant capital expenditure) all of which are used to operate several routes from that airport. On the basis of the evidence collected through the market investigation, the market participants generally support this definition (534).

(553) Generally speaking, base operations provide significant advantages to airlines as they allow for flexible deployment of assets, ability to react to demand shifts and generate operational cost savings. In the 2007 Decision, the Commission already noted that, "concentrating aircraft and traffic at certain bases can generate considerable cost savings and increased flexibility. In comparison to providing a point-to-point service on a single route without a base at either end, base operations provide numerous competitive advantages for the routes operated from that base. There are mainly two effects which arise from base operations. First, the base allows realising cost savings due to economies of scale and scope. Secondly, there is a "competition effect", in that the airline can react more easily and more quickly to changes in supply and demand on routes out of this base. Dynamic reaction to the behaviour of competitors is therefore easier when operating from a base than outside a base" (535).

(554) Ryanair claims that the Parties have no significant competitive advantage as a result of having bases in Ireland. According to Ryanair, this is illustrated amongst other things by recent events involving Aer Lingus, whereby Aer Lingus closed its Shannon base only to reopen it, and opened Belfast and Gatwick bases, to subsequently scale them back or close them, thereby reducing the so-called "base advantages" (536).

(555) It is acknowledged that large carriers can also enjoy certain advantages thanks to their overall size, for instance when purchasing aircraft. Advantages that result from an airline's size, while not necessarily being similar to advantages resulting from base operations, can to some extent overlap with or complement the base advantages. This however does not preclude the fact that operating from a base leads to specific cost efficiencies arising from an increased ability to react to changes in demand or supply as well as from economies of scale and scope.

(556) The evidence collected during the market investigation largely confirmed this and the advantages resulting from base operations are explained in detail below.

- Increased flexibility in reacting to demand-side or supply-side shocks

(557) Large operations at base airports allow faster responses to unexpected shocks or shifts in demand-side, but also supply-side incidents, on individual routes. The dynamic nature of competition in the airline industry is evidenced by the fact that airlines constantly adjust the capacity and routes operated in their network. In this respect both Ryanair and Aer Lingus appear to shift their frequencies from one route to another and open, close or change frequencies on routes out of Ireland regularly. This is evidenced inter alia by the higher number of overlap routes in this Decision compared to the 2007 Decision. Also the weekly frequencies between seasons vary significantly for both Ryanair and Aer Lingus, who also have a significant number of seasonal routes operated ex-Ireland.

(558) Two types of demand shocks (537) may be distinguished in the air transport industry (538). First, certain shifts in demand, such as changes in customer preference or a change in the overall income, can affect all routes out of an airport. The second type of changes in demand is route specific and does not affect all routes in the same way. Routespecific demand fluctuations can arise notably due to different seasonality patterns but also due to important one-time events or festivities. In the case of route specific demand shocks, having a wider route network at a base allows the carriers to adjust capacity to match such fluctuations in demand. Carriers can more easily reallocate planes, slots and crew, between the different routes in response to changes in relative demand or profitability without incurring significant additional costs or risks. A carrier which does not have a significant presence at the airport would find it more difficult and time-consuming to react to such an increase in demand on a certain route. DAA further explains that "only a based carrier with large scale [...] has the flexibility to so alter capacity. Where capacity is fragmented, as between different non-based carriers, each carrier has much less flexibility, and faces a different decision: Fly to Dublin or not, rather than where best to fly to from Dublin with that aircraft at that time" (539).

(559) DAA further claims that "As based carriers have a higher proportion of overall slots at peak times, they have greater flexibility as regards their deployment across the range of route and services offered form that base. This enables them to act quickly in response to changing market circumstances and to counter competitive entry. For example Ryanair has for a number of years based a second aircraft in Cork for the summer months. The airline has now elected to continue to deploy that aircraft in Cork to compete directly with Wizz commencing Winter 2012" (540).

(560) Also a number of competitors noted that Ryanair, in particular, adjusted its frequency in routes (or even entered the route) that these competitors entered ex-Dublin or ex- Cork (where Ryanair had based aircraft) following their entry into the market. This illustrates the increased ability of based companies to react flexibly to changes in the competitive conditions.

(561) Some competitor airlines also considered that carriers with an established base in an airport are better positioned to reallocate aircraft to adjust capacity to match fluctuations in demand than non-base carriers (541). For example, Swiss explains that "Airlines with an established base at an airport are better positioned because of an existing infrastructure and presence at the airport which allows quicker/easier frequency reductions/increases and aircraft shifts" (542).

(562) Aer Lingus adds that "Aer Lingus' sizeable base in Dublin airport allows for changes in route selection, schedule, and aircraft integration and enables Aer Lingus to react to market changes quickly. A route that performs strongly or poorly on a particular day of the week is likely to have more options for optimising frequency as the number of based aircraft increases." Aer Lingus also regularly "switches the deployment of its A321 and A320 aircraft on Dublin routes within a 4-6 week window to match growing or shrinking demand on individual routes" (543).

(563) Furthermore, on the basis of the evidence collected during the market investigation, the Commission considers that carriers with significant city presence have an advantage when opening new routes compared to a carrier with no such activities (544). If a carrier operates several routes out of an airport, it becomes more likely that consumers become generally aware of the carrier and will be more likely to fly with that carrier. This reflects the city presence effect that reduces the marketing cost of introducing new routes. Wizz Air explains that "The higher the market share of a particular airline, generally the better known the airline will be, and the more likely that this airline will be included in the relevant set of travel options. Total exposure of the consumer population to the airline and its network will also be higher (e.g. because more people consult the airline's website to look for travel options), which will in turn lead to better awareness of new routes (e.g. because the airline website can reach more people when it advertises a new route on its own website)" (545).

- Cost savings arising from economies of scale and scope

(564) The evidence collected through the market investigation indicates that there are significant economies of scale and scope from operating a base. Ryanair and Aer Lingus also benefit from supply-side economies of scale and scope stemming from spreading marketing, sales and catering, technical or maintenance, airport costs, disruption and crew deployment costs (reduced costs for crew hotel accommodation and increased crew usage) as well as from optimal infrastructure use at their bases. These cost categories are the following:

(565) Disruption costs: a based operation can more easily re-allocate aircraft across the network when delays or incidents occur minimising the impact for consumers. For example, stand-by aircraft and crew reduces the likelihood of cancellations and gives the carrier greater schedule recovery options (546). This is important given the legislative protection afforded to passengers (as set out in Regulation (EC) N° 261/2004 of the European Parliament and of the Council of 11 February 2004 establishing common rules in compensation and assistance to passengers in the event of denied boarding and of cancellation or long delay of flights, and repealing Regulation (EEC) N°295/91) in cases of denied boarding, cancellation or delay.

(566) Sales and Marketing: the costs associated with sales and marketing are largely fixed. Recovering these costs over a larger fleet of aircraft and a wider portfolio of routes therefore also translates into lower unit costs. Examples of such marketing costs are Ryanair and Aer Lingus adverts with slogans such as "Sales from EUR 25.99 - over 100 routes reduced" which may cover many individual route options (547). Also knowledge of the market can ensure more targeted and efficient sales and marketing expenditure. A strong base presence also leads to greater awareness and may allow based carriers to bypass costs associated with distribution channels such as travel agents and attract passengers directly to their websites. This is all the more relevant as Irish air travel distribution is heavily internet based. Wizz Air also explains that "Marketing at a particular end of the route involves a 'setup cost' to get in consumers' minds" (548).

(567) Customer service and provision of own ground-handling facilities: sufficient scale also allows carriers to operate their own dedicated facilities at an airport (such as own check-in and customer information counter), thereby providing higher levels of service at a lower cost. Air Berlin for example explains that "scale allows carriers to operate their own dedicated facilities (check-in, customer information counters etc) " (549).

(568) Slots and stands: in coordinated airports airlines that operate many routes from the airport will be able to substitute slots from one route to another as they will benefit from 'grandfather' rights over their slots. Aer Lingus and Ryanair account for the "bulk of slot capacity at Dublin Airport and 90% of peak capacity" (550). As explained in Section 8.5.2, at Dublin airport, runway slot capacity is constrained in the peak morning departure period. A feature of the slot co-ordination process is that demand for departures in this period exceeds supply. "However, a based carrier with a large number of existing slots nevertheless enjoys a significant amount of flexibility in the peak period through being able to switch the slot allocations between different routes. For instance, looking at the third week in August, as between 2011 and 2012, Ryanair added 10 routes to the peak departures period and withdrew 9 routes; further adjustments were made to the frequency/timing of other departures" (551)

(569) Large scale air carriers with a base may also benefit from a stronger negotiation position for example in discussions with groundhandlers, airports and regulators and suppliers.

(570) Ground-handling: As ground-handling is typically a volume business a base with multiple aircraft allows a carrier to obtain better terms for ground-handling.

(571) Negotiations with airports and regulators: Based operators, due to the large overall traffic they bring to the airport have leverage in different negotiations. For example, the Commission also notes that Ryanair has a significant track record in negotiating favourable agreements with airports, in particular regarding the setting of airport charges.

(572) However, the stronger negotiation position may not necessarily be reflected in terms of the airport charges that are negotiated between airports and carriers (notably in large or congested airports) and therefore it may be more difficult to obtain better terms from airport managers. In terms of airport charges DAA explains that "The airport charges levied at DAA Airports are not dependant on the size of the air carrier" and that "DAA does not engage in direct negotiations with individual carriers in relation to airport charges" (552).

(573) However, based carriers typically obtain preferential facilities and represent their interests in various user forums in airports ultimately leading to an increase in product quality at little or no additional cost (553). This is also reflected in DAA's response "However to encourage based operations, DAA offers a standby aircraft scheme for based carriers at Cork and Dublin airports. Under this scheme the aircraft operator is not charged aircraft parking charges in relation to a Standby Aircraft for periods on which it is parked on a Standby Aircraft Stand. A standby aircraft is one which is on standby to be used exclusively by a Relevant Operator for the purpose of temporarily replacing one or more aircraft (operated by the same Relevant Operator to or from Dublin or Cork Airports) which is prevented from flying into or out of Dublin or Cork Airports by reason of technical, mechanical and/or operational problems" (554).

(574) Moreover, based carriers would be in a more favourable position to represent their interests directly in various user forums, including the Dublin Airport Coordination Committee ("DACC"), while carriers that operate from bases outside Ireland may have no company personnel on the ground and would have to rely on their local ground-handler or agent to represent their views at such meetings, if indeed they engage in them at all. For example, DAA explained that "in the consultation exercise carried out with users in respect of changes to stand allocation rules in Summer 2011, ALL based carriers submitted views on the proposed rules, in contrast DAA received no direct replies from non based scheduled operators (despite issuing a request for comment to all AOC members). In relation to the slot co-ordination committee Dublin-based carriers have consistently attended meetings since 2007. Lufthansa and Turkish Airlines are the only non based carriers that have attended periodically. These examples demonstrate that based carriers have the necessary resources and vested interest to ensure that their airline's views are aired at consultation forums" (555).

(575) DAA provides an example where "in the recent consultation regarding stand allocation rules, Aer Lingus as a based operator was supportive of proposals to enable clustering of operations on particular stands. This minimises walking distances for terminal based groundhandling personnel between check-in, gates and ramp accommodation and enables more efficient operation of ground services equipment" (556).

(576) Therefore, even if based carriers do not necessarily obtain better monetary terms these examples indicate that they are in a position to extract better facilities from the airport to encourage their activities.

(577) Procurement: Based carriers also have greater power in negotiations with suppliers such as fuel, maintenance and catering companies and are thereby able to extract better terms. For example, Swiss notes that "Due to higher contract volumes (economies of scale), base carriers are likely to enjoy discounted rates" (557). DAA also notes that "In the event that a new route is not successful a based aircraft can simply be redeployed to other routes without incurring penalties that may be imposed on single operators that face difficulties with sustainability on individual routes (no ground-handling contract break fees, no surplus office or other property lease time etc)" (558).

Cost advantages arising from economies of scope

(578) By having a number of aircraft in a given airport, carriers benefit from economies of scope derived from the ability to easily switch assets between routes, increasing the competitiveness of the routes operated from a base.

(579) A base can also allow better aircraft utilisation by reducing the overall fleet downtime by reducing the fleet downtime as any temporary delays can be absorbed by using the ground time for other aircraft in the base (for example standby aircraft).

(580) In particular, the higher the number of aircraft the easier it will be for a carrier to maximize the aircraft utilisation because there are more combinations of flights to fill the 10, 12 or even 16 hour schedule. As operating rules set an hour limitation on crew utilisation, a based carrier can maximize the working hours of its crew and aircraft and extend the working day of its aircraft beyond the 10 hour limitation on crew by exchanging staff during the day at their base.

(581) A larger base also makes it possible to pool reserve requirements for pilots and cabin crew. Airlines operating a number of aircraft at a base can spread reserve crew costs over a wide number of operations as their crews are shared across more aircraft. For example Air Berlin explains that a base makes it possible to pool reserves of pilots and cabin crew thus increasing operational flexibility (559).

(582) Additionally, airlines operating from a base can use the early morning slots at their base airport which are economically important since the demand for flights is particularly high at that time, in particular for routes where a significant majority of passengers originate at this end of the route. For example, Vueling notes that "A base implies having more operation, so more options for interchanging slots are possible and better scheduling of each of the routes" as well as "Morning and night flights are needed to serve the demand so these flights (that otherwise you would not operate) increase your utilisation" (560). Furthermore, operating a base allows operating early morning flights and more generally providing a better schedule.

Conclusion

(583) Competitors have generally confirmed that the operation of an aircraft base confers all the categories of cost advantages and ability to have a better schedule described above. The category that most competitors consider as providing significant advantages is the ability to have a better schedule due to the flexibility of allocating aircraft to early morning slots and the possibility to adapt schedules to more profitable routes (561).

(584) The Commission has attempted to quantify the size of these advantages. However, this has proven a very complex exercise for all competitors (562). However, the different cost savings categories described in this Section 8.4.3.1 have been recognised by most competitors as providing at least some advantages.

(585) Therefore, the Commission concludes that base operations provide significant cost advantages and the ability to react flexibly to market conditions.

8.4.3.2. Ryanair and Aer Lingus operate significant bases in Ireland

(586) Both Ryanair and Aer Lingus operate significant bases in Ireland. In particular, Aer Lingus and Ryanair currently both have by far the largest bases at Dublin airport with respectively 32 (mostly Airbus 320 and A321, 174-212 seats) and 18 (Boeing 737-800, 189 seats) based aircraft in the summer season 2012, which is significantly more than any other competitor (see Table 30 below). Furthermore, Ryanair submits that in addition to Irish based aircraft, 12 routes identified in the decision opening the proceedings (on an airport-to-airport basis) are served by non-Dublin/Cork based aircraft, while 13 routes were served by a mix of Dublin-based and non-Dublin/Cork based aircraft (563). Therefore a figure of 18 Dublin-based aircraft significantly underestimates Ryanair's position in Dublin airport, in particular since no other airline competing on the routes concerned by this Decision has such a wide network of bases across Europe (564).

(587) The other Dublin-based carriers have bases of very limited size. Specifically, Aer Arann has a base at Dublin airport with 8 low capacity aircraft (all turboprops, 42 or 68 seats). As far as carriers independent from the Parties are concerned, Air France (CityJet) has 3 regional aircraft (Avro RJ 85 with 95 seats) based at Dublin airport. In addition, IAG/British Airways, TUI/Thomson, Lufthansa and Europe Airpost have one aircraft based at Dublin airport or keep it overnight.

Table 30: Carriers with bases or overnighting aircraft at Dublin airport

Airlines / No. of aircraft 2012 / Passenger estimate 2012 ('000) / No. of routes 2012

Aer Lingus / 32 (565) / 7,780 / 73

Ryanair / 18 (566) / 7,150 / 70

Aer Arann / 8 / 596 / 11

Air France / 3 (567) / 470 / 3

IAG (British Airways) / 1 / 310 / 1

Lufthansa / 1 / 290 / 3

TUI/Thomson / 1 (568) / 80 / 12

Europe AirPost (569) / 1 / unknown / unknown

Source: Form CO, annex 7.3(f), modified according to responses provided by competitors.

(588) The 2010 General Court decision acknowledged that "in relation to the impact of scheduled airlines or charter airlines having an aircraft parked overnight at Dublin for a season or throughout the year (see paragraphs 256 and 257 above), the Commission rightly noted in the contested decision that the advantages which those companies enjoy were not comparable to those inherent in operating a base, in particular as regards the flexibility of switching between routes, redeployment of aircraft, minimising disruption costs, exchange of crews, customer care and brand awareness (recital 560). It is those advantages of operating a base which are important in this case and not the possibility of offering flights to any particular destination by operating an aircraft. " (570)

(589) It is also noted that Aer Lingus has a base at Cork (4 aircraft) and Shannon (2 aircraft). Ryanair also has a base at Cork and Shannon, with 1 aircraft at each airport. (571) In addition Aer Arann has a base at Cork with two small capacity aircraft and a base at Shannon with one small capacity aircraft. No other airlines operate bases at Cork, Shannon, or Knock airports.

(590) The fact that Aer Arann also has aircraft based at Dublin, Cork and Shannon airports, does not materially impact the Commission's assessment, in particular given that as described in Section 8.2, it is considered that Aer Arann is a competitor of Ryanair, but not of Aer Lingus, and that the market shares of Aer Arann (operating under the Aer Lingus Regional brand) have to be attributed to Aer Lingus for the purposes of the competitive assessment.

(591) It is therefore concluded that Ryanair and Aer Lingus are the only carriers that operate significant bases in Ireland.

8.4.3.3. Base advantages increase with the size of the base and therefore the limited presence of the current based competitors is unlikely to constrain the Parties significantly

(592) On the basis of the evidence collected during the market investigation it is concluded that the cost savings related to having a base increase with the size of the base - at least until a certain threshold. In particular, several competitors do not consider that a one aircraft base generates significant cost savings, either because this would not lead to critical mass or because they have no knowledge of single-aircraft bases (572) that exist. However the clear majority of competitors consider that benefits are significant for 5, 10 and 15 aircraft bases compared to having no base (573). Moreover, the negative answers relate to competitors that do not have the experience of 5 and 10 aircraft bases.

(593) Also the vast majority of competitors consider that the economies of scale increase with the number of aircraft in a given base (574). For example, TUI Travel considers that "The synergies from scale are greater in operations of more than 6 aircraft. There are a number of steps that will drive synergies as aircraft numbers increase, but these thresholds will differ for each category of costs. Generally speaking, the larger the operation, the greater the cost and operational synergies that can be leveraged" (575).

(594) On the basis of the evidence in the market investigation, it is considered that the advantages of having a base follow decreasing returns once a certain number of aircraft has been attained. For example, Air France/CityJet considers that "the benefits from having a base do not increase much further when an air carrier has more than 30 aircraft in a given base" (576).

(595) The only competitor to Ryanair and Aer Lingus which may be considered as having bases of some scale is a subsidiary of Air France, CityJet, which operates with 3 lower capacity aircraft (a fleet of RJ 85 with 95 seats) from Dublin. However, in particular due to lower seat capacity, CityJet has higher costs per passenger than Ryanair and Aer Lingus. Moreover, the limited size of CityJet's operations at Dublin airport would not allow the carrier to redeploy capacity from one route to another in reaction to a shift in demand as is the case of Aer Lingus and Ryanair.

(596) Besides, the carriers that have single-aircraft bases or simply overnight aircraft at Dublin, Cork, and Shannon airports neither benefit from the same scale and scope of advantages of a base nor from the flexibility of switching between the routes and the opening of new routes. The 2010 Judgement acknowledged that "in relation to the impact of scheduled airlines or charter airlines having an aircraft parked overnight at Dublin for a season or throughout the year (see paragraphs 256 and 257 above), the Commission rightly noted in the contested decision that the advantages which those companies enjoy were not comparable to those inherent in operating a base, in particular as regards the flexibility of switching between routes, redeployment of aircraft, minimising disruption costs, exchange of crews, customer care and brand awareness (recital 560). It is those advantages of operating a base which are important in this case and not the possibility of offering flights to any particular destination by operating an aircraft. " (577)

(597) Therefore it is concluded that competitors of Ryanair and Aer Lingus with a base, or overnighting aircraft, at Dublin, Cork and Shannon airports are not equally close competitors.

8.4.3.4. Competitors with a base at the destination airports are not equally close competitors

(598) In 2007, it was also considered that carriers that have a base at the other end of a route did not exert the same competitive constraint on Ryanair compared to the constraint from Aer Lingus.

(599) Furthermore, the 2010 Judgement explained that: "It is apparent from the analysis set out in the contested decision that it is rare for carriers which operate on routes not to use a base airport; it grants them an economic advantage. The fact that both Ryanair and Aer Lingus have an important base at Dublin airport thus had to be taken into consideration in so far as that enables these two companies, among others, to benefit from similar advantages (sections 7.3.4.1 and 7.3.4.2). For the reasons set out in the contested decision, this situation is not comparable to that of competitors which base their aircraft overnight at Dublin airport, in particular given the significant differences in terms of economies of scale and scope which a base airport provides (section 7.3.4.3), nor to that of competitors which have a base at the destination airport, because of the particularities of Dublin airport (section 7.3.4.4)." (578) (emphasis added)

(600) Generally, air carriers with bases at the destination airport can only constrain the merged entity on the specific routes from these bases, while Aer Lingus and Ryanair exert dynamic competition on each other (that is to say on a variety of routes) as the same based carriers have increased flexibility in reacting to demand-side (or supplyside) shocks.

(601) In the 2007 Decision the issue was raised that both Ryanair and Aer Lingus have an advantage in gaining Irish originating customers who represent a significant share on certain routes (579). According to the information submitted by Ryanair on a significant number of the routes the proportion of passengers originating in Ireland is over 50% (580). There is in particular a significant imbalance between Ireland and non- Ireland originating passengers on routes to the Mediterranean (581).

(602) In this respect, Ryanair claims that on certain routes, the majority of Ryanair's passengers originate at the non-Irish end of a route [...]*. Accordingly, airlines with bases at the non-Irish end of the route would arguably enjoy advantages (such as lower marketing costs) associated with Ryanair and Aer Lingus having bases in Dublin, Cork, and Shannon. However, the evidence collected in the investigation does not fully support this view. Even on those routes, where the Ryanair's passengers originating from Ireland range between 30-50%, the evidence collected in the market investigation shows that both Ryanair and Aer Lingus carry a higher ratio of Irish originating passengers compared to incumbent scheduled airlines (582). Therefore both Ryanair and Aer Lingus have an advantage over their competitors in gaining Irish originating customers who represent a significant share on routes out of Ireland.

(603) The evidence collected in the market investigation also suggests that most competitors without bases in Ireland would suffer from certain disadvantages in terms of brand recognition (583). Nearly all carriers consider that they do not enjoy the same kind and level of brand recognition at the Irish end of the route as Ryanair and Aer Lingus (584).

(604) Besides, unlike the Parties, carriers with bases on the other end of a route are likely to only operate one or very few routes from Dublin, Cork, or Shannon. The destination-based carriers are therefore unlikely to show the same degree of commitment to routes to and from Dublin, Cork, and Shannon or the same ability and incentives to adjust capacity to react to shifts in demand as Ryanair and Aer Lingus. For such carriers, Dublin, Cork, and Shannon represent just one route from their destination base or hub. It would have many other alternative routes to consider opening or expanding operations, without facing aggressive competition from the merged entity. Furthermore, unlike Ryanair and Aer Lingus (primarily) (585), most competitors active on the routes of concern are full service/network carriers, focused on connecting traffic to their "hub" destinations outside Ireland.

(605) Therefore it is considered that competitors with a base at the destination airports are not equally close competitors.

8.4.4. Price regression analysis: Ryanair and Aer Lingus are close and closest competitors

8.4.4.1. Approach followed

(606) The Commission has used regression analysis technique, which is a statistical tool for understanding the relationship between two and more variables, to test the competitive interaction between the Parties (586). Using this regression analysis the Commission has investigated whether the presence of one of the Parties has an impact on the fares of the other, to quantify the magnitude of such an effect and to help predict the average increase in fares post-Transaction (if any). In more detail, this quantitative work has tested the following:

- whether the presence of one of the Parties on the route is associated with a statistically and economically significant reduction in the fares of the other;

- whether the Parties exert on each other a stronger competitive constraint than any other existing competitor;

- whether the existence of an actual or potential competitor with a significant presence at the destination airport on a route originating in Dublin has an impact on the Parties' prices;

- whether a stronger presence of one of the Parties (in terms of number of frequencies) has a more pronounced effect on the other's fares.

(607) The Commission in the 2007 Decision used different regression techniques (crosssection and fixed-effects analysis) but concluded that the fixed-effects analysis provides the most appropriate approach in the case at hand.

(608) The principal advantage of the fixed-effects analysis is that it takes into account important but unobserved (or unmeasured) influences on price that do not vary over time on any single route. For this reason, the Commission has followed the same fixed-effects methodology in this Decision.

(609) Ryanair's comments on the regression analysis (587) have been assessed in detail by the Commission in Annex III to this Decision.

8.4.4.2. Fixed-effects regressions

(610) The fixed-effect procedure compares the level of Aer Lingus prices on a route after Ryanair entered, with the level before Ryanair entered. This before-and-after comparison is done systematically over time and for all routes where Aer Lingus operates. In this way the regression generates the average effect of Ryanair's presence on Aer Lingus fares.

(611) A regression analysis with route specific fixed-effects accounts for specificities on given routes (588). When an analysis does not control for such specificities the results of the analysis may be biased due to the so-called omitted variable bias (that affects alternative regression analysis techniques, namely cross-sectional regressions). By controlling for route specificities the omitted variable bias risk is mitigated because unobservable cost or demand factors, whose variation across routes would be likely to affect fares, are more likely not to vary over time on any single route. The Commission regards this methodology as the most suitable for assessing the competitive constraint exerted by Ryanair on Aer Lingus.

8.4.4.3. Aer Lingus' fixed-effects regression analysis

(612) A fixed-effects regression analysis, following largely the methodology of the 2007 Decision, has been provided by Aer Lingus based on their own data (589).

(613) Aer Lingus used average net fare data and load factors of Aer Lingus from January 2007 to June 2012 on all short- and mid-haul routes out of Dublin and in addition extracted schedule data from Diio Mi (590) on the total number of departures and seats per carrier on all routes for the same period. Aer Lingus employed a fixed-effect methodology and estimated the effect of Ryanair's entry or exit on the net fares and load factors of Aer Lingus. It included route and month fixed-effects and a capacity variable in certain specifications. The main variable of interest has been the presence of Ryanair.

(614) The analysis showed that Aer Lingus' fares are, on average, lower as a result of Ryanair's entrance on the route and as a result of Ryanair's entrance on the city-pair. Aer Lingus claims that these findings remain true regardless of whether they include the capacity variable in the regression or not. According to Aer Lingus, the analysis also shows that Aer Lingus' load factor is on average lower on routes that are also served by Ryanair for the airport-pair specification and also lower for the city-pair specification. These findings would be statistically significant and remain true irrespective of whether the capacity variable is included in the regression or not.

(615) While Aer Lingus' analysis closely follows the methodology employed by the Commission, it does not, however, take into account several other variables of interest (for example the presence of competitors). These further variables, as discussed in the Commission's fixed-effects regression analysis, are used partly to confirm the robustness of the results but also to test whether other competitors constrain the Parties.

8.4.4.4. Commission's fixed-effects regression analysis

(616) The Commission has undertaken a fixed-effects regression analysis which has largely confirmed that, as in the 2007 Decision, there is significant competitive interaction between the Parties. The analyses are performed both at airport pairs (that is to say, in instances where both carriers fly to exactly the same airport) and at city pairs (that is to say, in instances where both carriers fly to substitutable airports; city pairs are defined on the basis of the market definition as set out in this Decision (591)).

(617) The Commission has requested data from Aer Lingus and Ryanair on fares and fuel cost. The Commission has also requested DAA data for frequencies, passengers, and capacity. The data was requested on a monthly basis for the period from November 2004 to July 2012.

(618) Using merged data sets, the Commission has used a standard fixed-effects procedure to assess the competitive constraint exerted by the Parties on each other. The Commission's initial specification tests whether the presence of one of the Parties in a given route is negatively related with the fares charged by the other Party (this is labelled as the presence specification). An alternative specification tests whether the number of frequencies of one Party in a given route is associated with the other Party charging lower fares (this is labelled as the frequency specification).

(619) The presence specification includes controls for the presence of Ryanair, one or more flag carriers and one or more non-flag carriers. It also controls for the presence of Aer Arann and CityJet (through the use of dummy variables), which according to Ryanair's claim in the 2007 Decision also exercise a strong constraint on the Parties. The Commission extends the baseline specification to include various control variables for demand and costs as well as variables to indicate whether flag or nonflag carriers have a strong presence at the destination airport (as a proxy for a base at destination airport). In all cases, the presence of Ryanair is statistically significant and is correlated to Aer Lingus charging lower prices. Similarly, Aer Lingus also constrains Ryanair's fares, albeit to a smaller extent and in particular less on city pairs.

(620) The fixed-effects regressions with Aer Lingus's fares as the dependent variable show that Ryanair exerts a significant competitive constraint on Aer Lingus's fares. In particular the Commission's analysis show the following results:

- First, depending on the specification, Ryanair's presence is associated with Aer Lingus charging around 9-14 % lower prices when considering airport-pairs and around 7 - 11% lower prices when considering citypairs. This effect is economically and statistically significant in all tested regressions. It is also highly robust to the use of alternative specifications including alternative demand and supply controls. Also, in most cases the control variables in the different regressions have the expected signs and are statistically significant. The explanatory power of the regression (that is to say, the fit of the model) is also high (R2, a measure of the fit of the model, is consistently around 80%). Ryanair also appears to impose a more significant constraint on Aer Lingus when it serves the same airport.

- Secondly, comparing the coefficients of Ryanair with that of flag-carriers and non-flag carriers, as well as Aer Arann and CityJet, Ryanair's presence or number of frequencies has a much stronger economic impact than that of any other type of carrier. This also holds true for charter carriers. In fact, in most cases the regressions indicate that the presence of other carriers has no economic or statistically significant effect on Aer Lingus' fares. The presence of flag carriers has a negative effect on the average fares of Aer Lingus but at the same time this effect seems to be significant only if Ryanair is also present on the route and is typically significantly smaller than the effect of Ryanair. Therefore, Ryanair appears to be Aer Lingus' closest competitor.

- Thirdly, neither destination-based flag carriers nor destination-based nonflag carriers exert a significant constraint (that is to say, a negative effect). Moreover a flag carrier base at the destination airport has a significant positive effect on Aer Lingus prices, limiting the effect of flag carriers. Thus it cannot be expected that the merged entity would be effectively constrained by flag or other non-flag carriers post- Transaction.

- Fourthly, measuring the strength of Ryanair's presence using the number of frequencies on the route as a proxy provides further confirmation that Ryanair constrains Aer Lingus. Depending on the specification, the price effect of a 1 % increase in frequencies of Ryanair on Aer Lingus prices is negative at around 5-6 %. This adds to the robustness of the results derived from the presence specifications.

(621) The fixed-effects regressions with Ryanair's fares as the dependent variable show that Aer Lingus exerts a significant competitive constraint on Ryanair's fares. In particular the Commission's analysis shows the following results:

- First, the presence of Aer Lingus is associated with Ryanair charging around 3-10 % lower prices when considering airport-pairs and around 3- 14 % lower prices when considering city-pairs. This effect is economically and statistically significant in all tested regressions. It is also highly robust to the use of alternative specifications including alternative demand and supply controls. The explanatory power of the regression is also high with R2 consistently above 80% (592).

- Secondly, comparing the coefficients of Aer Lingus with that of flagcarriers and non-flag carriers, as well as Aer Arann and CityJet, Aer Lingus's presence or number of frequencies has stronger economic impact than that of any other type of carrier. In fact, in most cases the regressions indicate that the presence of flag-carriers, as well as Aer Arann and CityJet, has no economic or statistically significant effect on Ryanair fares. The presence of non-flag carriers has a significant negative effect on Ryanair's prices but this effect is not robust in the different specifications. Therefore, Aer Lingus appears to be the closest competitor to Ryanair on routes from Dublin, Cork and Shannon.

- Thirdly, neither destination-based flag nor destination-based non-flag carriers exert a constraint on Ryanair (that is to say, a negative effect). Moreover a non-flag carrier base at the destination airport has a significant positive effect on Aer Lingus prices, limiting the effect of non-flag carriers. Thus it cannot be expected that the merged entity would be effectively constrained by flag or other non-flag carriers after the merger.

- Fourthly, measuring the strength of the presence of Aer Lingus using the number of frequencies on the route as a proxy provides further confirmation that Aer Lingus constrains Ryanair. Following the Commission's frequency regressions, a 1 % increase in frequencies of Aer Lingus leads to around a 2 % price decrease in Ryanair prices. This adds to the robustness of the results derived from the presence specifications. Also, the frequencies of non-flag carriers do not appear to constrain Ryanair's pricing.

(622) The Commission also notes that the estimated effects of Ryanair on Aer Lingus prices (and of Aer Lingus on Ryanair's prices) are likely to be underestimated as the presence of Ryanair and Aer Lingus in Dublin exerts a potential competitive constraint on Aer Lingus and Ryanair. For example, on routes out of Dublin where Ryanair is the only carrier, it can be expected that it sets prices which are lower than what it would charge if Aer Lingus had no base in Dublin (as there is the threat that Aer Lingus can enter such a route). Since the regression analysis only considers fares' overtime variations within each route and only captures price reductions subsequent to Aer Lingus' entry, this potential competition constraint does not show up in the empirical results (see also Section 8.10).

(623) Ryanair has raised some criticisms on the regression analysis put forward by the Commission and Aer Lingus has performed some robustness checks. (593) The Commission discusses Ryanair's criticisms and Aer Lingus' comments in detail in Annex III to this Decision. However, the Commission considers that these comments do not alter the Commission's findings.

8.4.4.5. Conclusion

(624) The Commission's price regression analysis confirms significant competitive interaction between Ryanair and Aer Lingus. It confirms and complements the conclusion derived from qualitative evidence that Ryanair and Aer Lingus are close and the closest competitors.

8.4.5. Conclusion on closeness of competition

(625) The activities of Ryanair and Aer Lingus overlap on a number of routes and they compete very closely.

(626) On the majority of the overlap routes, Ryanair and Aer Lingus are the only operating carriers. On these routes, Ryanair and Aer Lingus are, by definition, each other's closest competitor. Even on those routes on which Ryanair and Aer Lingus face one or more competitors, the Parties are most often by far the largest operators in terms of market shares.

(627) The assertion that Aer Lingus is the closest competitor of Ryanair and that Ryanair is very close (if not the closest) competitor of Aer Lingus is further confirmed by the fact that they have similar business models, which differ from those of most of their competitors. They also enjoy very high brand recognition in Ireland, which is a distinctive feature in contrast to all their competitors. Ryanair and Aer Lingus are particularly well-established in Irish airports and, unlike other operators, they both operate significant bases in Ireland, in particular at Dublin airport. Having important bases at the same airport corroborates the finding that Ryanair and Aer Lingus are very close (if not the closest) competitors.

(628) Therefore, the Commission concludes that Aer Lingus and Ryanair are very close competitors, if not each others closest competitor, on all routes subject to this Decision. The Transaction is therefore likely to eliminate the important competitive constraint that both carriers exert upon each other pre-Transaction on the routes. Customers' choices of travelling options would be substantially reduced and it is unlikely that other operators could sufficiently constrain the merged entity, especially in terms of fare levels.

(629) Further route specific assessment is made in Sections 8.9.4-8.9.6.

8.5. Entry is unlikely to eliminate the anti-competitive effects of the Transaction

8.5.1. Introduction

(630) In the 2007 Decision, it was extensively examined whether the entry of new competitors onto the market or the expansion of existing competitors was likely to eliminate the anti-competitive effects of the concentration (594). While it was found that regulatory barriers did not play as important a role as barriers to entry (595), the Commission found that barriers to entry did exist on account of Ryanair's and Aer Lingus' 'strong position' with large bases in Ireland (596). It also found that entry costs and risks would be significant in a market already served by two strong airlines with well-established brands (597), that the risk of 'aggressive retaliation' by the merged entity was high (598), that competitors considered other markets more attractive than the small Irish market (599), that airport congestion constituted an additional barrier to entry, both at Dublin airport and certain destination airports (600) and that the strong position of the merged entity at Dublin airport might hinder further expansion by competitors (601).

(631) It was thus concluded that "there are a number of important entry barriers to operating flights from or to Dublin in competition with Ryanair/Aer Lingus. These barriers go well beyond the problem of the partly congested airport and are in particular linked to Ryanair's and Aer Lingus' well established position at their home base. The investigation showed that, as a result of these barriers, entry is not likely, but even unlikely on almost all overlap routes. In the absence of potential entrants on the vast majority of overlap routes and given that competitors have unanimously indicated that they would not even consider entering in direct competition with the merged entity on a significant scale (in particular by opening a base at Dublin), the Commission concludes that entry is not likely, timely and sufficient to constitute a sufficient competitive constraint on the merged entity and defeat the likely anti-competitive effects of the proposed merger." (602)

(632) These findings were upheld by the General Court in the 2010 Judgement. With respect to the entry analysis, the General Court confirmed that "What counts is the prospect of an entrant which offsets the anti-competitive effects specifically established in the contested decision at that stage of the assessment" (603). The General Court considered in this respect that the mere 'threat" of an entry or the explanation that there are no entrants because of Ryanair's efficiency on the relevant routes and because of customer satisfaction are not sufficient (604). The General Court further found that as regards the analysis of barriers to entry, what is relevant is not the current situation (a situation in which Aer Lingus is present as a competitor of Ryanair or represents the most likely potential competitor) "but the situation which would result from the concentration on the routes dominated by Ryanair-Aer Lingus combined" (605). In this regard, the General Court found that Ryanair's argument that the lack of entry would show that its fares and capacity are so competitive that no competitor considers that it would make sense to compete, is not relevant for the analysis. It also rejected Ryanair's argument that it does not envisage charging prices higher than required to remain competitive following the implementation of the concentration since the control of concentrations focuses on the control of market structures and not on the control of companies' conduct thereby also noting that the price criterion is not the only one which may be taken into consideration (606).

(633) Ryanair argues that the overlap routes are highly contestable. (607) Carriers can and do move aircraft between different routes quickly to maximize profits and take advantage of growth opportunities. According to Ryanair there are virtually no barriers to entry and expansion for an existing carrier on almost any Irish route, since services can be commenced on short notice with little or no capital expenditure, given that slots are available.

(634) In Ryanair's view, a carrier with an aircraft base at Dublin airport could immediately enter and offer services on any or all of the overlap routes. According to Ryanair, Air France, British Airways and Lufthansa already have established bases at Dublin airport and are well placed to compete.

(635) Ryanair also submits that the Parties have no significant competitive advantage from having a base in Ireland and that the scale advantages are the same at both ends of a given route. (608) Ryanair also submits that the threat of competition from existing bases at destination cities on overlap routes and/or the presence of strong competitors which could easily establish such bases will effectively constrain the merged entity. (609)

(636) Generally, if entering a market is sufficiently easy, a merger is unlikely to pose any significant anti-competitive risk. (610) Therefore, entry analysis constitutes an important element of the overall competitive assessment, in particular in airline cases.

(637) In line with the provisions of the Horizontal Merger Guidelines, the Commission examines whether entry is likely to eliminate the anti-competitive effects of the Transaction. For entry to be considered a sufficient competitive constraint on the merging parties, it must be shown to be likely, timely and sufficient to deter or defeat any potential anti-competitive effects of the Transaction. (611)

(638) It is examined whether entry is likely and whether potential entry is likely to constrain the behaviour of the merged entity post-Transaction. It is recalled that, for entry to be likely, it must be sufficiently profitable taking into account the price effects of injecting additional output into the market and the potential responses of the incumbents. (612)

(639) It is also examined whether there are barriers to entry that potential entrants would encounter. Barriers to entry are specific features of the market, which give incumbent firms advantages over potential competitors. (613) Barriers to entry can take various forms such as legal advantages, technical advantages or may result from the established position of the incumbent firms on the market. (614)

(640) Moreover, it is examined whether entry would be timely, that is to say sufficiently swift and sustained to deter or defeat the exercise of market power (615) and whether it would be of sufficient scope and magnitude to deter or defeat the anti-competitive effects of the merger. (616)

(641) As will be set out below, the Commission's analysis and the market investigation show that a number of important barriers to entry and expansion exist on the overlap routes.

(642) Furthermore, the evidence collected in the market investigation indicates that no other competitor would have post-Transaction entry or expansion projects which could be considered as timely, likely and sufficient enough to constitute a competitive constraint on the merged entity and defeat the anticompetitive effects of the Transaction.

8.5.2. Barriers to entry

8.5.2.1. Introduction

(643) During the market investigation, competitors were asked to provide their views on the existence of barriers to entry in relation to the overlap routes. (617) The evidence collected during the market investigation points to the existence of various barriers to entry. The most commonly-identified barriers to entry are the position of competitors, the frequency advantage of competitors, brand recognition, availability of slots, airport charges, and availability of aircraft. The fear of retaliation, network effects, need for a base, and to a lesser extent constraints on terminal capacity are also identified as barriers to entry. It is generally agreed that there are no regulatory barriers.

(644) However, some other competitors consider that there are no barriers to entry. Lufthansa, which is currently operating on the Dublin - Frankfurt and Dublin - Munich routes considers that there are currently (pre-Transaction) no barriers to entry whatsoever at Dublin airport (nor at Shannon, Cork and Knock) (618). It nevertheless indicates that there could possibly be barriers should the Transaction take place "since it cannot be excluded that the dominance of a combined FR/EI could lead to a potential barrier for some carriers" (619). IAG indicates that there are no barriers at Dublin "other than the general economic climate and the weak demand" (620). Flybe, while identifying a number of barriers to entry that it considers as 'significant' (621), nevertheless indicates that the existence of these barriers would not, as such, prevent it from entering on the routes (622) post-Transaction.

(645) The competitors were also asked whether the entry barriers at Dublin airport have evolved over the last five years (623). A number of respondents consider that the entry barriers had either not changed since 2007 or had worsened. For instance, Aer Arann indicates that "Yes Dublin airport has become a more expensive place to operate as an airline. As a result the bar required for new routes to make money has become higher over the last 5 years" (624). Jet2.com indicates that "Increase in airport charges makes this airport less attractive" (625). Aer Lingus considers that "None of the entry barriers identified in the EC's Prohibition Decision in 2007 have become less relevant in 2012. Indeed, with regard to some of them an entrant would be in a worse position now then in 2007" (626). LOT indicates that "Yes, the strength of Ryanair has grown over the last 5 years" (627). Air France on the other hand indicates that "(...) stand capacity, which was a huge issue in 2007 and therefore a significant obstacle (...) is no longer relevant (...)" (628) and easyJet indicates that "it has become easier to acquire slots in peak time at Dublin, but we don't think this has had a significant effect on barriers to entry, as the entry costs would remain very high" (629).

(646) The competitors overall consider that if the Transaction takes place, these barriers to entry would be the same, or would increase. (630)

(647) The views expressed by competitors during the market investigation show that a majority of the respondents having expressed an opinion considers that if the Transaction took place, the combined market position of Ryanair and Aer Lingus at Irish airports may act as a barrier to entry. For instance, Aer Arann indicates that "The relative size of the combined entity is too large for most commercial airlines to take on in Ireland and the United Kingdom." (631) Aer Lingus refers to the merged entity becoming by far the largest purchaser of services from third parties at Dublin airport. In Aer Lingus' view, "This may result in the merged entity obtaining preferential treatment over potential entrants or provide it with the opportunity to influence any plans on expansion of facilities in a way which may frustrate potential new entrants at Dublin airport." (632) Lufthansa (633) and Swiss (634) indicate that jointly Ryanair and Aer Lingus "will have a greater competitive clout and can optimize their offer on specific routes and in Ireland overall. They will be the biggest customer (at airports) in Ireland and be able to exert increased purchasing power (esp. A/A handling, fee, flexibility in slot use etc.)". They further indicate that the merged entity "will be the largest airline in Ireland and therefore will have a strong sales power and this might be a barrier for other entrants in the Irish market." (635) Germanwings indicates that "The incumbents combined market share, pricing power and slot flexibility throughout the day would make it even harder for a new entrant." (636) Vueling indicated that "If the Transaction takes place, the new airline will have more than 80% of the total seats at DUB. This gives the new company a great power in many fields: commercial, network, operational". (637)

(648) However, some other respondents answer that the strength of the merged entity would not be a barrier to entry. In particular, IAG indicate that they are used to operating into the hubs of their competitors. (638) Wizz Air indicates that "While the combined market-strength of the combined carrier may increase, the Transaction itself may trigger for DUB to become more active in attracting new carrier to diversify. There may also be some capacity consolidation which could open up new opportunities for other carriers." (639)

(649) The existence and extent of factors that could deter entry or expansion by competitors on the overlap routes is examined in detail below.

8.5.2.2. Historical overview of past entry and exit events by competitors

(650) Historical examples of entry and exit in the industry may provide useful information about the size of entry barriers. (640)

(651) It is recalled that in the 2010 Judgement, the General Court found that there were only a marginal number of entry events in competition with Ryanair on routes out of Dublin, Cork and Shannon (other than by Aer Lingus) and that most of them were failed entry attempts insofar as the airline was no longer present on the route the same month a year later (641). The General Court further found that the existence of barriers to entry as identified by the Commission in the 2007 Decision is not contradicted by the lively entry and exit which characterises the air transport sector in Europe since it was deregulated. The General Court thereby upheld the Commission's approach that consisted in basing the evaluation of competition and the entries on a targeted assessment of the affected markets and not on the air transport sector in general (642). Entry and exit events on the routes subject to this Decision (643)

(652) The routes subject to this Decision have seen little sustained entry by scheduled airlines other than Ryanair and Aer Lingus in the last 5 years: (644)

- Centralwings, the former low cost subsidiary of LOT Polish airlines, entered in 2007 on the Dublin - Warsaw route, however exited the route the following year. (645)

- Clickair entered the Dublin - Barcelona route in 2007 and exited in 2008. (646)

- British Airways started services on the Dublin - London City route in 2008, however it stopped its Dublin - London City and Dublin - London Gatwick services in 2009. (647) In 2012 it took over bmi's services on the Dublin - London route, on which it still operates.

- Austrian operated a weekly service between Dublin and Vienna from May 2009 to September 2009. (648)

- Malev entered the Dublin - Budapest route in March 2010, however stopped operating on the route in February 2012. (649)

- Lufthansa entered the Dublin - Berlin route in April 2008 only to exit it in September 2008. (650)

- Lufthansa re-entered the Dublin - Munich route in 2011. (651) It had previously entered the Dublin - Munich route in July 2008, but exited it in August 2008. (652)

(653) There have been a number of entry events by charter companies: Thomson entered the Dublin - Lanzarote route, the Dublin - Gran Canaria route, the Dublin - Malaga route and the Dublin - Palma route in 2008. It entered the Dublin - Barcelona route in 2010 and the Dublin - Faro route in 2012. (653) Germania entered the Dublin - Faro and the Dublin - Malaga routes in 2012. (654) Orbest entered the Dublin - Faro route in 2012. (655)

(654) Finally, there have been some expansions by operating scheduled airlines on the routes subject to this Decision: Flybe increased capacity on the Dublin - Bristol/Cardiff/Exeter route in summer 2010 and summer 2012, Lufthansa increased capacity on the Dublin - Frankfurt route in winter 2010/2011, summer 2011, winter 2011/2012 and summer 2012; IAG (bmi) recently increased capacity on the Dublin - London route (it however reduced capacity in winter 2009/2010, in summer 2010 and in winter 2010/2011); Air France/CityJet increased capacity in winter 2009/2010 and summer 2010 on the Dublin - London route (however it also reduced capacity in winter 2010/2011 and summer 2011); SAS increased capacity on the Dublin- Stockholm route in summer 2010, summer 2011 and summer 2012. (656)

(655) The above shows that there have only been limited entry events on routes subject to this Decision, especially by scheduled airlines other than the Parties. Moreover, the limited entry events by scheduled carriers were often followed shortly thereafter by an exit, thus showing that many of these entries were not sustainable. The only scheduled airlines, other than the Parties, that entered in the last five years and that still operate on any of the routes subject to this Decision are Lufthansa on the Dublin-Munich route with a seasonal service (summer only) and very low frequencies and British Airways that took over services on the Dublin-London route following its acquisition of bmi in 2012.

(656) On the contrary, several scheduled airlines exited the routes subject to this Decision (657): TAP exited the Dublin - Faro route in 2007; SkyEurope exited the Dublin - Bratislava route in 2008; Flynordic exited the Dublin - Stockholm route in 2007; LOT exited the Dublin - Warsaw route in 2007; Spanair exited the Dublin - Malaga route in 2007; Luxair exited the Dublin -Manchester route in 2007; Iberia exited the Dublin - Barcelona route in 2007; Clickair exited the Dublin - Barcelona route in 2008; BA exited the Dublin - London City and Dublin - London Gatwick route in 2009; bmibaby exited the Cork - Manchester route in 2010 and the Knock - Birmingham route in 2012; Malev exited the Dublin -Budapest route in 2012;

Lufthansa exited the Dublin - Berlin route in 2008. Several charter companies also exited routes subject to this Decision: MyTravel exited the Dublin - Ibiza route in 2007; XL Airways exited the Dublin - Faro route in 2008; Centralwings exited the Dublin - Warsaw route in 2008.

Entry events on routes to and from Dublin, Shannon and Cork

(657) The Commission also examined entry events by scheduled carriers on routes to and from Dublin, Shannon and Cork that took place since 2008 up to the end of August 2012. (658)

(658) Such analysis shows that there have been relatively few entry events by scheduled airlines other than Ryanair and Aer Lingus at the airports of Dublin, Cork and Shannon. Ryanair and Aer Lingus (including Aer Arann) represent for more than half of all entry events by scheduled carriers on routes to or from Dublin, Shannon and Cork. Moreover, most of the entry events by other scheduled carriers occurred on routes that were not yet operated by Ryanair or Aer Lingus (including Aer Arann). Finally, a considerable number of the entry events by other scheduled carriers were not sustainable (meaning that the route was exited again).

(659) The limited number of sustained entry events resulted in a significant drop in the number of airlines that operate from Dublin airport. As illustrated in Table 31, while there were 46 airlines operating from Dublin airport in 2007, there are just 32 airlines operating from the airport in 2012.

Table 31: evolution of number of airlines operating at Dublin airport

Year / # Airlines / Entry/Exit

2007 / 46 / Entry (3): Etihad, SATA, Akria Israel / Exits (10): My Travel, LOT, Spanair, Gulf Air, Finnair, Flynordic, TAP, Blue1, Bulgaria Air, Air Malta

2008 / 41 / Entry (5): Air Canada, Air Transat, Flyglobespan, Pegasus, S7 Airlines / Exits (8): Centralwings, Clickair, Sky Europe, XL Airways, LTU, Eurocypria, Pegasus, FlyLAL

2009 / 36 / Entry (3): Austrian, Thomson, Star1 / Exits (3): Flyglobespan, Austrian, British Airways

2010 / 35 / Entry (2): Blue Air, Cimber Sterling / Exits (1): Estonian,

2011 / 36 / Entry (2): Sunwing, Pegasus / Exits (6): Pegasus, Sunwing, Cimber Sterling, Adria Airways, Air Southwest, CSA,

2012 / 32 / Entry (2): Air Moldova, Emirates / Exits (2): Malev,Luxair

Source: Ryanair - reply to question 41 of the Commission's request for information of 11 July 2012.

8.5.2.3. Base operations provide significant advantages to the merged entity that would constitute a barrier to entry at Dublin, Cork and Shannon airports while the relevant conditions for competitors to establish a base at Irish airports are generally not satisfied

(660) In the 2007 Decision, the Commission examined the entry barriers related to Ryanair's and Aer Lingus' strong position with large bases in Ireland (659). It found that the fact that Ryanair and Aer Lingus operate from the same base at Dublin airport and that they are also present in Cork and Shannon is one of the reasons why the two airlines are closest competitors on the routes to and from Ireland (660). The Commission further found that the role of the base-related barriers to entry was even more significant than in past cases, given the fact that both airlines has a strong presence at the same airport (661).

The level of market presence of the Parties at Irish airports, and in particular at Dublin airport

(661) Ryanair claims that base operations are not a barrier to entry for new competitors and in its response to the Statement of Objections, it claims that the Commission failed to conduct a correct analysis of base operations as a barrier to entry (662).

(662) The individual position of the Parties in Ireland, at Irish airports, and at Dublin airport in particular (see Section 8.4.3) is very strong.

(663) During the market investigation, competitors most frequently mentioned the position of their competitors at Dublin airport as a barrier to entry. (663) Competitors also frequently indicated frequency advantage of competitors referring to the advantage of the two Irish incumbent carriers. The vast majority of competitors found that post- Transaction, the combined market position of Ryanair and Aer Lingus at Irish airports would act as a barrier to entry or expansion. (664) Brussels Airlines explain that "Coverage of all market segments make it very hard to penetrate" given the strong combined market position of Ryanair and Aer Lingus at Irish airports. (665) Germanwings considers that "The incumbents combined market share, pricing power and slot flexibility throughout the day would make it even harder for a new entrant". (666) These conclusions are equally relevant for the airports of Cork, Shannon and Knock. (667)

(664) Moreover, the Commission considers that, if threatened by entry on a particular route, thanks to its very strong base at Dublin airport as well as at its bases in Cork, and Shannon, the merged entity could quickly re-allocate flights in order to offer the most attractive schedules on the challenged route (for example, by re-arranging departure times close to those of the entrant). Previous entry events in the Irish market have been followed by an aggressive response by Ryanair as explained in Section 8.5.2.6.

(665) Therefore, it appears that the strong market presence of the Parties at Irish airports, and at Dublin airport in particular, constitutes a very significant barrier to entry for competitors.

The decision to establish a base is dependent on the market position of competitors

(666) The level of competition and airlines concentration at an airport is considered by competing carriers as a very important parameter in their decision for establishing a base. For example, Flybe considers this to be the second most important parameter (after the size of demand in the relevant catchment area) (668). Brussels Airlines considers that "Given the market position of Ryanair/Aer Lingus, establishing a base at Dublin airport is not sustainable" (669). Similar conclusions also apply for the other Irish airports.

(667) Wizz Air also considers that "if competing carriers have a base at a given airport, this makes entry harder, especially in a capacity constrained airport (where historical advantages will accrue to based competitors)". (670)

(668) Air France/CityJet currently already "finds it hard to compete with the two large based companies in Dublin given [that] it operates with smaller aircraft" (671). At the same time some carriers such as Flybe considers that "the existence of bases as such is not decisive, but rather the size of the operations conducted from the airport. Therefore, to compete effectively with Ryanair and Aer Lingus in Dublin, an air carrier would also need to have scale". (672)

(669) From the above it appears that the level of competition and airlines concentration at a given airport is considered by possible new entrants as a very important parameter in their decision to establish a base at such airport. This fact, combined with the very strong market position of the Parties at Irish airports and in particular at Dublin airport, is likely to deter a new entrant from establishing a base at the Irish airports of Dublin, Shannon and Knock.

Establishing a base requires a significant upfront investment and involves significant commercial risks

(670) The likelihood of entry on a given route would depend on the potential profitability of the given route. Also, as explained in the Horizontal Merger Guidelines: "a high risk and costs of failed entry may make entry less likely. The costs of failed entry will be higher, the higher is the level of sunk cost associated with entry" (673).

(671) Ryanair argues that sunk costs of establishing a base are minimal [...]* and that any competing airline planning to open a base would be able to do so in a very short period of time (674). Ryanair claims that the Commission has included certain non-sunk costs as costs for opening a base and that the Commission has failed to conduct a correct analysis of base operations as an entry barrier.

(672) The Commission considers that it is complex to accurately quantify the sunk costs involved in setting up bases as this would depend on the identity and business model of the airline, as well as on the level of its activities from that base. However, the results of the market investigation have indicated that setting up a base is a risky activity that indicates an airline's commitment to a particular airport and therefore involves a significant level of sunk costs as these costs cannot be fully recouped.

(673) The marketing costs but also costs for installing and operating necessary ground services in an airport where no facilities existed before represent a risk of incurring additional sunk costs (675). Regarding brand recognition for instance, entering into competition with Ryanair/Aer Lingus in Ireland would require a significant investment in marketing, brand-building, and other start-up costs for any potential competitor who would wish to compete head-to-head with the merged entity. Since these costs cannot be recouped should the attempt to enter the market fail, any potential competitor takes a material financial risk. This is likely to deter entry.

(674) Wizz Air considers that establishing a base involves "high start up costs - marketing" (676). Flybe considers that they are "Approximately £3m pa per aircraft." (677)

(675) Even according to Ryanair press releases recently announcing base openings, the investment per aircraft required to open a base would appear to be approximately EUR 54 million (678). While part of these costs may not be sunk, the level of expenses reflects the important costs involved and the associated commitment of this investment, and is inconsistent with the figures quoted by Ryanair [...]*. (679) Besides, Ryanair has not provided a break down of which costs are sunk and which are nonsunk for the particular cases indicated in the press releases.

(676) Therefore, on the basis of the evidence collected during the market investigation, and despite Ryanair's contention to the contrary (680), the Commission considers that the costs for establishing a base are significant.

(677) The significant costs involved in establishing a base coupled with the significant financial risk that this investment entails, in particular in the case of the Irish airports of Dublin, Cork and Shannon, where the Parties have their own established bases, are a deterrant for possible new entrants and make entry at these airports unlikely. Furthermore, the Parties also benefit from their established brand and experience in the Irish market (see for instance Section 8.5.2.4).

Barriers are likely to increase post-Transaction

(678) The strong individual position of the Parties in Ireland, at the Irish airports, and at Dublin airport in particular would be significantly reinforced after the Transaction. The existence of this strong base at Dublin airport post-Transaction would be likely to give the Parties an unrivalled position. Indeed, a merger may lead to price increases but not attract entry because entrants would suffer a significant cost disadvantage relative to incumbents. A reason for a cost disadvantage could indeed be the presence for competitor(s) of significant economies of scale and scope.

(679) Overall competitors do not consider that barriers to entry will decrease post- Transaction. If anything several competitors have claimed that higher barriers to entry could result from the Transaction. For example, Lufthansa is of the view that "it cannot be excluded that the dominance of a combined FR/EI could lead to a potential barrier for some carriers" (681) while Lufthansa did not consider that there would be barriers to entry at Dublin airport pre-Transaction. Lufthansa further explains that post-Transaction the "Cost of entry would probably be higher if there is one large competitor. FR and EI will be the biggest customer and therefore be able to exert increased purchasing and sales power and will therefore be at an advantage to other airlines especially new (and smaller) entrants." (682)

(680) Thomas Cook also explains that "the fact that the merged entity would have post- Transaction an even more important base in Ireland would constitute a further barrier to entry given that a competitor or new entrant would need to compete with the merged entity in its home base" (683).

(681) Furthermore, the great majority of competitors explain that they would not reconsider their decision regarding opening a base in Ireland post-Transaction (684). Aer Arann for example claims that "the scale of the competition post-transaction that would exist at the airport would make opening a base at Dublin airport impossible financially. The potential losses that would be incurred in the initial years would be too large." (685) Flybe claims that it would consider establishing a base post- Transaction, however, this would take place in the context of the remedies (686).

Conclusion

(682) The Commission concludes that the strong market position of the Parties in Dublin, Cork and Shannon inter alia as a result of their established bases, and the position of the merged entity post-Transaction constitute a significant barrier to entry. In addition, the Commission concludes that the relevant conditions for competitors to establish a base at Irish airports are generally not satisfied, inter alia because competitors would be deterred by the strong market position of the merged entity in Ireland, by the significant upfront investment and commercial risks.

8.5.2.4. Strong brand recognition

(683) The Commission considered in 2007 that Ryanair and Aer Lingus have a well-known brand in Ireland which provides them with specific advantages and which means that entry by potential competitors would require high marketing costs and entail a high risk of sunk costs.

(684) The 2010 Judgement confirmed that "the Commission was right to state in the contested decision that an entry in Ireland required considerable expenditure in order to establish a brand capable of competing with Ryanair and Aer Lingus and to gain access to new customers." (687)

(685) It remains the case, as in 2007, that no other competing airline enjoys the same high level of brand recognition as Ryanair and Aer Lingus across a large number of routes to and from Ireland.

(686) Ryanair contests the importance of a strong brand recognition. It argues that competition is driven either by product differentiation (which may include branding), in which case consumers are less-price sensitive, or by price, in which case competition is between more homogeneous products. It argues that in an environment where consumers are said to be price-sensitive, and to be purchasing tickets predominantly online (where all prices can be easily compared), brand awareness is a much less relevant factor, while value-for-money is the key criterion for consumers. (688)

(687) Ryanair also claims that the Commission should distinguish between Ryanair's and Aer Lingus' brand awareness at the Irish end and the non-Irish end of the routes, notably on routes where a strong competitor, such as BA, Lufthansa, Air France or SAS, is currently present. Brand awareness of these competitors at the non-Irish end of the route is, according to Ryanair, at least as high as Ryanair's or Aer Lingus' brand recognition. (689)

(688) Ryanair further argues that there are other airlines that are well-known brands in Ireland (for example Flybe, Wizz Air, easyJet, Alitalia, SAS and Iberia). (690)

(689) Ryanair finally argues that marketing expenditures are insignificant for any existing airline and that the recent expansion of numerous carriers demonstrates that airlines can easily establish strong brand presences in relatively short periods of time (691). It refers to the examples of various Middle East carriers such as Qatar Airways, Etihad and Emirates which have recently begun operations on routes from various destinations to and from Europe. It argues in this respect that the new, non-facilitated entries of Emirates (entry in 2011) and Etihad (entry in 2007) are examples of establishment of strong brand recognition in a relatively short period of time through intensive marketing. (692)

Importance of brand recognition

(690) Ryanair's argument that brand awareness is a much less relevant factor in an environment where consumers are said to be price sensitive and purchase tickets predominantly online, cannot be sustained.

(691) While the core Irish market may be price sensitive, price is not the only factor that drives competition. The Commission does not consider the market to be homogeneous but acknowledges that carriers are highly differentiated and that there are differences in services offered. Ryanair itself notes that there are "significant differences in service levels between Aer Lingus and Ryanair" (693) and states that "value-for-money" is the key criterion for consumers (694), which is not the same as a pure price factor.

(692) The Irish market for short-haul passenger flights is characterised by (i) a high level of direct sales and more in particular (ii) of online sales.

(693) Direct sales present the advantage that they allow airlines to avoid paying a commission to travel agents. Both Ryanair and Aer Lingus, the most important providers of shorthaul scheduled passenger air transport services to and from Ireland, distribute their shorthaul flights in Ireland exclusively through direct channels, without the intervention of travel agents.

(694) The Irish market is furthermore characterised by a high level of online sales. As set out in Section 8.4.2. Ryanair and Aer Lingus both realised most of their bookings in Ireland from the web.

(695) In a market where an important part of the sales are direct sales (that is to say, without the intervention of a professional travel agent who could advise customers of the airlines operating on a particular route) and where an important part of the sales occur via airline websites, it is important for airlines to enjoy brand awareness amongst customers. Customers - also price sensitive customers - are more likely to visit the websites of airlines that they know than websites of airlines with which they are less familiar.

(696) Aer Lingus and Ryanair both have very strong websites. Their internet pages are, according to independent third party research, the most popular airline internet pages in Ireland (ranking number 61 and 80 respectively of Ireland's most visited websites). No web page from any actual or potential competitor appears in the "top 100" list of Irish web pages nor do any websites which collect fares of different competitors. Other airlines are ranked much lower, for example British Airways is ranked at number 572, easyJet at number 874, Lufthansa at number 2 627, Flybe is ranked at number 2 657 and Air France at number 4 895. Skyscanner ranks at number 226, Ebookers at 394 and Expedia at number 459 (695).

(697) As noted in the 2007 Decision (696), it remains the case that while the importance of the internet distribution channel has certainly reduced distribution cost, new competitors still face the problem that their web pages must be visited and used by Irish customers in order to win such customers.

(698) Consequently, brand awareness - also in an environment where customers are said to be price sensitive and purchase tickets predominantly online - is an important factor and any entrant to the Irish market will need to develop a very strong brand if it wants to be able to compete effectively with the merged entity for customers (and avoid a disadvantageous position compared to the merged entity with respect to the cost of sale).

(699) The result of the market investigation showed that competitors broadly recognise the significant role of brand names and good reputation for the competitive process with respect to short haul routes to and from Ireland (697).

(700) The history of Aer Arann / Aer Lingus Regional provides a good example of the importance of a recognised brand and of the difficulty of sustaining a viable Irish operation against the Ryanair and Aer Lingus brands. Aer Arann explains that "as the Aer Arann brand was not recognised in the United Kingdom (compared with the Aer Lingus brand), Aer Arann faced serious difficulties and could not compete with other carriers on short-haul flights out of Ireland." Then, Aer Arann entered into a franchise agreement with Aer Lingus in January 2010 which enabled it to use the Aer Lingus Regional brand and to have access to distribution through the Aer Lingus website. Aer Arann confirmed that since the franchise agreement has been put in place, it has increased its load factor (from 55% in 2009 to 65% in 2012 and it expects a further increase) and registered a significant growth every year, in terms of passengers and thus yield. (698)

Importance of brand recognition in Ireland

(701) The Commission acknowledges that certain competitors that are active on routes to and from Ireland may enjoy a strong brand recognition in their home territory. For instance, Sections 8.4.2 and 8.9.5 discuss in this respect the brands IAG, and Air France/CityJet, Lufthansa, SAS, and Flybe. The Commission also acknowledges that potential entrants that currently do not operate on any routes affected by the Transaction may enjoy a strong brand at the non-Irish end of the route.

(702) The Commission acknowledges that a strong brand at a destination may to some extent be of use when competing with the Parties on certain routes to and from Ireland. The Commission nevertheless considers that the benefits of such brand recognition outside Ireland are limited compared to the advantages that Ryanair and Aer Lingus enjoy thanks to their strong brand recognition in Ireland.

(703) In this respect, the Commission first notes that the fact that a competitor or potential entrant would have a strong brand at the non-Irish end of the route does not imply that they also have a strong brand in Ireland (for example as set out in Section 8.9.5.2 Lufthansa considers that, although it has a strong brand across Europe, it is weak compared to Ryanair and Aer Lingus when it comes to traffic ex-Ireland).

(704) Moreover, the benefits of having a strong brand at the non-Irish end of the route would be limited to any such route, while Ryanair and Aer Lingus would benefit from their strong brand with respect to all routes to and from Ireland. This is reflected by the fact that the competitors referred to by Ryanair are only active on a very limited number of routes to and from Ireland (mostly routes connecting their home territory to Ireland).

(705) Moreover, the Commission notes that on many of the routes affected by the Transaction, a majority of customers originate from Ireland (the share of Irish originating passengers is in particular high on routes from Ireland to holiday destinations).

(706) As noted in the 2007 Decision (699), in particular on routes where the majority of customers originate in Ireland, competitors face a significant disadvantage vis-à-vis Ryanair and Aer Lingus. In order to successfully operate such routes, it is important for competitors to be able to sell tickets to Irish customers. It is therefore crucial for competitors to have access to Irish customers. Since the customers originating from the non-Irish end of the route would account only for a minority of the passengers on this route, it would not be sufficient to rely only on such customer base. This is especially the case given that Ryanair is not only strong in Ireland, but also Europewide.

(707) Even on routes where only half or fewer of the passengers originate from Ireland, it would still be important for competitors to have access to Irish customers given Ryanair's strength not only in Ireland but also elsewhere in Europe.

Difficulty of creating a strong brand in Ireland.

(708) Both Ryanair and Aer Lingus enjoy strong brand recognition on the routes from/to Ireland. No other competing airline enjoys the same high level of brand recognition across a large number of routes to and from Ireland. The airlines that, according to Ryanair, are well-known brands in Ireland (for example, Flybe, Wizz Air, easyJet, Alitalia, SAS and Iberia) mostly did not share this view:

- Flybe, Wizz Air, easyJet and Alitalia identified "brand recognition" as a significant barrier to entry on routes subject to this Decision while SAS and Iberia did not.

- Flybe identified brand awareness as a significant barrier to entry and indicated in this respect that "it's important to promote the brand and new services when developing new routes (...)" (700). It considered that while Ryanair and Aer Lingus enjoy a strong brand on the routes subject to this Decision, its own brand is considered as a medium brand by passengers on the routes subject to this Decision. The main reasons for this is that it only has a presence on two routes ex-Dublin and that only moderate levels of marketing have been undertaken in Ireland. It estimates that it would take two years to upgrade its brand into a strong brand. Flybe considers that as such Ryanair's and Aer Lingus' brand presence would not be a barrier for entry into operations ex- Dublin as it has a strong presence at the other end of the route, namely in the United Kingdom market.

- Wizz Air, which also identified brand recognition as a significant barrier to entry, did not share Ryanair's view about the strength of its brand in Ireland. Wizz Air (701) considered that while Ryanair and Aer Lingus enjoy strong brand recognition on the routes subject to this Decision, its own brand is considered as a "weak" brand by passengers. In the Irish market, Wizz Air's brand awareness is low, due to the fact that Irish operations represent a small part of Wizz Air's network and thus Wizz Air undertakes limited marketing in Ireland. In Ireland, Wizz Air only operates ex Cork. It explains that it works with direct distribution (not through agents) and that therefore both marketing and low fares are important for its business to be successful. It further considers that a period of two years would not be sufficient to upgrade its brand into a strong brand for routes to and from Ireland and that investment will be required to raise brand awareness which it considers can only be achieved through discounting fares and advertising spend. In conclusion, it considered that "The large number of destinations already served combined with the brand strength of Ryanair and Aer Lingus would be significant barriers to entry." (702)

- easyJet identified "brand recognition" as a significant barrier to entry for the routes subject to this Decision. It indicates that, while having strong brand recognition in Northern Ireland (Belfast), the fact that it does not operate to the Republic of Ireland means that its brand recognition is weaker there. It expects that it would take significant time (more than two years) and cost to upgrade its brand to a strong brand for routes to and from Ireland. (703) easyJet indicates in this respect that it believes that even more than a strong brand, market presence is important. As easyJet has a lower market presence in Ireland than Ryanair, its brand is also less known. easyJet thus believes that Ryanair and Aer Lingus would be the first brands travellers would think of in connection with Ireland and that new entrants in Ireland, in addition to having a strong presence in Ireland, would need to invest in significant marketing campaigns (704).

- Alitalia also identified "brand recognition" as a significant barrier to entry for the routes subject to this Decision. While indicating that it considers it has a strong brand, Alitalia indicates that in Ireland it only has a medium brand due to the lack of operation on the relevant route. (705)

- Iberia indicated that it has strong brand recognition in Spain which it generally perceives to be stronger than the brand recognition of Aer Lingus and Ryanair. (706) For the size of the Irish market it expects that a significant investment on brand marketing would be needed in order to establish itself as a recognised brand.

(709) Furthermore, the Commission takes into account the difficulty and cost of establishing a recognised brand in Ireland. The result of the market investigation indicated that brand recognition is amongst the most frequently identified barriers to entry. (707) Moreover, several competitors are of the opinion that at least two years or more would be required to create a strong brand or to upgrade their own brand on routes to and from Ireland, notably due to an increased marketing costs expenditure. (708) Furthermore, the cost of marketing and promotion required to upgrade the brand would be significant. (709)

(710) The examples raised by Ryanair of new, non-facilitated entries such as by Qatar Airways, Emirates and Etihad (710) do not contradict the Commission's finding. First, these entries took place on routes where they were not in competition with the Parties and the routes on which these entry events took place were not routes subject to this Decision. Moreover, Etihad and Emirates for example conducted important marketing campaigns. Etihad has been sponsoring the All Ireland Hurling Championship since its entry on the Abu Dhabi-Dublin route in 2007 and recently renewed its sponsorship undertaking for a further 5 years, a deal which is believed to be worth EUR 5 million (711). Emirates, when launching its new Dubai - Dublin route last year, launched an extensive marketing campaign in Ireland to promote its new service. It declared in this respect that the campaign represents a significant investment for the airline (712). As acknowledged by Ryanair, these airlines conducted an intensive marketing campaign (which most airlines would not be in a position to do). This confirms rather than contradicts the Commission's finding that marketing and promotion costs required to upgrade a brand would be significant.

Conclusion

(711) The Commission concludes that on the short-haul routes to and from Ireland, brand recognition and good reputation play a significant role for the airline industry in the competitive process. Considering that a new entrant would face competition from the merged entity which has the two strongest brands in Ireland and that it would take considerable time and investment for an entrant without a strong brand to upgrade its brand in Ireland, brand recognition constitutes a barrier to entry. This entry barrier is equally relevant for routes from Dublin, Shannon, Cork and Knock.

8.5.2.5. Dual branding strategy and strong market position

(712) Ryanair intends to maintain Aer Lingus as a separate company operating as a viable and scalable competitor to legacy carrier short-haul operations, as well as carriers such as easyJet. Ryanair indicates that it intends to maintain the individual brand and identity of both Ryanair and Aer Lingus.

(713) Ryanair believes that, following the Transaction, it will be able to reduce Aer Lingus' operating costs considerably in some areas. It further indicates that Aer Lingus will be committed to a growth strategy, which will facilitate securing volume and growth driven cost reductions at primary airports such as Copenhagen and Brussels (Zaventem). It argues that as part of the Ryanair Group, Aer Lingus will enjoy lower unit costs and will be better placed to increase competition, thereby offering an alternative to the Air France, British Airways, and Lufthansa groups at their "home" airports.

(714) In the 2007 Decision, it was considered that "If anything, a dual branding strategy may enable Ryanair to exploit its market power more effectively and, thus, extract additional profits as compared to a single-branding strategy from its customers. Dual branding can make it easier for a firm to charge higher prices to customers with a higher willingness to pay and lower prices to more price-elastic customers, increasing the firm's profits." (713)

(715) The results of the market investigation showed mixed views amongst competitors as to whether the dual branding strategy would make entry less likely (714). Alitalia for instance indicated that "Double brand operations could create more consolidation and make entry by another player less attractive" (715). Loganair (716) considers that it would make entry less likely and explains that the "[m]arket perception would be of two independent operators and that fares would be competitive" (717). TAP considers that it might make entry less likely "if such dual-branding operations perceived as, or effectively reflecting, a stronger market presence and visibility" (718). Vueling indicates that a dual branding strategy would "be more of a deterrant to new entrants". It argues that it has experience in this area and that "two "selling" brands typically generate a more competitive sales platform" (719). Wizz Air indicates that customers consider flight options based on a 'relevant set' of airlines offering services and that "a third entrant is less likely to be included in consumers' relevant set of choices, especially [i]f both incumbent airlines are already well established" (720).

(716) Germanwings (721) and IAG (722) do not consider that the dual branding strategy would make entry less likely. Thomas Cook (723) believes that it makes little difference. Lufthansa indicates that this would not be the case and explained that "(...) there are already several merged airline entities that operate under two or even more brands (AF, KL, IAG) etc" (724). Flybe (725) considers that there are other more important factors such as market size, potential future market share and pricing.

(717) The dual branding strategy announced by Ryanair post-Transaction would provide the merged entity with a wide coverage of the different segments of the short-haul point-to-point air transport.

(718) The merged entity would benefit from a strong brand and a strong position both in the no-frills segment with the lowest fares (Ryanair) as well as in the segment with higher service level (Aer Lingus). In addition, in both segments the merged entity would have an advantageous cost base structure compared to its competitors.

(719) Therefore, the Commission considers that the merged entity's strong market position in the two segments resulting from the announced dual branding strategy would constitute an additional barrier to entry. This entry barrier is equally relevant for routes to and from Dublin, Shannon, Cork and Knock.

8.5.2.6. The risk of aggressive retaliation Introduction

(720) In its 2007 Decision, the Commission found that Ryanair has a reputation of engaging in aggressive competition in case of new entry to Ireland, notably by temporarily lowering prices and expanding its capacity in order to drive out the new entrant on routes to or from Ireland. (726)

(721) The 2010 Court decision confirmed that "the risk of aggressive retaliation would be even greater following the concentration, since Ryanair-Aer Lingus combined would be the dominant operator on literally all routes from and to Ireland. In building up a reputation for deterring the entry of competitors, Ryanair creates a de facto barrier to entry for new competitors." (727)

(722) Ryanair argues that its competitive behaviour is not an entry barrier. It claims that it does not engage in retaliatory behaviour designed to deter entry and that it simply aims to deliver low fares to customers in line with its business model as Europe's lowest fare airline. Ryanair considers this to be a healthy competitive environment which benefits consumers and is consistent with new entry and innovation (728). It claims that the views expressed by competitors that Ryanair, as Europe's lowest fare airline, responds to competition by offering low fares, is unsurprising and indicative of nothing more than that Ryanair has the lowest cost base and the lowest fares in Europe. (729) Ryanair further claims that the Commission's 2007 finding that Ryanair responds to entry by "temporarily lowering prices and expanding capacity in order to drive out new entrants on routes ex Ireland" was incorrect and that it has continued to offer low fares and increased capacity to or from ex-Ireland since 2007. (730) Ryanair finally claims that the fact that it has identified upfront buyers in the context of the proposed remedies package clearly shows that they will be able to compete successfully, which undermines, according to Ryanair, the conclusion on aggressive retaliation (731).

(723) In line with the General Court's findings in the 2010 Judgement, it is noted from the outset that the debate does not concern Ryanair's pricing policy as such but the classification of its aggressive reputation as a barrier to entry (732). The Commission thus examines whether Ryanair's past conduct (even if such conduct would result from its business model) is liable to dissuade potential competitors from entering a market on which it is present. Moreover, Ryanair's claim that it has continued to offer low fares and increased capacity ex Ireland since 2007 is not relevant for the analysis since - even if it were well founded (quod non as shown by Table 27: Ryanair and Aer Lingus share of European departing seats as % of European departing seats at Dublin) - such claim refers to a competitive situation in which Aer Lingus is present, while the Transaction would eliminate Aer Lingus as a competitive constraint on Ryanair.

Risk of aggressive retaliation as a barrier to entry

(724) Fear of retaliation was amongst the commonly referred to barriers to entry. (733) Several respondents indicated that they would expect an aggressive reaction by Ryanair in terms of pricing and marketing. (734)

(725) According to Aer Lingus "Ryanair has an established reputation for aggressive retaliation to entry". (735) Similarly, Vueling believes "...that Ryanair and Aer Lingus will commercially react with all the tools available to them to prevent any new additional airline at Dublin" (736). It finds it very difficult to compete with Ryanair when Ryanair has a strong position in a market and it considers Ryanair to be a very aggressive competitor. (737) Aer Arann indicates that "The primary barrier to route expansion by Aer Arann at Dublin airport is the potential competitive reaction from Ryanair." (738) In the same way, Air Berlin would expect very aggressive pricing. (739) According to easyJet, "Ryanair is an aggressive competitor as has been shown by its reactions to other airlines trying to enter that market. Thanks to its financial strength, its low cost model and its size it can easily enter into long-lasting price wars and push competitors out of the market". (740)

(726) During the market investigation, several references were made to Ryanair reacting to new entry on its routes by increasing frequencies and capacity, thus forcing competitors out of the market.

(727) Some examples were already described in the 2007 Decision and include the failed attempt by easyJet to enter the United Kingdom-Ireland market (741), MyTravelLite's attempt to enter the Dublin - Birmingham route or GoFly's failed entry attempts on the Dublin - Glasgow and Dublin Edinburgh routes (742).

(728) Aer Arann refers to its operations on the Dublin-Cork route as an exemple of Ryanair's strategy towards competitors. Aer Arann had been operating on this route for a number of years before Ryanair entered. When Ryanair did enter, it significantly undercut the prices, and Aer Arann could not sustain its level of frequencies on the route. Aer Arann reduced its frequencies (from 6 to 4 to 2) and then ceased operating on the route altogether as it was incurring significant losses. After Aer Arann exited the route, Ryanair stopped operating on the Dublin-Cork route. (743)

(729) Ryanair's reaction towards Wizz Air with respect to routes to and from Modlin airport provides another example of the aggressive strategies used by Ryanair towards competitors (744). When Wizz Air announced in the beginning of February 2012 that it would move its entire Warsaw based operations to the new Warsaw Modlin Airport, this triggered an immediate reaction by Ryanair who announced the opening of eight Warsaw Modlin routes from July 2012 (Brussels, Budapest, Dublin, London, Milan, Oslo, Rome and Stockholm). In its press announcement concerning the opening of the new routes, Ryanair indicated releasing launch fares "less than half Wizz's lowest fares to/from Warsaw in July 2012". (745) The fact that Ryanair's behaviour was designed to deter entry and force Wizz Air to exit the routes is evidenced by statements in Ryanair's internal documents. [...]*.

(730) Later that same year, Ryanair also engaged in an aggressive strategy against Wizz Air on the routes operated by Wizz Air from Cork (746). Ryanair entered 5 routes from Cork (to Gdansk, Krakow, Warsaw, Wroclaw and Vilnius) in Winter 2012. Ryanair actively promoted these new routes by referring to the fact that its fares would be "more than 50% cheaper than Wizz's lowest fares". (747) Wizz Air has meanwhile announced its exit from three of these routes (namely from Cork to Warsaw, to Wroclaw and to Vilnius), whereby it notably referred to overcapacity on the routes which made it sensible to focus on other routes with higher demands instead (748).

(731) Ryanair's claim that the upfront buyers that it has identified to take up part of the operations to be divested in the context of the commitments clearly conclude that they will be able to compete successfully in this environment is unconvincing and does not undermine the Commission's findings. Any such entry by upfront buyers would take place in the context of the commitments, and thus in the context of an agreement between the upfront buyer and Ryanair. Such circumstances cannot be assimilated with a typical competitive environment.

(732) The above shows that Ryanair's claims that it does not engage in retaliatory behaviour designed to deter entry and that it simply aims to deliver low fares to customers cannot be sustained.

(733) Ryanair has a reputation for and the ability to engage in aggressive competition where airlines try to enter a route on which it operates ex-Ireland (or elsewhere), in particular by lowering its prices temporarily and expanding its capacity. Ryanair's past conduct is likely to dissuade potential competitors from entering on a market on which it is present since the risk of an aggressive retaliation by the merged entity is perceived as high.

(734) Building up a reputation of deterring entry creates a factual barrier to entry for new competitors as set out in the Horizontal Merger Guidelines. (749)

(735) The risk of aggressive retaliation would be even greater following the Transaction given the strong market position that the merged entity would have on the routes subject to this Decision and given its even stronger position on the Irish market.

(736) In the light of the above, the Commission concludes that the fear of aggressive retaliation by the merged entity post-Transaction constitutes a barrier to entry. This entry barrier is equally relevant for routes to and from Dublin, Shannon, Cork and Knock.

8.5.2.7. Level of airport charges at Dublin airport

Introduction

(737) In the recent decision in case M.5830 Olympic / Aegean Airlines the Commission considered that high airport charges may constitute a deterrent for domestic and international airlines (in particular for low cost carriers) to set up a base at Athens International Airport. (750)

(738) Airport charges at Dublin airport have increased over the past number of years (751). Ryanair itself indicates that in recent years, airport charges at Dublin airport have doubled. (752) It refers to the introduction by the Irish Government of air travel tax in March 2009 and the doubling of airport charges at Dublin airport since 2007. (753)

(739) In February 2011, Ryanair brought a complaint before the Commission against DAA and against Aer Lingus regarding alleged violations of Articles 101 and/or 102 of the Treaty (754). With respect to the alleged violation of Article 102 of the Treaty by DAA, Ryanair argued inter alia that the increase in airport charges (including their nondifferentiation between terminals), the Transfer Incentive Scheme and the policy on remote stands represented excessive pricing. In its letter of 23 November 2012 pursuant to Article 7(1) of Regulation No 773/2004 in Case No COMP/39.886 - Ryanair/DAA-Aer Lingus, the Commission adopted the provisional conclusion that Ryanair's claim regarding excessiveness of DAA's pricing could not be supported given that there are no indications that the charges would not reflect the actual costs incurred and given that the aeronautical revenue per passenger at Dublin airport is below the Union average.

(740) Referring to the Commission's provisional conclusions in Case No COMP/39.886 - Ryanair/DAA-Aer Lingus, where the Commission considered that "even considering the recent airport charges increases, the aeronautical revenue per passenger at Dublin airport is below the EU average", Ryanair argues that the Commission cannot interpret the same facts in diametrically opposite ways in order to arrive at conflicting conclusions. (755)

(741) Ryanair also argues that airport charges are borne equally by incumbents and new entrants and thus do not constitute barriers to entry under the Horizontal Merger Guidelines. It claims that if indeed the level of airport charges at Dublin airport were a barrier to entry, it would be in Ryanair's interest, as an established airline, to support high airport charges in order to deter competitors from entering the market, while it has campaigned against the increased in airport charges at Dublin airport (756).

(742) Ryanair further argues that, if the Commission's claim that a combined Ryanair/Aer Lingus would be able to exert some market power against the DAA monopoly is correct, then the claimed entry barrier in the form of high airport charges at Dublin should naturally be removed post-Transaction. As Wizz Air highlights, the Transaction can deliver a pro-competitive effect in that it may force the DAA "to become more active in attracting new carrier[s]" (757).

(743) Ryanair finally argues that the fact that as part of the proposed remedies package, at least two upfront buyers are willing to set up substantial base operations at Dublin airport, makes it clear that Dublin airport charges do not constitute a barrier to entry. (758)

(744) Ryanair's view that the level of airport charges at Dublin airport cannot be considered as a barrier to entry is shared by the DAA. The latter make to following arguments: (i) that airport charges at Dublin airport are amongst the lowest when benchmarked against comparable airports, (ii) that there are other more predominant considerations such as for example, the economic climate, affecting entry and expansion at Dublin airport, (iii) that similarly, traffic volumes at Dublin airport are affected by economic growth, (iv) that the government's decision to reduce the air travel tax (by an amount that exceeded the level of increase in airport charges) has not led to a material increase in passengers and (v) that Ryanair has devoted considerable marketing efforts to generating a negative perception of airport charges at Dublin airport (759).

The level of airport charges as a deterrent to entry and expansion

(745) It is noted from the outset that the Commission's provisional conclusion that DAA's pricing is not excessive (as set out in its letter of 23 November 2012 pursuant to Article 7(1) of Regulation No 773/2004 in Case No COMP/39.886 - Ryanair/DAA-Aer Lingus) does not mean that the level of airport charges at Dublin airport could not be a factor deterring competitors from entering or expanding on routes to and from Dublin airport. Also, the finding that the level of airport charges would deter potential entrants does not necessarily imply that DAA's pricing is excessive and constitutes a violation of Article 102 of the Treaty.

(746) For the excessive pricing analysis under Article 102 of the Treaty, the Commission assesses whether the price bears a reasonable relation to the economic value of the product supplied. For the present analysis, what is relevant is the question of whether the level of airport charges (regardless of whether they constitute excessive pricing or not) is perceived by potential entrants as a factor deterring entry on routes to and from Dublin airport (for example, compared to the level of profitability they expect on the route).

(747) Also, the fact that the aeronautical revenue per passenger at Dublin airport is below the EU average does not necessarily mean that for routes to and from Dublin the level of airport charges (even if below EU average) could not act as a deterrent to entry.

(748) Ryanair used the increased airport charges at Dublin airport to justify its decision to reduce capacity at Dublin Airport in 2009. (760) It consistently maintained that growth at Dublin airport is only possible, and therefore contingent upon, significant reductions in airport charges. (761) When DAA confirmed the airport charges for 2012 at the end of last year, Ryanair called for government action, asking the government to force DAA to reduce Dublin airport's charges by at least 40%. It confirmed in this context that "Dublin Airport is the most expensive of the 160 airports it operates to/from, which is why the DAA monopoly's Dublin traffic has plummeted by 20% to just 18.4m (2010) from its peak of 23.4m pax in 2008" and that "tourist taxes and the DAA monopoly's high fees continue to damage Irish tourism and jobs". (762)

(749) Moreover, Ryanair referred to quotes made following the increase of airport charges after the opening of Terminal 2 by Aer Lingus' CEO (with a similar 40% share of traffic at Dublin airport) described as "insane" (763) or by Etihad's CEO which has also described Dublin airport's price increases as "too excessive" (764).

(750) The foregoing shows that Ryanair itself considers the high level of airport charges and taxes as reasons for the decrease in traffic at Dublin airport and a barrier to growth in traffic at Dublin airport.

(751) Several respondents to the market investigation, in particular low cost companies, indicated that the level of airport charges makes routes to and from Dublin less attractive for potential new entrants. (765)

(752) Aer Lingus indicated that "there has been a significant increase in airport charges since the 2007 Decision which makes it much less attractive for entry by third party carriers, particularly low cost entrants." (766)

(753) Jet2.com indicated that "Increase in airport charges makes this airport less attractive" (767) and Aer Arann indicated that "Yes Dublin airport has become a more expensive place to operate as an airline. As a result the bar required for new routes to make money has become higher over the last 5 years." (768)

(754) Flybe indicated that they consider airport charges as a 'significant' barrier to entry at Dublin airport (769).

(755) Wizz Air stated that "key to a decision [to enter any of the routes ex-Dublin] would also be a reduction in DUB airport charges. Only then could a meaningful analysis of route opportunities be made" (770). It identified the level of airport charges at Dublin airport as a significant barrier to entry (771) and explained that the level of airport charges was one of the factors making entry at Dublin difficult (772), thereby noting that the transaction may force DAA to become more active in attracting new airlines.

(756) The level of airport charges is a factor which co-determines the level of profitability that a potential competitor can achieve on a route. In line with Aer Arann's above mentioned response, the higher the airport charges, the higher the bar required for new routes to generate a profit.

(757) Ryanair makes the point that airport charges are borne equally by the incumbents and by potential new entrants. This however does not mean that the level of airport charges could not deter new entrants from operating on routes to and from Dublin and would thus be irrelevant for the assessment of the question of whether entry is likely to eliminate the anti-competitive effects of the Transaction. Moreover, Ryanair itself identified the level of airport charges as a barrier to entry for low cost carriers in the context of the Commission's investigation in the recent case n°M.5830 - Olympic/Aegean. As noted in Recital 582 of the Decision n°M.5830 - Olympic/Aegean, Ryanair indicated that "(...) the level of airports charges, in particular at AIA, as the primary barrier to entry for low cost companies. According to Ryanair, entry can occur only if there is a reduction of the airports costs."

(758) Ryanair's argument that, if indeed the level of airport charges at Dublin airport were a barrier to entry it would be in its interest to support high airport charges in order to deter competitors, is not convincing. Ryanair operates on a large number of routes to and from Dublin and it is thus affected by the level of airport charges on all these routes. If entry were to take place, this would only affect one or a limited number of routes on which Ryanair operates. The level of airport charges on the other hand affects Ryanair's operations on all routes ex Dublin, which would be even more in number post-Transaction. It is thus unlikely that it would be in Ryanair's interest to deter entry on one or more routes by supporting high airport charges which would affect it on all its operations to and from Dublin. Moreover, as set out in this Section, the level of airport charges is just one factor contributing to deterring entry and expansion.

(759) Ryanair further argues that the claimed entry barrier in the form of high airport charges at Dublin would naturally be removed since the DAA may become more active in attracting new carriers in response to the alleged market power of the merged entity against the DAA. This argument is unconvincing and unfounded. In line with its provisional conclusions in Case No COMP/39.886 - Ryanair/DAA-Aer Lingus, it is noted that there are no indications that the current charges set by the DAA would not reflect the actual cost incurred. There are no elements that indicate that the DAA would be in a position to proceed to a significant reduction of the airport charges in the near future. Moreover, the DAA has already put in place incentive programmes to promote traffic at Dublin airport (773). However, these incentive schemes have had only limited success in inducing entry or expansion on European short-haul routes. Even if - as Ryanair claims - the Transaction would lead to DAA becoming more active in attracting new carriers at Dublin airport, such strategy to attract new carriers would only be relevant for the present assessment insofar as it would lead to entry or expansion on routes which are already served by the merged entity. It is thereby noted that incentive schemes such as the current short-haul route support scheme only apply to new routes (that is to say, routes that are not yet served).

(760) Finally, the mere fact that upfront buyers would be willing to enter routes to and from Dublin does not contradict the finding that the level of airport charges deters entry. Any such entry by upfront buyers would only take place in the context of the proposed Commitment package.

Conclusion

(761) In the light of the foregoing, the Commission therefore considers that the level of airport charges at Dublin airport is likely to deter new entrants, in particular low cost companies, from opening routes from or to Dublin.

8.5.2.8. The Irish market for passenger flight is not considered as very attractive

Introduction

(762) In the 2007 Decision (774), it was concluded that the Irish market is not considered by many competitors as a particularly attractive market and that many airlines indicated that they would rather seek to open new routes to other destinations than routes to Ireland. The reasons for this were as follows:

- that entering this market where there are already two strong low-cost carriers with a strong base would require significantly higher marketing costs than entry on other markets;

- that the Irish market is regarded as a relatively small market with limited turnover;

- that the small Irish market is in general not expected to be the fastest growing and most profitable market in the future by many competitors, at least not on the main routes between Ireland and the United Kingdom, and

- that the specific geographic situation of Ireland does not favour new entry but rather discourages it because Dublin is not a "pass-by" destination.

(763) It was also noted that when it comes to serving Dublin, the incentives of a network carrier operating a route between Dublin and its hub are different from those of a point-to-point carrier. It was finally considered as a disadvantage, in particular for low-frills operators, that there are no secondary airports in Ireland.

(764) The Commission's finding that many competitors considered that the Irish market was not particularly attractive and that many potential competitors rather seek to open new routes to destinations other than Ireland was upheld by the General Court in the 2010 Judgement, and this despite the fact that the General Court recognised the then prevailing buoyancy of the Irish economy (775).

(765) Ryanair argues that the Irish economic situation affects incumbents and new entrants equally, and it therefore cannot properly be characterized as an entry barrier within the meaning of the Commission's own Horizontal Merger Guidelines. Ryanair further argues that the analysis is not grounded on the business realities of the low-fare airline industry and that relatively weak economies have often served as springboards for low-cost airlines to begin operations as consumers become more price-sensitive and look for cheaper ways to travel (776).

(766) It underlines that both Ryanair and easyJet have seen sustained growth across Europe in the last four years in spite of the ongoing economic difficulties in the area. A relatively weak domestic economy does not only represent an opportunity for future growth, but it also ensures that airlines have an incentive to lower prices in order to increase consumer demand (777).

(767) Ryanair finally argues that the fact that the Irish market is attractive to competing airlines is self-evident from Ryanair's proposed remedies package and the upfront buyer's commitment to entering the market. Unattractivess of the Irish market deters entry and expansion

(768) Most of the circumstances reflected in the 2007 Decision remain unchanged: the Irish market is still a small market with limited turnover and is already served by two strong carriers. Due to its geographic situation, Ireland is not a pass-by destination and there are no secondary airports.

(769) The most important change affecting the attractiveness of the Irish market compared to 2007 is the economic situation in Ireland. While the 2007 situation was characterised by a buoyancy of the Irish economy, Ireland is now dealing with a severe economic downturn.

(770) The fact, as argued by Ryanair, that the Irish economic situation affects incumbents and new entrants equally, does not mean that it would not deter potential new entrants and would thus be irrelevant for the assessment of the question of whether entry is likely to eliminate the anti-competitive effects of the Transaction.

(771) As set out in Section 8.1, it is clear that passenger numbers, which is one of the important factors affecting the profitability that a potential entrant could achieve on a route are affected by the Irish economic downturn.

(772) Flybe, one of the upfront buyers identified by Ryanair, notes that due to the financial crisis, operations at Dublin are still down by 20-25%. It believes however that growth will slowly return to Dublin also due to the increase of growth in the aviation sector overall and it indicates that Dublin is attractive in terms of air passenger duty charges which are lower than for example in the United Kingdom. (778)

(773) The results of the market investigation showed however that many other airlines still consider the Irish market as an unattractive market and competitors referred to the economic situation and drop in passenger numbers. (779)

(774) British Airways refers to the "general economic climate and the weak demand" which mean that "BA and IB can make better returns from their assets being deployed on routes other than those in Annex 1". (780)

(775) Vueling indicates that the current economic situation is worse than in 2007 and it does not expect that demand will grow in either Ireland or Spain due to the economic situation of both countries. It also believes that it would not be possible to operate a profitable operation at Dublin. (781)

(776) Aer Arann considers that the Irish market has suffered a lot as a result of the economic crisis. However, Aer Arann notes that in 2011 average fares started to rise for the Irish to United Kingdom routes. In Aer Arann's views, the factor that limits the attractiveness of the Irish market is the stalling economic growth. (782)

(777) easyJet considers that "the overall economic situation in Ireland does not make the market attractive either. Ireland's economic outlook does not look great. easyJet was already not very interested in Ireland even pre-2008 and certainly now that the economic situation has gone worse". (783) For this reason and because of the strong presence of Ryanair on the Irish market, easyJet does not consider Ireland to be a market that is worth entering. It is therefore clear that easyJet does not share Ryanair's view that the weak economy would serve as a springboard and that it would represent an opportunity for future growth.

(778) Ryanair itself indicates that a weak economy ensures that airlines have an incentive to lower prices in order to increase consumer demand. However, the prospect of lower prices, especially on routes already served by Europe's largest low cost carrier, is not likely to incentivise entry, rather to the contrary.

(779) As set out in Section 8.1, Ireland has suffered a large reduction in economic output in recent years. This has translated into a contraction of air passenger movements to and from Ireland (which according to Ryanair are further explained by the introduction by the Irish Government of air travel tax in March 2009 or the doubling of airport charges at Dublin airport since 2007). In Ireland, the economic outlook for the medium term projects a continuation of modest GDP growth over the 2012-2015 period, while domestic demand is set to make a negative contribution to growth (of about 1 percentage point) over the period as the adjustment in the domestic economy continues. At the same time, a number of other Member States such as Spain are also suffering from an economic downturn. Such economic circumstances do not favour entry on routes to and from these countries. Moreover, other circumstances referred to in the 2007 Decision with respect to the attractiveness of the Irish market remain unchanged.

(780) Finally, the mere fact that upfront buyers would be willing to enter routes to and from Dublin does not contradict the finding that the Irish market is not considered by many competitors as an attractive market. Any such entry by upfront buyers would take place in the context of the proposed Commitments packages.

Conclusion

(781) In the light of the above it is concluded that the Irish market is not considered by many competitors as an attractive market. The economic downturn in Ireland, but also in other Member States such as Spain, has further deteriorated the attractiveness of the Irish market compared to in 2007.

8.5.2.9. Airport congestion

(782) Different degrees of capacity constraints exist at relevant airports. These constraints could make entry or expansion by airlines on routes to/from these airports more difficult. This is because in order to operate a flight to/ from a given airport an airline needs access to airport facilities, such as a runway, stands, and terminals.

(783) In the 2007 Decision the Commission concluded with respect to Dublin airport that potential entrants were deterred by congestion problems at certain peak times which were not likely to disappear in the near future. (784) The Commission considered that there were runway congestion problems during peak hours (785) and that there was insufficient (contact-) stand availability (786). It did however disregard congestion problems at terminal level as a potential entry barrier at Dublin airport. (787) The Commission further considered that at many of the airports at the other end of the route (such as London Heathrow / Gatwick or Frankfurt), congestion problems were even more serious than in Ireland since some of these airports were congested throughout the whole day. (788) The Commission finally concluded that airport congestion did play a role as a deterrent factor for potential entrants, despite Ryanair's argument that airport congestion plays a more limited role than in other cases since many of the routes relate to holiday or leisure routes. (789)

(784) The Commission's findings were upheld by the General Court in the 2010 Judgement (790).

(785) In the present case, Ryanair disagrees that there are capacity constraints at Dublin airport. It argues that the congestion problem indicated in the 2007 Decision was eliminated and that Dublin airport is currently operating at just over 50% capacity. Ryanair also argues that there are no congestion problems at other Irish airports or at the destination airports.

(786) The Commission has analysed to what extent the conclusions of the 2007 Decision regarding airport congestion remain valid and whether and to what extent airport congestion constitutes an entry barrier in the present case.

Airport congestion at Dublin airport

1. Runway congestion

- Introduction

(787) The right to use the runway for a departure or arrival of an aircraft is ensured through the allocation of a 'slot' to a particular airline. Runway capacity constraints together with security requirements regarding the time interval between aircraft movements put a limit on the maximum number of available slots during a given time interval, which might be less than the slots demanded by airlines. In such a case a mechanism to allocate slots is needed and, according to the level of congestion, different rules for slot allocation apply. Airports with limited level of congestion are designated as schedules facilitated airports and operations are facilitated by a scheduled facilitator appointed by airlines on the basis of voluntary cooperation. Highly congested airports are designated as coordinated and slots there are allocated by a slotcoordinator.

(788) The Commission has found in previous cases that the limited availability of slots can be an important barrier to entry for potential competitors. (791) Without access to a sufficient number of slots at commercially attractive times throughout the day, a new entrant is often unable to operate a regular service on a sufficient scale, and as a result, would not be able to provide an effective competitive constraint on the merged entity. Landing and take-off slots are therefore critical inputs for any entrant wishing to operate a new service or expanding an existing service to or from Ireland.

(789) The results of the market investigation show that early morning peak hour slots are considered by competitors as important for the viability of potential entry plans at Dublin airport (792). Respondents point out that these slots are of particular significance for business travel, which is identified as more time sensitive, as well as to reach connections in the United Kingdom and Continental Europe. For example, according to Lufthansa: "It is important for business passengers and for connecting traffic to have a slot in the morning at DUB" (793) and Alitalia: "It's important both for local traffic which needs to move early in the morning and for connections" (794). Competitors mostly describe peak hour slots as being between 06.00 and 08.00 (795).

- Current situation at Dublin airport

(790) Dublin airport is a level 3 coordinated airport. It has been a coordinated airport since 2007, following a decision by the Commission for Aviation Regulation ("CAR"). (796) Prior to 2007, Dublin airport was a schedules facilitated airport.

(791) Asked how the limitations on terminal, aircraft parking stand, runway capacity and other limitations (as relevant) have evolved since Dublin airport's designation as a coordinated airport in March 2007, the CAR (797) indicated that "Overall, total runway capacity has increased slightly, especially between hours 10 am to 4 pm. Small capacity decreases can be observed between 5-6 am, 8-9am, 5-6pm and 8-9pm for the total numbers. In the critical morning hours between 6 and 7 am, runway capacity for departing passengers increased by one slot". (798)

(792) Ryanair claims that Dublin airport is a level 3 coordinated airport for historical reasons only and that it suffers no slot constraints even at peak hours either in early morning or late evening. (799) Accordingly, Ryanair has argued in recent years that since there is no capacity constraint at Dublin airport, it should no longer be Level 3 coordinated, but has been outvoted (800) at the DACC (801) meetings.

(793) Aer Lingus (including Aer Lingus Regional) and Ryanair hold up to 90% of slots allocated between 06.00 and 08.00 local time. (802)

- A number of slot requests have not been accommodated

(794) Airport Coordination Limited ("ACL") (803) indicates that Dublin airport faces slot constraints. (804) In the previous four IATA seasons, there were slot requests that could not be accommodated within a +/- 20 minutes time window due to limitations/constraints at the airport. (805) These requests were mainly related to early morning departure slots.

(795) DAA also submits that "Dublin Airport is runway constrained during the peak morning hours between 0600 and 0755hrs local time where initial demand (during the first stage of the slot filings process) from airlines to use the runway cannot be satisfied. This constraint is particularly acute in the 0630 to 0730 period." (806)

(796) From ACL's "Start of Season Report" for Dublin for the summer season 2012, it appears that Dublin airport operates close to, or at its maximum capacity in the morning peak from 05.00 to 07.00 UTC. (807) The Report also shows that the demand for departure slots in certain segments of the time period from 05.00 to 07.00 UTC exceeds maximum capacity. (808)

(797) Ryanair submits that airlines (including Ryanair) routinely over-request slots to facilitate route planning and schedule optimisation. It considers that it is both normal and irrelevant that some slot requests cannot be accommodated and that these requests are therefore not indicative of real or actual demand. (809) Ryanair also argues that the report shows that only 29 out of a total of 101 046 slot requests at Dublin airport for the entire summer season 2012 could not be facilitated within 15 minutes of the requested time. (810)

(798) The Commission acknowledges that the number of final slot requests may be lower than the number of initial slot requests. However, rather than attributing this difference solely to the fact, as Ryanair claims, that airlines routinely overstate their requests, the Commission considers that this difference also results from the fact that airlines limit their requests to those that they know could be accommodated. ACL indicates in this respect that "For the initial slot requests made by carriers, it is not uncommon that there is oversubscription in the 5:00-6:55 am (UTC) interval in the Summer Season and in the 6:10-7:00 am (UTC) interval in the Winter Seasons. Based on the current demand (2012) most of the requests however eventually can be satisfied within R10 (as some requests may drop out), also as carriers limit their requests to what they know would be available. Airlines typically do not necessarily get their optimal time for the morning departure wave." (811).

(799) The Commission considers that Dublin airport operates in the early morning peak hour at or close to its maximum capacity and that a limited number of slot requests cannot be accommodated. The Commission further considers that this might be a deterrent for competitors to request early morning peak slots.

- Early morning congestion exist despite the decline in traffic

(800) Ryanair also refers to the decline in traffic at Dublin airport over the past 5 years from 23.3 million passengers per annum ("MPPA") (2007) to 18.7 MPPA (2011) while the 2012 traffic, up to July 2012, has declined again by 1%. (812) It submits that Dublin airport is operating at 50% capacity. It further claims that the fact that Dublin airport used to process 30% more passengers and air traffic movements than it currently does, is conclusive and irrebuttable evidence that Dublin airport suffers no congestion issues whatsoever. (813)

(801) This decline in traffic is acknowledged by the Commission. However, as explained by the CAR, it appears that "while overall traffic has fallen in recent years, at peak times demand remains high relative to capacity." (814) CAR's most recent annual report shows that while overall traffic has fallen in recent years, at peak times demand remains high relative to capacity. (815)

(802) Figure 4 shows how the number of movements by scheduled hour of departure has evolved since 2007 up to 2012.

Figure 4 - Movements by Scheduled Hour of Departure

<Graphic 4>

Source: DAA (816)

(803) Figure 4 shows that there has been little change in the number of departure movements during the peak hours (between 06.00 and 07.59) since 2007 despite a decline in overall movements at Dublin airport. While there has been a 23% decline in total departure movements since 2007, in the peak hours from 06.00 to 07.59 there has only been a 9% decline. Moreover, while in 2007 17% of all departures left between 06.00 and 07.59, in 2012 21% of all departures left during this peak time. The situation is even more pronounced during the months of July and August, where there has only been a 3% decline in weekly departures since 2007 during the 06.00 to 07.59 peak time (817).

(804) Therefore, the Commission considers that early morning congestion exists despite the decline in passenger traffic at Dublin airport over recent years. - It is unlikely that the maximum limit of peak morning runway movements will be increased in the foreseeable future.

(805) Ryanair submits in this respect that ACL and the DAA appear to unnecessarily constrain, or artificially understate peak, morning runway movements at Dublin to 39 per hour compared to the declared peak of 48 movements per hour during the 18.00 to 18.55 hour period. Moreover, according to Ryanair, an efficient single runway like Dublin should and could operate at a maximum limit of 51 (as is the case at London Gatwick). (818)

(806) Regarding the question why runway capacity is lower at peak time than at later times in the day, DAA explains that "it is not possible to operate the airport at peak conditions throughout the day due to the mixed profile of activity during the day (i.e. arrival peaks, departure peaks and mixed traffic peaks represent different characteristics)." (819) Given that departures can follow each other more quickly than arrivals, allocations between arrival and departure slots are flexed to ensure that more departures are allowed when departure capacity is required, reducing the number of arrivals to compensate, and more arrivals are allowed when arrival capacity is required, reducing the number of departures to compensate. Therefore, the declared capacities for departures and arrivals are different at different times of day. (820)

(807) Asked whether there is room to increase the total movement capacity at peak time at Dublin airport, ACL indicated that "The total capacity figures at any one time at Dublin are based on the available infrastructure. Changes to the capacity are based on a wish list combining previous demand, potential growth of carriers and significant planned retimes. NATS regularly performs simulations for DAA to test the impact of increased capacity. These simulations show that it is not possible to increase the number of movements in the peak early morning hours (between 05:00 and 06:55 UTC) without exceeding the 10 minute delay criteria which is the agreed limit at Dublin Airport". (821)

(808) Asked whether Dublin airport is comparable, for example concerning the total peak hour flights, to other airports having a single runway, such as Stansted or Gatwick, ACL explained that "it would be impossible to compare two airports only by looking at the declared runway capacity in isolation. When determining capacity, while the runway capacity is one item, it is important to also consider other items like the infrastructure surrounding the runway and in general what space is available and how it its utilised. For example, the facilities at Gatwick airport are such that the space around the runway can be maximised; for example there is a greater staging area which enable optimum delivery of departing aircraft to the runway and more rapid exits for arrival aircraft. Further aspects which are also relevant in addition to the organisation of the movements are the air space constraints. These might render a different picture on the runway capacity at two airports, even if they both have only one single runway." (822)

(809) DAA highlighted that "other single runway airports, such as Gatwick, Luton, and Stansted are comparable to Dublin airport in their declared departure and arrival movement limits (for example 30-31 departure limit at peak hours of 05.00 to 06.55 am, UTC). However, in other airports the movements might be spread more equally throughout the day or there might be more movements overall depending on particular circumstances, infrastructure, ATC rules, departure flight paths, agreed separation distances etc." (823)

(810) Article 6 of Council Regulation (EEC) No 95/93 of 18 January 1993 on common rules for the allocation of slots at Community airports (824) shows that it is too simplistic to simply compare airports on the basis of the number of runways they have. A whole set of constraints need to be taken into account when determining the coordination parameters.

(811) In the light of the above, the Commission considers that it is unlikely that the maximum limit of peak morning runway movements would be increased in the foreseeable future.

- The second runway cannot be taken into account for the calculation of runway capacity.

(812) Runway capacity at Dublin airport is calculated on the basis of the main runway, namely runway 28. In practice, a second runway (a crosswind runway or runway 34) is also regularly being used. Nevertheless, this second runway is not taken into account for calculating the runway capacity.

(813) The second runway cannot be used under certain meteorological conditions. DAA declares that "the use of the cross runway during dual runway operation periods (i.e. runways 28 & 34) is restricted to those occasions when weather conditions allow airline flight crew to select the runway based on their aircraft operating limits in the context of wind/speed direction and condition of runway pavement (wet/dry)." (825)

This is confirmed by ACL who states that "it cannot be used at all times as its utilisation depends on weather conditions (such as wind)." (826)

(814) Also, the second runway is not used by all types of aircraft. The DAA indicates that aircraft that tend to use the runway have sector lengths less than approximately two hours duration. (827) ACL indicates that this runway is shorter and that it is up to each pilot to decide whether to use it or not. (828)

(815) In IATA Season Summer 2012, the second runway was used 62.7% of the days. (829)

(816) According to DAA "There is very little prospect that this runway will be taken into account for the declared capacity in the future. (...) The inclusion of Runway 34 in a capacity declaration would result in significant delays. Runway capacity experts NATS UNITED KINGDOM advised the Co-ordination Group as recently as 24th September 2012 that approximately 4 minutes extra delay would accrue with just 2 extra departure movements being added to the peak hour departure capacity and this delay would become exponential as extra departure movements are added. Delay to the first wave of departing flights could persist for a significant part of the day. NATS therefore concluded that dual runway operations [are] not a capacity enhancing tool." (830)

(817) The Irish Aviation Authority (the statutory body responsible for air traffic control) advised at the DAAC meeting that took place on 17 April 2012 that "it would not be prudent to base runway capacity on dual runway operations when it is not available 100% of the time." (831)

(818) Notwithstanding the fact that the second runway is regularly used in practise, because of the operational restrictions attached to it, it appears unlikely that it will be taken into account for the calculation of declared runway capacity in the foreseeable future and it is thus unlikely that this second runway would lead to an increase of the declared runway capacity in the near future.

- Plans for the construction of a new runway are in an early stage

(819) The DAA has plans to build a new northern parallel runway. The CAR has made an allowance for capital expenditure on a new runway to be included in the calculation of airport charges subject to a threshold for the years 2010-13 which would be activated if annual passenger numbers at Dublin airport exceeded 23.5 million per annum. This threshold has not been met. Passenger numbers in 2011 were 18.7 million. (832)

(820) DAA specified that the parallel runway project is under way but still at a very early stage (DAA plans to resubmit its planning permission, but the project, if successful, would not start until 2017 or 2018). (833)

(821) Therefore, it appears that plans for the construction of a new runway are at a very early stage and will not lead to an increase of the declared runway capacity in the foreseeable future.

- Views of competitors

(822) Respondents to the market investigation indicate frequently that the availability of slots would be a significant barrier to entry at Dublin airport.

(823) Ryanair claims to the contrary that the evidence resulting from the market investigation shows that there are no capacity constraints at Dublin airport. (834) Ryanair also claims that the fact that a number of carriers expressly decided not to identify the availability of slots as a barrier to entry at Dublin airport (when the option was clearly presented to them at question 33 of the Commission's Questionnaire - Q1) is strong evidence that no such barrier exists. (835)

(824) A number of carriers did not identify the availability of slots as a barrier to entry at Dublin airport. Nevertheless, the fact that such carriers did not identify slot constraints as a barrier to entry does not necessarily mean that there are no slot constraints. It must thereby be noted that a carrier that wishes to start operating one route to/from Dublin will have a different need for slots at Dublin than a carrier that wishes to start operating a number of routes out of Dublin. The first type of carrier will have a more limited need for slots at Dublin, while the operations of the second type of carrier would critically depend on obtaining a sufficient number of slots in the morning peak time.

(825) Also, carriers that operate a route to/from Dublin from their exiting hub or base (outside Ireland) will have a different need for slots than carriers that wish to establish a base at Dublin. For instance, given that their first morning flight will depart from their home base or hub, they do not necessarily need a very early morning departure slot at Dublin airport.

(826) Moreover, it is noted that Dublin airport is not slot constrained during the entire day, and slots are available outside peak hours. Should a carrier therefore be satisfied with a slot outside peak hours, it would not be facing slot constraints. (836)

(827) Aer Lingus indicates in this respect that the availability of slots outside of the morning peak hour would not help a prospective entrant which intends to operate a Dublin base with a substantial number of aircraft and that were its aircraft unable to leave in the first morning departure wave, the operation of a Dublin base would not be economically sustainable. By way of illustration, Aer Lingus indicated that in the summer season 2012 all of its short-haul aircraft based at Dublin airport (apart from its standby aircraft) left the base before the end of the morning peak.

- Conclusion

(828) Therefore, the Commission concludes that there is runway congestion at Dublin airport in morning peak hours. This congestion would in particular affect new entrants which would like to operate a base on several routes ex- Dublin, and would contribute to the already high barriers to entry.

2. Parking stands

(829) In the 2007 Decision the Commission considered that there was insufficient stand capacity at Dublin airport. (837)

(830) Since 2007, a number of changes took place at Dublin airport with respect to parking stand availability. The main change was the construction of Terminal 2, Piers D and E and some minor changes in stand layouts around Pier A and Pier B. (838)

(831) In order to manage the demand for contact stands (839), formal Stand Allocation Rules were instituted in 2002 and updated in 2006 and 2011. These confirm that "Stands will be allocated to ensure efficient usage of airport infrastructure. In particular the usage of airbridge-served and walk-in walk-out Contact Stands will be maximised". (840)

(832) The responses to the market investigation show that airlines generally consider it important to have access to contact stands at Dublin airport in the early morning peak hours for European short-haul or mid-haul routes. (841)

(833) In theory, pier served contact stands may be supplemented by use of remote stands. (842) Nevertheless, the majority of competitors would not be willing to use remote stands or would only be willing to do so under certain circumstances. (843)

(834) According to DAA, despite the changes that took place since 2007, there are nevertheless still constraints with regard to parking stand capacity at Dublin airport at the early morning peak hours. (844) DAA submits that Pier served parking stands in piers A, B and D are constrained during the first wave of departures in the morning (06.00 to 07.30). All of the piers (A, B, D and E) are constrained from 22.00 to 06.00 due to aircraft parking on stands overnight for the first wave departures. (845)

(835) Ryanair submits to the contrary that pier served and remote parking stands are freely available during the early morning peak period.

(836) The market investigation showed mixed results with respect to the availability of parking stands at Dublin airport. (846) Several competitors indicate that contact stands are restricted to some extent in their availability in the morning peak. (847) Air France on the other hand, while indicating that contact stands are available in early morning peak hours only in Pier A, indicates at the same time that "(...) stand capacity , which was a huge issue in 2007 and therefore a significant obstacle (...) is no longer relevant (...)". (848)

(837) Data provided by the DAA show daily contact stand usage for the month of July 2012, at 06.00 (local time). (849) These data show that while there is a maximum of 57 contact stands available (in narrow body equivalents) usage is often at 55 or more and never drops below 50. DAA also presented data on stand usage by pier for 6 July 2012, a day which they consider represents a typical day. These data show that at 06.00 local time, Piers E, B and D are 100% occupied while at Pier A 9 out of 11 stands were used.

(838) Ryanair on the other hand produced 'dynamic' data (based on a 35 minute turnaround time plus a 10 minute buffer) showing that 88% of pier-served stands and 92% of remote stands are available for departures and turnaround between 06.30 and 08.30 in summer 2012. Ryanair also argues that there is substantial stand availability at Dublin airport in morning peak hours from 05.00 to 07.00, where 28% of stands remain available. Ryanair, when asked to provide utilisation ratio of contact stands by pier (for piers A, B, D and E) at various points in time in the morning (that is to say, at 04.00 , at 05.00 and at 06.00) on a typical weekday of the summer season, argued that stand utilisation rates cannot be assessed with reference to "a precise moment". (850)

(839) The dynamic argument put forward by Ryanair is not without substance. Airlines could for example be willing to use remote stands or after their passengers have disembarked, the aircraft can be towed to remote stands and later towed back to contact stands.

(840) Nevertheless, the data on static stand capacity (851) show that in early morning peak hours there are only a very limited number of contact stands available. Despite the fact that there appear to be some stands available at Pier A in the morning, it must be noted that stands are not easily interchangeable for airlines which cluster their activities in particular terminals and piers. Moreover, DAA indicates that Pier A is regarded as unsuitable by some airlines because arriving and departing passengers are not segregated, because there are some constraints as regards aircraft size, because it is not air-bridge-served and because it is the oldest pier. (852)

(841) It may thus be difficult for a competitor starting up a route from Dublin airport to have access to a contact stand in early morning peak hours and as indicated above the results to the market investigation show that airlines generally consider it important to have access to contact stands in the early morning peak hours for European shorthaul or mid-haul routes.

(842) While the DAA does not exclude the technical possibility in the current situation for an airline to enter Dublin airport with a small base or without a base, it indicates that this would nevertheless require some flexibility from the airlines already present (and it considers that this would be more difficult if there is one big intransigent operator such as Ryanair or the merged entity). (853)

(843) Therefore the Commission considers that competiting airlines may face some difficulties in having access to contact stands for early morning peak hour flights, which can contribute to the already high barriers to entry.

3. Terminal constraints

(844) As regards terminal capacity, there has been a recent major change, namely the construction of a new terminal, Terminal 2 ("T2"). Terminal 2 opened for its first full season in Sumer 2011 and this resulted in a substantially improved terminal capacity. Terminal 1 (and some of its piers) have been extended and modernized in recent years. (854)

(845) Terminals 1 and 2 are both classified as IATA level of service 'C' and offer equivalent services with equivalent service levels. (855)

(846) All long-haul services at Dublin airport are concentrated in Terminal 2 which has a United States Customs and Border Protection pre-clearance facility. Aer Lingus' short-haul and long-haul operations are located in Terminal 2 but some of its flights depart from Pier B (which is connected to Terminal 1 but linked to Terminal 2 via a walkway) during peak times. In addition, all Aer Lingus Regional services currently depart from Pier A which is linked to Terminal 1.

(847) DAA submits that there are constraints with regard to check-in desk capacity at Dublin airport in the early morning peak hours. (856) In particular, DAA submits that Terminal 2 check- in Desks 1 to 28 are currently fully utilised from 06.00 to 10.15 and the remaining desks 29 to 56 are full from 04.00 to 07.00. At the same time DAA is investigating a range of options to increase terminal operational capacity and reduce constraints over the next few years. At the same time, DAA nevertheless acknowledges that there are no constraints in Terminal 1. (857)

(848) Ryanair claims that the question of whether Terminal 2 is constrained is misleading and irrelevant in view of the fact that Terminal 1 check-in desks are less than 25% utilized during that period. It further submits that its own data on Terminal 1 and Terminal 2 check-in desk utilisation for 05.00 , 06.00 , and 07.00 on September 11, 2012, show that during these times 77%, 68%, and 70% of all Terminal 1 and Terminal 2 check-in desks are not utilized. (858)

(849) DAA itself does not contest that there is, and will be in the near future, spare capacity in Terminal 1.

(850) ACL, who monitors whether there are terminal capacity constraints as part of the coordination process, declared that "currently the terminal is not a constraining factor at Dublin". (859)

(851) The routes subject to this Decision are European short-haul routes for which United States Customs and Border Protection pre-clearance facilities are irrelevant. Therefore, in general both Terminal 1 and 2 should be suitable for operations on such routes. This is confirmed by the results of the market investigation (860) which show that only a few airlines consider Terminal 2 to be the only suitable terminal for their operations.

(852) Aer Lingus referred to the fact that all long-haul services operate from Terminal 2 where pre-clearance facilities are located and that its operations are therefore based in Terminal 2 as it would be inefficient to spread operations over two terminals (it being noted that some of its flights and Aer Lingus Regional services (861) continue to depart from piers which are linked to Terminal 1.) (862)

(853) Flybe considers that Terminal 2 would be the only suitable terminal for its operations because "the business traveller product is enhanced flying out of T2 and competition with Aer Lingus might be constrained for business oriented routes if services were developed using T1" (863).

(854) Given that the routes subject to this Decision are European-short-haul routes for which the US Customs and Border Protection pre-clearance facilities which are only available in Terminal 2 are not required, and given that there is spare capacity in Terminal 1, it can be considered that a new entrant would in principle find the necessary terminal capacity in Terminal 1. Constraints in Terminal 2 would constitute a barrier to entry only for entrants such as Flybe which would consider access to Terminal 2 as a pre-requisite for their operations, however such competitors represent a minority.

(855) Therefore, the Commission considers that terminal capacity constraints do not appear to be a barrier to entry relevant for the Transaction.

4. Conclusion on congestion at Dublin airport

(856) The Commission concludes that there is runway congestion at Dublin airport during morning peak hours and carriers may face some difficulties in having access to contact stands for early morning peak hour flights. This would contribute somewhat to the already high barriers to entry. The Commission concludes on the other hand that terminal capacity constraints do not appear to be a barrier to entry relevant for the Transaction.

Airport congestion at Shannon, Cork and Knock

(857) The airports of Cork and Shannon are not coordinated, their status is level 1. According to ACL, there are no capacity constraints at these airports.

(858) The operator of these airports, DAA, also indicated the absence of any constraints at these airports. According to DAA capacity problems would arise only if the traffic at Shannon were to reach around 5 MPPA. (864) The current passenger number at Shannon reaches 1.6 MPPA. (865)

(859) Knock airport is not coordinated, nevertheless it is stand restricted due to the fact that the airport has only 3 aircraft stands. Ryanair submits nevertheless that aircraft stand constraints at Knock Airport have not in any way impeded the airport's development and could be readily remedied by the construction of additional stands. (866)

(860) According to the slot coordinator, "[t]he airport authority has made applications to the Irish Department of Transport, Tourism and Sport for capital funding to increase its size so to permit additional stands and thus an increase in capacity at peak times" (867), which is currently under review.

(861) It therefore seems that, subject to the issue of aircraft parking stand constraints at Knock airport, there are no capacity constraints at the airports of Shannon, Cork and Knock.

Airport congestion at other airports

(862) Ryanair submits that there is spare capacity at one or more airports at each end of the overlap city pairs. (868)

(863) While it is acknowledged that secondary airports are generally not congested, this does not necessarily mean that the congestion at the main airport would not constitute a barrier to entry for potential new entrants. Indeed, this would depend on the willingness of the new entrant to operate on routes to and from secondary airports which is not an option for many airlines, in particular for those airlines whose service or operating model does not allow secondary airports to be used.

(864) Ryanair further submits that the fact that an airport is level 3 coordinated does not mean that such airport is constrained and refers notably to the airports of Warswaw Chopin, London Stansted, Barcelona El Pratt and Manchester which it claims are operating significantly below their maximum capacity and have ample room for growth. (869)

(865) The fact that an airport operates below its maximum capacity does not mean that that is the case at all times during the day, and it is quite possible that an airport, even though it operates in general well below its maximum capacity, faces constraints during peak hours.

(866) The following relevant airports are level 3 coordinated airports:

Table 32: Coordinated airports on the routes of concern

Warsaw Chopin / London Stansted / London Gatwick / London Heathrow / London City

Barcelona El Prat / Nice / Tenerife Norte Los Rodeos / Tenerife Sur Reina Sofia / Manchester

Paris CDG / Frankfurt / Munich / Berlin / Alicante

Malaga / Madrid / Palma / Ibiza / Rome Fiumicino

Rome Ciampino / Milan Linate / Milan Malpensa / Bergamo Orio al Serio / Gran Canaria

Brussels / Bilbao / Venice / Lanzarote / Fuertaventura

Stockholm Arlanda / Faro / Paris Orly

(867) The results of the market investigation show that airport congestion remains an issue at several other airports on the routes of concern.

(868) At Heathrow there are constraints on airline access to slots throughout the whole operating day. Also in LGW, demand exceeds capacity throughout most or all of the day. In London City, there are slot constraints during part of the day and in addition, there are a number of other limitations, for example not all aircraft models can use this airport.

(869) ACL indicates runway and terminal constraints for STN, LGW and LTN (870), in addition to the known congestion situation of LHR regarding lack of slots and runway, terminal and stand constraints. BAA however considers that STN airport generally does not have capacity constraints (871).

(870) Spanish airports may present constraints with regard to access to slots, in particular at peak hours, although due to the traffic decrease of recent years the situation has improved compared to 2007. Aeropuertos Españoles y Navegación Aérea ("AENA"), the Spanish slot coordinator, indicated that at all the spanish coordinated airports (Level 3) and at some of the schedules facilitated airports (Level 2) may be or may have been very recently some limitations/constraints for airlines to get access to slots. (872) These limitations would apply to all the Spanish airports on the routes of concern. However, it was specified that "...constraints present today much less difficulties than they presented during the preceding years" (873) and that "[i]n most cases, limitations/constraints only exist during "peak times" and/or "peak days" (874). The AENA indicates that "the situation of congestion at these airports, that used to be very important during the years before 2008, has changed dramatically and where it used to be a lack of capacity available now airlines have good choices to obtain slots (perhaps with the exception of Ibiza and Palma de Mallorca airports on peak days)" (875). Moreover, at several of the Spanish airports, plans would be developed to implement capacity expansion during the next 3 years.

(871) Congestion problems are present or may arise at the airports of Rome (876) (Fiumicino, Ciampino), Milan (877) (Linate, Malpensa, Bergamo), Frankfurt (878), Munich (879), Paris (Charles de Gaulle, Orly), Brussels (880), Stockholm Arlanda (881), Warsaw Chopin (882) and Manchester (883).

(872) Faro Airport is classified as a coordinated airport in IATA summer season however it appears that there are currently no constraints with regard to the capacity of Faro Airport (884).

(873) The same is the case for the Aeroport de la Cote d'Azur in Nice (885). In Germany, at Berlin Schonefeld, despite being designated a fully coordinated airport, there are, according to the airport authorities, no constraints (886). Moreover, the opening of a new airport is planned for spring 2013. For Venice, no constraints were identified by the slot coordinator Assoclearance. (887)

(874) The results of the market investigation show that competitors mentioned the availability of slots as a barrier to entry in particular at London Heathrow, London Gatwick, London City, Barcelona El Prat, Paris CDG, Frankfurt, Munich, Milan Linate, Amsterdam, Ibiza, Alicante, Madrid Barajas. (888)

(875) Therefore, the Commission concludes that some capacity constraints appear to exist at various airports at the other end of the routes of concern, especially at peak hours, constituting a possible barrier to entry and expansion post-Transaction.

(876) The level of congestion at the other end of each analysed route will be discussed individually as relevant on a case by case basis in the context of the analysis of the competitive situation on the respective routes.

8.5.2.10. The merged entity's position as largest user at Dublin, Shannon and Cork airports

(877) In the 2007 Decision, it was concluded that the merger would allow the merged entity to increase its weight in the consultation process for airport charges, airport facility allocation or expansion plans and that the merged entity's strong position would be likely to hinder further expansion by competitors. (889)

(878) The General Court upheld the Commission's findings on this point in the 2010 Judgement (890).

(879) Airlines are represented, at the airports where they operate, on a number of airport users committees and user consultation interfaces.

(880) At Dublin airport, given its status as coordinated airport, there is a coordination committee, the DACC. The DACC assists the coordinator in a consultative capacity and in particular advises on possibilities for increasing the capacity of the airport, improvements to traffic conditions, complaints about the facilitation of air traffic movements, guidelines for the facilitation of air traffic movements and on methods for monitoring air traffic movements' data quality. (891) Moreover, the DACC also agrees, by majority vote, the capacity limits at Dublin airport. (892)

(881) In Dublin, Shannon and Cork, there is an Airport Operators Committee ("AOC"). The AOC is primarily focussed on operational issues such as terminals and systems installed, conditions of apron, taxiways and runways, local air traffic and ground circulation, airport development progress from a user's point of view, access to airport and airport buildings. (893) It meets regularly to discuss these operational issues and to liaise with the airport authority on behalf of the members.

(882) A large majority of the competitors consider that the Transaction would provide the merged entity with advantages as regards the DACC and the AOC. (894) The market investigation, however, showed mixed views with respect to the question of whether the Transaction would have an impact on the entry or expansion opportunities due to the merged entity's influence in these bodies (895): For instance, Aer Lingus indicates that "In the case of a combined Ryanair/Aer Lingus entity, such an entity would be the single dominant voice in the various consultation bodies (...) and would be in a strong position to determine issues relating to the operation of and developments at the various airports." It further submits that the "strengthened position [within the DACC] would provide Ryanair with the opportunity to control the decision making on important issues such as slot availability and possibly ensure that there is limited slot availability for any potential new entrants" and that "There is no doubt therefore that a combined Ryanair/Aer Lingus entity would dominate the AOC and would have the greatest influence on operational decisions at the airport." (896) British Airways indicates that the merged entity would have 80% of the votes and "could have some influence over decisions in these meetings" (897). LOT considers that the merged entity would have a stronger voice. (898) However, Lufthansa considers that the Transaction would not have an impact on the extry or expansion opportunities due to the mergerd entity's influence in the DACC and AOC (899); Air France indicates that "No once there is capacity available and the EU Commission should deal with this issue in the remedies package that will be offered by the merging parties." (900).

(883) Ryanair argues that the merged entity will not be able to influence airport infrastructure investments at Irish airports to their advantage (901). It submits in this respect that Dublin airport is a government-owned monopoly with no competitor or substitutable airport.

(884) Ryanair also claims that it is factually untrue that the CAR has to take the views of users into account and that in recent years the views of users have been consistently overridden by binding directions from the Irish Minister for Transport. It refers for example, to its longstanding opposition to the development of Terminal 2 which it argues proceeded based on a binding directive by the Irish Minister for Transport issued to the CAR.

(885) It further underlines that the DACC is a consultative body, not a decision-making body and that any majority by the merged entity will therefore have no effect on runway capacity.

(886) It also underlines that the real decision making body at Dublin airport is the DAA, supported by binding legal directions issued by the Minister for Transport.

Dublin airport - decisions on airport charges, capital expenditure and infrastructure investments

(887) The maximum level of airport charges and the level of capital expenditure for Dublin airport is set by the CAR (902). Airport users are consulted in the process and their view is one of the factors the CAR considers when making its determination on airport charges. (903)

(888) The CAR submits that Ryanair and Aer Lingus have been historically the two users that were most likely to respond to its consultation on price reviews. The Transaction would consolidate the views of two currently separate users and so, all other things being equal, increase the weight of the merged entity in the CAR's assessment of the reasonable interests of current users of Dublin airport. (904) The CAR further indicates that "The merged entity might be expected to have a larger voice in capital expenditure consultations, which influence new investments made at the airport, such as runways, terminals or other aeronautical and commercial investments. In the end, this could influence the possibilities for entry or expansion at the airport." (905) DAA claims that the merged entity would have the ability to influence capital expenditure and the use and development of airport infrastructure (906).

(889) Ryanair and Aer Lingus currently are the two largest customers of DAA accounting for 82% of total traffic, while the remaining 18% are divided over 60 other airlines (907). The Transaction would thus combine the two airlines which are by far the largest airport users.

(890) Ryanair has often had views and expectations in the past with respect to airports and airport facilities that were different from those of other airlines (including from Aer Lingus). For example, Ryanair was opposed to the construction of the new Terminal 2.

(891) Such different views and expectations with respect to airports and airport facilities result in particular from the fact that Ryanair is a low-fares airline with a no-frills business model and thus seeks only basic, low cost, airport services, while many other airlines present at Dublin airport have a different business model for which they seek a higher service level.

(892) The fact that, in the words of Ryanair, Dublin airport is a government owned monopoly with no competitor or substitutable airport or that according to Ryanair the views of users at Dublin airport have been overridden by binding directions from the Irish Minister for Transport does not mean that the merged entity would not have leverage vis-à-vis the DAA or CAR. Ryanair has significantly cut its operations from Dublin airport in the past in response to a decision by the Irish Government to impose a tourist tax (908). The Transaction would even increase the merged entity's leverage and this would likely deter to some extent potential new entry at Dublin.

Dublin airport - Slot coordination issues

(893) In the past, there have been several examples of divergent views between Ryanair and the other airlines represented in the DACC. For example, Ryanair was strongly opposed to Dublin airport's classification as a Level 3 - coordinated airport, and it continues to be opposed to this status. (909) Also the voting pattern at the recent meeting of the DACC's Executive Committee showed that Ryanair took a different position than the other airlines as regards for example the runway capacity limits and the terminal limits (910).

(894) A large majority of competitors consider that the Transaction would provide the merged entity with advantages as regards the DACC (and the AOC). The market investigation showed however mixed views with respect to the question of whether the Transaction would have an impact on the entry or expansion opportunities due to the merged entity's influence in these bodies.

(895) The DACC is a consultative and not a decisional body. It nevertheless has an important role in assisting the airport's coordinator and advises on various issues such as possibilities for increasing the capacity of the airport, improvements to traffic conditions, complaints about the facilitation of air traffic movements, guidelines for the facilitation of air traffic movements and on methods for monitoring air traffic movements' data quality. Moreover, the DACC also agrees (by majority) the capacity limits at Dublin airport.

(896) As a result of the Transaction, the merged entity would have the majority of the voting rights within the DACC (911) while the remaining votes would be fragmented amongst other airlines with a much smaller presence at Dublin airport and amongst the airport operator, air traffic control and other organisations. (912)

(897) The Transaction would thus give the merged entity increased weight within the DACC and the merged entity, thanks to its strong position with a majority of voting rights, would be able to influence the decision making within the DACC. Nevertheless, because of the role of the slot coordinator and of the Irish State, the impact of such influence seems to be limited.

(898) According to the DAA, the merged entity could use its voting power within the DACC to adopt a proposal to reverse Dublin airport's status as a coordinated airport to a schedules facilitated airportv (913). However, the decision to change the airport's status from level 3 (coordinated airport) to level 2 (schedules facilitated airport) is a decision of the Irish State, not of the DACC. The Irish State must comply in this respect with the provisions of Regulation (EEC) No 95/93, in particular Article 3 of that Regulation. It therefore does not seem likely that the merged entity, even though it would gain weight in the DACC, would be in a position to influence a possible change in Dublin airport's status as coordinated airport.

(899) The DAA also claims that the merged entity could use its voting power within the DACC to ensure that slot allocation is made on the most favourable basis to it. (914) However, decisions regarding slot allocation are made by the slot coordinator ACL (915). ACL must act in an independent, neutral, non-discriminatory and transparent manner (916). It therefore does not seem likely that the merged entity, eventhough it would gain weight in the DACC, would be able to ensure that slot allocation by ACL is done in a manner favourable to it.

(900) Also, according to the DAA, Ryanair's use of slots will no longer be effectively controlled by the DACC and the DACC will not be able to prevent it from abusing the air traffic movements to the detriment of the passengers and the other air carriers. While the DACC shall make proposals to or advise ACL on the methods of monitoring the use of allocated slots (917), the monitoring of airlines' operations is nevertheless the responsibility of ACL. Pursuant to Article 4.6 of Regulation (EEC) No 95/93, the coordinator shall monitor the conformity of air carriers' operations with the slots allocated to them. ACL shall carry out the conformity checks in cooperation with the DAA and the air traffic control authorities. It therefore does not seem likely that the merged entity, even though it would gain weight in the DACC, would be in position to avoid control of the use of its slots.

(901) The above shows that while the Transaction would provide the merged entity with a majority of the voting rights within the DACC, which would give it more weight in the decision making within the DACC, because of the roles of the Irish State (for example, with respect to the status of the airport ) and of ACL (for example, with respect to slot allocation and monitoring) who must act in an independent, neutral, non-discriminatory and transparent manner, it does not seem likely that the merged entity would be able to influence to a significant extent slot issues at Dublin airport so as to deter potential new entrants or competitors with expansion plans.

Dublin Airport, Shannon Airport and Cork Airport - Operational issues

(902) As set out above, the AOC in Dublin, Shannon and Cork meets regularly to discuss operational issues and to liaise (on behalf of the members) with the airport authority.

(903) As set out above, a large majority of competitors consider that the Transaction would provide the merged entity with advantages as regards the DACC and the AOC. The market investigation showed, however, mixed views with respect to the question of whether the Transaction would have an impact on the entry or expansion opportunities due to the merged entity's influence in these bodies.

(904) While within the AOC, each member airline is entitled to one vote on each issue, motion or proposition placed at meetings (918), it is however, not unlikely that the merged entity will have an important say in the decision-making of the AOC. As set out above, the merged entity will be the largest airport user in Dublin, Shannon and Cork airports. It is not unlikely that this strong position would enable the merged entity to influence operational matters at the airports, possibly to its own advantage and to the detriment of competitors or new entrants. It is however unclear to what extent this in itself would deter potential new entry at Dublin, Cork and Shannon.

8.5.3. Conclusion on barriers to entry and expansion

(905) The barriers to entry with respect to the routes subject to this Decision are high.

(906) These barriers to entry relate primarily to Ryanair's and Aer Lingus' strong positions in Ireland.

(907) A new entrant would need to compete with the merged entity which would have a strong market position in Dublin, Cork and Shannon as a result of its established bases. Also, the merged entity would have the two strongest brands in Ireland and a new entrant would need considerable time and investment to upgrade its brand. Due to the announced dual branding strategy, the merged entity would have a strong market position in two different market segments, namely in the no-frills segment with the lowest fares as well as in the segment with higher service level. In addition, competitors would fear aggressive retaliation by the merged entity in case of entry. These considerations constitute significant barriers to entry which are equally relevant for routes to and from Dublin, Shannon, Cork and Knock.

(908) The entry barriers also derive from the specific situation at Dublin airport or in Ireland. The level of airport charges and taxes at Dublin airport is likely to deter new entrants, in particular low cost companies, from opening routes from or to Dublin. Also, the Irish market is not considered by many competitors as an attractive market and the economic downturn in Ireland but also in other Member States such as Spain has further deteriorated the attractiveness of the Irish market compared to the situation in 2007.

(909) Also, the fact that new entrants may face difficulties in obtaining early morning peak hour slots and parking stands at Dublin airport, coupled with the fact that there are capacity constraints at certain destination airports would contribute further to the already high barriers to entry. Finally, the merged entity's increased weight in the decision making process regarding the use and development of airport infrastructure at Dublin airport and operational matters at Dublin, Shannon and Cork airport could further deter to some extent potential new entry at Dublin, Cork and Shannon.

(910) The conclusion that the entry barriers in this case are high is corroborated by the fact that in recent years only limited entry events took place (by airlines other than Ryanair of Aer Lingus) on the routes subject to this Decision.

8.6. Forms of entry/expansion

(911) Entry or expansion may take three different forms: First, a possible entrant may establish itself with a base in Dublin, or at other Irish airports to compete on routes ex-Ireland. Secondly, an entrant may use a base which it has established at an airport at the other end of the route. Thirdly, a carrier could enter on a point-to-point basis without having a base at either end of the route (a non-base entrant).

8.6.1. Entry with a base in Ireland

(912) Ryanair and Aer Lingus are the only carriers which operate significant bases in Ireland, and the only other competitor which may be considered as having bases of some scale is Air France/CityJet which operates from Dublin airport with 3 lower capacity jets.

(913) While the Commission does not dismiss the possibility that airlines with a base at the other end of the route would in theory be in a position to constrain the merged entity to some extent (see Section 8.6.2) the results of the market investigation showed that competitors often considered that in order to be capable and have sufficient incentives to effectively constrain the merged entity post-Transaction on routes ex- Dublin (and ex-Shannon and ex-Cork), an air carrier would need to establish a base at such airports (919).

8.6.1.1. Increased flexibility, cost savings and advantages

(914) As set out in Section 8.4.3, operating with a base provides advantages such as increased flexibility in reacting to demand-side or supply-side shocks, cost savings arising from economies of scale and scope and cost advantages arising from economies of scope.

(915) The importance of economies of scale and scope arising from base presence at Dublin airport, as well as in Cork and Shannon airports emerges from competitors' responses to the Commission's market investigation. Several competitors confirmed the advantages of economies of scale and scope, increased operational flexibility and other advantages provided by base operations (920).

(916) In Aer Lingus' view, "Operating from a base allows an airline to spread costs over its various routes. These costs include ground crew and pilot/inflight crews, ground equipment, and maintenance efficiencies (staff, parts, facilities). Operating from a base means the same ground crew and equipment can be used for one flight or multiple flights in each shift, with benefits accruing for the airline as utilisation and productivity are maximised." (921)

(917) According to Lufthansa "an operation out of a base airport is always more favorable than an operation out of a non-base airport due to economies of scale and scope and additional operational flexibility." (922) Lufthansa further argues that "[..]the airline business is a business in which economies of scale play an important role. Against this background bases provide for higher economic efficiencies (...) In terms of comparing the effects of having a base with the effects of being a large company, LH contends that base effects have more impact as they influence a larger proportion of costs" (923).

(918) A base is needed inter alia to benefit from a cost position that would allow a competitor to constrain the merged entity. Lufthansa and Swiss indicate in this respect that "Only a carrier with an established base in Dublin would have the cost position and market power to constrain the merged entity in ex-Dublin routes." (924) A base also allows for a better utilisation of aircraft (925).

8.6.1.2. A base is a prerequisite for having sufficient operations

(919) A base is also a pre-requisite for having sufficient operations from a given airport. According to Austrian Airlines "Due to the larger number of routes the combined entity offers, only a (new) base airline could effectively constrain the merged entity." (926) Swiss indicates that in order to be an effective competitor, an airline has to have a base to enjoy the same benefits in terms of cost savings but also to offer a comparable product in terms of wide network portfolio with a competitive schedule which includes morning departure times (927).

8.6.1.3. Brand awareness

(920) A base and accordingly market presence is also important to have sufficient brand awareness at the base end of the route. According to Air Berlin, a new entrant should have a base because of the "Need of a certain size to compete in Dublin / brand awareness." (928) According to Lufthansa, "an airline needs to have a certain number of routes to increase customer awareness. The airline needs to be present on more than one or two destinations so it can attract customers, especially business customers". (929)

8.6.1.4. A base allows for early morning departures and late arrivals

(921) Moreover, a base allows the new entrant to offer early morning departures and late arrivals. Flybe indicates in this respect that "it would be essential to have a Dublin base to ensure the new entrant could serve the Dublin market with early departures and late arrivals." (930) It also indicates that while early morning departures can be achieved by a 'night-stopped' fleet, establishing a base brings benefits from cost efficiencies (931). In the same line, Vueling indicates that "A base allows having an early morning flight out, which is an important criterion for efficiency." (932) According to Aer Arann, a base is necessary because "(...) you need to be able to operate the morning slots to provide frequencies to the United Kingdom." (933) According to BA, "Having a base ensures a convenient slot portfolio which is a requirement to be a real competitive alternative." (934)

8.6.1.5. The size of the base

(922) The size of the base that a new entrant would require depends on a number of parameters, including inter alia the business model of the new entrant, the size of demand for flights to and from the airport and the level of operations of the merged entity (that is to say, more aircraft would be required for Dublin than for Shannon and Cork). Given the fact that Aer Lingus' and Ryanair's bases at Shannon and Cork airports are smaller, it is considered that an entrant could have a base with less aircraft there (935).

8.6.2. Entry from a base outside Ireland

(923) The results of the market investigation showed that some competitors considered that it could be sufficient for a new entrant to have a base at the destination airport (936).

(924) Loganair indicated that a financially sound operator with a strong home customer base and a powerful brand could mount a potential challenge on routes to Dublin, with the marketing expenditure to build some brand awareness in the Irish market. It further indicates that for example on the Dublin - London route, a United Kingdombased operator would be a strong competitor providing home market advantages for flights to Dublin (937). LOT indicates that "a new entrant needs a base or a hub at the other end of a relevant route to have connecting traffic. It will be very difficult to compete on point-to-point traffic with the merged Ryanair/Aer Lingus." (938) TAP goes on to indicate that "Being a relatively small point-to-point market, it is unlikely that any scheduled carrier will operate a route if it doesn't have a base on one of its ends." (939) It also indicates that there may be a competitive advantage at the other end of the route that surpasses the advantage at the Dublin end of the route. Alitalia makes a distinction according to the plans of the new entrant. It indicates that "In order to be an effective competitor on all routes, a carrier should establish a base at Dublin airport. In order to be a competitor on a few routes it is not necessary to do so, even if it should be convenient from an economic point of view (cost savings, economies of scale, etc...)." (940) Wizz Air indicates that it currently serves six intra- European routes from Cork (941) and that all these routes are served from the other end of the route. It indicates nevertheless that to operate longer flights (such as to Alicante) it would need a base in Ireland.

(925) The Commission does not dismiss the possibility that airlines with a base at the other end of the route would in theory be in a position to constrain the merged entity to a certain extent.

(926) However, as set out in Section 8.4.3.8, it is considered that competitors with a base at the destination airports are not equally close competitors. In particular, the constraint on the merged entity by an entrant with a base at the other end of the route would be limited inter alia for the following reasons:

(927) An airline that operates flights to Dublin, Cork and Shannon from a base outside Ireland would only offer a constraint on the merged entity on that specific route while Aer Lingus currently competes with Ryanair on a variety of routes out of Ireland.

(928) Moreover, an airline operating from a base outside of Ireland would likely only have a limited market presence in Ireland and would thus lack a strong brand in the Irish market, while it would often have a strong brand in its home base. This would be a disadvantage compared to Ryanair and Aer Lingus, which have strong brands in the Irish market (942).

(929) Also, given that airlines with a base at the other end of a route are likely to operate only one or very few routes from Ireland, they would be unlikely to show the same level of commitment to routes to and from Ireland.

(930) They would also not have the same ability and incentives to adjust capacity to react to shifts in demand as Ryanair and Aer Lingus have on routes ex Ireland, and if they do, this would typically be limited to the route or the small number of routes they currently operate to Ireland.

(931) Finally, network carriers operating flights to Ireland from their hub in their home country would likely be interested in connecting passengers to feed flights from their hub, while they would compete for point-to-point passengers only to a lesser extent.

(932) The 2010 Judgement already noted that Ryanair's arguments were unable to contradict the conclusion that "potential entrants with a base at the other end of a route face disadvantages." (943)

(933) It is thus unlikely that an airline with a base at the other end of the route would in general be an equally close competitor to the merged entity as Ryanair and Aer Lingus currently are.

8.6.3. Non base-related entry

(934) The need to have base operations in order to be an effective competitor to the Parties is also corroborated by the limited activities of competitors on non-base routes.

(935) As set out in the 2007 Decision, the most common option (944) of non-base point-topoint flights are so-called "W" or "triangular" flights. On such a W-flight, the aircraft first flies from its base to a non-base destination. Instead of returning to the base, the aircraft then flies one or several times to another non-base destination (for example a destination a shorter distance away to use the weak demand over lunchtime) during the day and only returns to the base in the evening (945).

(936) Out of a total of 1,318 routes in summer 2012, Ryanair operates 26 (946) routes on a "W" or triangular pattern (that is to say, less than 2%). These routes account for approximately 1.4% of Ryanair's flights in the summer 2012 (947). A paper prepared in November 2011 by RBB Economics at the request of Ryanair on the subject of "airline bargaining power at the airport of Stansted" shows that Ryanair considers that there are significant costs associated with operating such routes. In particular it is indicated that one of the main operational costs associated with this model relates to the difficulties of changing crews between rotations. Another cost derives from the fact that there is less scope to react when things go wrong, which can translate into significant delays and complications due to crew flight time limits. As a result of these significant costs, Ryanair operates W routes only as a strategic investment in building up potential base airports (948).

(937) In the merger case Olympic/Aegean, Ryanair stated: "It is Ryanair's policy to avoid routes that do not have a base at either end of the route, given that such routes introduce complexity to operations and increase the risk of so-called knock-on delays to the system." (949) (emphasis added)

(938) The results of the market investigation correspond to the above and show that competitors in general tend not to operate to a significant extent routes without a base at either end of the route.

(939) In particular while several competitors state that they operate on routes with no base at either side, (950) the proportion of routes and of passengers transported on routes without a base at either end exceeds 5% for only two airlines (Volotea and Wizz Air) (951).

(940) Companies typically operate on these routes only when it could not justify incurring the costs involved in establishing a base in either airport, for example where the market is niche or where there is only seasonal demand for the route.

(941) Wizz Air considers that W flights allow market development and testing (952).

(942) On the other hand it was indicated by carriers that do not operate routes in the Union without having a base at either end that operating on a route without a base at either end would lead to significantly higher costs. Reference is made to a number of factors, including increased operational risk whereby it is more difficult to handle unforeseen disruptions without a base, higher complexity and cost of crew changeover, higher costs because of the impossibility of self-supplying some of the related ground services or getting volume discounts and higher marketing and sales cost in general (953).

(943) Swiss explains that "Operating routes with no base at either end of the route result in higher production costs and operational complexity (954)". Also, Thomas Cook considers that there are two main disadvantages of having W flights: "they imply a significant crew cost and increase impact when delays occur (A delay on one leg causes delays on the subsequent legs and such delays are difficult to recover)." (955) SAS explains that "it is difficult to operate on a route without having a base at one end of the route." (956)

(944) Competitors did not consider that triangular or W flights would be an economically viable way for their company to enter routes to/from Dublin (957), Cork, Shannon or Knock (958). The results of the market investigation did not reveal that entry or entry on a sufficient scale could be expected through the use of triangular or W flights.

(945) The above shows that carriers use base airports and therefore that point-to-point entry on a non-base route is unlikely to be sufficient to constrain the merged entity. On the contrary, the market investigation showed that there are significant disadvantages attached to operating such flights and that triangular or W flights would not be an economically viable way for competitors to enter routes to or from Dublin, Cork, Shannon or Knock.

8.6.4. Conclusion

(946) Therefore, the Commission concludes that because of the various disadvantageous attached to it, non-base related entry is unlikely to sufficiently constrain the merged entity.

(947) Furthermore, the Commission acknowledges that while an entrant with a base outside Ireland might in theory be in a position to constrain the merged entity to some extent, such constraint would however be limited to the specific route or routes on which the competitor would enter, and the entrant would likely be a less close competitor to Ryanair and Aer Lingus (due to factors such as the limited market presence and brand awareness in Ireland, the lack of early morning and late evening flights, amongst others).

(948) The Commission considers that in order to act as an effective constraint of sufficient scale on the merged entity on the routes to and from Dublin, Cork and Shannon airports, a new entrant would need to establish a base at these airports, which would allow it to enjoy the advantages resulting therefrom, to have a sufficient market presence and brand awareness and to have an appropriate schedule with early morning departures and late arrivals.

8.7. Entry plans by actual and potential competitors

8.7.1. Analytical framework

(949) Since new entrants would be confronted, post-Transaction, with a significant number of barriers to entry in competition with the merged entity on the routes of concern, it is necessary to examine to what extent post-Transaction entry would constrain the merged entity.

(950) Ryanair claims that whether or not a new carrier commences service on an overlap route, the credible threat of entry is sufficient to offset any incentive that the merged entity might otherwise have to raise fares (959).

(951) British Airways indicates that "(...) it is often easy to switch assets from one route to another for network and p2p carriers providing there is a more profitable opportunity. Therefore the threat of entry is always a competitive discipline." (960) Flybe considered that "The industry has good visibility of acheived load factors and can assess competing carriers fare performance by route through web monitoring; this provides the mechansims for market entry in the event of 'super-profits' being achieved by incumbents." (961) However, a majority of competitors responding to the question consider that a mere possibility of entry (that is to say, without the threat of entry materialising) at Dublin, Cork, Shannon or Knock by competing airlines would not discipline the merged entity if the Transaction took place (962).

(952) For entry to be a sufficient competitive constraint on the merged entity, it must be shown to be likely, timely and sufficient in order to offset the anti-competitive effects of the Transaction.

(953) According to the 2010 Judgment, "the mere 'threat' of an entry, on which the applicant relies, is not sufficient [...] What counts is the prospect of an entrant which offsets the anti-competitive effects specifically established in the contested decision at that stage of the assessment" (963).

(954) Hence, the Commission must ensure that entry is not only a theoretical and remote possibility, but constitutes an immediate and actual threat to the merged entity. Entry must exert a competitive constraint on the merged entity to such an extent that the latter would refrain from any transaction-induced anti-competitive behaviour. It is not necessary to show that entry is impossible on a given route to dismiss the likelihood of entry as a countervailing factor. It is sufficient to establish that the degree of likelihood of Transaction-induced entry is low or that the scope of the expected entry is insufficient to counteract the anti-competitive effects of the Transaction in all relevant markets. This is applicable in particular in the present case, as Ryanair and Aer Lingus hold very large market shares and are considered as very close (if not closest) competitors on all relevant domestic markets.

(955) Furthermore, in order to be a countervailing factor, entry should compensate for the loss of competitive constraint that Ryanair and Aer Lingus would have exerted on each other in the absence of the Transaction. A selective entry whereby competing airlines either avoided competing with the merged entity or engaged in limited selection of particular routes would not compensate for the loss of the competitive constraint.

(956) In markets in which Ryanair and Aer Lingus reach very large market shares or even create monopoly, a remote likelihood of entry cannot be sufficient to dismiss the Commission's competition concerns. In such a situation, the significant impediment to effective competition cannot be dismissed without clear indications that actual or potential entry would constrain the merged entity's behaviour on relevant city-pairs.

8.7.2. General context

(957) The economic and financial crisis has affected Member States and in particular Ireland. The sluggish economic environment in Ireland and also in other member States relevant for the assessment of the Transaction, in particular Spain and the United Kingdom, does not appear as conducive to entry (964). Moreover, the results of the market investigation show that the Irish market is still considered by many carriers as an unattractive market. This constitutes a barrier to entry and it is against this background that any entry plans must be assessed (see in particular Sections 8.1 and 8.5.2.8).

8.7.3. Analysis of entry projects by actual and potential competitors

(958) In this context, the Commission analysed whether potential competitors have entry and expansion plans on an individual basis or an aggregate level, which would be sufficient to offset the anti-competitive effects of the Transaction on the routes of concern.

(959) Airlines constantly monitor their existing routes and explore possible opportunities on new routes that would complement their portfolio. For example, Alitialia indicates that its subsidiary Air One "would, in principle, be interested in operating flights on some of the routes indicates in the Commitments" (965). Such monitoring or general, remote and inconcrete interest in itself is not sufficient to establish that timely and likely entry of a carrier would take place on a given route. There are overall no indications that other airlines, wheter they be full service, low cost or charter airlines, have concrete and timely entry or expansion plans.

(960) It is worth noting in this regard that, in the context of the remedies discussions, IAG mentioned that, if it manages to acquire full control over Vueling, which it hopes to do in 2013, it "(...) would look across the Group for growth opportunities [...]*" (966). IAG was however not in a position to provide any specific details in this regard and indicated that the decision to enter a given route will depend on a number of factors, in particular the supply and demand features of the route and the return that could be expected (967). It moreover indicated that at least one additional condition for IAG to start operating with its own aircraft is that "current market and demand conditions in the Irish market would have to improve" (968) .

(961) In its response to the market investigation Vueling indicated that if the Transaction were to take place, it "would not enter on any of these routes. When Vueling analyses a possible entry on a route, it looks in particular at the existing competitors. Ryanair has an extreme low cost model. In addition, the merged entity will have a strong market position in Dublin where it will represent about 80% of total traffic. This will allow it to compete harsh with a new entrant by putting an interesting schedule and low prices. Vueling feels that it cannot compete with this and that Ryanair could push competitors out of the market". Vueling added that "even though being the second lowest cost carrier in Europe, (Vueling) finds it very difficult to compete with Ryanair where Ryanair has a strong position in a market. It considers it a very aggressive competitor" (969). Vueling also explained that it does not expect a growth in demand between Ireland and Spain and that it would not be possible to perform a profitable operation at Dublin Airport (970). It also indicated that while brand plays a significant role (971), its own brand is perceived as weak on Irish routes. Furthermore, in the context of the first market test, Vueling indicated that "As main Spanish operator in the short-mid haul, all the routes from/to Spain are under evaluation permanently. At this moment, we cannot set a timeframe for the launching of these routes, but the deicision will be based on profitability analysis" (972). Vueling clarified in this context that it "expressed a mere potential interest in operating said routes. This cannot be construed as any kind of commitment on our side to do so" (emphasis added) (973).

(962) To conclude, the Commission considers that the launch by IAG, including its subsidiaries of its own services on the routes subject to this Decision is unlikely and would in any event not take place in a timeframe relevant for the assessment of the Transaction.

(963) Therefore, actual or potential entry by another carrier is unlikely to sufficiently compensate for the substantial loss of competition between the Parties.

(964) It should also be noted that competing entry would have to occur on 32 overlap routes where Ryanair have bases at both ends of the routes. (974) Entry in such circumstances would appear even more difficult. On the other hand, the mere possibility of entry, would not discipline the merged entity according to the majority of competitors. (975)

8.7.4. Conclusion

(965) The Commission considers that the anti-competitive effects of the Transaction would not be sufficiently compensated for by timely entry on an isolated route or even buy expanded entry by carriers operating some routes to and from their home bases.

8.8. Treatment of routes exited by Ryanair and Aer Lingus after the announcement of the Transaction

(966) Given the dynamic nature of competition in the airline industry, Ryanair and Aer Lingus regularly open and close routes according to their performance and to their overall strategy. In particular, Ryanair and Aer Lingus exited several routes, on which they were competitors, since the public announcement of the Transaction in June 2012.

(967) The Commission analyses whether, for all routes, the exit of Ryanair or Aer Lingus was related to the Transaction ("Transaction specific"). If the exit of a given route is not found to be Transaction specific, that route is not be treated as an overlap between the Parties for the purpose of the present competitive assessment.

8.8.1. Routes exited by Ryanair following the announcement of the Transaction

(968) Following the failure to resolve a contractual dispute with Verona airport, Ryanair closed all flights from Verona on 13 October 2012 (976). Accordingly, it can be concluded that Ryanair's exit from the Dublin-Verona route is not Transactionspecific. As such, for the assessment of the Transaction, there is no competitive overlap between Ryanair and Aer Lingus on those routes.

(969) For the purposes of the assessment of the Transaction, the Commission concludes therefore that there is no actual competition between Ryanair and Aer Lingus on the Dublin-Verona route.

(970) As exit took place very recently, in a cautious approach, no potential competition issues will be considered on this route.

8.8.2. Routes exited by Aer Lingus following the announcement of the Transaction

(971) Having cited 'business reasons' as the basis of its decision, Aer Lingus is exiting the Dublin-Krakow, and Dublin-Vilnius (977) routes following the announcement of the Transaction.

(972) Accordingly, it can be concluded that Aer Lingus' exit from these four routes is not Transaction-specific. As such, for the purposes of the assessment of the Transaction, there is no competitive overlap between Ryanair and Aer Lingus on those routes.

(973) Therefore, for the purposes of the assessment of the Transaction, the Commission concludes that there is no actual competition between Ryanair and Aer Lingus on the Dublin-Krakow and Dublin-Vilnius routes.

(974) As exits took place very recently, the Commission concludes that no potential competition issues will arise on these routes.

8.9. Actual competition

8.9.1. Methodology

8.9.1.1. General approach

(975) The route-by-route analysis is based on the data for the following IATA seasons: summer 2011, winter 2011/2012, and summer 2012. In general, the Commission considers that as operating conditions may vary significantly between winter and summer seasons, in particular in terms of frequencies, analys must be conducted and conclusions must be drawn on a seasonal basis.

(976) Ryanair provided Market Information Data Tape ("MIDT") data. MIDT data is based on bookings through booking systems, known as Computerised Reservation Systems ("CRS") or Global Distribution Systems ("GDS") as traditionally used by travel agents. MIDT data thus does not include directs sales made by the airlines through their website. Further MIDT data does not include connecting passengers, but comprises only point-to-point passengers. Ryanair has further provided data from the Central Statistics Office ("CSO") (978). This data includes all airlines, both scheduled and charter, without distinguishing between connecting and non-connecting passengers. Ryanair provided also frequency data from OAG. With regard to weekly flight frequency, Ryanair provided the number of weekly flights for the respective first week of the following three months of a summer season: April, July and October, and the first week of January for the winter season (979). These figures are intended to represent a typical week of the summer peak period (July), the shoulder period (April/October) and the winter peak period (January). In the route-by-route analysis, the Commission has taken the frequency for the respective peak months of the seasons, namely January and July as a general benchmark for conducting its assessment.

(977) With regard to seat capacity, Ryanair provided the number of weekly seats for the respective first week of the following three months of a summer season: April, July and October, and the first week of January for the winter season (980). These figures are intended to represent a typical week in terms of capacity of the summer peak period (July), the shoulder period (April/October) and the winter peak period (January). Ryanair calculated market shares based on seat capacity on the basis of OAG data for January and July. In the route-by-route analysis, the Commission has taken the seat capacity for the respective peak months of the seasons, namely January and July as a general benchmark for conducting its assessment.

(978) In addition, Ryanair, in its comments regarding the Commission letter of facts of December 14, 2012, provided for a limited number of routes figures based on the capacity data from Innovata (981) and passenger numbers provided by the DAA for the summer season 2012. Innovata is a dataset which primarily focuces on data for scheduled carriers. Only some, but not all, charter airlines are included in its schedules. Innovata is based on planned schedules and is not updated to reflect actual operations following, for instance, last-minute cancellations or weather closures. Innovata information reflects the information available at the point in time that the schedule is run; any subsequent changes of schedules by airlines are not taken into account (982).

(979) The Commission regularly (983) endorses the use of MIDT data as the best available proxy to estimate market shares. Accordingly, the Commission has relied on the MIDT data for the purposes of market share analysis.

(980) However, Ryanair has contested the use of MIDT data (984). Ryanair has argued that as MIDT data only includes passengers that fly on the point-to-point route in question with no connection flights, the shares of network carriers on point-to-point routes was likely to be understated. Ryanair generally considers that frequency and capacity data are a better indicator of market position.

(981) In order to take into account Ryanair's arguments and, in addition, the fact that MIDT does not take into account direct sales made by the airlines through their website, the market shares computed on the basis of the MIDT data were cross-checked by the Commission with data provided by DAA, tracking the number of passengers flown on the respective routes. With regard to this data, a distinction between point-to-point passengers and connecting passengers was made in order to provide for true comparability. However, in the DAA data, the distinction between connecting and non-connecting passengers refers only to passengers that are making a connection at Dublin, Cork or Shannon (985). For this reason, the Commission has recomputed market shares on the basis of the total passengers on major airlines taking into account the share of connecting passengers for their routes ex-Dublin that were provided to the Commission for the routes from Dublin to Frankfurt, Munich, Madrid, Paris, London and Stockholm (986). When material discrepancies were detected for capacity and passenger data for IATA summer season 2011 and for IATA winter season 2011/2012, both the market shares provided by Ryanair (based on MIDT) and those based on the DAA data are indicated in the route-by-route analysis. For the IATA summer season 2012, the Commission has indicated market shares based on passengers relying on DAA's data, recomputed as explained above, in particular because Ryanair did not provide data on passengers carried by the other airlines active on all the overlapping routes of concern.

(982) As regards indirect services, the Commission relied on the MIDT data (987) and crosschecked its accuracy during the market investigation. Considering the small numbers of passengers using indirect services, the analysis shows that the Parties' combined market shares including indirect services would be similar to their combined market shares taking into account only direct services.

(983) The Commission also acknowledges that frequencies constitute a relevant aspect to be considered in the competition assessment of the overlap routes. For instance, frequencies are an indicator for the quality of the services offered, and of the targeted customers (that is to say, business passengers would in general require higher frequencies than leisure passengers).

(984) However, the Commission considers that it is the assessment of market shares based on point-to-point passengers that reflects most accurately the market power of the competitors on a certain route as they mirror the interest of passengers on that route and in the services provided by the chosen carrier.

(985) Moreover, the Commission has also considered market shares based on seat capacity to be an additional indicator for the competition assessment. With regard to market shares based on seat capacity, the Commission has taken into account DAA's data based on real seat capacity offered by airlines for the full relevant season. For that reason, DAA's data are more accurate than the data provided to Ryanair by OAG which merely stem from frequency data and thus does not capture variations within a season.

(986) In order to respect the confidentiality of DAA's data, market shares based both on seat capacity and on passengers have been replaced by the standard ranges used by the Commission. These ranges are adequate for the competition assessment.

(987) Finally, the Commission, in its assessment, also considered whether the different types of information available to it indicated the same outcome and corroborated each other. In that respect, market shares based on seat capacity may differ from market shares based on passengers, as the load factor may differ between carriers.

8.9.1.2. Charter services

(988) Ryanair has not determined seat capacity and passenger volumes for individual charter companies (988). Moreover, no meaningful distinction could be made by Ryanair between dry seats and other seats sold to tour operators and for package holidays.

(989) The DAA dataset reports the full passenger numbers and capacity data of carriers irrespective of whether the offered service is a "dry seat" or otherwise. Therefore, in the computation of market shares based on capacity and passengers for charter companies on the basis of DAA data, the Commission applied the very conservative estimate that 20% of the passengers and seats (989) attributed to charter companies are for the market for "dry-seats" on the basis of the responses of charter companies in question (990). This approach very likely overestimates the share of dry seats sold by charter companies on any given route.

(990) For routes where a charter company operates, market shares based on DAA data have been replaced by the standard ranges used by the Commission. These ranges are adequate for the competition assessment.

(991) Ryanair provided imprecise charter frequencies only for the "peak months" of January and July. On each route where according to Ryanair competition of charter services is relevant, DAA data regarding the precise level of frequencies of the Parties and of charter operators were gathered by the Commission.

8.9.2. Overview

(992) In general, the following 3 main groups of overlap routes with distinct characteristics can be distinguished:

- monopoly routes;

- routes where other scheduled carriers operate;

- routes where charter carriers operate.

(993) Even though these main groups could be further subdivided according to a number of criteria, they have common features.

8.9.3. No significant constraints exercised by indirect services

(994) Ryanair argues that depending on price and total travel time (including flight duration and connecting time), indirect flights can be a feasible alternative for passengers as they provide one-stop flights to the relevant destinations. Ryanair stressed that this assessment is in line with the Commission's previous Decisions, and Ryanair considers that this assessment applies equally to all routes where the duration of a direct flight is around 3 hours or more (991).

(995) According to Ryanair (992), the ultra-low fares that it offers make these alternative services less attractive to Ryanair's typical consumer base than direct routes, but they nonetheless are commonly offered to and purchased by travellers as an alternative to direct flights, and thus will continue to represent a constraint on Ryanair and the merged entity.

(996) However, the Commission considers that this argument is not supported by the evidence collected during the market investigation.

(997) Ryanair and Aer Lingus both offer point-to-point scheduled air transport services. Neither of them is member of an airline alliance and Ryanair does not have interlining agreements with any other airline. Indirect services are not a matter of business focus for them.

(998) [Ryanair's data and monitoring practices on indirect services]* (993).

(999) Aer Lingus does not monitor indirect services [...]*. According to Aer Lingus, indirect flight availability between two cities is not taken into account in Aer Lingus yield management operations when setting the prices for its own direct flights between the two cities (994). Aer Lingus "does not consider the possibility of connecting traffic to be a competitive constraint on the pricing of its own direct flights and believes that essentially no passengers would substitute to indirect flights in the face of an increase in the price of Aer Lingus' or Ryanair's direct flights. As a matter of more general planning, if Aer Lingus identifies a significant volume of passengers following an indirect itinerary, this would indicate unmet customer demand and may prompt it to add capacity on the direct route or launch a direct route; but third party prices on the indirect routing are not among the competitive parameters used by the yield management team in setting Aer Lingus prices. Indeed the Aer Lingus team does not monitor pricing on indirect services on any of the overlap routes using Sabre or QL2 or otherwise." (995)

(1000) Furthermore, there are only 11 routes where Ryanair argues specifically that the merged entity would be constrained by indirect competitors: Dublin-Budapest, Dublin-Fuerteventura, Dublin-Gran Canaria, Dublin-Lanzarote, Dublin-Malaga, Dublin-Palma, Dublin-Rome, Dublin-Tenerife, Dublin-Warsaw/Modlin, Cork- Lanzarote and Cork-Tenerife (996). As regards, Cork-Lanzarote and Cork-Tenerife, the data provided by Ryanair (997) does not show evidence of any indirect passengers. On 8 of these routes (Dublin-Budapest, Dublin-Fuerteventura, Dublin-Gran Canaria, Dublin-Lanzarote, Dublin-Malaga, Dublin-Rome, Dublin-Tenerife and Dublin- Warsaw Modlin), the percentage represented by indirect passengers is limited, namely below [5-10]*%. The Dublin-Palma route is, according to the data provided by Ryanair, the only route where indirect passengers would represent [10-20]*% of all passengers. However, this estimate of [10-20]*% is likely not be fully accurate as in particular it only concerns the IATA winter season 2011/2012 (compared to [0- 5]*% for the IATA summer season 2011) and corresponds to only 326 passengers. In addition, further data provided by British Airways show no indirect passengers on this particular route (998). Furthermore, out of these 11 routes, the British Airways data mention indirect passengers for only the Dublin-Budapest, Dublin-Rome and Dublin- Warsaw/Modlin routes with a percentage of [0-5]*% or less for each of these 3 routes (999). More generally, it is also British Airways' and Iberia's view that for the Dublin-Budapest, Dublin-Fuerteventura, Dublin-Gran Canaria, Dublin-Lanzarote, Dublin-Malaga, Dublin-Palma, Dublin-Rome, Dublin-Tenerife, and Dublin- Warsaw/Modlin, and for European short-haul routes ex-Ireland in general, indirect flights do not constrain direct flights (1000).

(1001) In addition, as indicated by Ryanair (1001), travel flight duration for indirect services is much longer than direct flights on these 11 routes. These flights present the disadvantage of being longer as a result of stopovers. Taking the Dublin-Budapest route as an example, a direct flight (outbound) lasts 2 hours and 50 minutes with Ryanair and 2 hours and 55 minutes with Aer Lingus. An indirect flight (outbound) operated by Air France through CDG lasts 7 hours and 25 minutes (with a stopover of 3 hours and 25 minutes) or, at the minimum, 4 hours and 25 minutes with Lufthansa via Frankfurt (with a stopover of 45 minutes (1002)). As another example, the same feature applies to the Dublin-Gran Canaria route, where an Iberia flight (outbound) through Madrid lasts 11 hours and 15 minutes with a stopover time of 5 hours and 50 minutes. Ryanair's direct flight lasts 3 hours and 25 minutes while Aer Lingus' direct flight takes 3 hours and 30 minutes.

(1002) Indirect flights appear also to be more expensive. According to Ryanair (1003), as an example, the price for a Dublin-Budapest direct flight with Ryanair is EUR 74.98, compared to a direct flight with Aer Lingus at EUR 123, whereas the prices of the other scheduled carriers with indirect services range from EUR 212.90 for Lufthansa to EUR 570.75 for Air France. On the Dublin-Warsaw/Modlin route, the fares range from EUR 51.98 for a direct flight with Ryanair, EUR 113.00 for a direct flight with Aer Lingus to and indirect flight with LOT/Lufthansa with the highest fares at EUR 404.88. A similar situation occurs also on the Dublin-Gran Canaria route: Ryanair's ticket is EUR 152.98 for a direct flight, Aer Lingus' is EUR 182.38 for a direct flight and Iberia's ticket is EUR 415.90 for an indirect flight (1004). Spot checks confirm such finding: on the Dublin-Budapest route (1005) and the Dublin-Gran Canaria route (1006): direct flights are much cheaper than indirect flights. This is further confirmed by Aer Lingus: "due to the intense competition between Aer Lingus and Ryanair, fares are low and Aer Lingus believes that the prices of indirect flights would likely significantly exceed those of direct flights, particularly given the cost disadvantages faced by an airline in offering indirect flights (including multiple sets of airport charges, ground-handling, etc.)" (1007).

(1003) Moreover, as emphasised by Aer Lingus (1008), the inconvenience of connecting through a hub airport would be particularly significant for families with children, as well as other categories of passengers such as the elderly for whom the experience of connecting flights may be particularly burdensome. This is likely to be a relevant factor on leisure routes such as the Dublin-Fuerteventura, Dublin-Tenerife, Dublin- Gran Canaria, Dublin-Lanzarote, Dublin-Malaga, and Dublin-Palma routes.

(1004) On relatively short flights such as the Dublin-Budapest, Dublin-Rome, Dublin- Warsaw/Modlin, Dublin-Malaga, and Dublin-Palma routes, all of which are below 3.5 hours, the additional time required for an indirect flight will be particularly significant in proportion to the travel time of the direct flight.

(1005) It is therefore very unlikely that a customer would find an indirect flight more convenient than any direct flight offered by Aer Lingus or Ryanair (1009).

(1006) To conclude, the Commission considers that it is unlikely that indirect flights would be able to sufficiently constrain the merged entity in its market behaviour, especially concerning fare setting.

8.9.4. Monopoly routes

(1007) The Transaction would create a monopoly on 28 routes (1010): Dublin-Alicante/Murcia; Dublin-Berlin; Dublin-Bilbao/Santander; Dublin-Birmingham/East Midlands; Dublin-Brussels/Charleroi; Dublin-Budapest; Dublin-Edinburgh/Glasgow; Dublin- Fuerteventura; Dublin-Glasgow /Prestwick; Dublin-Manchester/Liverpool/Leeds; Dublin-Marseille; Dublin-Milan/Bergamo; Dublin-Nice; Dublin-Rome Fiumicino/Ciampino; Dublin-Tenerife; Dublin-Toulouse/Carcassonne; Dublin- Venice/Treviso; Dublin-Vienna/Bratislava; Dublin-Warsaw/Modlin; Cork- Alicante/Murcia; Cork-Faro; Cork-London; Cork-Manchester/Liverpool; Cork- Tenerife; Knock-Birmingham/East Midlands; Knock-London; Shannon- Manchester/Liverpool; and Shannon-London.

(1008) Ryanair disagrees with the fact that Aer Lingus is its closest competitor on the basis that the product offering of Ryanair and Aer Lingus would be significantly differentiated in terms of fares, services and airports served (1011).

(1009) However, the concept of closeness of competition is of limited added value for routes on which from the outset only two competitors are active, since both competitors in a duopoly are by definition each other's closest competitors (1012).

(1010) On most of these routes, Ryanair and Aer Lingus operate all year round. Only 5 of these monopoly routes (Dublin-Bilbao/Santander; Dublin-Marseille; Dublin- Venice/Treviso; Cork-Alicante and Cork-Faro) can be considered as seasonal.

(1011) Despite Ryanair's contention to the contrary (1013), the Commission does not consider the Cork-Tenerife route as seasonal as both parties have operated during the IATA summer seasons 2011 and 2012 as well as during the IATA winter season 2011/2012 (1014). The Commission also does not consider the Dublin-Budapest route as seasonal as the number of passengers is significant enough on this route in both seasons. As regards the Dublin-Nice route, the Commission treats this route as a year round route because Aer Lingus operates in both the summer and the winter season. In any event, the seasonality of a monopoly route cannot change the outcome of the competitive assessment as the Commission has to ensure that customers of seasonal routes would not be harmed. Such harm to consumers would stem from the elimination of the only remaining competitor in the case of a merger leading to monopoly.

8.9.4.1. Monopoly routes on airport-to-airport basis

(1012) Out of these 28 monopoly routes, the activities of Ryanair overlap with Aer Lingus on 6 routes on an airport-to-airport basis on which the Commission has not identified other relevant alternative airports.

Table 33: Monopoly routes for which there is no airport substitutability issue

Route / Total pax winter season 2011/12 / Total pax summer season 2011 / Seasonality

1. Dublin-Berlin [...]* [...]* Year round service

2. Dublin- Budapest (1015) [...]* [...]* Year round service

3. Dublin- Fuerteventura [...]* [...]* Year round service

4. Dublin-Marseille [...]* [...]* Summer only

5. Dublin-Nice [...]* [...]* Year round service (1016)

6. Cork-Faro [...]* [...]* Summer only (1017)

Source: Annex 7.3.b and Annex 10 as updated on 22 August 2012, Annex 7.3(a)(i) of the Form CO of 24 July 2012

(1013) The Transaction would lead to a reduction in the number of carriers active on all these routes from 2 to 1 and would give the merged entity a 100% market share. The Transaction would therefore lead to a monopoly which would eliminate all existing competition on these 6 routes.

(1014) On an additional 7 routes out of these 28 monopoly routes, although there is more than one airport in the destination city, Ryanair and Aer Lingus fly to the same destination airport ("airport pair overlap"). For some of them, the other airport is also served by one of the Parties.

Table 34: Monopoly routes for which there is an airport pair overlap

Route / Total pax winter season 2011/12 / Total pax summer season 2011 / Seasonality / Airport to which both Ryanair and Aer Lingus fly / Name of the Party or third party who flies to an additional airport

1. Dublin-Alicante/Murcia (1018) [...]* [...]* Year round service Alicante Ryanair (Murcia)

2 Dublin-Birmingham/East Midlands (1019) [...]* [...]* Year round service Birmingham Ryanair (East Midlands)

3 Dublin-Edinburgh/Glasgow (1020) [...]* [...]* Year round service Edinburgh Aer Lingus(Glasgow)

4. Dublin-Manchester/Liverpool/Leeds (1021) [...]* [...]* Year round service Manchester Ryanair (Leeds and Liverpool)

5. Dublin-Tenerife [...]* [...]* Year round service Tenerife South None

6. Cork-Alicante/Murcia [...]* [...]* Summer only Alicante None

7 Cork-Tenerife [...]* [...]* Year round service Tenerife South None

Source: Annex 7.3.b and Annex 10 as updated on 22 August 2012, Annex 7.3(a)(i) of the Form CO of 24 July 2012

(1015) The question of airport substitutability (such as substitutability of Alicante for Murcia airport) could be relevant to entry projects by the Parties' potential competitors. However, according to the outcome of the market investigation, entry/expansion by competitors on these routes post-Transaction would not be likely, timely and sufficient to constitute a competitive constraint on the merged entity and offset the anticompetitive effects of the Transaction on these routes.

(1016) Therefore, on these 7 'merger-to-monopoly' routes, the question of airport substitutability can be left open.

(1017) Indeed, irrespective of the precise market definition (that is to say, the actual overlap on the Dublin-Alicante, Dublin-Birmingham, Dublin-Edinburgh, Dublin-Manchester, Dublin-Tenerife South, Cork-Alicante or Cork-Tenerife South routes or the alternative comprising also the other airports(s) as indicated in Table 34), the Transaction would in any event lead to a monopoly which would eliminate all existing competition on these 7 routes.

8.9.4.2. Monopoly routes on city basis

(1018) On 15 out of the 28 monopoly routes, Ryanair and Aer Lingus fly from Ireland to different airports at same destination cities as shown in Table 35:

Table 35: Monopoly routes for which there is a destination city overlap

Route / Total pax winter season 2011/12 / Total pax summer season 2011 / Seasonal / Airport to which Ryanair flies / Airport to which Aer Lingus flies

1. Dublin-Bilbao/Santander [...]* [...]* Summer only Santander Bilbao

2. Dublin-Brussels/Charleroi [...]* [...]* Year round service Charleroi Brussels

3. Dublin-Glasgow/Prestwick [...]* [...]* Year round service Prestwick Glasgow International

4. Dublin-Milan/Bergamo [...]* [...]* Year round service Bergamo Malpensa and Linate

5. Dublin-Rome Fiumicino/Ciampino [...]* [...]* Year round service Ciampino Fiumicino

6. Dublin-Toulouse/Carcassonne [...]* [...]* Summer only (1022) Carcassone Toulouse

7. Dublin-Venice/Treviso [...]* [...]* Summer only Treviso Venice

8. Dublin-Vienna/Bratislava [...]* [...]* Year round service Bratislava Vienna

9. Dublin-Warsaw/Modlin (1023) [...]* [...]* Year round service Modlin Warsaw

10. Cork-London (1024) [...]* [...]* Year round service Gatwick, Stansted Heathrow

11. Cork-Manchester/Liverpool [...]* [...]* Year round service Liverpool Manchester

12. Knock-Birmingham/East Midlands (1025) [...]* [...]* Year round service East Midlands Birmingham

13. Knock-London (1026) [...]* [...]* Year round service Luton, Stansted Gatwick

14. Shannon-London (1027) [...]* [...]* Year round service Gatwick, Stansted Heathrow

15. Shannon-Manchester/Liverpool 38 837 65 619 Year round service Liverpool Manchester

Source: Annex 7.3.b and Annex 10 as updated on 22 August 2012, Annex 7.3(a)(i) of the Form CO of 24 July 2012

(1019) However, as explained in depth in Section 7.3.3 above, air transport services from Dublin, Cork, Shannon or Knock respectively to each of the airports listed above belong to the same relevant market.

(1020) The Dublin-Glasgow/Prestwick, Cork-Manchester, Knock-Birmingham and Shannon-Manchester routes are operated by Aer Arann under the Aer Lingus Regional brand. As explained in Section 8.2, Aer Arann's operations under the Aer Lingus Regional brand must be attributed to Aer Lingus.

(1021) Therefore, the Transaction leads to an overlap between the Parties on these 15 routes. The Commission concludes that the Transaction is likely to eliminate all existing competition on these 15 routes and lead to a monopoly on each of these routes.

8.9.4.3. Entry

(1022) First, the Commission notes that there has been no sustained entry on these routes in recent years. Centralwings entered the Dublin - Warsaw route in 2007, but exited the route the following year. Also, Lufthansa entered the Dublin - Berlin route in April 2008 only to exit it again in September 2008. Moreover, there have been a number of exit events by scheduled airlines. SkyEurope exited the Dublin - Bratislava route in 2008. LOT and Norwegian airlines exited the Dublin - Warsaw route in 2007. Luxair exited the Dublin-Manchester route in 2007. Bmibaby exited the Cork - Manchester route in 2010 and the Knock - Birmingham route in 2012. Malev exited the Dublin - Budapest route in 2012.

(1023) Secondly, after assessing the evidence collected in the market investigation, the Commission is of the view that no competitor would be likely to have post- Transaction entry projects which could be considered as timely, likely and sufficient enough to constitute a competitive constraint on the merged entity so as to offset the anticompetitive effects of the Transaction.

8.9.4.4. Conclusion on the monopoly routes

(1024) The Commission therefore concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position on the following 28 routes, where the merged entity would enjoy a monopoly post-Transaction: Dublin-Alicante/Murcia; Dublin-Berlin; Dublin- Bilbao/Santander; Dublin-Birmingham/East Midlands; Dublin-Brussels; Dublin- Budapest; Dublin-Edinburgh/Glasgow; Dublin-Fuerteventura; Dublin-Glasgow International/Prestwick; Dublin-Manchester/Liverpool/Leeds; Dublin-Marseille; Dublin-Milan/Bergamo; Dublin-Nice; Dublin-Rome Fiumicino/Ciampino; Dublin- Tenerife; Dublin-Toulouse/Carcassonne; Dublin-Venice/Treviso; Dublin- Vienna/Bratislava; Dublin-Warsaw/Modlin; Cork-Alicante/Murcia; Cork-Faro; Cork-London; Cork-Manchester/Liverpool; Cork-Tenerife; Knock-Birmingham/East Midlands; Knock-London; Shannon-Manchester/Liverpool, and Shannon-London.

(1025) This conclusion also applies to the 7 merger to monopoly routes where the question of airport substitutability can be left open (that is to say, Dublin-Alicante/Murcia; Dublin-Birmingham/East Midlands; Dublin-Edinburgh/Glasgow; Dublin- Manchester/Liverpool/Leeds; Dublin-Tenerife South; Cork-Alicante/Murcia and Cork-Tenerife South) as the Transaction would in any event lead to a monopoly which would eliminate all existing competition on these 7 routes.

8.9.5. Routes where other scheduled carriers operate

(1026) The Parties' activities lead to further overlaps on 7 routes on which other scheduled carriers operate: Dublin-Bristol/Cardiff/Exeter; Dublin-Frankfurt; Dublin-London; Dublin-Madrid; Dublin-Munich; Dublin-Paris and Dublin-Stockholm.

8.9.5.1. Dublin - Bristol/Cardiff/Exeter

(1027) The Commission left open the question of whether scheduled point-to-point passenger air transport services between Dublin and BRS airport or CWL airport and between Dublin and BRS airport or EXT airport belong to the same market (1028). For the sake of convenience, the Commission will present the various features of this route by taking into account air transport services from Dublin to all these airports (BRS airport, CWL airport and EXT airport). Such an approach does not impact the Commission's assessment on the airport substitutability on this route (1029).

- Route characteristics

(1028) On the Dublin-Bristol/Cardiff/Exeter route, [...]* passengers travelled in IATA summer season 2011 while [...]* passengers travelled by air in IATA winter season 2011/12 (1030).

(1029) According to Ryanair (1031), the nature of the route is a mix of business and leisure.

(1030) The route has no significant seasonal pattern (1032).

(1031) [...]* of Ryanair's passengers on the Dublin-Bristol/Cardiff/Exeter route originate in Ireland (1033).

- Parties' and competitors' operations

(1032) Both Parties (1034) operate on the route with non-stop services. Ryanair and Aer Lingus/Aer Arann overlap on an airport-to-airport basis. Both Ryanair and Aer Lingus/Aer Arann operate from Dublin to Bristol (BRS).

(1033) Aer Lingus started to operate from Dublin to BRS before 2004 and then exited in March 2007. Aer Arann entered Dublin-Cardiff (CWL) in May 2010 and Dublin - BRS in March 2011 and since then has been operating both segments under the Aer Lingus Regional brand. Ryanair has been operating on the route since May 1997 (1035). Ryanair operates a fleet of Boeing 737-800 aircraft (1036). Aer Arann operates turboprops ATR42-500 (1037).

(1034) Flybe has been operating from Dublin to Exeter (EXT) for around 15 years. Flybe operates a fleet of Q400 aircraft with 78 seats and Embraer jet 88 aircraft with 112 seats (1038).

- Frequencies

(1035) According to Ryanair, the weekly flights operated by the Parties and the only other operating carrier on this route are as follows:

Table 36: Weekly flights between Dublin and Bristol/Cardiff/Exeter

Summer season 2011: Weekly frequency (1 way) July

Aer Lingus/Aer Arann 33

Ryanair 21

Combined 54

Flybe 6

Winter season 2011/2012:Weekly frequency (1 way) January

Aer Lingus/Aer Arann 28

Ryanair 21

Combined 49

Flybe 4

Source: Annex 7.3 (b) as updated on 22 August 2012 and Ryanair's response to the Commission's request for information of 25 and 26 October 2012

(1036) In IATA summer season 2011, Ryanair operated 21 weekly frequencies to BRS, while Aer Lingus/Aer Arann operated 33 weekly frequencies to BRS and CWL combined (namely, 20 weekly frequencies to BRS and 13 to CWL). In IATA winter season 2011/2012, Ryanair operated 21 weekly frequencies to BRS, while Aer Lingus/Aer Arann operated 28 weekly frequencies to BRS and CWL combined (namely, 15 weekly frequencies to BRS and 13 to CWL). The Parties would therefore have had combined 54 and 49 weekly frequencies for IATA summer season 2011 and IATA winter season 2011/2012 respectively.

(1037) Apart from the Parties, Flybe operated a service between Dublin and EXT with 6 weekly frequencies in summer season 2011 and with 4 weekly frequencies in winter season 2011/2012.

(1038) The weekly frequencies for the summer season 2012 for Ryanair amounted to 21 to BRS, to 27 for Aer Lingus/Aer Arann to BRS and CWL combined (namely, 14 weekly frequencies to BRS and 13 to CWL), whereas for Flybe amounted to 6 to EXT.

- Market shares - capacity

(1039) Table 37 indicates the market shares of the Parties and Flybe on the route on the basis of seat capacity in IATA summer season 2011 and in IATA winter season 2011/12:

Table 37: Market shares for seat capacity

Carrier /Summer season 2011 (July) : Weekly Seats available (1 way) / Summer season 2011 (July) : Parties' and competitor's market shares

Aer Lingus/Aer Arann / 2 376 / 34%

Ryanair / 3 969 / 57%

Combined / 6 345 / 92%

Flybe / 573 / 8%

Carrier /

Winter season 2011/2012 (January) : Weekly Seats available (1 way) / Winter season 2011/2012 (January) : Parties' and competitor's market shares

Aer Lingus/Aer Arann / 2 016 / 32%

Ryanair / 3 969 / 63%

Combined / 5 985 / 95%

Flybe / 304 / 5%

Source: Annex 7.3 (b) as updated on 22 August 2012 and Ryanair's response to the Commission's request for information of 25 and 26 October 2012

(1040) According to Ryanair's data updated on 22 August 2012, the weekly capacity for the summer season 2012 for Ryanair remained unchanged, with 3 969 weekly seats (60%) to BRS. The weekly capacity for the summer season 2012 for Aer Lingus/Aer Arann amounted to 1 944 weekly seats (29.5%) to BRS and CWL combined (namely, 1 008 to BRS and 936 to CWL), whereas the weekly capacity of Flybe was 666 weekly seats (10.1%) to EXT. If only weekly seats between Dublin and BRS and EXT are taken into account, Flybe's market share would be 12%.

(1041) According to DAA's data pertaining to the whole IATA summer season 2012, the market shares based on seat capacity are the following: Ryanair [60-70%], Aer Lingus [20-30%] and Flybe [5-10%] (1039).

(1042) In its comments regarding the Commission Letter of Facts (1040), Ryanair argued that the ranges provided for capacity shares misrepresented the actual capacity shares of the relevant airlines on this route for the IATA summer season 2012. However, as stated in the Methodolgy Section of this Decision, the market shares provided by DAA are replaced by the standard ranges used by the Commission. The suggestion made by Ryanair to use 5% ranges instead of 10% ranges would in any event not alter the Commission's conclusion (1041).

(1043) It follows that the Parties would have had a very large market share in terms of seat capacity of more than 80% in each of the three previous IATA seasons.

- Market shares based on passengers

(1044) Table 38 indicates the market shares of the Parties and the only other competitor active on the route on the basis of passengers carried in IATA summer season 2011 and in IATA winter season 2011/12:

Table 38: Market shares for air services

Season / Total pax Annex 7.3.b / Ryanair market share / Aer Lingus Market share / Combined market share / Flybe / Total

Summer season 2011 / [...]* (1042) / [50-60]*% / [30-40]*% / [90-100]*% / [5-10]*% / 100%

Winter season 2011/2012 / [...]* (1043) / [60-70]*% / [30-40]*% / [90-100]*% / [5-10]*% / 100%

Source: Annex 7.3 (b) as updated on 22 August 2012 and Ryanair's response to the Commission's request for information of 25 and 26 October 2012

(1045) Furthermore, on the basis of DAA's data for IATA summer season 2012, the combined market share of the Parties would amount to [90-100]% and of Flybe to [5- 10]% (1044).

(1046) It follows that the Transaction would lead to a monopoly on the Dublin- Bristol/Cardiff route and to a reduction in the number of carriers operating on the Dublin-Bristol/Exeter route from three to two and would give the merged entity a very high combined market share of around 90% in terms of point-to-point passengers carried on the route.

(1047) The Commission considers that such a high combined share on this route is in itself evidence of a dominant market position.

- Closeness of competition

(1048) Ryanair has not identified any other competitor apart from Aer Arann on the Dublin- Bristol route, in particular because Ryanair does not consider Exeter (where Flybe operates) to be in the same product market (1045).

(1049) Ryanair also argues that this route should not be treated as an overlap route because Aer Arann should be considered as an independent carrier that competes not only with Ryanair, but also with Aer Lingus. However, as explained in Section 8.2, it is considered that Aer Arann is a competitor of Ryanair, but not of Aer Lingus and that the market shares of Aer Arann have to be attributed to Aer Lingus for the purposes of the competitive assessment.

(1050) Given that the market definition has been left open with respect to Bristol, Cardiff and Exeter, the Commission will assess the closeness of competition between all three carriers active on Dublin-Bristol/Cardiff/Exeter route that is to say Ryanair, Aer Lingus and Flybe.

(1051) The Parties are the only carriers operating from Dublin to Bristol and Cardiff. They are the closest competitors on the market defined as the Dublin-Bristol/Cardiff route.

(1052) It can also be observed from the frequencies, seat capacity and passenger data that both Ryanair and Aer Lingus/Aer Arann are by far the largest competitors on this route, even on the basis of a market definition, comprising flights from Dublin to Bristol and Exeter. This is in itself a strong presumption that they are closest competitors.

(1053) As regards closeness of competition on this route, the evidence collected in the market investigation clearly confirmed that Ryanair and Aer Lingus are each other's closest competitors. This was specified by the vast majority of competitors (1046), airports, (1047) travel agents (1048) and corporate customers (1049).

(1054) Flybe is a large regional European airline operating scheduled, short-haul services on a point-to-point and hub basis. Flybe operates a mid-frills business model, which combines low-frills features, such as low-fare options, high levels of internet sales, on-board sales including purchased food and beverages with full-service characteristics, such as pre-assigned seating, business schedules, IATA membership and interlining and thus overall is closer to Aer Lingus/Aer Arann than to Ryanair. Flybe also operates smaller aircraft (Q400 with 78 seats and E series jets with 88-112 seats), which are much smaller than Ryanair's (or Aer Lingus') aircraft (1050).

(1055) According to Flybe: "For example, Ryanair has limited seat allocation, Flybe allocates every seat, with a paid for choice of a specific seat if required. Ryanair charge for check in, Flybe doesn't. Flybe has a frequent flyer loyalty programme and executive lounges, Ryanair does not. Flybe has a business class product, Economy Plus and a premium product, New Economy, where additional benefits such as baggage, lounge access and free changes to tickets are allowed, as well as free vouchers for in-flight catering, Ryanair sells only one class of ticket. Flybe operates aircraft where there is only 2x2 seating, with significantly more cabin space per passenger than Ryanair's 3x3 seating where 33% of passengers have to take a middle seat" (1051). Flybe also operates at lower load factors than Ryanair (1052).

(1056) Furthermore, according to Flybe its business model is differentiated from Aer Lingus': "Aer Lingus's regional operations are conducted through Aer Arann as a franchise of Aer Lingus. These are delivered by relatively old ATR turboprop aircraft on routes that cannot justify Airbus operations due to their small scale. Again the main proposition to the customer is price. From Flybe's perspective, although both Flybe and Aer Lingus operate several of the features such as frequent flyer programme, business lounges and assigned seating, Flybe has a greater focus on customer convenience, offering relatively high frequency flights to both major, but importantly also smaller local airports where there is less emphasis on price" (1053). Internal documents of Flybe also indicate that Flybe considers that "In many ways, there is little to differentiate between Ryanair and Aer Lingus" (1054), while Flybe "is able to offer a modern jet based product which is particularly relevant to the business and time-sensitive customer segments which represent a significant portion (c45%) of curent Flybe customer base" (1055). The Commission's regression analysis confirms and complements the conclusion derived from qualitative evidence that Ryanair and Aer Lingus are close and closest competitors as described in more detail in Annex III.

(1057) The Commission furthermore considers, as set out in Section 8.4, that the Parties are each other's closest competitors on the Dublin-Bristol/Cardiff/Exeter route with respect to Irish originating passengers, based on their high brand recognition in Ireland, strong overall market positions and bases in Ireland.

(1058) Flybe has confirmed that Ryanair and Aer Lingus/Aer Arann are the closest competitors to each other on this route (1056) and has perceived Aer Lingus and Ryanair to be the strongest brands in Ireland. Flybe has stated that it "does not have a very strong brand in Ireland, as it only has a presence on two routes ex-Dublin, albeit for many years" (1057). The evidence collected in the investigation also shows that both Ryanair and Aer Lingus enjoy strong brand recognition on the routes to and from Ireland (1058).

(1059) Flybe has no aircraft based in Dublin and therefore does not enjoy the same base advantages as Ryanair and Aer Lingus on the Irish side of the route. Similarly, Flybe does not exert the same level of constraint from its base in Exeter, as Aer Lingus and Ryanair do on each other. Ryanair and Aer Lingus exert dynamic competition on each other (that is to say, on a variety of routes) as they both have significant bases in Dublin and a similar degree of flexibility to react to demand-side (or supply-side) shocks. For Flybe, Dublin represents only one route to and from its base in Exeter, and it is therefore less committed to adjusting its capacity to match fluctuations in demand on the variety of routes from and to Dublin.

(1060) Furthermore, both Ryanair and Aer Lingus have the strongest city presence in Dublin and therefore have an advantage in winning Irish-originating customers who represent a significant share of passengers on the Dublin-Bristol/Cardiff/Exeter route. The evidence collected in the market investigation shows that both Ryanair and Aer Lingus carried a higher ratio of passenger originating in Ireland than Flybe over the summer 2011 and winter 2011/2012 seasons: [...]* % for Ryanair (1059), [40- 50] % for Aer Lingus (1060) and [30-40] % for Flybe (1061).

(1061) Therefore the Commission is of the view that, Ryanair and Aer Lingus are each other's closest competitor on the Dublin-Bristol/Cardiff/Exeter route, irrespective of the precise market definition. The Transaction would lead to the elimination of the competitive rivalry existing between Ryanair and Aer Lingus, which has been an important source of competition on this route, in particular for point-to-point passengers.

- Entry and expansion

(1062) Ryanair argues (1062) that easyJet, Thomson and Eastern have a base at Bristol airport and can enter the route at any time they wish and compete with the merged entity.

(1063) The route is subject to the general barriers to entry described in Section 8.5 and to the assessment described in Sections 8.6 and 8.7.

(1064) Moreover, ACL, the slot coordinator for British and Irish airports and Warsaw Chopin airport, indicates runway and terminal constraints for Bristol airport (1063).

(1065) The evidence collected in the market investigation indicates that no competitor would have post-Transaction entry or expansion projects which could be considered as timely, likely and sufficient enough to constitute a competitive constraint on the merged entity and offset the anticompetitive effects of the Transaction on the Dublin - Bristol/Cardiff/Exeter route, irrespective of the precise market definition.

- Conclusion

(1066) The Commission therefore concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-Bristol/Cardiff/Exeter route, irrespective of the precise market definition (that is to say, whether (i) scheduled point-to-point passenger air transport services between Dublin and BRS airport or CWL airport belong to the same market, in which case the Transaction would lead to a monopoly or (ii) air transport services between Dublin and BRS airport or EXT airport belong to the relevant market in which case the Transaction would lead to a duopoly in which the Parties would have very high combined market shares).

(1067) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus/Aer Arann and thus to eliminate the important competitive constraint that both carriers exert upon each other pre-Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that Flybe would be able to constrain the merged entity's market behaviour sufficiently, especially with regard to fare setting, on the Dublin-Bristol/Cardiff/Exeter route under any plausible market definition.

8.9.5.2. Dublin-Frankfurt

- Route characteristics

(1068) As set out in Section 7.3.3.6, the Commission takes the view that scheduled point-topoint passenger air transport services between Dublin and FRA airport and Dublin and HHN airport belong to the same market. On the Dublin-Frankfurt route, [...]* passengers travelled by air in IATA summer season 2011 while [...]* passengers travelled by air in IATA winter season 2011/12 (1064).

(1069) According to Ryanair, the nature of the route is a mix of business and leisure: passengers flying to Frankfurt airport (FRA) were mainly business travellers, thus TS passengers, while passengers flying to Frankfurt Hahn airport (HHN) were mainly leisure travellers, thus NTS passengers (1065).

(1070) The route has no significant seasonal patterns (1066).

(1071) [...]*. [...]* of Ryanair's passengers originate from Ireland (1067).

- Parties' and competitors' operations

(1072) Both Parties operate on the route with non-stop services. Ryanair and Aer Lingus overlap on a city-to-city basis. Ryanair operates from Dublin to HHN and Aer Lingus from Dublin to FRA.

(1073) Aer Lingus started to operate on the route before 2004, while Ryanair has been operating on it since May 2005 (1068). Ryanair operates a fleet of Boeing 737-800 aircraft. For short-haul routes, Aer Lingus operates essentially a fleet of Airbus A 320 aircraft (1069). Lufthansa has been operating on this route since before 2004 and currently uses Airbus A 321 aircraft.

- Frequencies

(1074) According to Ryanair, the weekly flights operated by the Parties and the only other operating carrier on this route are as follows:

Table 39: Weekly flights between Dublin and Frankfurt

Summer season 2011

Carrier / Weekly Frequency (1 way) : July

Aer Lingus / 14

Ryanair / 4

Combined / 18

Lufthansa / 21

Winter season 2011/2012

Carrier / Weekly Frequency (1 way) : January

Aer Lingus / 13

Ryanair / 4

Combined / 17

Lufthansa / 18

Source: Annex 7.3(b)(i) as updated on 22 August 2012 and Ryanair's response to the Commission's request for information of 25 and 26 October 2012

(1075) In the IATA summer season 2011, Ryanair operated 4 weekly frequencies to HHN, while Aer Lingus operated 14 weekly frequencies to FRA. In the IATA winter season 2011/12, Ryanair operated 4 weekly frequencies to HHN, while Aer Lingus operated 13 weekly frequencies to FRA. On the Dublin-Frankfurt route the Parties would therefore have had combined 18 weekly frequencies for IATA summer season 2011 and 17 weekly frequencies for IATA winter season 2011/12.

(1076) Lufthansa operated a service between Dublin and FRA with 21 weekly frequencies in the IATA summer season 2011 and 18 weekly frequencies in the IATA winter season 2011/12.

(1077) The frequencies for the IATA summer season 2012 were 5 for Ryanair and 14 for Aer Lingus, thus 19 weekly frequencies combined. The frequencies for Lufthansa remained at 21.

- Market shares - capacity

(1078) Table 40 indicates the market shares of the Parties and the only other competitor active on the route on the basis of seat capacity in the IATA summer season 2011 and the IATA winter season 2011/12:

Table 40: Market shares for seat capacity

Summer season 2011

Carrier / Weekly Seats available (1 way) : July / Market shares

Aer Lingus / 2 436 / 37%

Ryanair / 756 / 11%

Combined / 3 192 / 48%

Lufthansa / 3 458 / 52%

Winter season 2011/2012

Carrier / Weekly Seats available (1 way) : January / Market shares

Aer Lingus / 2 262 / 36%

Ryanair / 756 / 12%

Combined / 3 018 / 48%

Lufthansa / 3 276 / 52%

Source: Annex 7.3(b) as updated on 22 August 2012 and Ryanair's response to the Commission's request for information of 25 and 26 October 2012

(1079) The weekly capacity for the IATA summer season 2012 indicated that Aer Lingus offered 2 436 seats (34%) while Ryanair offered 945 seats (13%). The Parties' combined weekly capacity thus increased to 3 381 seats (47%). Lufthansa's weekly seat capacity for the IATA summer season 2012 also increased to 3 822 seats (53%). The market share as regards the seat capacity has therefore not changed significantly.

(1080) It follows that post-Transaction the Parties would control in terms of shares of capacity nearly half of the available seats in both IATA summer and winter seasons.

- Market shares based on passengers

(1081) Table 41 indicates the market shares of the Parties and of the only other competitor active on the route on the basis of passengers carried in the IATA summer season 2011 and in the IATA winter season 2011/12:

Table 41: Market shares for air services

Season / Total pax Annex 7.3.b / Ryanair market share / Aer Lingus Market share / Combin ed market share / Lufthansa / Total

Summer season 2011 / [...]* / [20- 30]*% / [40- 50]*% / [60- 70]*% / [30-40]*% / 100%

Winter season 2011/2012 / [...]* / [20- 30]*% / [50- 60]*% / [70- 80]*% / [20-30]*% / 100%

Source: Annex 7.3(b) as updated on 22 August 2012 and Ryanair's response to the Commission's request for information of 25 and 26 October 2012

(1082) However, the Commission notes that the market shares for passengers as calculated on the basis of the information provided by Ryanair are not in line (to the disadvantage of Ryanair) with the market shares for passengers calculated on the basis of the more accurate information provided by DAA. The market shares calculated on the basis of the information provided by DAA would be the following:

Table 42: Market shares for air services on the Dublin-Frankfurt route

Season / Combined market share / Lufthansa

Summer season 2011 / [60-70]% / [30-40]%

Winter season 2011/2012 / [50-60]% / [40-50]%

Source: DAA's response to the Commission's request for information of 26 July 2012

(1083) Moreover, on the basis of DAA's data for IATA summer season 2012, the market shares would remain similar to those of the previous winter season, namely combined market share of the merged entity of [50-60%] and of Lufthansa of [40- 50%] (1070).

(1084) Such a high combined share on this route is in itself evidence of a dominant market position.

- Closeness of competition

(1085) Ryanair argues that Aer Lingus and Lufthansa are much closer competitors of each other than each is to Ryanair on this route, in particular because they both operate services to FRA, while Ryanair operates to HHN (1071). Ryanair also argues that product offering of Ryanair and Aer Lingus is significantly different in terms of fares and services (1072).

(1086) However, the finding that Aer Lingus, not Ryanair, is a closer competitor to Lufthansa in certain respects would not in itself exclude the finding that on the Dublin-Frankfurt route Ryanair and Aer Lingus are very close competitors, if not each other closest competitors.

(1087) As regards closeness of competition, the responses to the market investigation are divided in relation to the Dublin-Frankfurt route (1073).

(1088) Lufthansa considers Aer Lingus to be a closer competitor to itself than Ryanair "inasmuch as Aer Lingus' business model is closer to LH's business model" (1074). Lufthansa also considers that it is in direct competition with Aer Lingus as they serve the same airports and the same customer segment (1075). It is also apparent that Aer Lingus is closer to Lufthansa in terms of the number of frequencies and overall seat capacity (1076). Lufthansa considered itself also to be the closest competitor of Ryanair, but also of Aer Lingus based on its significant seat capacity (1077). It further considered itself to be competing with Ryanair in particular with regard to its NTS customers on this route (1078) and with Aer Lingus for the same customer segment and the same airports (1079).

(1089) An equal number of competitors have indicated Aer Lingus and Lufthansa as the closest competitor to Ryanair and slightly more competitors have considered Lufthansa as the closest to Aer Lingus. The responses of airports were also largely divided as to whether Aer Lingus or Lufthansa is a closer competitor to Ryanair, while the majority considered that Ryanair is the closest to Aer Lingus. Most travel agents considered Aer Lingus as the closest competitor to Ryanair, but the views were equally divided as to whether Lufthansa or Ryanair is the closest to Aer Lingus. The majority of corporate customers considered that Aer Lingus is the closest competitor to Ryanair; the majority of corporate customers also indicated that Lufthansa is the closest competitor to Aer Lingus.

(1090) It appears that the replies were largely split because the respondents based their views on different criteria. For instance, they considered the following: (i) Lufthansa's frequency of flights (1080), (ii) the fact that Lufthansa has a significant seat capacity (1081) and (iii) that Lufthansa has a strong network brand, a hub in Frankfurt and that it provides connectivity from Frankfurt (1082). On the other hand those who specified Aer Lingus and/or Ryanair as the closest competitors to each other have based their views for example on the following: (i) the fact that several airlines exited from the Irish market (1083), (ii) on Ryanair's and Aer Lingus' frequency of flights on the route (1084), (iii) the fact that Lufthansa is more focused on business and connecting traffic than the Parties (1085), (iv) the fact that both Ryanair and Aer Lingus are Irish operators (1086). It can be observed that a number of respondents did not give any explanations and many have specified all three carriers as the closest to each other, despite Aer Lingus and Ryanair flying to different airports (1087). Some specified that Lufthansa's pricing is much more expensive than Aer Lingus' (1088).

(1091) Despite these diverging responses to the market investigation, the Commission considers that the Parties are very close competitors if not each other's closest competitors on the Dublin-Frankfurt route. Although the Parties do not fly to the same airport, it is apparent that Aer Lingus is a closer competitor to Ryanair in comparison to Lufthansa, a full service carrier. In addition, Ryanair and Aer Lingus also appear as very close competitor if not the closest competitors in terms of services and business models.

(1092) Lufthansa's business model is that of a traditional network carrier operating a hub and spoke system as opposed to the models of Ryanair and Aer Lingus. Even though Lufthansa serves the same airport in Frankfurt as Aer Lingus, Lufthansa is primarily focused on feeding its medium and long-haul services in Frankfurt. The point-topoint passengers carried on the route represent a smaller part of its activities, while connecting passengers accounted for over 50% of total passengers for Lufthansa in the last three IATA seasons (1089). This compares to only [10-20] % for Aer Lingus (1090). Unlike Lufthansa, Ryanair does not carry connecting passengers.

(1093) In addition, Lufthansa concentrates more on business customers than does Aer Lingus and Ryanair. Lufthansa operates a business cabin and offers refundable and freely exchangeable classes of tickets. It has estimated that [80-90] % of its passengers in the summer 2012 season were TS on this route (1091). For Aer Lingus the sales of flexi-fare and plus-fare tickets still represents a very small portion of its overall sales. Aer Lingus does not operate business class cabins and confirms that it "does not collect or possess data which enables it to distinguish between different categories of passengers on its short-haul routes" (1092). Similarly, Ryanair does not specifically target business passengers.

(1094) Moreover, Lufthansa has a different model in terms of pricing. It sells only a very limited amount of one-way tickets, including on the route at issue, as its pricing and business models are based on return tickets (1093). Ryanair and Aer Lingus both operate true one-way pricing, and primarily sell a one class restricted ticket. According to Aer Lingus on the Dublin -Frankfurt route in the event of a 5-10% price increase of Ryanair's fares only "few passengers would switch to the scheduled airline operating direct services, whilst significant numbers of passengers would switch to Aer Lingus. [...] the scheduled carriers on these routes (all "flag carriers") are often significantly more expensive than Aer Lingus or Ryanair" (1094).

(1095) According to Aer Lingus, Ryanair is also the closest competitor to Aer Lingus as regards the number of frequencies on the days when both Ryanair and Aer Lingus operate. Aer Lingus further explained that while it operates to the same airport as Lufthansa, the Lufthansa operation includes double the frequency per day and is considered a premium product. For Aer Lingus, due to the variances in frequency, its primary competitor is Ryanair on the same day of operation, but analysts will follow Lufthansa more closely on the days which Ryanair do not operate. Aer Lingus monitors both and keeps fares closer to Ryanair where possible (1095).

(1096) The Commission's regression analysis shows that the conclusion derived from qualitative evidence that Ryanair and Aer Lingus are close and closest competitors as described in more detail in Annex III.

(1097) The Commission furthermore considers, as set out in Section 8.4, that the Parties are each other's closest competitors on the Dublin-Frankfurt route with respect to Irish originating passengers, based on their high brand recognition in Ireland, strong overall market positions and bases in Ireland.

(1098) While Ryanair and Aer Lingus both enjoy strong brand recognition on the routes to and from Ireland, (1096) according to Lufthansa its strong brand as a network carrier does not give it an important advantage on the Irish side of a market: "for traffic ex- Ireland [Lufthansa's brand is a] rather weak brand compared to Ryanair and Aer Lingus as [there is a] rather small network to/from Ireland" (1097). Similarly, according to Aer Lingus, Ryanair remains its closest competitor on the Dublin-Frankfurt route: "[t]his remains unchanged from the 2007 decision: as the strongest Irish carriers Ryanair and Aer Lingus clearly enjoy the best brand recognition on routes into and out of Ireland" (1098).

(1099) Both Ryanair and Aer Lingus have the strongest city presence in Dublin and therefore have an advantage in winning Irish-originating customers who represent a significant share of passengers on the Dublin-Frankfurt route. The evidence collected in the market investigation shows that both Ryanair and Aer Lingus carried a higher ratio of passenger originating in Ireland compared to Lufthansa over the summer 2011 and winter 2011/2012 seasons: that is to say [...]* % for Ryanair (1099), [40-50] % to [50-60] % (1100) for Aer Lingus and [20-30] % for Lufthansa. (1101)

(1100) Furthermore, Lufthansa has only one overnight aircraft at Dublin airport, and therefore does not enjoy the same base advantages as Ryanair and Aer Lingus on the Irish side of the route. Similarly, Lufthansa does not exert the same level of constraint from its hub in Frankfurt, as Aer Lingus and Ryanair do on each other. Ryanair and Aer Lingus exert dynamic competition on each other on a variety of routes as they have significant bases in Dublin and a similar degree of flexibility to react to demand-side or supply-side shocks. For Lufthansa, Dublin represents only one route from its destination hub, and it is therefore less committed to adjusting its capacity to match fluctuations in demand on the variety of routes from and to Dublin. As a network carrier, Lufthansa is more focused on connecting traffic to its hub airports.

(1101) Therefore it is considered that Ryanair and Aer Lingus are very close competitors, if not each other's closest competitors, on the Dublin-Frankfurt route. The Transaction would lead to the elimination of the competitive rivalry existing between Ryanair and Aer Lingus, which has been an important source of competition on this route, in particular for point-to-point passengers.

- Entry and expansion

(1102) According to Ryanair (1102), given Lufthansa's overall size and its base in FRA, Lufthansa is particularly well placed to compete with Aer Lingus and Ryanair on that route, including by adding frequencies.

(1103) [...]* (1103), Ryanair argues that airlines with bases at the non-Irish end of the route would enjoy every advantage associated with Ryanair and Aer Lingus having bases in Dublin (1104). This would apply to Lufthansa in FRA.

(1104) Furthermore, according to Ryanair (1105), slots are freely available at both FRA and HHN, so that any airline can begin operating services on this route. Moreover, Air Berlin, Condor and Thomson operate bases at FRA and could thus freely enter the route at any time they wish.

(1105) The route is subject to the general barriers to entry described in Section 8.5 and to the assessment described in Sections 8.6 and 8.7.

(1106) There are no examples of an attempted new entry on the Dublin-Frankfurt route by any other carrier than Lufthansa in recent years.

(1107) Moreover, FRA is a level 3 coordinated airport. While the airport operator Fraport AG indicates that by opening a new runway in October 2011, there are presently no airside restrictions (1106), the Slot Coordinator FHKD nevertheless indicates that capacity is reached during the arrival or departure peaks, while arrival slots would still be available during departure peaks and departure slots would still be available during arrival peaks, total capacity would be close to full during morning hours (1107).

(1108) Furthermore, the evidence collected in the market investigation indicates that no competitor would have post-Transaction entry or expansion projects which could be considered as timely, likely and sufficient enough to constitute a competitive constraint on the merged entity and defeat the anticompetitive effects of the Transaction on the Dublin - Frankfurt route.

- Conclusion

(1109) The Commission therefore concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-Frankfurt route.

(1110) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that Lufthansa would be able to constrain sufficiently the merged entity's market behaviour, especially with regard to fare setting, on the Dublin-Frankfurt route.

8.9.5.3. Dublin - London

- Route characteristics

(1111) As set out in Section 7.3.3.7, the Commission takes the view that scheduled point-topoint passenger air transport services between Dublin and the five London Airports (LHR, LGW, LTN, STN, LCY) belong to the same market, while the question whether SEN airport should also be treated as substitutable with the other London Airports is left open.

(1112) On the Dublin-London route, [...]* passengers travelled by air in the IATA summer season 2011 while [...]* passengers travelled by air in the IATA winter season 2011/12 (1108).

(1113) According to Ryanair (1109), the nature of the route is a mix of business and leisure.

(1114) The route has no significant seasonal pattern (1110).

(1115) [...]* (1111). - Parties' and competitors' operations

(1116) Both Parties operate on the route with non-stop services. Ryanair and Aer Lingus overlap on a city-to-city basis and on one airport pair. Ryanair operates from Dublin to London Gatwick (LGW), to Stansted (STN) and to Luton (LTN) and Aer Lingus from Dublin to London Heathrow (LHR) and Gatwick (LGW). Aer Arann operates from Dublin to London Southend (SEN) under the Aer Lingus Regional brand.

(1117) Aer Lingus has been operating from Dublin to LHR since before 2004, from Dublin to LGW since October 2007 and from Dublin to SEN (Aer Arann) since May 2012, while Ryanair has been operating from Dublin to LGW since October 1994, from Dublin to STN since June 1989 and from Dublin to LTN since May 1986 (1112). Ryanair operates a fleet of Boeing 737-800 aircraft. For short-haul routes as Dublin-London, Aer Lingus operates essentially a fleet of Airbus A 320 aircraft (1113). Aer Arann operates a fleet of turboprops ATR42-500 (1114).

(1118) Air France/CityJet has been operating from Dublin to LCY since 2003. Air France/CityJet has a base at Dublin airport and operates two 85 seater Avro RJ aircraft on this route (1115).

(1119) British Airways, one of the operating carriers of IAG, started operating the Dublin- London route in March 2012, following the acquisition of bmi, which was operating on the route. Before being acquired by British Airways, bmi, an independent competitor controlled since 2009 by Lufthansa, operated from Dublin to LHR and LGW before 2004 and to LCY since June 2008. bmi exited LGW and LCY in March 2009. On the Dublin-London route, British Airways operates a fleet of Airbus A320 and A319 (1116).

- Frequencies

(1120) According to Ryanair, the weekly flights operated by the Parties and the other operating carriers on this route are as follows:

Table 43: Weekly flights between Dublin and the five London airports

Summer season 2011

Carrier / Weekly Frequency (1 way) : July

Aer Lingus / 124

Ryanair / 99

Combined / 223

Bmi / 34

CityJet / 27

Winter season 2011/2012

Carrier / Weekly Frequency (1 way) : January

Aer Lingus / 125

Ryanair / 97

Combined / 222

bmi / 27

CityJet / 26

Source: Annex 7.3 (b) as updated on 22 August 2012

(1121) During the IATA summer season 2011, Ryanair operated 47 weekly frequencies to STN, 31 weekly frequencies to LGW, 21 weekly frequencies to LTN, while Aer Lingus operated 89 weekly frequencies to LHR and 35 weekly frequencies to LGW. During the IATA winter season 2011/2012, Ryanair operated 47 weekly frequencies to STN, 31 weekly frequencies to LGW, 19 weekly frequencies to LTN, while Aer Lingus operated 89 weekly frequencies to LHR and 36 weekly frequencies to LGW. On the Dublin-London route, the Parties would therefore have had combined 223 and 222 weekly frequencies for the IATA summer season 2011 and the IATA winter season 2011/2012 respectively.

(1122) Apart from the Parties, bmi operated a service between Dublin and LHR with 34 weekly frequencies in IATA summer season 2011 and 27 weekly frequencies in IATA winter season 2011/2012 and Air France/CityJet operated from Dublin to LCY with 27 weekly frequencies in IATA summer season 2011 and with 26 weekly frequencies in IATA winter season 2011/2012.

(1123) The weekly frequencies for the summer season 2012 for Ryanair and Aer Lingus combined amounted to 226 weekly frequencies (to the five London airports), whereas the frequencies for Air France/CityJet amounted to 31 and for IAG (British Airways) to 37. In addition, in IATA summer season 2012, Aer Arann under the Aer Lingus Regional brand started operating to London Southend with 21 weekly frequencies.

- Market shares - capacity

(1124) Table 44 indicates the market shares of the Parties and the other competitors active on the route on the basis of seat capacity in IATA summer season 2011 and in IATA winter season 2011/12:

Table 44: Market shares for seat capacity on the Dublin-London route

Summer season 2011

Carrier / Weekly Seats available (1 way) : July / Parties' and competitors' market shares

Aer Lingus / 21 918 / 47%

Ryanair / 18 711 / 40%

Combined / 40 629 / 87%

Bmi / 4 010 / 9% (1117)

Air France/CityJet / 2 214 / 5%

Winter season 2011/2012

Carrier / Weekly Seats available (1 way) : January / Parties' and competitors' market shares

Aer Lingus / 23 308 / 49%

Ryanair / 18 333 / 39%

Combined / 41 641 / 88%

Bmi / 3 734 / 8%

Air France/CityJet / 2 132 / 4%

Source: Annex 7.3 (b) as updated on 22 August 2012

(1125) According to Ryanair's data updated on 22 August 2012, the capacity for the summer season 2012 for Ryanair amounted to 18 900 (38.2%) and for Aer Lingus amounted to 22 890 (46.3%). For the merged entity this gives a total of 41 790 (84.5%), while the capacity for Air France/CityJet amounted to 2 287 (4.6%) and for IAG (British Airways) amounted to 5 354 (10.8%). In addition, in the summer season 2012, Aer Arann started operating to London Southend with a 966 seat capacity (2%) (1118).

(1126) According to DAA's data pertaining to the whole IATA summer season 2012, the market shares based on seat capacity are the following: Ryanair [30-40%], Aer Lingus together with Aer Arann [40-50%], IAG [10-20%] and Air France [5- 10%] (1119).

(1127) It follows that the Parties would have had a large combined market share in terms of seat capacity of at least 70% or more in each of the three previous IATA seasons.

- Market shares based on passengers

(1128) Table 45 indicates the market shares of the Parties and the other competitors active on the route on the basis of passengers carried in IATA summer season 2011 and in IATA winter season 2011/12:

Table 45: Market shares for air services on the Dublin-London route

Season / Total pax Annex 7.3.b / Ryanair market share / Aer Lingus Market share / Combined market share / Other airlines' market share / Total

Summer season 2011 / [...]* / [50- 60]*% / [30- 40]*% / [90- 100]*% (1120) / [5- 10]*% / 100%

Winter season 2011/2012 / [...]* / [40- 50]*% / [40- 50]*% / [90- 100]*% / [5- 10]*% / 100%

Source: Annex 7.3 (b) as updated on 22 August 2012

(1129) However, the Commission notes that the market shares for passengers as calculated on the basis of the information provided by Ryanair are not in line (to the disadvantage of Ryanair) with the market shares for passengers calculated on the basis of the more accurate information provided by DAA. The market shares when calculated on the basis of the information provided by DAA would be the following:

Table 46: Market shares for air services on the Dublin-London route

Season / Combined market share / Other airlines' market share

Summer season 2011 / [80-90]% / [10-20]%

Winter season 2011/2012 / [80-90]% / [10-20]%

Source: DAA's response to the Commission's request for information of 26 July 2012

(1130) Moreover, on the basis of DAA's data for IATA summer season 2012, the market shares of the merged entity would amount to [80-90] % and the market share of the other competitors active on the route would amount to [10-20] % (1121).

(1131) Such a high combined share on the route is in itself evidence of a dominant market position.

- Closeness of competition

(1132) Ryanair argues (1122) that British Airways/bmi and Air France/CityJet are two significant competitors on the Dublin-London route based on their combined market shares of around 15% and of around 10% in terms of frequency and point-to-point passengers respectively. Ryanair also claims that Aer Lingus is much closer to IAG (British Airways), which also flies to LHR (1123) and that the product offering of Ryanair and Aer Lingus is significantly different in terms of fares and services (1124).

(1133) However, it can be observed from the seat capacity and passenger data that both Ryanair and Aer Lingus are by far the largest competitors, which in itself is a strong presumption that they are the closest competitors. Furthermore, as explained in Section 8.9.1, market shares in terms of frequency are not the most accurate proxy for assessing the market power of airlines, and all the more so if these airlines have different business models and use very different types of aircraft in terms of capacity - Air France/CityJet operates a fleet of RJ 85 aircraft with 95 seats, while Ryanair and Aer Lingus both operate large aircraft.

(1134) As regards closeness of competition, the results of the market investigation also show that Aer Lingus and Ryanair are indeed each other's closest competitors on the Dublin-London route (1125). The majority of competitors who responded to the market investigation have indicated that Aer Lingus and Ryanair are the closest competitors to each other (1126). The same was confirmed by the majority of airports (1127), travel agents (1128), and corporate customers (1129). For example BAA Airports Limited considers Ryanair and Aer Lingus to be each other's closest competitors on the Dublin-London route by reason of the frequency offered on the route and the closeness of the ticket price. (1130) The same was confirmed by the National Consumer Agency of Ireland (1131).

(1135) The Commission does not contest the view that Aer Lingus, not Ryanair, is a closer competitor to IAG (British Airways) and Air France in view of their business models, especially given that Aer Lingus positions itself between Ryanair and full service carriers as a value-carrier. However, this does not exclude the finding that Aer Lingus remains the closest competitor to Ryanair.

(1136) It is also apparent that Aer Lingus is a closer competitor to Ryanair in comparison to IAG (British Airways) and Air France/CityJet. In addition and despite the fact that both IAG (British Airways) and Aer Lingus operate to London Heathrow airport, Ryanair and Aer Lingus appear as very close competitors if not each other's closest competitors in terms of services and business models on this route.

(1137) Although IAG (British Airways) indicated both Ryanair and Aer Lingus as competitors on the Dublin-London route (1132), IAG (British Airways) is, unlike the Parties, a full service network airline operating a hub-and-spoke model. IAG (British Airways) offers both economy and business class on the Dublin-London route and provides an integrated passenger product with additional services such as ticket flexibility, complementary luggage transport, access to airports lounges, frequent flyer programmes, on-board meals, seat allocation.

(1138) Furthermore, IAG (British Airways) operates a large network of medium and longhaul services out of LHR and uses short-haul routes including Dublin-London to feed its hub at LHR. The market shares of Aer Lingus' connecting passengers on this route (1133) were in the range of [10-20] % over the last three IATA seasons, which is lower than the share of connecting passengers for IAG (British Airways) on this route, that is to say [30-40] % (1134). While both Ryanair and Aer Lingus operate true one-way pricing policy, IAG (British Airways) also sells return tickets and applies "fare fencing" (1135). Bmi used to operate a single cabin on this route (as do Ryanair and Aer Lingus); however, IAG (British Airways) has now introduced a Club Europe cabin (business class) in line with its other European routes (1136). Generally, IAG considers that both Ryanair and Aer Lingus "offer a practically identical service in terms of product (typical low cost product / service). Both companies follow the same business model [...]" (1137).

(1139) Air France/CityJet considers Aer Lingus, IAG (British Airways) and Ryanair as competing on the Dublin-London route, the closest to Air France being Aer Lingus followed by IAG (British Airways) (1138). However, unlike the Parties, Air France/CityJet is a scheduled full service carrier and point-to-point carrier operating on short-haul routes (1139). The services offered by CityJet on the Dublin-London route are very distinct from those of Ryanair and Aer Lingus; for instance CityJet flies with smaller aircraft (95 seats) (1140). Although CityJet does not carry a significant number of connecting passengers on this route, it is a specialised service provider, operating to LCY only. As LCY is the closest airport to London city centre, CityJet targets mostly business travelers (1141). However, the Commission also takes into account that CityJet sells one-way tickets on the Dublin-London route.

(1140) Aer Lingus considers that it operates in direct competition with Ryanair on the Dublin-LGW airport pair, but also reviews Ryanair fares on STN and LTN as secondary comparisons (1142). On the Dublin-LHR airport pair, where a close competitor to Aer Lingus is IAG (British Airways) (1143), Aer Lingus still monitors Ryanair's operations from Dublin to LGW, from Dublin to STN and from Dublin to LTN against its operations from Dublin to LHR (1144). This indicates that Ryanair represents an effective constraint on Aer Lingus' operations from Dublin to LHR. Furthermore, according to Aer Lingus, CityJet is a distant competitor, as its fares can be considered as one of the highest in the market and therefore not relevant for lowfare monitoring (1145).

(1141) Similarly, in Aer Arann's view (1146) Ryanair and Aer Lingus are the closest competitors to each other on the Dublin-London route. Furthermore, Aer Arann considers Ryanair as a competitor and monitors all competing routes with Ryanair.

(1142) These findings are in line with the situation that was prevailing in 2007: "British Airways is a full service network carrier and CityJet is a carrier offering flights aimed at business passengers to an airport situated near London city centre. Those specific features mean that the business model and competitive capacity of those undertakings is unlike those of Ryanair-Aer Lingus combined." (1147)

(1143) The Commission's regression analysis confirms and complements the conclusion derived from qualitative evidence that Ryanair and Aer Lingus are close and closest competitors as described in more detail in Annex III.

(1144) The Commission furthermore considers, as set out in Section 8.4, that the Parties are each other's closest competitors on the Dublin-London route with respect to Irish originating passengers, based on their high brand recognition in Ireland, strong overall market positions and bases in Ireland.

(1145) While Ryanair and Aer Lingus both enjoy strong brand recognition on the routes from and to Ireland, (1148) CityJet considers indeed that its brand is mostly a business focused brand on the London City network (1149). Air France considers that Ryanair and Aer Lingus both dominate the Irish market (1150). IAG (British Airways) has strong brand recognition (1151), but it has very limited activities on the routes out of Ireland (that is to say, only Dublin-London, while Dublin-Madrid is operated by Iberia).

(1146) Both Ryanair and Aer Lingus have the strongest city presence in Dublin and therefore an advantage in winning Irish-originating customers who represent a significant share of passengers on Dublin-London route. [...]* for Ryanair (1152), [40- 50] % - [50-60]% for Aer Lingus (1153), [20-30]%- [30-40]% for IAG (British Airways) (1154) and [40-50]% for Air France (1155).

(1147) Furthermore, IAG (British Airways) has only one overnight aircraft at Dublin airport, (1156) and therefore does not enjoy the same base advantages as Ryanair and Aer Lingus on the Irish side of the route. Air France/CityJet has based 3 lower capacity aircraft (a fleet of three 95 seater RJ 85 aircraft) at Dublin airport, of which 2 serve the Dublin-London route (1157). However due to lower seat capacity, CityJet would have higher costs per passenger than Ryanair and Aer Lingus. The limited size of CityJet's operations at Dublin airport would not allow the CityJet to redeploy capacity from one route to another in reaction to demand shift as in the case of Aer Lingus and Ryanair. Ryanair and Aer Lingus exert dynamic competition on each other (that is to say, on a variety of routes) as they have significant bases in Dublin and have a similar degree of flexibility to react to demand-side or supply-side shocks. For IAG (British Airways) and Air France/CityJet (1158), Dublin represents only one route to and from their destination hub, and they are therefore less committed to adjusting their capacity to match fluctuations in demand on the variety of routes from or to Dublin. Furthermore, network carriers are typically more focused on connecting traffic to their hub airports.

(1148) Therefore it is considered that Ryanair and Aer Lingus are each other's closest competitors on the Dublin-London route. They are both airlines operating essentially point-to-point air transport services, and are not strongly specialised in business traffic, as is the case for CityJet. Furthermore, Ryanair and Aer Lingus compete head-to-head on the Dublin-LGW route, where the Transaction would lead to a monopoly. Generally, the Transaction would lead to the elimination of the competitive rivalry existing between Ryanair and Aer Lingus, which has been an important source of competition on the Dublin-London route, in particular for pointto- point passengers.

- Entry and expansion

(1149) According to Ryanair (1159), IAG (British Airways), bmi regional, Flybe, easyJet, Wizz Air, Virgin Atlantic and Air France have a base at one or more of the various London airports or at Dublin, so they can enter or expand their services on the route at any time.

(1150) [...]* (1160) carriers based there would enjoy the same advantages as Ryanair and Aer Lingus having bases in Dublin.

(1151) According to Ryanair (1161), while LHR is slot constrained, slots are generally available at the other relevant London airports, in particular at LTN and STN, so that entry would not be constrained to a significant degree due to availability of slots at the London end of the route.

(1152) The route Dublin-London is subject to the general barriers to entry described in Section 8.5 and to the assessment described in Sections 8.6 and 8.7.

(1153) Moreover, ACL indicates runway and terminal constraints for STN, LGW and LTN airports (1162), in addition to the known congestion situation of LHR regarding lack of slots and runway, terminal and stand constraints. BAA however considers that STN generally does not have capacity constraints (1163).

(1154) London Southend Airport indicates that the passenger terminal is constrained at peak times due to the size of the building, in particular between 06.30 and 07.30 (1164).

(1155) In the recent Case M. 6447 IAG/bmi, it was concluded that LHR, LGW and LCY, as fully coordinated airports, are slot constrained. In particular, LHR and LGW suffer capacity constraints across the whole operating day, while LCY, although it is partly slot constrained, suffers from other limitations, such as runway capacity as only short-haul aircraft certified to operate at LCY are admitted (1165).

(1156) The availability of slots has often been identified by competitors as a barrier to entry at LHR, LGW and LCY (1166). IAG (British Airways) considers that there are barriers to entry or to expansion at LHR, LGW and LCY due to lack of slots available at those airports (1167). The fact that there is availability at other London airports would only be relevant insofar as a potential entrant would be interested in flying to such other airports.

(1157) IAG (British Airways) entered the route in March 2012 following it's acquisition of bmi. British Airways increased its frequencies on this route for IATA winter season 212/2013: "We have now planned 1061 LHR-DUB departures and 1059 DUB-LHR departures in W12" (1168).However, IAG has indicated (1169) that even if "BA has since acquiring bmi re-entered the LHR/DUB and LHR/BFS routes", it "would not currently devote more LHR slots to these routes as there are more profitable options on other parts of the network. Therefore absent a slot remedy on LHR BA would not expand significantly its service on these routes nor enter other Irish routes from LHR." IAG also mentioned (1170) that Iberia "is not interested in entering any other DUB routes." However, this new player on the Dublin-London route cannot exert the same level of competitive constraint as that exerted by the Parties, who are each other's closest competitors. Besides, the Commission considers that even with this increased level of services in the winter season 2012/13, the competition constraint exercised by competitors on the merged entity would not be sufficient after the Transaction.

(1158) Beside British Airways' entry, there are no other examples of an attempted new entry to this route by any other carrier in recent years (1171).

(1159) Furthermore, the evidence collected in the market investigation indicates that no other competitor would have post-Transaction entry or expansion projects which could be considered as timely, likely and sufficient enough to constitute a competitive constraint on the merged entity and defeat the anticompetitive effects of the Transaction on the Dublin - London route.

- Conclusion

(1160) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-London route irrespective of whether flights between Dublin and SEN airport are included or not.

(1161) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that BA and CityJet would be able to constrain the merged entity's market behaviour sufficiently, especially with regard to fare setting, on the Dublin-London route.

8.9.5.4. Dublin-Madrid

- Route characteristics

(1162) On the Dublin-Madrid route, [...]*passengers travelled by air in IATA summer season 2011 while [...]* passengers travelled by air in IATA winter season 2011/12 (1172).

(1163) According to Ryanair, the nature of the route is a mix of business and leisure (1173).

(1164) The route has no significant seasonal patterns (1174).

(1165) [...]*. [...]*% of Ryanair's passengers originate from Ireland (1175).

- Parties' and competitors' operations

(1166) Both Parties operate the route with non-stop services. Ryanair and Aer Lingus overlap on an airport-to-airport basis, namely Dublin - Barajas (MAD).

(1167) Iberia has been operating on this route since before 2004. From May 2011 to May 2012 the operations were conducted by Air Nostrum as Iberia's franchisee; tickets were still sold exclusively by Iberia (1176). Since June 2012 Iberia has started to operate the service itself again under the brand name "Iberia Express" (1177). Spanair operated the route between IATA summer season 2004 and IATA summer season 2008, when it exited the route.

- Frequencies

(1168) According to Ryanair, the weekly flights operated by the Parties and the only other operating carrier on this route are as follows:

Table 47: Market shares for air services on the Dublin-Madrid route

Summer season 2011

Carrier / Weekly Frequency (1 way) : July

Aer Lingus / 10

Ryanair / 7

Combined / 17

Iberia / 14

Winter season 2011/2012

Carrier / Weekly Frequency (1 way) : January

Aer Lingus / 10

Ryanair / 7

Combined / 17

Iberia / 7

Source: Annex 7.3(b)(i) as updated on 22 August 2012

(1169) In IATA summer season 2011, Ryanair operated 7 weekly frequencies to MAD, while Aer Lingus operated 10 weekly frequencies to MAD. In IATA winter season 2011/12, Ryanair operated 7 weekly frequencies and Aer Lingus 10 weekly frequencies to MAD. On the Dublin-Madrid route the Parties would therefore have had combined 17 weekly frequencies in both IATA season.

(1170) Apart from the Parties, only Iberia operates a service between Dublin and Madrid with 14 weekly frequencies during summer season and 7 weekly frequencies during winter season.

(1171) The frequencies for the IATA summer season 2012 for Ryanair and Aer Lingus combined amounted to around 17 weekly frequencies, whereas the frequencies for Iberia amounted to 9 (1178).

- Market shares - capacity based

(1172) Table 48 indicates the market shares of the Parties and the only other competitor active on the route on the basis of seat capacity in IATA summer season 2011 and IATA winter season 2011/12:

Table 48: Market shares for seat capacity

Summer season 2011

Carrier / Weekly Seats available (1 way) : July / Parties' and competitor's market shares

Aer Lingus / 1 740 / 42%

Ryanair / 1 323 / 32%

Combined / 3 063 / 74%

Iberia / 1 060 / 26%

Winter season 2011/2012

Carrier / Weekly Seats available (1 way) : January / Parties' and competitor's market shares

Aer Lingus / 1 740 / 48%

Ryanair / 1 323 / 36%

Combined / 3 063 / 84%

Iberia / 566 / 16%

Source: Annex 7.3(b)(i) as updated on 22 August 2012

(1173) It follows that post-Transaction the Parties would control, in terms of shares in capacity, around 74% of the available seats in IATA summer season 2011 and around 84% in IATA winter season 2011/12.

(1174) According to Ryanair's data updated on 22 August 2012, the weekly capacity for IATA summer season 2012 for Aer Lingus amounted to 1 740 seats (39%) while for Ryanair it amounted to 1 323 seats (29.8%). Thus, the Parties' combined weekly capacity remained at 3 063 seats (69%). Iberia's weekly seat capacity, as Iberia Express, increased to 1 350 seats (31%).

(1175) However, the Commission notes that the market shares for seat capacity as calculated on the basis of the information provided by Ryanair are not in line with the market shares for seat capacity calculated on the basis of the more accurate information provided by DAA. The market shares would then be the following:

Table 49: Market shares for seat capacity on the Dublin-Madrid route

Season / Combined market share / Iberia

Summer season 2011 / [80-90]% / [10-20]%

Winter season 2011/2012 / [80-90]% / [10-20]%

Source: DAA's response to the Commission's requests for information of 26 July 2012 and of 3 December 2012

(1176) According to DAA's data pertaining to the whole IATA summer season 2012, the market shares based on seat capacity are the following: Ryanair [30-40%], Aer Lingus [40-50%], Iberia [20-30%] (1179).

(1177) It follows that the Parties, if merged, would have had a very large market share in terms of seat capacity in each of the three previous IATA seasons.

(1178) Such a high combined market share on the route is in itself evidence of a dominant market position.

- Market shares - passengers

(1179) Table 50 indicates the market shares of the Parties and the only other competitor active on the route on the basis of passengers carried in IATA summer season 2011 and in IATA winter season 2011/12:

Table 50: Market shares for air services

Season / Total pax Annex 7.3.b / Ryanair market share / Aer Lingus Market share / Combined market share / Iberia / Total

Summer season 2011 / [...]* / [50- 60]*% / [30- 40]*% / [90-100]*% / [0- 5]*% / 100%

Winter season 2011/2012 / [...]* / 50- 60]*% / [40- 50]*% / [90-100]*% / [0- 5]*% / 100%

Source: Annex 7.3(b) as updated on 22 August 2012

(1180) The market shares for passengers, as calculated on the basis of the information provided by Ryanair, are not in line (to the disadvantage of Ryanair) with the market shares for passengers calculated on the basis of the more accurate information provided by DAA.

(1181) The market shares calculated on the basis of the information provided by DAA would then be the following:

Table 51: Market shares for air services on the Dublin-Madrid route

Season / Combined market share / Iberia

Summer season 2011 / [80-90]% / [10-20]%

Winter season 2011/2012 / [80-90]% / [10-20]%

Source: DAA's response to the Commission's request for information of 26 July 2012

(1182) Moreover, on the basis of DAA's data, for IATA summer season 2012 the market shares of the Parties amounted to [80-90%] and of Iberia amounted to [10-20%] (1180).

(1183) The very high combined market share on this route is itself evidence of a dominant market position.

- Closeness of competition

(1184) Ryanair argues that the merged entity would face a strong constraint from Iberia, which is the largest Spanish airline and has a share based on frequencies of 16-18% in winter and 26-31% in summer. Ryanair also argues that product offering of Ryanair and Aer Lingus is significantly different in terms of fares and services.

(1185) However, it can be observed from the seat capacity and passenger data that both Ryanair and Aer Lingus are by far the largest competitors, which in itself is a strong presumption that they are the closest competitors. In any event, a finding that Aer Lingus may be closer in certain respects to Iberia, does not in itself exclude the finding that Ryanair and Aer Lingus are each other's closest competitors on this route.

(1186) As regards closeness of competition, a significant majority of competitors that replied to the market investigation have indicated that Aer Lingus is the closest competitor to Ryanair, while an equal number has considered that Ryanair and/or Iberia are the closest to Aer Lingus (1181). A slight majority of airports considered that Iberia is closer to Ryanair than Aer Lingus is, while the majority indicated that Ryanair is the closest to Aer Lingus on this route (1182). For example, according to the Spanish Airports and Air Navigation authority ("AENA"), Iberia would present only a limited competitive alternative to Ryanair (1183). Most travel agents considered Aer Lingus as the closest competitor to Ryanair, but the views were equally divided as to whether Iberia or Ryanair is the closest to Aer Lingus (1184). The majority of corporate customers considered that Ryanair and Aer Lingus are each other's closest competitors (1185).

(1187) Despite divided answers to the market investigation, the Commission considers that the Parties are however each other's closest competitors on the Dublin-Madrid route. The Commission considers that Aer Lingus is a closer competitor to Ryanair in comparison to Iberia, which is a full service carrier. In addition, Aer Lingus and Ryanair appear as very close if not each other's closest competitor in terms of services and business models.

(1188) Iberia indicated both itself and Aer Lingus to be the closest competitor to Ryanair on the route, but without giving any explanation based on which grounds. Iberia monitors the prices of both Ryanair and Aer Lingus. However, Iberia is distinct from Ryanair and Aer Lingus in terms of its business model. Iberia operates the route mostly to feed its medium and long-haul services in MAD. Iberia's business model is the one of a traditional network carrier operating a hub and spoke system as opposed to the models of Ryanair and Aer Lingus. The point-to-point passengers carried on the route by Iberia is not the most significant part of its activities. In particular, over the last three IATA seasons, Iberia carried from [30-40]% to [50-60]% of connecting passengers, while Aer Lingus only [0-5]% of connecting passengers on this route. Ryanair does not carry connecting passengers. The different business model is also reflected in the fact that Iberia in addition to one-way tickets, offers return fares and operates business cabins and fares without conditions (1186), while Ryanair and Aer Lingus have a true one-way pricing policy and operate single cabins. Iberia considers this route to be interesting from a feed perspective but indicates that it is very difficult to compete effectively with Ryanair and Aer Lingus (1187).

(1189) Aer Lingus monitors Ryanair with considerably more attention on this route than it monitors Iberia. According to Aer Lingus, in this highly competitive route Ryanair and Aer Lingus compete almost exclusively on point-to-point traffic, while Iberia is the clear secondary competitor for Aer Lingus (1188).

(1190) These finding are in line with the situation that was prevailing in 2007: "[....] the Commission set out the reasons why Iberia, so far as concerns the Dublin-Madrid route (recital 984) [...], would not be capable of competing effectively with the merged entity. The Commission stated that Iberia would not be considered to be a close competitor to the parties to the merger due to its business model based on fullservice network operations, including a significant portion of connecting passengers on that route." (1189)

(1191) The Commission's regression analysis confirms and complements the conclusion derived from qualitative evidence that Ryanair and Aer Lingus are close and closest competitors as described in more detail in Annex III.

(1192) The Commission furthermore considers, as set out in Section 8.4, that the Parties are each other's closest competitors on the Dublin-Madrid route with respect to Irishoriginating passengers, based on their high brand recognition in Ireland, strong overall market positions and bases in Ireland.

(1193) Furthermore, while Ryanair and Aer Lingus both enjoy strong brand recognition on a large number of routes from and to Ireland (1190), Iberia has a strong brand (1191) but only very limited activities on the routes from and to Ireland (that is, only Dublin-Madrid, while Dublin-London is operated by British Airways).

(1194) Both Ryanair and Aer Lingus have the strongest city presence in Dublin and therefore have an advantage in winning Irish-originating customers who represent a significant share of passengers on the Dublin-Madrid route. The evidence collected in the market investigation shows that both Ryanair and Aer Lingus carried a higher ratio of passengers originating in Ireland compared to Iberia over the summer 2011 and winter 2011/2012 seasons: that is to say, [...]*% for Ryanair (1192), [40-50] % for Aer Lingus (1193) and [10 -20]%-[20-30]% for Iberia (1194).

(1195) Furthermore, British Airways has only one overnight aircraft at Dublin airport, and therefore does not enjoy the same base advantages as Ryanair and Aer Lingus at the Irish side of the route. Ryanair and Aer Lingus exert dynamic competition on each other (that is to say, on a variety of routes) as they have significant bases in Dublin and have a similar degree of flexibility to react to demand-side or supply-side shocks. For British Airways, Dublin represents only one route from and to its destination hub, and it is therefore less committed to adjusting its capacity to match fluctuations in demand on the variety of routes from and to Dublin. Furthermore as a network carrier, Iberia is more focused on connecting traffic to its hub airports.

(1196) Therefore it is considered that Ryanair and Aer Lingus are each other's closest competitor on the Dublin-Madrid route. The Transaction would lead to the elimination of the competitive rivalry existing between Ryanair and Aer Lingus, which has been an important source of competition on this route, in particular for point-to-point passengers.

- Entry and expansion

(1197) According to Ryanair (1195), given Iberia's overall size and its base in MAD, Iberia is particularly well placed to compete with Aer Lingus and Ryanair on that route, including by adding frequencies. Ryanair further refers to the bases of Vueling and Air Europa in MAD.

(1198) [...]* (1196), Ryanair argues that airlines with bases at the non-Irish end of the route would enjoy every advantage associated with Ryanair and Aer Lingus having bases in Dublin (1197). This would apply to Ryanair in MAD.

(1199) According to Ryanair (1198), slots are freely available at MAD so that any airline can begin operating services on this route.

(1200) The route is subject to the general barriers to entry described in Section 8.5 and to the assessment described in Sections 8.6 and 8.7.

(1201) In addition, MAD is a level 3 coordinated airport. According to AENA, MAD airport is not constrained as such apart from during particular peak hours. However, according to the AENA, for the cases in which there may be some excess demand at particular peak hours, there is infrastructure capacity to increase traffic (1199). The Commission came to the same finding in its recent decision IAG/bmi (1200).

(1202) The only scheduled airline that entered and still operates on the overlap route is Iberia, which operates with lower frequencies than the Parties.

(1203) There are no examples of an attempted new entry on this route by another carrier in the recent years (1201). Iberia declared in particular not to have an entry plans with regard to the route at issue or any routes to and from Ireland for the moment. Because of the strong competition from Aer Lingus and Ryanair, Iberia explained that it has even been forced to downsize its operations on the route at issue (1202).

(1204) However, Iberia Express very recently decided to expand its services on the Dublin- Madrid route for the 2013 summer season, adding five extra return flights per week in July (which means it will have a double daily service throughout the month). For the month of August, Iberia Express will increase its frequency from 9 flights per week to 10 flights per week. Such increase in frequencies, essentially limited to the month of July, is not sufficient to constrain the merged entity post-Transaction during most of the year (1203).

(1205) Vueling explicitly stated that without the Transaction it would consider entry on some routes, however, that should the proposed Transaction be cleared it would certainly refrain from any new entry on the routes concerned by the Transaction (1204).

(1206) Furthermore, the evidence collected in the market investigation indicates that, post- Transaction, no other competitor would have entry or expansion projects which could be considered as timely, likely and sufficient enough to constitute a competitive constraint on the merged entity and defeat the anticompetitive effects of the Transaction on the Dublin - Madrid route.

- Conclusion

(1207) The Commission therefore concludes that the Transaction is likely to significantly impede effective competition as a result of the creation of a dominant position in relation to the Dublin-Madrid route.

(1208) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that Iberia would be able to constrain the merged entity's market behaviour sufficiently, especially with regard to fare setting, on the Dublin-Madrid route.

8.9.5.5. Dublin-Munich - Route characteristics

(1209) As set out in Section 7.3.3.10, the Commission takes the view that scheduled pointto- point passenger air transport services between Dublin and MUC airport and between Dublin and FMM airport belong to the same market.

(1210) On the Dublin-Munich route, [...]*passengers travelled by air in IATA summer season 2011 while [...]*passengers travelled by air in IATA winter season 2011/12 (1205).

(1211) According to Ryanair (1206), the nature of the route is a mix of business and leisure. According to Aer Lingus (1207), the leisure segment is highly competitive between the two airlines on this route.

(1212) The route has no significant seasonal pattern (1208).

(1213) [...]*of Ryanair's passengers on the Dublin-Munich route originate in Ireland (1209).

- Parties' and competitors' operations

(1214) Both Parties operate on the route with non-stop services. Ryanair and Aer Lingus overlap on a city-to-city basis. Ryanair operates from Dublin to Memmingen Airport (FMM) and Aer Lingus from Dublin to Munich Aiport (MUC).

(1215) Aer Lingus started to operate on this route before 2004 while Ryanair has been operating on it since May 2009 (1210). Ryanair operates a fleet of Boeing 737-800 aircraft. For short-haul routes, Aer Lingus operates essentially a fleet of Airbus A 320 aircraft (1211).

(1216) Lufthansa started operating on this route in April 2011 and only for the IATA summer season. Lufthansa operates essentially a fleet of Airbus A 319 aircraft (1212)

- Frequencies

(1217) According to Ryanair, the weekly flights operated by the Parties and Lufthansa on this route are as follows:

Table 52: Weekly flights between Dublin and Munich

Summer season 2011

Carrier / Weekly Frequency (1 way) : July

Aer Lingus / 11

Ryanair / 4

Combined / 15

Lufthansa / 1

Winter season 2011/2012

Carrier / Weekly Frequency (1 way) : January

Aer Lingus / 11

Ryanair / 4

Combined / 15

Lufthansa / 0

Source: Annex 7.3 (b) as updated on 22 August 2012.

(1218) In IATA summer season 2011, Ryanair operated 4 weekly frequencies to FMM and Aer Lingus operated 11 weekly frequencies to MUC. In IATA winter season 2011/12, Ryanair operated 4 weekly frequencies to FMM and Aer Lingus operated 11 weekly frequencies to MUC. On the Dublin-Munich route, the Parties would therefore have combined 15 weekly frequencies, for each of IATA summer season 2011 and winter season 2011/2012.

(1219) Apart from the Parties, only Lufthansa operates a service between Dublin and MUC with only 1 weekly frequency operating on Sunday and only during the summer season.

(1220) The weekly frequencies for the IATA summer season 2012 amounted to 4 for Ryanair and 11 for Aer Lingus. The weekly frequency for Lufthansa remained unchanged with 1 frequency.

- Market shares - capacity

(1221) Table 53 indicates the market shares of the Parties and the other only competitor active on the route on the basis of seat capacity in IATA summer season 2011 and in IATA winter season 2011/12:

Table 53: Market shares for seat capacity

Summer season 2011

Carrier / Weekly Seats available (1 way) : July / Parties' and competitor's market shares

Aer Lingus / 1 914 / 68%

Ryanair / 756 / 27%

Combined / 2 670 / 94% (1213)

Lufthansa / 156 / 6%

Winter season 2011/2012

Carrier / Weekly Seats available (1 way) : January / Parties' and competitor's market shares

Aer Lingus / 1 914 / 72%

Ryanair / 756 / 28%

Combined / 2 670 / 100%

Lufthansa / 0 / 0%

Source: Annex 7.3(b) as updated on 22 August 2012

(1222) The weekly capacity for IATA summer season 2012 showed that Aer Lingus offered 1 914 seats (68%) while Ryanair offered 756 seats (27%). Thus, the Parties' combined weekly capacity remained unchanged in the IATA summer season 2012. Lufthansa's weekly capacity for IATA summer season 2012 remained unchanged with 156 seats (6%).

(1223) It follows that post-Transaction, the Parties would hold more than 90% market share in capacity in the IATA summer season. Furthermore, this would be a monopoly route in the IATA winter season, as Lufthansa does not operate in that season.

- Market shares - passengers

(1224) Table 54 indicates the market shares of the Parties and the only other competitor active on the route on the basis of passengers carried in IATA summer season 2011 and in IATA winter season 2011/12:

Table 54: Market shares for air services

Season / Total pax Annex 7.3.b / Ryanair market share / Aer Lingus Market share / Combined market share / Lufthansa / Total

Summer season 2011 / [...]* / [30- 40]*% / [60- 70]*% / [90-100]*% / [0-5]*% / 100%

Winter season 2011/2012 / [...]* / [20- 30]*% / [70- 80]*% / [90-100]*% / [0-5]*% / 100%

Source: Annex 7.3(b) as updated on 22 August 2012

(1225) It follows that post-Transaction, the Transaction would lead to a reduction in the number of carriers operating on the Dublin-Munich route from three to two in the IATA summer season 2011 and would give the merged entity a very high combined market share of 97% in terms of point-to-point passengers carried on the route.

(1226) Moreover, on the basis of DAA's data for the IATA summer season 2012, the market share would remain similar to those of the previous summer season, namely a combined market share of the merged entity of [90-100%] and of Lufthansa of [0- 5%] (1214).

(1227) Such a high combined share on this route is in itself evidence of a dominant market position. As regards the IATA winter season, this would be a merger to monopoly situation which would eliminate all existing competition on this route.

- Closeness of competition

(1228) Ryanair argues (1215) that Aer Lingus and Lufthansa are much closer competitors to each other than they are to Ryanair on this route, in particular because they both operate services to MUC, while Ryanair operates to FMM. Ryanair argues that the combined entity would face a strong constraint from Lufthansa, given its overall size and base in MUC, despite its modest market share. Ryanair also argues that the product offering of Ryanair and Aer Lingus is significantly different in term of fares and services (1216).

(1229) However, the Commission observes from the frequencies, seat capacity and passenger data that both Ryanair and Aer Lingus are by far the largest competitors, which in itself is a strong presumption that they are the closest competitors. In any event, although Aer Lingus may be closer in certain respects to Lufthansa, this would not in itself exclude the finding that Ryanair and Aer Lingus are the closest competitors on this route.

(1230) As regards closeness of competition, the responses to the market investigation were largely divided in relation to the Dublin-Munich route. A significant majority of competitors have indicated that Aer Lingus is closest to Ryanair, while a larger number has considered that Lufthansa is the closest to Aer Lingus (1217). A slight majority of airports considered that Lufthansa is the closest both to Ryanair and Aer Lingus (1218). However, most travel agents (1219) and corporate customers (1220) considered Aer Lingus and Ryanair as each other's closest competitors.

(1231) It appears that the replies were largely split because the respondents based their views on different criteria. A number of respondents did not give any explanations and many have specified all three carriers as the closest to each other, despite Aer Lingus and Ryanair flying to different airports (1221).

(1232) Despite divided answers to the market investigation and the fact that Aer Lingus and Ryanair fly to different airports, the Commission considers that the Parties are each other's closest competitors on the Dublin-Munich route. It is apparent that Aer Lingus is a closer competitor to Ryanair in comparison to Lufthansa, a full service carrier. In addition, Aer Lingus and Ryanair appear as very close if not each other's closest competitors in terms of services and business models.

(1233) Lufthansa's business model is the one of a traditional network carrier operating a hub and spoke system as opposed to the models of Ryanair and Aer Lingus. Lufthansa is primarily focused on feeding its medium and long-haul services in MUC and the point-to-point passengers carried on the route represent a smaller part of its activities (for example, Lufthansa's connecting passengers accounted for [30-40]% of total passengers compared to only [0-5]% for Aer Lingus in the last three IATA seasons) (1222). Ryanair does not carry connecting passengers.

(1234) In addition, Lufthansa concentrates more on business customers than is the case for Aer Lingus and Ryanair (1223). Lufthansa offers, amongst others, refundable and change free tickets and estimates that [80-90]% its passengers were TS passengers on this route in the IATA Summer season 2012 (1224). For Aer Lingus the sales of flexi-fare and plus-fare tickets still represents a very small portion of overall sales and it "does not collect or possess data which enables it to distinguish between different categories of passengers on its short-haul routes" (1225). Similarly, Ryanair does not target business passengers.

(1235) Moreover, Lufthansa is different in terms of pricing and ticket model. It sells only a very limited amount of one-way tickets, including on the route at issue, as its pricing and business model are based on return tickets (1226), as opposed to Ryanair and Aer Lingus that operate true one-way pricing and primarily sell one class restricted tickets.

(1236) According to Aer Lingus, Ryanair is its primary competitor on this route due to lack of a real alternative in terms of frequency with Lufthansa only operating on a Sunday. The leisure segment is highly competitive between Aer Lingus and Ryanair on this route (1227).

(1237) The Commission's regression analysis confirms and complements the conclusion derived from qualitative evidence that Ryanair and Aer Lingus are close and closest competitors as described in more detail in Annex III.

(1238) The Commission furthermore considers, as set out in Section 8.4, that the Parties are each other's closest competitors on the Dublin-Munich route with respect to Irish originating passengers, based on their high brand recognition in Ireland, strong overall market positions and bases in Ireland.

(1239) While Ryanair and Aer Lingus both enjoy a strong brand recognition on the routes to and from Ireland, (1228) according to Lufthansa its strong brand as a network carrier does not give it an important advantage on the Irish side of a market: "for traffic ex- Ireland [Lufthansa's brand is] rather weak brand compared to Ryanair and Aer Lingus as [there is] rather small network to/from Ireland" (1229). Similarly, according to Aer Lingus, Ryanair remains its closest competitor on the Dublin-Munich route: "[t]his remains unchanged from the 2007 decision: as the strongest Irish carriers Ryanair and Aer Lingus clearly enjoy the best brand recognition on routes into and out of Ireland" (1230).

(1240) Both Ryanair and Aer Lingus have the strongest city presence in Dublin and therefore an advantage in winning Irish-originating customers who represent a significant share of passengers on the Dublin-Munich route. The evidence collected in the market investigation shows that both Ryanair and Aer Lingus carried a higher ratio of passenger originating in Ireland compared to Lufthansa over the summer 2011 and winter 2011/2012 seasons: that is to say, [...]*% for Ryanair (1231), [40-50] % to [60-70] % (1232) for Aer Lingus and only [5-10] % for Lufthansa (1233).

(1241) Furthermore, Lufthansa has only one overnight aircraft at Dublin airport, and therefore does not enjoy the same base advantages as Ryanair and Aer Lingus at the Irish side of the route. Similarly, Lufthansa does not exert the same level of constraint from its hub in Frankfurt, as Aer Lingus and Ryanair do on each other. Ryanair and Aer Lingus exert dynamic competition on each other (that is to say, on a variety of routes) as they have significant bases in Dublin and a similar degree of flexibility to react to demand-side or supply-side shocks. For Lufthansa, Dublin represents only one route from its destination hub, and it is therefore less committed to adjusting its capacity to match fluctuations in demand on the variety of routes from and to Dublin. Furthermore, as a network carrier, Lufthansa is more focused on connecting traffic to its hub airports.

(1242) Therefore it is considered that Ryanair and Aer Lingus are each other's closest competitor on the Dublin-Munich route. The Transaction would lead to the elimination of the competitive rivalry existing between Ryanair and Aer Lingus, which has been an important source of competition on this route, in particular for point-to-point passengers.

- Entry and expansion

(1243) According to Ryanair (1234), given Lufthansa's overall size and its base in MUC, Lufthansa is particularly well placed to compete with Aer Lingus and Ryanair on that route, including by adding frequencies.

(1244) [...]* (1235), Ryanair argues that airlines with bases at the non-Irish end of the route would enjoy every advantage associated with Ryanair and Aer Lingus having bases in Dublin (1236). This would apply to Ryanair in Munich.

(1245) Moreover, according to Ryanair (1237), slots are freely available at both MUC and FMM, so that any airline can begin operating services on this route.

(1246) The route is subject to the general barriers to entry described in Section 8.5 and to the assessment described in Sections 8.6 and 8.7.

(1247) In addition, MUC is a level 3 coordinated airport, and is slot constrained. According to Flughafen München GmbH (1238), the Terminal that is used by Lufthansa is constrained both in terms of runway capacity during peak hours and of passenger terminal capacity. Aer Lingus does not operate during these peak times and their flights are operated from to Terminal 1, which has no capacity constraints. The construction of a new terminal satellite is due for 2015. In addition, planning and approval procedures for a third independent runway are under way. Flughafen München GmbH expects a deterioration of capacity constraints regarding runway capacity until completion of that third runway. According to FHKD, the airport coordinator, there are hourly movement restrictions due to runway limitations and capacity is reached during the arrival and departure peaks (1239).

(1248) As regards FMM which has no slot restrictions, Memmingen Airport has indicated to the Commission that there is a limitation relating to the runway and the available aircraft stand during peak hours. If the planning approval under process is not granted, Memmingen Airport expects a deterioration of the currently existing limitations (1240).

(1249) Beside Lufthansa's entry, there are no other examples of an attempted new entry to this route by any other carrier in the recent years (1241).

(1250) Furthermore, the evidence collected in the market investigation indicates that no competitor would have post-Transaction entry or expansion projects which could be considered as timely, likely and sufficient enough to constitute a competitive constraint on the merged entity and defeat the anticompetitive effects of the Transaction on the Dublin - Munich route.

- Conclusion

(1251) Therefore the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-Munich route.

(1252) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that Lufthansa would be able to constrain the merged entity's market behaviour sufficiently, especially with regard to fare setting, on the Dublin-Munich route.

8.9.5.6. Dublin-Paris

- Route characteristics

(1253) As set out in Section 7.3.3.11, the Commission takes the view that scheduled pointto- point passenger air transport services between Dublin and CDG airport, BVA airport and ORY airport belong to the same market. On the Dublin-Paris route, [...]* passengers travelled by air in IATA Summer season 2011 while [...]* passengers travelled by air in IATA Winter season 2011/12 (1242).

(1254) According to Ryanair (1243), the nature of the route is a mix of business and leisure: passengers flying to Paris-Beauvais (BVA) would be NTS passengers whereas passengers flying to Paris Charles de Gaulle (CDG) would include both TS and NTS customers.

(1255) The route has no significant seasonal pattern (1244).

(1256) [...]* of Ryanair's passengers on the Dublin-Paris route originate in Ireland. (1245)

- Parties' and competitors' operations

(1257) Both Parties operate on the route with non-stop services. Ryanair and Aer Lingus overlap on a city-to-city basis. Ryanair operates from Dublin to BVA and Aer Lingus from Dublin to CDG.

(1258) Aer Lingus started to operate on this route in 2001, while Ryanair has been operating on it since May 1997 (1246).

(1259) Ryanair operates a fleet of Boeing 737-800 aircraft. For short-haul routes, including this route Aer Lingus operates a fleet of Airbus A 320 aircraft (1247).

(1260) Air France, through CityJet, has been operating on this route since 2001. Air France/CityJet operates essentially a fleet of Avro RJ85 aircraft (1248). No services are currently offered from Dublin to ORY.

- Frequencies

(1261) According to Ryanair, the weekly flights operated by the Parties and the only other operating carrier on this route are as follows:

Table 55: Weekly flights between Dublin and Paris

Summer season 2011

Carrier / Weekly Frequency (1 way) : July

Aer Lingus / 21

Ryanair / 10

Combined / 31

Air France/CityJet / 42

Winter season 2011/2012

Carrier / Weekly Frequency (1 way) : January

Aer Lingus / 21

Ryanair / 8

Combined / 29

Air France/CityJet / 41

Source: Annex 7.3 (b) (i) as updated on 22 August 2012.

(1262) In IATA summer season 2011, Ryanair operated 10 weekly frequencies to BVA and Aer Lingus operated 21 weekly frequencies to CDG. In IATA winter season 2011/12, Ryanair operated 8 weekly frequencies to BVA and Aer Lingus operated 21 weekly frequencies to CDG. On the Dublin-Paris route, the Parties combined would therefore have 31 weekly frequencies, based on IATA summer season 2011 and 29 weekly frequencies, based on IATA winter season 2011/2012.

(1263) Apart from the Parties, only Air France/CityJet operates on the route, to CDG, with around 41 weekly frequencies in the winter season 2011/2012 and 42 in the summer season 2011.

(1264) During the IATA summer season 2012, Ryanair operated 13 weekly frequencies to BVA and Aer Lingus operated 21 weekly frequencies to CDG (34 frequencies combined). Air France/CityJet operated 41 weekly frequencies to CDG.

- Market shares - capacity

(1265) Table 56 indicates the market shares of the Parties and the only other competitor active on the route on the basis of seat capacity in IATA summer season 2011 and in IATA winter season 2011/12:

Table 56: Market shares for seat capacity

Summer season 2011

Carrier / Weekly Seats available (1 way) : July / Parties' and competitor's market shares

Aer Lingus / 3 654 / 41%

Ryanair / 1 890 / 21%

Combined / 5 544 / 62%

Air France/CityJet / 3 444 / 38%

Winter season 2011/2012

Carrier / Weekly Seats available (1 way) : January / Parties' and competitor's market shares

Aer Lingus / 3 654 / 43%

Ryanair / 1 512 / 18%

Combined / 5 166 / 61%

Air France/CityJet / 3 362 / 39%

Source: Annex 7.3(b) (i) as updated on 22 August 2012

(1266) According to Ryanair data, as updated on 22 August 2012, Ryanair had 2,457 weekly seats (26%) available to BVA in IATA summer season 2012. Aer Lingus had 3,645 weekly seats (39%) available to CDG. The Parties' combined weekly capacity has thus increased to 6 102 weekly seats (65%). Air France/CityJet weekly seats' availability to CDG remained unchanged with 3362 seats (36%). In its' comments regarding the Commission Letter of Facts of 14 December 2012, Ryanair adjusted the figures on the basis of the Innovata database, indicating that Ryanair, Aer Lingus and Air France/CityJet had respectively 23%, 37%, and 39% share of capacity. According to DAA, the Innovate estimates are confirmed by its own dataset (1249).

(1267) It follows that the Parties would have had a very large combined market share of at least 60% in terms of shares in capacity in each of the three previous IATA seasons.

- Market shares - passengers

(1268) Table 57 indicates the market shares of the Parties and the only other competitor active on the route on the basis of passengers carried in IATA Summer season 2011 and in IATA Winter season 2011/12:

Table 57: Market shares for air services

Season / Total pax Annex 7.3.b / Ryanair market share / Aer Lingus Market share / Combined market share / Air France/CityJet / Total

Summer season 2011 / [...]* / [30- 40]*% / [40- 50]*% / [70-80]*% / [20-30]*% / 100%

Winter season 2011/2012 / [...]* / [20- 30]*% / [50- 60]*% / [80- 90]*% (1250) / [10-20]*% / 100%

Source: European Commission on the basis of information provided by Ryanair in Annex 7.3(b) as updated on 22 August 2012

(1269) However, the Commission notes that the market shares for passengers as calculated on the basis of the information provided by Ryanair are not in line (to the disadvantage of Ryanair) with the market shares for passengers calculated on the basis of the more accurate information provided by DAA. The market shares calculated on the basis of the information provided by DAA would then be the following:

Table 58: Market shares for air services on the Dublin-Paris route

Season / Combined market share / Air France/CityJet

Summer season 2011 / [60-70]% / [30-40]%

Winter season 2011/2012 / [60-70]% / [30-40]%

Source: DAA's response to the Commission's request for information of 26 July 2012

(1270) Moreover, on the basis of DAA's data, in IATA summer season 2012 the market share of the merged entity increased to [70-80%] and of Air France/CityJet decreased to [20-30%] (1251).

(1271) In its' comments regarding the Commission Letter of Facts of 14 December 2012, Ryanair considered that the market shares based on passengers on this route deviates considerably from the figures Ryanair calculated based on capacity numbers and DAA traffic statistics. However, the capacity data do not necessarily have to be similar to passenger data, as for instance the percentage of point to point passengers (1252) and the load factor may differ significantly between carriers.

(1272) The very high combined market share on this route is itself evidence of a dominant market position.

- Closeness of competition

(1273) Ryanair argues that Aer Lingus and Air France/CityJet are much closer competitors to each other than they are to Ryanair on this route, in particular because they both operate services to CDG, while Ryanair operates to BVA (1253). Ryanair also argues that the product offering of Ryanair and Aer Lingus is significantly different in terms of fares and services (1254).

(1274) The Commission does not contest the view that Aer Lingus, not Ryanair, is a closer competitor to Air France in view of its business model, especially given that Aer Lingus positions itself between Ryanair and full service carriers as a value-carrier. However, this does not exclude the finding that Ryanair and Aer Lingus would remain very close competitors, if not each other's closest competitors on the Dublin- Paris route.

(1275) As regards closeness of competition, the responses to the market investigation are largely divided in relation to the Dublin-Paris route (1255). A majority of responding competitors have indicated Aer Lingus as the closest competitor to Ryanair, but a comparable majority of competitors have also considered Air France as the closest competitor to Aer Lingus (1256). Most of the airports considered Air France as the closest competitor to Ryanair, while Ryanair was more often indicated as the closest competitor to Aer Lingus (1257). Most travel agents considered Aer Lingus and Ryanair as each other's closest competitors (1258). The replies of corporate customers were equally split between all three carriers (1259).

(1276) It appears that the replies were largely split because the respondents based their views on different criteria. For example those who specified on the one hand both Air France and Ryanair as the closest competitor to Aer Lingus and on the other hand both Air France and Aer Lingus as the closest competitor to Ryanair, either gave no explanation or based their views on the following: (i) high seat capacities and/or market shares (for example, "AF No.2 in seats per week" (1260) or "Air France is the main competitor" or Air France (the biggest)" (1261)) or(ii) Air France's strong network brand recognition (1262). It can be observed that many respondents have specified Ryanair and Aer Lingus or all three carriers as the closest competitor to each other, despite Aer Lingus and Ryanair flying to different airports (1263). Although it is true that all three carriers enjoy relatively high market shares, they differ in terms of brand recognition, business models and the extent of operations in Ireland.

(1277) Despite split answers to the market investigation and the fact that Aer Lingus and Ryanair fly to different airports, the Commission considers that the Parties are very close competitors, if not each other's closest competitors, on the Dublin-Paris route. It is apparent that Aer Lingus is a closer competitor to Ryanair in comparison to Air France/CityJet, a full service carrier. In addition, Aer Lingus and Ryanair appear as very close if not each other's closest competitor in terms of services and business models.

(1278) Even though Air France/CityJet serves the same airport in Paris as Aer Lingus, its business model is clearly different from Aer Lingus and Ryanair. CityJet, which is a 100% subsidiary of Air France, is primarily focused on feeding the medium and long-haul services of Air France hub at CDG (1264). Connecting passengers accounted for [30-50]% of total passengers of Air France on this route compared to only [5- 10]% - [10-20]% for Aer Lingus in the last three IATA seasons (1265). For this purpose, CityJet operates a high frequency service with schedules accommodated according to the departure and arrival waves of Air France at CDG.

(1279) As regards point-to-point passengers, CityJet concentrates on higher-yield business customers. CityJet operates substantially smaller aircraft than Ryanair or Aer Lingus and thus would not be able to compete with them efficiently for the lower fares segments of customers (1266). Air France/CityJet offers business class tickets (1267) and estimates that [60-70]% of passengers are TS. On the contrary, Ryanair does not target business passengers and for Aer Lingus the sales of flexi-fare and plus-fare tickets still represents a very small portion of overall sales and it "does not collect or possess data which enables it to distinguish between different categories of passengers on its short-haul routes" (1268).

(1280) The difference in business model is also reflected in the fact that Air France/CityJet sells mainly return tickets on the route at issue (1269). Air France/CityJet also offer fenced fares, determined by rules such as "Saturday night away" or "same day return" (1270). Ryanair and Aer Lingus both operate one-way pricing, and primarily sell one class restricted tickets.

(1281) Although Aer Lingus flies to the same airport as Air France/CityJet, Aer Lingus compares its fares primarily against Ryanair and secondarily against Air France (CDG). According to Aer Lingus, "there are greater similarities in product and service between Ryanair and Aer Lingus than exist between Aer Lingus and Air France (all economy service, one-way pricing, LCC ethos) which renders Ryanair Aer Lingus' closest competitor on the Dublin-Paris route. The Air France product is not a comparable product and is not suitable for competitive tracking: Air France operates a business class cabin; offers highly fenced fares [...]; operates a higher pricing structure; and is a "full-service" airline" (1271).

(1282) It is to be noted that these findings are in line with the situation that was prevailing in 2007 as acknowledged in paragraph 377 of the 2010 Judgement: "[....] the Commission set out the reasons why [...] City Jet, so far as concerns the Dublin- Paris route (recital 1017), would not be capable of competing effectively with the merged entity. [...] the Commission stated that CityJet is primarily focused on feeding the medium and long-haul services of Air France at Roissy-Charles-de- Gaulle Airport."

(1283) The Commission's regression analysis shows that the conclusion derived from qualitative evidence that Ryanair and Aer Lingus are close and closest competitors as described in more detail in Annex III.

(1284) The Commission furthermore considers, as set out in Section 8.4, that the Parties are each other's closest competitors on the Dublin-Paris route with respect to Irish originating passengers, based on their high brand recognition in Ireland, strong overall market positions and bases in Ireland

(1285) Furthermore, both Ryanair and Aer Lingus enjoy the strongest brand recognition on the routes to and from Ireland (1272). While Air France on Paris CDG is also a strong brand, "Air France considers that the Ryanair and Aer Lingus brands dominate the Irish market. There are no other strong brands in the Irish airline market" (1273). Similarly, according to Aer Lingus, Ryanair remains its closest competitor on the Dublin-Paris route: "[t]his remains unchanged from the 2007 decision: as the strongest Irish carriers Ryanair and Aer Lingus clearly enjoy the best brand recognition on routes into and out of Ireland" (1274).

(1286) Both Ryanair and Aer Lingus have the strongest city presence in Dublin and therefore an advantage in winning Irish-originating customers who represent a significant share of passengers on the Dublin-Paris route. The evidence collected in the market investigation shows that both Ryanair and Aer Lingus carried a higher ratio of passengers originating in Ireland compared to Air France/CityJet over the summer 2011 and winter 2011/2012 seasons: that is to say, [...]*% for Ryanair (1275), [40-50] % - [60-70] % for Aer Lingus (1276) and [30-40]% - [40-50] % for Air France/CityJet (1277).

(1287) Furthermore, Air France/CityJet has based 3 lower capacity aircraft (RJ 85 with 95 seats) at Dublin airport, of which 1 serves the Dublin-Paris route. However due to lower seat capacity, CityJet would have higher costs per passenger than Ryanair and Aer Lingus. The limited size of CityJet's operations at Dublin airport would not allow CityJet to redeploy capacity from one route to another in reaction to demand shift as in the case of Aer Lingus and Ryanair. Ryanair and Aer Lingus exert dynamic competition on each other (that is to say, on a variety of routes) as they have significant bases in Dublin and have a similar degree of flexibility to react to demand-side (or supply-side) shocks. For Air France/CityJet, Dublin represents only one route to/from its destination hub, and it is therefore less committed to adjusting its capacity to match fluctuations in demand on the variety of routes from and to Dublin. Furthermore, as a network carrier, Air France/CityJet is more focused on connecting traffic to its hub airports.

(1288) Therefore it is considered that Ryanair and Aer Lingus are very close, if not each other's closest competitor, on the Dublin-Paris route. The Transaction would lead to the elimination of the competitive rivalry existing between Ryanair and Aer Lingus, which has been an important source of competition on this route, in particular for point-to-point passengers.

- Entry and expansion

(1289) According to Ryanair (1278), slots are freely available at both Dublin airport and BVA, so that any airline can begin operating services on this route.

(1290) The route is subject to the general barriers to entry described in Section 8.5 and to the assessment described in Sections 8.6 and 8.7.

(1291) In addition, CDG is a level 3 coordinated airport. According to Aéroport de Paris (ADP), CDG is constrained due to night operation restrictions. In order to reduce aircraft noise pollution, the number of night movements is limited between 00.30 and 05.29 local time for arrivals and between 00.00 and 04.59 local time for departures from parking stands. There are no limitations or constraints which would only apply at peak times (1279).

(1292) There are no examples of an attempted new entry to this route by other carriers in recent years (1280).

(1293) Furthermore, the evidence collected in the market investigation indicates that no competitor would have post-Transaction entry or expansion projects which could be considered as timely, likely and sufficient enough to constitute a competitive constraint on the merged entity and defeat the anticompetitive effects of the Transaction on the Dublin - Paris route.

- Conclusion

(1294) Therefore, the Commission concludes that the Transaction would significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-Paris route.

(1295) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that Air France/CityJet would be able to constrain the merged entity's market behaviour sufficiently, especially with regard to fare setting, on the Dublin-Paris route.

8.9.5.7. Dublin- Stockholm

- Route characteristics

(1296) As set out in Section 7.3.3.13, the Commission takes the view that scheduled pointto- point passenger air transport services between Dublin and Arlanda (ARN) airport and Dublin and Skavsta (NYO) airport belong to the same market. On the Dublin- Stockholm route, [...]* passengers travelled by air in IATA Summer season 2011 while [...]* passengers travelled by air in IATA Winter season 2011/12 (1281).

(1297) According to Ryanair (1282), the nature of the route is a mix of business and leisure.

(1298) Ryanair contends that this route is seasonal (1283). Ryanair operates on this route in both seasons (1284) whereas Aer Lingus only started to operate on this route in March 2012. However, Aer Lingus has confirmed that tickets for flights on the Dublin-Stockholm route are on sale for the Winter 2012/2013 and the proposed schedule extends for the full duration of the Winter 2012/2013 season (1285). Therefore, the Commission considers that this route could be considered as seasonal, depending on the evolution of the numbers of passengers. The precise qualification does not affect the conclusion drawn in this Decision about this route.

(1299) [...]* of Ryanair's passengers on the Dublin-Stockholm route originate in Ireland (1286).

- Parties' and competitors' operations

(1300) Both Parties operate on the route with non-stop services. Ryanair and Aer Lingus flights overlap on a city-to-city basis. Ryanair operates from Dublin to (NYO) and Aer Lingus from Dublin to ARN.

(1301) Aer Lingus started to operate on this route only in March 2012, while Ryanair has been operating it since October 2007. Ryanair operates a fleet of Boeing 737-800 aircraft. For short-haul routes, Aer Lingus operates essentially a fleet of Airbus A 320 aircraft (1287).

(1302) Since April 2004, SAS offers direct services on the Dublin-Stockholm city pair, namely to ARN. SAS operates essentially a fleet of Boeing 737-800 aircraft.

- Frequencies:

(1303) According to Ryanair, the weekly flights operated by the Parties and the only other operating carrier on this route are as follows (1288):

Table 59: Weekly flights between Dublin and Stockholm

Summer season 2011

Carrier / Weekly Frequency (1 way) : July

Aer Lingus / 0

Ryanair / 5

Combined / 5

SAS / 5

Winter season 2011/2012

Carrier / Weekly Frequency (1 way) : January

Aer Lingus / 0

Ryanair / 3

Combined / 3

SAS / 2

Source: Annex 7.3 (b) (i) as updated on 22 August 2012

(1304) Aer Lingus started to operate on this route (to ARN) only in March 2012, with 4 weekly frequencies both in the shoulder season starting the first week of April and in the peak season starting the first week of July.

(1305) Ryanair operated an average of 5 weekly frequencies in the summer season 2011 and an average of 3 weekly frequencies in the winter season 2011/2012 to NYO. Apart from the Parties, SAS operated 5 weekly frequencies in the summer season 2011 and 2 weekly frequencies in the winter season 2011/2012 to ARN.

(1306) The weekly frequencies for the summer season 2012 for Ryanair and Aer Lingus combined amounted to 7 weekly frequencies, whereas for SAS remained the same as in summer season 2011. Therefore, in the summer season 2012, the Parties had more combined weekly frequencies than SAS.

- Market shares - capacity

(1307) Table 60 indicates the market shares of the Parties and the other only competitor active on the route on the basis of seat capacity in IATA summer season 2011 and in IATA winter season 2011/12 (1289):

Table 60: Market shares for seat capacity

Summer season 2011

Carrier / Weekly Seats available (1 way) : July / Parties' and competitor's market shares

Aer Lingus / 0 / 0%

Ryanair / 945 / 56%

Combined / 945 / 56%

SAS / 730 / 44%

Winter season 2011/2012

Carrier / Weekly Seats available (1 way) : January / Parties' and competitor's market shares

Aer Lingus / 0 / 0%

Ryanair / 567 / 67%

Combined / 567 / 67%

SAS / 280 / 33%

Source: Annex 7.3(b) (i) as updated on 22 August 2012

(1308) Ryanair had a market share of more than 50% in each of the two above presented IATA seasons.

(1309) Aer Lingus entered the Dublin-Stockholm route in March 2012 and took market share away from Ryanair and SAS. According to Ryanair's data, updated on 22 August 2012, as regards IATA summer season 2012, Ryanair had 567 weekly seats (28.7%) available to NYO. Aer Lingus had 696 weekly seats (35%) available to ARN. SAS weekly capacity to ARN amounted to 712 seats (36%). According to DAA data pertaining to the whole IATA summer season 2012, the market shares based on seat capacity are the following: Ryanair [30-40%], Aer Lingus [30-40%] and SAS [30-40%] (1290).

(1310) In its' comments regarding the Commission letter of facts of December 14, 2012, Ryanair argued that the ranges provided for capacity shares misrepresented the actual capacity shares of the relevant airlines on this route for the IATA summer season 2012. However, as stated in the Section 8.9.1., the market shares provided by DAA are replaced by the standard ranges used by the Commission. The suggestion made by Ryanair to use 5% ranges instead of 10% ranges would not alter the Commission's conclusion in any event (1291).

- Market shares - passengers

(1311) Table 61 indicates the market shares of the Parties and the only other competitor active on the route on the basis of passengers carried in IATA summer season 2011 and in IATA winter season 2011/12:

Table 61: Market shares for air services

Dublin - Stockholm

Season / Total pax Annex 7.3.b / Ryanair market share / Aer Lingus Market share / Combined market share / SAS / Total

Summer Season 2011 / [...]* / [60- 70]*% / [0-5]*% / [60-70]*% / [30- 40]*% / 100%

Winter Season 2011/2012 / [...]* / [60- 70]*% / [0-5]*% / [70- 80]* (1292) % / [20- 30]*% / 100%

Source: European Commission on the basis of information provided by Ryanair in Annex 7.3(b) as updated on 22 August 2012

(1312) However, the Commission notes that the market shares for passengers as calculated on the basis of the information provided by Ryanair are not in line (to the disadvantage of Ryanair) with the market shares for passengers calculated on the basis of the more accurate information provided by DAA. The market shares calculated on the basis of the information provided by DAA would then be the following:

Table 62: Market shares for air services on the Dublin-Stockholm route

Season / Combined market share / SAS

Summer season 2011 / [50-60]% / [40-50]%

Winter season 2011/2012 / [50-60]% / [40-50]%

Source: DAA's response to the Commission's request for information of 26 July 2012

(1313) Moreover, on the basis of DAA's data for IATA summer season 2012, the market share of the merged entity, taking Aer Lingus entry into account, would increase to [60-70%] (1293).

(1314) Such high combined market share on this route is itself evidence of a dominant market position.

(1315) It is to be noted that SAS' current very difficult financial situation would indicate that SAS would be a weak competitor to the merged entity in the short to mediumterm (1294).

- Closeness of competition

(1316) Ryanair argues that Aer Lingus and SAS are much closer competitors to each other than they are to Ryanair on this route, in particular because they both operate services to ARN, while Ryanair operates to NYO (1295).

(1317) Ryanair also argues that the product offering of Ryanair and Aer Lingus is significantly different in term of fares and services (1296). The Commission does not contest the view that Aer Lingus is a closer competitor to SAS than Ryanair, given that Aer Lingus positions itself between Ryanair and full service carriers, such as SAS as a value-carrier. In any event, this does not in itself exclude the finding that on the Dublin-Stockholm route Ryanair and Aer Lingus would remain very close if not each other's closest competitor.

(1318) As regards closeness of competition, the responses to the market investigation were largely divided, in favour of SAS, in relation to the Dublin-Stockholm route (1297). A majority of competitors have indicated SAS as the closest competitor to Ryanair and the majority of competitors have considered SAS as the closest competitor to Aer Lingus (1298). The majority of airports considered SAS as the closest competitor to Ryanair, while Ryanair was more often indicated as the closest competitor to Aer Lingus (1299). Most travel agents considered Aer Lingus as the closest competitor to Ryanair, but the views were equally divided as to whether SAS or Ryanair is the closest competitor to Aer Lingus (1300). The majority of corporate customers considered that SAS is the closest to both Ryanair and Aer Lingus (1301).

(1319) However, those who specified both SAS and/or Aer Lingus as the closest competitor to Ryanair either gave no explanation or based their views on either of the following: (i) comparable seat capacities and/or market shares of SAS, Aer Lingus and Ryanair (ii) SAS's strong network brand recognition. Although it is true that all three carriers have had similar market shares in summer 2012, they differed in terms of brand recognition, business models and the extent of operation in Ireland.

(1320) Despite diverging responses to the market investigation and the fact that Aer Lingus and Ryanair fly to different airports, the Commission considers that the Parties are very close, if not each other's closest competitor on the Dublin-Stockholm route. It is apparent that Aer Lingus is a closer competitor to Ryanair in comparison to SAS, a full service carrier. In addition, Aer Lingus and Ryanair appear as very close if not each other's closest in terms of services and business models.

(1321) SAS does not consider that either Aer Lingus or Ryanair is a very close competitor to SAS. SAS deems itself to be "a network carrier, flying to primary airports and targeting passengers that travel mainly for business purposes" (1302). SAS' business model is the one of a traditional network carrier operating a hub and spoke system as opposed to the models of Ryanair and Aer Lingus. It can also be observed from the evidence collected in the market investigation that even on the Dublin-Stockholm route Aer Lingus is less focused on connecting passengers than SAS (SAS carried [10-20]% of connecting passengers, while Aer Lingus only [0-5]% of connecting passengers over the last two IATA seasons) (1303).

(1322) Even though SAS serves the same airport in Stockholm as Aer Lingus, SAS "network is mainly dimensioned according to business travellers' needs, but leisure travel is a growing segment [...]" (1304). SAS offers business class tickets and estimates that [10-20]% are TS passengers on this route (1305). On the contrary, Ryanair does not target business passengers and for Aer Lingus the sales of flexi-fare and plus-fare tickets still represents a very small portion of overall sales and it "does not collect or possess data which enables it to distinguish between different categories of passengers on its short-haul routes" (1306).

(1323) SAS uses a one-way pricing model as does Ryanair and Aer Lingus. However, unlike for Ryanair and Aer Lingus, SAS sells a vast majority of its tickets indirectly (such as, via travel agents) (1307). Ryanair sell all tickets online, while Aer Lingus also sells [80-90]% of tickets directly to end-customers.

(1324) The Commission's regression analysis confirms that the conclusion derived from qualitative evidence that Ryanair and Aer Lingus are close and closest competitors as described in more detail in Annex III.

(1325) According to Aer Lingus, Ryanair remains its primary competitor on the Dublin- Stockholm route, since " [t]he other carrier (SAS) is focused on connecting and business traffic" (1308). Aer Lingus confirms that it offers low fares and a choice of unbundled products, similar to Ryanair. Although SAS operates true one-way pricing, as do the Parties, SAS prices much higher and is monitored less closely by the parties (1309). Aer Lingus also considers that on the Dublin-Stockholm route in case of 5-10% price increase of Ryanair fares only "few passengers would switch to the scheduled airline operating direct services, whilst significant numbers of passengers would switch to Aer Lingus. [...] the scheduled carriers on these routes (all "flag carriers") are often significantly more expensive than Aer Lingus or Ryanair" (1310).

(1326) Further evidence on the closeness of competition between Ryanair and Aer Lingus can be observed from the changes in the seat capacity and passenger market share data in response to a recent entry by Aer Lingus in late Winter 2011/2012. Since entry, in the summer of 2012 Aer Lingus has gained a significant market share of 35% in terms of seat capacity and [20-30]% in terms of point-to-point passenger numbers (DAA data) largely at the expense of Ryanair, rather than SAS. In particular, the market share of Aer Lingus has increased from [0-5]% to [20-30]%, while the market share of Ryanair decreased from [50-60]% to [30-40]%. The market share of SAS changed only slightly by [0-5]%.

(1327) The Commission furthermore considers, as set out in Section 8.4, that the Parties are each other's closest competitors on the Dublin-Stockholm route with respect to Irish originating passengers, based on their high brand recognition in Ireland, strong overall market positions and bases in Ireland.

(1328) While Ryanair and Aer Lingus both enjoy a strong brand recognition on the routes to and from Ireland (1311), according to SAS, its brand is stronger in Sweden than in Ireland and primarily in "segment(s) requiring service, punctuality and reliability" (1312), not low fares (1313). Similarly, according to Aer Lingus, Ryanair remains its closest competitor on the Dublin-Stockholm route: "[t]his remains unchanged from the 2007 decision: as the strongest Irish carriers Ryanair and Aer Lingus clearly enjoy the best brand recognition on routes into and out of Ireland" (1314).

(1329) Both Ryanair and Aer Lingus have the strongest city presence in Dublin and therefore an advantage in winning Irish-originating customers who represent a significant share of passengers on the Dublin-Stockholm route. The evidence collected in the market investigation shows that both Ryanair and Aer Lingus carried a higher ratio of passenger originating in Ireland compared to SAS over the last three IATA seasons: that is to say, [...]* % for Ryanair (1315), [40-50] % for Aer Lingus (1316) and only [20-30]* % for SAS (1317).

(1330) Furthermore, SAS has no based aircraft at Dublin airport, and therefore does not enjoy the same base advantages as Ryanair and Aer Lingus at the Irish side of the route. Similarly, SAS does not exert the same level of constraint from its hub in Stockholm, as Aer Lingus and Ryanair do on each other. Ryanair and Aer Lingus exert dynamic competition on each other (that is to say, on a variety of routes) as they have significant bases in Dublin and have a similar degree of flexibility to react to demand-side (or supply-side) shocks. For SAS, Dublin represents only one route from its destination hub, and it is therefore less committed to adjusting its capacity to match fluctuations in demand on the variety of routes from and to Dublin. As a network carrier, SAS has more focus on connecting traffic to its hub.

(1331) Therefore it is considered that Ryanair and Aer Lingus are very close competitors, if not each other's closest competitor, on the Dublin-Stockholm route. The Transaction would lead to the elimination of the competitive rivalry existing between Ryanair and Aer Lingus, which has been an important source of competition on this route, in particular for point-to-point passengers.

- Entry and expansion

(1332) SAS started to operate the Dublin-Stockholm route in 2004. Ryanair entered the route in October 2007. Aer Lingus entered in March 2012.

(1333) Flynordic started operations on this route in April 2006 but exited in September 2007, following its acquisition by Norwegian. (1318)

(1334) Ryanair argues that Norwegian has a base at ARN and is a potential competitor that can enter the route at any time (1319).

(1335) The route is subject to the general barriers to entry described in Section 8.5 and to the assessment described in Sections 8.6 and 8.7.

(1336) In addition, ARN is a level 3 coordinated airport, and is slot constrained. According to the ACS (1320), the airport coordinator and Swedavia AB (1321), there are runway constraints during peak hours, in particular in the morning from 06.55 local time to 09.15 local time. However, Swedavia AB indicates that "even though there are some constraints, we do not see future problems based on today's prognosis" (1322).

(1337) Finally, the evidence collected in the market investigation indicates that no competitor would have post-Transaction entry or expansion projects which could be considered as timely, likely and sufficient enough to constitute a competitive constraint on the merged entity and defeat the anti-competitive effects of the Transaction on the Dublin - Stockholm route.

- Conclusion

(1338) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-Stockholm route.

(1339) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that SAS would be able to constrain the merged entity's market behaviour sufficiently, especially with regard to fare setting, on the Dublin-Stockholm route.

8.9.5.8. Conclusion on the routes where other scheduled carriers operate

(1340) In view of the foregoing the Commission concludes that the Transaction is likely to significantly impede effective competition on the following 7 routes where other scheduled carriers operate: Dublin-Bristol/Cardiff/Exeter; Dublin-Frankfurt; Dublin- London; Dublin-Madrid; Dublin-Munich; Dublin-Paris and Dublin-Stockholm.

8.9.6. Routes where charter companies operate

(1341) There are 11 routes on which charter companies and the Parties operate: Dublin- Malaga; Dublin-Faro; Dublin-Barcelona; Dublin-Palma; Dublin-Lanzarote; Dublin- Gran Canaria; Dublin-Ibiza; Cork-Barcelona; Cork-Lanzarote; Cork-Malaga and Cork-Palma.

8.9.6.1. Closeness of competition

(1342) As assessed in particular in Section 8.4.2, charter companies are only distant competitors to the Parties and therefore cannot be considered to constitute a significant competitive constraint to Ryanair and Aer Lingus on the overlap routes. The Parties are each other's closest competitors considering that each of the Parties is by far the largest competitor of the other Party on all the routes where charter companies are active.

(1343) Charter companies are not close competitors to the Parties as they often operate on a seasonal basis with relatively low flight frequencies set in response to the requirements of tour operators. In particular, on seven out of the 11 routes, charter companies operate lower flight frequencies than Ryanair or Aer Lingus (Dublin- Barcelona, Dublin-Faro, Dublin-Gran Canaria, Dublin-Ibiza, Dublin-Malaga, Cork- Barcelona, Cork-Malaga). On eight out of the 11 routes (Dublin-Barcelona, Dublin- Faro, Dublin-Ibiza, Dublin-Malaga, Dublin-Palma, Cork-Barcelona, Cork-Malaga, and Cork-Palma), charter companies do not operate, or operate negligibly, in winter season, while both Aer Lingus and Ryanair offer scheduled services.

(1344) In this respect, Ryanair considers that a number of charter companies offer dry seats on the 11 routes where charter companies are active, and that they have proven ability to expand capacity (1323). Ryanair also submits that by dismissing charter companies the Commission "ignores the fact that many of the overlap routes are seasonal routes and, although charter carriers' average frequencies during an entire season may be low, actual frequencies flown during periods of significant customer demand likely exceed the season-wide average charter frequencies" (1324).

(1345) As set out in Section 8.9.1 regarding methodology, the Commission has found that the frequency data appears adequate for the competitive assessment. In any event, even on the routes where charter companies offer a somewhat comparable number of frequencies to Ryanair and Aer Lingus, the sales of dry seats still represent only a small part of charter companies' business activities. In fact, based on the market share data, on all overlap routes the share of dry seats did not exceed [10-20] % in the summer season, and [5-10]% in winter season, both in terms of weekly seat capacity and number of passengers. In most cases the market shares of charter companies were even lower. In this respect the business models of all the charter companies that are active on the 11 routes is materially different from that of the Parties.

(1346) The market investigation has also confirmed that charter companies are not the closest competitors of either of the Parties. The travel agencies that sell package holidays named mostly Ryanair and Aer Lingus to be each other's closest competitor on the charter routes at issue (1325). Similarly, the tour operators, to which charter companies sell seats or entire flights, named mostly Ryanair and Aer Lingus to be each other's closest competitor on the charter routes at issue (1326). This was confirmed by the charter companies themselves (1327) as well as by the corporate customers (1328).

(1347) Finally, Ryanair submits that the Commission has wrongly dismissed charter companies as actual and potential constraints on the merged entity [...]* and because no data on passenger volumes was readily available. Ryanair states that it cannot monitor charter companies simply because there is no source of pricing evidence to monitor. However, this does not appear consistent with the information provided by Ryanair. For instance, in Annex 7.7 to the Form CO, Ryanair specified that it monitors a number of carriers, including Thomson, Thomas Cook, Monarch and Germania which are charter companies. However, Ryanair does not monitor these airlines on the charter routes concerned by this Decision despite their presence (1329). On the contrary, Ryanair actively monitors Aer Lingus on the 11 routes, where charter operators are active.

(1348) Similarly, Aer Lingus does not take into account the prices of charter companies and does not believe that charter companies impose a competitive constraint on it (1330).

(1349) The Commission's regression analysis confirms and complements the conclusion derived from qualitative evidence that Ryanair and Aer Lingus are close and each other's closest competitors as described in more detail in Annex III.

(1350) In any event, although Ryanair's entry on certain leisure routes may have had an impact on charter operations, it does not mean that charter companies constrain Ryanair to the same extent as Aer Lingus, that essentially offers point-to-point services on these routes.

(1351) Therefore it is considered that Ryanair and Aer Lingus are each other's closest competitor on the 11 routes on which charter companies compete with the Parties. The Transaction would lead to the elimination of the competitive rivalry existing between Ryanair and Aer Lingus, which has been an important source of competition on those routes, in particular for point-to-point passengers.

8.9.6.2. No significant constraints exercised by indirect services

(1352) The absence of significant constraints by indirect services has been developed in Section 8.9.3. The same conclusion applies to routes on which charter companies operate.

8.9.6.3. Entry/exit

(1353) Evidence collected in the market investigation indicates that no charter company would have post-Transaction entry projects which could be considered as timely, likely and sufficient enough to constitute a competitive constraint on the merged entity and offset the anti-competitive effects of the Transaction on the routes where charter companies operate.

8.9.6.4. Dublin-Barcelona

- Route characteristics

(1354) On the Dublin-Barcelona route, [...]* passengers travelled with the Parties in IATA summer season 2011, while [...]* passengers travelled in IATA winter season 2011/12. These figures include Ryanair passengers from Girona (GRO) and Reus (REU), as Ryanair not only operates to Barcelona city airport (BCN), but also these latter airports (1331). However, as Ryanair did not provide numbers for the passengers travelling on charter flights on this route (1332), the total number of passengers including those charter passengers cannot be indicated (1333).

(1355) According to Ryanair, Dublin-Barcelona is a mixed business - leisure route (1334).

(1356) The route overlaps are not only seasonal. Both Ryanair and Aer Lingus operate this route to BCN in both seasons. Moreover, Ryanair also operates its services to Girona and Reus in both seasons (1335).

(1357) With regard to the airport pairs DUB-BCN and DUB-GRO [...]*% of Ryanair's passengers originate from the Irish end of the route. On DUB-REU this percentage is higher, namely [...]*% (1336).

- Parties' operations

(1358) Both Parties operate on the route with non-stop services. Ryanair and Aer Lingus overlap on an airport-to-airport basis with regard to DUB-BCN and moreover on a city-to-city basis with regard to Ryanair's additional operations to GRO and REU.

(1359) Aer Lingus entered this route in March 2002. Ryanair first entered DUB-GRO in 2003, then added DUB-REU in 2004 and finally also entered DUB-BCN in September 2012. According to Ryanair, Thomson entered this route in 2010. Iberia operated the airport pair DUB-BCN from 2004 to 2007. Clickair was active between this airport pair from March 2007 to September 2008 (1337).

- Frequencies

(1360) Table 63 indicates the weekly flights operated by the Parties on this route. The average frequency for the charter flights is based on DAA data.

Table 63: Weekly flights between Dublin and Barcelona

Summer season 2011

Carrier / Weekly Frequency (1 way) : July

Aer Lingus / 13

Ryanair / 7 (1338)

Combined / 20

All charters / 3

Winter season 2011/2012

Carrier / Weekly Frequency (1 way) : January

Aer Lingus / 11

Ryanair / 7 (1339)

Combined / 18

All charters / 0

Source: Annex 7.3 (b) as updated on 22 August 2012 and Annex "Question 1.6" to DAA's response to the Commission's request for information of 20 September 2012 (1340)

(1361) In the IATA summer season 2011 Ryanair operated 7 weekly frequencies between the airports DUB-BCN; Aer Lingus operated 13 weekly frequencies between the same airport pair. Three charter companies each operated on average 1 weekly frequency between the airport pair DUB-REU.

(1362) In the IATA winter season 2011/12 Ryanair operated 7 weekly frequencies on DUBBCN, and Aer Lingus operated 11 weekly frequencies on DUB-BCN. No charter company operated the route at issue during winter seasons 2011/12.

(1363) In the IATA summer season 2012 Ryanair operated a total of 18 weekly frequencies on the route Dublin-Barcelona (7 on DUB-BCN, 4 on DUB-GRO and 7 on DUB-REU), while Aer Lingus increased its frequencies to 14 weekly frequencies. Two charter companies each operated on average 1 weekly frequency between the airport pair DUB-REU.

(1364) Charter companies thus operated only on DUB-REU and only during the summer seasons.

- Market shares - capacity

(1365) Table 64 indicates the market shares of the Parties and the charter companies active on this route on the basis of seat capacity in the IATA summer season 2011 and in the IATA winter season 2011/12. As Ryanair did not provide data in this respect for charter companies, the market shares were computed on data made available by DAA.

Table 64: Market shares for seat capacity

Summer season 2011

Carrier / Market shares

Aer Lingus / [40-50%]

Ryanair / [50-60%]

Combined / [90-100%]

Charters / [0-5%]

Winter season 2011/2012

Carrier / Market shares

Aer Lingus / [50-60%]

Ryanair / [40-50%]

Combined / 100%

Charters / 0

Source: DAA's response to the Commission's request for information of 26 July 2012.

(1366) As regards the IATA summer season 2012, Ryanair had a market share of [50-60%] and Aer Lingus a market share of [30-40%]. The charter companies represented [0- 5%] of the seat capacity on this route (1341).

(1367) It follows that post-Transaction, the Parties would have a quasi-monopoly on the route in the summer season and a monopoly in the winter season.

- Market shares - passengers

(1368) Table 65 indicates the market shares of the Parties and of the charter companies active on the route on the basis of passengers carried in the IATA summer season 2011 and in the IATA winter season 2011/12.

Table 65: Market shares for air services

Summer season 2011

Carrier / Market shares

Aer Lingus / [30-40%]

Ryanair / [60-70%]

Combined / [90-100%]

Charters / [0-5%]

Winter season 2011/2012

Carrier / Market shares

Aer Lingus / [50-60%]

Ryanair / [40-50%]

Combined / 100%

Charters / 0

Source: DAA's response to the Commission's request for information of 26 July 2012

(1369) As regards the IATA summer season 2012, Ryanair had a market share of [60-70%] and Aer Lingus of [30-40%]. The charter companies carried a share of [0-5%] of the passengers on this route (1342).

(1370) It follows that post-Transaction, the Parties would have a quasi-monopoly on the route in the summer season and a monopoly in the winter season.

- Conclusion

(1371) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-Barcelona route.

(1372) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that charter companies would be able to constrain sufficiently the merged entity's market behaviour, especially with regard to fare setting, on the Dublin-Barcelona route.

8.9.6.5. Dublin-Faro - Route characteristics

(1373) On the Dublin-Faro route, [...]* passengers travelled in IATA summer season 2011 while [...]* passengers travelled in IATA winter season 2011/12 (1343).

(1374) According to Ryanair, Dublin-Faro is predominantly a leisure route (1344).

(1375) However, the route overlaps are not only seasonal, as both Ryanair and Aer Lingus operate this route in the summer and the winter seasons (1345).

(1376) According to Ryanair [...]*% of its passengers originate from the Irish end of the route (1346).

- Parties' operations

(1377) Both Parties operate on the route with non-stop services and overlap on an airport-toairport basis.

(1378) Aer Lingus entered this route in 2002. Ryanair subsequently entered in March 2003.

- Frequencies

(1379) Table 66 indicates the weekly flights operated by the Parties on this route. The average frequency for the charter companies is based on DAA data.

Table 66: Weekly flights between Dublin and Faro

Summer season 2011

Carrier / Weekly Frequency (1 way) : July

Aer Lingus / 11

Ryanair / 18

Combined / 29

All charters / 4

Winter season 2011/2012

Carrier / Weekly Frequency (1 way) : January

Aer Lingus / 3

Ryanair / 3

Combined / 6

All charters / 0

Source: Annex 7.3 (b) as updated on 22 August 2012 and Annex "Question 1.6" to DAA's response to the Commission's request for information of 20 September 2012 (1347)

(1380) In the IATA summer season 2011 Ryanair operated 18 weekly frequencies and Aer Lingus operated 11 weekly frequencies on the route. Two charter companies operated each on average 1 weekly frequency, while a third charter company operated on average 2 weekly frequencies on the route.

(1381) In the IATA winter season 2011/12 Ryanair and Aer Lingus both operated 3 weekly frequencies. No charter company operated on the route during the winter season 2011/12.

(1382) In the IATA summer season 2012, Ryanair increased its weekly frequencies to 19, and Aer Lingus increased its weekly frequencies to 14. The Parties thus operated a total of 33 weekly frequencies on the route at issue. The number of frequencies by the charter companies decreased to a total of 3 weekly frequencies.

(1383) Charter companies thus operate on the Dublin-Faro route only during the summer seasons.

- Market shares - capacity

(1384) Table 67 indicates the market shares of the Parties and the charter companies active on this route on the basis of seat capacity in the IATA summer season 2011 and in the IATA winter season 2011/12. As Ryanair did not provide data in this respect for charter companies, the market shares were computed on data made available by DAA.

Table 67: Market shares for seat capacity

Summer season 2011

Carrier / Market shares

Aer Lingus / [40-50%]

Ryanair / [50-60%]

Combined / [90-100%]

Charters / [0-5%]

Winter season 2011/2012

Carrier / Market shares

Aer Lingus / [40-50%]

Ryanair / [50-60%]

Combined / 100%

Charters / 0

Source: DAA's response to the Commission's request for information of 26 July 2012

(1385) As regards IATA summer season 2012 Ryanair had a market share of [40-50%] and Aer Lingus a market share of [40-50%]. The charter companies represented between [0-5%] (1348).

(1386) It follows that post-Transaction, the Parties would have a quasi-monopoly on the route in the summer season and a monopoly in the winter season.

- Market shares - passengers

(1387) Table 68 indicates the market shares of the Parties and of the charter companies active on the route on the basis of passengers carried in the IATA summer season 2011 and in IATA winter season 2011/12.

Table 68: Market shares for air services

Summer season 2011

Carrier / Market shares

Aer Lingus / [40-50%]

Ryanair / [50-60%]

Combined / [90-100%]

Charters / [0-5%]

Winter season 2011/2012

Carrier / Market shares

Aer Lingus / [40-50%]

Ryanair / [50-60%]

Combined / 100%

Charters / 0

Source: DAA's response to the Commission's request for information of 26 July 2012

(1388) As regards IATA summer season 2012 Ryanair had a market share of [40-50%] and Aer Lingus a share of [40-50%]. The charter companies represented between [0- 5%] (1349).

(1389) It follows that post-Transaction, the Parties would have a quasi-monopoly on the route in the summer season and a monopoly in the winter season.

- Conclusion

(1390) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-Faro route.

(1391) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that charter companies would be able to constrain sufficiently the merged entity's market behaviour, especially with regard to fare setting, on the Dublin-Faro route.

8.9.6.6. Dublin-Gran Canaria

- Route characteristics

(1392) On the Dublin-Gran Canaria route, [...]* passengers travelled in the IATA summer season 2011 while [...]* passengers travelled in the IATA winter season 2011/12 (1350).

(1393) According to Ryanair (1351), the Dublin-Gran Canaria route is a leisure route.

(1394) [...]*% of Ryanair's passengers on this route originate at the Irish end of the route (1352).

- Parties' operations

(1395) Both Parties operate on the route with non-stop services. Ryanair and Aer Lingus overlap on an airport-to-airport basis. Both operate from Dublin to Las Palmas airport (LPA).

(1396) Aer Lingus started to operate on this route in winter 2004/2005 while Ryanair has been operating on it since winter 2009/2010 (1353).

- Frequencies

(1397) Table 69 indicates the weekly flights operated by the Parties on this route. The average frequency for the charter companies is based on DAA data.

Table 69: Weekly flights between Dublin and Gran Canaria

Summer season 2011

Carrier / Weekly Frequency (1 way) : July

Aer Lingus / 3

Ryanair / 4

Combined / 7

All charters / 2

Winter season 2011/2012

Carrier / Weekly Frequency (1 way) : January

Aer Lingus / 3

Ryanair / 3

Combined / 6

All charters / 1

Source: Annex 7.3 (b) as updated on 22 August 2012 and Annex "Question 1.6" to DAA's response to the Commission's request for information of 20 September 2012 (1354)

(1398) In the IATA summer season 2011, Ryanair operated 4 weekly frequencies and Aer Lingus operated 3 weekly frequencies to LPA. In the IATA winter season 2011/12, Ryanair operated 3 weekly frequencies and Aer Lingus operated 3 weekly frequencies to LPA. On the Dublin-Gran Canaria route, the Parties would therefore have combined 7 weekly frequencies for the IATA summer season 2011 and 6 weekly frequencies for the IATA winter season 2011/2012.

(1399) Two charter companies each operated 1 weekly frequency in the IATA summer season 2011, while in in the IATA winter season 2011/2012 the only charter company active on the route operated 1 weekly frequency.

(1400) The frequencies for the IATA summer season 2012 for Ryanair and Aer Lingus combined amounted to around 7 weekly frequencies, whereas the frequencies for charter companies amounted to 2 in total.

- Market shares - capacity

(1401) Table 70 indicates the market shares of the Parties on the basis of seat capacity in the IATA summer season 2011 and in the IATA winter season 2011/12. As Ryanair did not provide data in this respect for charter companies, the market shares were computed on data made available by DAA.

Table 70: Market shares for seat capacity

Summer season 2011

Carrier / Market shares

Air Lingus / [40-50%]

Ryanair / [50-60%]

Combined / [90-100%]

Charters / [0-5%]

Winter season 2011/2012

Carrier / Market shares

Air Lingus / [40-50%]

Ryanair / [50-60%]

Combined / [90-100%]

Charters / [0-5%]

Source: DAA's response to the Commission's request of information of 26 July 2012

(1402) As regards IATA summer season 2012 (1355), Ryanair had a market share of [50-60%] and Aer Lingus a market share of [40-50%]. The charter companies represented between [5-10%].

(1403) In its Comments regarding the Commission Letter of Facts of 14 December 2012, Ryanair stated that one charter company Thomson had a market share of [10-20]*% and therefore contested the capacity ranges based on DAA data for the IATA summer season 2012. However, the ranges indicated in the Decision for that season represent the market shares of the Parties and of charter companies active on this route, based on DAA data for the entire summer season and on the very conservative assumption that for charter companies the ratio of dry seats amounts to [...]*% of passengers. As apparent from Ryanair's response to the request for information of 7 January 2013, the Innovata figures used by Ryanair do not distinguish dry seats from the total seat capacity offered by Thomson. In any event, the use of Ryanair's data would not affect the Commission's conclusion, in particular as the Parties combined capacity market share would add up to [80-90]*%.

(1404) The very high combined market share on this route is itself evidence of a dominant market position.

- Market shares - passengers

(1405) Table 71 indicates the market shares of the Parties and of the charter companies on the route at issue on the basis of passengers carried in IATA summer season 2011 and in IATA winter season 2011/12. Table 71: Market shares for air services Summer season 2011 Carrier Market shares Air Lingus [40-50%] Ryanair [50-60%] Combined [90-100%] Charters [0-5%] Winter season 2011/2012 Carrier Market shares Air Lingus [40-50%] Ryanair [50-60%] Combined [90-100%] Charters [0-5%] Source: DAA's response to the Commission's request of information of 26 July 2012

(1406) As regards IATA summer season 2012, Ryanair had a market share of [50-60%] and Aer Lingus of [40-50%]. The charter companies represented between [5-10%] (1356).

(1407) Such high combined market share on this route is in itself evidence of a dominant market position.

- Conclusion

(1408) In view of the foregoing, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-Gran Canaria route.

(1409) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that charter companies would be able to constrain the merged entity's market behavior sufficiently, especially with regard to fare setting, on the Dublin-Gran Canaria route.

8.9.6.7. Dublin-Ibiza

- Route characteristics

(1410) On the Dublin-Ibiza route, [...]*passengers travelled in IATA summer season 2011 while [...]*passengers travelled in the IATA winter season 2011/12 (1357).

(1411) According to Ryanair (1358), the Dublin-Ibiza route is a leisure route.

(1412) The route is a summer only route (1359).

(1413) [...]*% of Ryanair's passengers on this route originate at the Irish end of the route (1360).

- Parties' operations

(1414) Both Parties operate on the route with non-stop services in the summer season only. Ryanair and Aer Lingus overlap on an airport-to-airport basis. Both operate from Dublin to Ibiza (IBZ).

(1415) Ryanair is the only airline that operates during the winter season.

(1416) Aer Lingus started to operate on this route in summer 2008 while Ryanair has been operating on it since the winter 2011 (1361).

- Frequencies

(1417) Table 72 indicates the weekly flights operated by the Parties on this route. The average frequency for the charter companies is based on DAA data.

Table 72: Weekly flights between Dublin and Ibiza

Summer season 2011

Carrier / Weekly Frequency (1 way) : July

Aer Lingus 3

Ryanair 2

Combined 5

All charters 1

Annex 7.3 (b) as updated on 22 August 2012 and Annex "Question 1.6" to DAA's response to the Commission's request for information of 20 September 2012 (1362)

(1418) In the IATA summer season 2011, Ryanair operated 2 weekly frequencies and Aer Lingus operated 3 weekly frequencies to Ibiza (IBZ). On the Dublin-Ibiza route, the Parties combined would therefore have 5 weekly frequencies for the IATA summer season 2011.

(1419) 1 weekly frequency was operated by a charter company in IATA summer season 2011.

(1420) In the IATA summer season 2012, the frequencies for Ryanair and Aer Lingus combined amounted to around 7 weekly frequencies, whereas for the charter company active on this route the figure was 1 weekly frequency.

e) Market shares - capacity

(1421) Table 73 indicates the market shares of the Parties on the basis of seat capacity in the IATA summer season 2011. As Ryanair did not provide data in this respect for charter companies, the market shares were computed on data made available by DAA.

Table 73: Market shares for seat capacity

Summer season 2011

Carrier / Market shares

Air Lingus / [40-50%]

Ryanair / [40-50%]

Combined / [90-100%]

Charters / [0-5%]

Source: DAA's response to the Commission's request for information of 26 July 2012

(1422) As regards IATA summer season 2012, Ryanair had a market share of [50-60%] and Aer Lingus a share of [40-50%]. Charter services represented a market share of between [0- 5%] (1363).

(1423) The very high combined market share of the Parties on this route is itself evidence of a dominant market position.

f) Market shares - passengers

(1424) Table 74 indicates the market shares of the Parties and of the charter companies on the route at issue on the basis of passengers carried in the IATA summer season 2011:

Table 74: Market shares for air services

Summer season 2011

Carrier / Market shares

Air Lingus / [40-50%]

Ryanair / [40-50%]

Combined / [90-100%]

Charters / [0-5%]

Source: DAA's response to the Commission's request for information of 26 July 2012.

(1425) As regards IATA summer season 2012, Ryanair had a market share of [50-60%] and Aer Lingus of [40-50%]. Charter services represented a market share of between [0- 5%] (1364).

(1426) Such a high combined market share of the Parties on this route is in itself evidence of a dominant market position.

- Conclusion

(1427) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-Ibiza route.

(1428) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that charter companies would be able to constrain the merged entity's market behavior sufficiently, especially with regard to fare setting, on the Dublin-Ibiza route.

8.9.6.8. Dublin-Lanzarote

- Route characteristics

(1429) On the Dublin-Lanzarote route, [...]* passengers travelled in IATA summer season 2011 while [...]* passengers travelled in IATA winter season 2011/12 (1365).

(1430) According to Ryanair (1366), the Dublin-Lanzarote route is a leisure route.

(1431) [...]* of Ryanair's passengers on this route originate at the Irish end of the route (1367).

- Parties' operations

(1432) Both Parties operate on the route with non-stop services. Ryanair and Aer Lingus overlap on an airport-to-airport basis. Both operate from Dublin to Lanzarote airport (ACF).

(1433) Aer Lingus started to operate on this route in October 2006 while Ryanair has been operating on it since October 2009 (1368).

- Frequencies

(1434) Table 75 indicates the weekly flights operated by the Parties on this route. The average frequency for the charter companies is based on DAA data.

Table 75: Weekly flights between Dublin and Lanzarote

Summer season 2011

Carrier / Weekly Frequency (1 way) : July

Aer Lingus / 6

Ryanair / 4

Combined / 10

All charters / 6

Winter season 2011/2012

Carrier / Weekly Frequency (1 way) : January

Aer Lingus / 3

Ryanair / 4

Combined / 7

All charters / 2

Annex 7.3 (b) as updated on 22 August 2012 and Annex "Question 1.6" to DAA's response to the Commission's request for information of 20 September 2012 (1369)

(1435) In the IATA summer season 2011, Ryanair operated 4 weekly frequencies and Aer Lingus operated 6 weekly frequencies to Lanzarote. In the IATA winter season 2011/12, Ryanair operated 4 weekly frequencies and Aer Lingus operated 3 weekly frequencies to Lanzarote. On the Dublin-Lanzarote route, the Parties combined would therefore have 10 weekly frequencies for the IATA summer season 2011 and 7 weekly frequencies for the IATA winter season 2011/2012.

(1436) In the IATA summer season 2011, charter companies operated on average 6 weekly frequencies (namely, four charter companies each operated on average 1 weekly frequency, while one charter company operated on average 2 weekly frequencies) and in IATA winter season 2011/2012 charter companies operated on average 2 weekly frequencies (namely, two charter companies each operated on average 1 weekly frequency).

(1437) The frequencies for the IATA summer season 2012 for Ryanair and Aer Lingus combined amounted to around 10 weekly frequencies, whereas the total frequencies for charter companies amounted to 5.

- Market shares - capacity

(1438) Table 76 indicates the market shares of the Parties on the basis of seat capacity in IATA summer season 2011 and in IATA winter season 2011/12. As Ryanair did not provide data in this respect for charter companies, the market shares were computed on data made available by DAA.

Table 76: Market shares for seat capacity

Summer season 2011

Carrier / Market shares

Air Lingus / [30-40%]

Ryanair / [50-60%]

Combined / [80-90%]

Charters / [10-20%]

Winter season 2011/2012

Carrier / Market shares

Air Lingus / [50-60%]

Ryanair / [40-50%]

Combined / [90-100%]

Charters / [0-5%]

Source: DAA's response to the Commission's request for information of 26 July 2012

(1439) As regards the IATA summer season 20121370, Ryanair had a market share of [50-60%] and Aer Lingus a market share of [40-50%]. The charter companies represented a market share of between [5-10%].

(1440) In its Comments regarding the Commission Letter of Facts of 14 December 2012, Ryanair stated that one charter company Thomson had a market share of 18% and therefore contested the capacity ranges based on DAA data for the IATA summer season 2012. However, the ranges indicated in the Decision for that season represent the market shares of the Parties and of charter companies active on this route, based on DAA's data for the entire summer season and on the very conservative assumption that for charter companies the ratio of dry seats amounts to 20% of passengers. As apparent from Ryanair's response to the request for information of 7 January 2013, the Innovata figures used by Ryanair do not distinguish dry seats from the total seat capacity offered by Thomson. In any event, the use of Ryanair's data would not affect the Commission's conclusion, in particular as the Parties combined capacity share would add up to 82% if calculated using Ryanair's data.

(1441) The very high combined market share on this route is itself evidence of a dominant market position.

- Market shares - passengers

(1442) Table 77 indicates the market shares of the Parties and of the charter companies on the route at issue on the basis of passengers carried in the IATA summer season 2011 and in the IATA winter season 2011/12.

Table 77: Market shares for air services

Summer season 2011

Carrier / Market shares

Air Lingus / [30-40%]

Ryanair / [50-60%]

Combined / [80-90%]

Charters / [10-20%]

Winter season 2011/2012

Carrier / Market shares

Air Lingus / [40-50%]

Ryanair / [40-50%]

Combined / [90-100%]

Charters / [0-5%]

Source: DAA's response to the Commission's request for information of 26 July 2012

(1443) As regards the IATA summer season 2012, Ryanair had a market share of [50-60%] and Aer Lingus of [40-50%]. The charter companies represented a market share of between [5-10%] (1371).

(1444) Such high combined market share on this route is in itself evidence of a dominant market position.

- Conclusion

(1445) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-Lanzarote route.

(1446) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that charter companies would be able to constrain the merged entity's market behavior sufficiently, especially with regard to fare setting, on the Dublin-Lanzarote route. 8.9.6.9. Dublin-Malaga - Route characteristics

(1447) On the Dublin-Malaga route, [...]* passengers travelled in the IATA summer season 2011 while [...]* passengers travelled in the IATA winter season 2011/12 (1372).

(1448) According to Ryanair, Dublin-Malaga is predominantly a leisure route (1373).

(1449) The route overlaps are not only seasonal. Both Ryanair and Aer Lingus operate on this route in the summer and the winter seasons.

(1450) [...]* of Ryanair's passengers originate from the Irish end of the route (1374).

- Parties' operations

(1451) Both Parties operate on the route with non-stop services and overlap on an airport-toairport basis.

(1452) Ryanair started operating on this route in March 2003 and Aer Lingus has been operating on it since before 2004.

- Frequencies

(1453) Table 78 indicates the weekly flights operated by the Parties on this route. The average frequency for the charter companies is based on DAA data. Table 78: Weekly flights between Dublin and Malaga

<TABLE>

Source: Annex 7.3 (b) as updated on 22 August 2012 and Annex "Question 1.6" to DAA's response to the Commission's request for information of 20 September 2012 (1375)

(1454) In the IATA summer season 2011, Ryanair operated 16 weekly frequencies, while Aer Lingus operated 14 weekly frequencies. Two charter companies each operated on average 1 weekly frequency.

(1455) In the IATA winter season 2011/12, Ryanair operated 4 weekly frequencies, while Aer Lingus operated 7 weekly frequencies. No charter company operated this route during the winter seasons 2011/12.

(1456) The frequencies for the IATA summer season 2012 for Ryanair amounted to 16 weekly frequencies, for Aer Lingus to 14 weekly frequencies and for charter companies to a total of 2 weekly frequencies.

(1457) Charter companies thus operated only during the summer seasons.

- Market shares - capacity

(1458) Table 79 indicates the market shares of the Parties and the charter companies active on this route on the basis of seat capacity in the IATA summer season 2011 and in the IATA winter season 2011/12. As Ryanair did not provide data in this respect for charter companies, the market shares were computed on data made available by DAA.

Table 79: Market shares for seat capacity

Summer season 2011

Carrier / Market shares

Aer Lingus / [50-60%]

Ryanair / [40-50%]

Combined / [90-100%]

Charters / [0-5%]

Winter season 2011/2012

Carrier / Market shares

Aer Lingus / [50-60%]

Ryanair / [40-50%]

Combined / 100%

Charters / 0

Source: DAA's response to the Commission's request for information of 26 July 2012.

(1459) As regards IATA summer season 2012, Ryanair had a market share of [40-50%] and Aer Lingus a market share of [50-60%]. Charter services represented a market share of between [0-5%] of passengers (1376).

(1460) It follows that post-Transaction, the Parties would have a quasi-monopoly on the route in the summer season and a monopoly in the winter season.

- Market shares - passengers

(1461) Table 80 indicates the market shares of the Parties and of the charter companies active on the route on the basis of passengers carried in the IATA summer season 2011 and in the IATA winter season 2011/12.

Table 80: Market shares for air services

Summer season 2011

Carrier / Market shares

Aer Lingus / [50-60%]

Ryanair / [40-50%]

Combined / [90-100%]

Charters / [0-5%]

Winter season 2011/2012

Carrier / Market shares

Aer Lingus / [50-60%]

Ryanair / [40-50%]

Combined / 100%

Charters / 0

Source: DAA's response to the Commission's request for information of 26 July 2012

(1462) As regards IATA summer season 2012, Ryanair had a market share of [40-50%] and Aer Lingus of [50-60%]. The charter companies carried between [0-5%] of the passengers travelling on this route (1377).

(1463) It follows that post-Transaction, the Parties would have a quasi-monopoly on the route in the summer season and a monopoly in the winter season.

- Conclusion

(1464) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-Malaga route.

(1465) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that charter companies would be able to constrain the merged entity's market behavior sufficiently, especially with regard to fare setting, on the Dublin-Malaga route.

8.9.6.10. Dublin-Palma

- Route characteristics

(1466) On the Dublin-Palma route, [...]* passengers travelled in IATA summer season 2011 while [...]* passengers travelled in IATA winter season 2011/12 (1378).

(1467) According to Ryanair, the Dublin-Palma route is predominantly a leisure route (1379).

(1468) The route overlaps are only seasonal. Both Ryanair and Aer Lingus have been operating this route only during summer seasons.

(1469) [...]* of Ryanair's passengers originate from the Irish end of the route (1380).

- Parties' operations

(1470) Both Parties operate on the route with non-stop services and overlap on an airport-toairport basis.

(1471) Aer Lingus started this route in April 2006. Subsequently Ryanair entered it in March 2008.

- Frequencies

(1472) Table 81 indicates the weekly flights operated by the Parties on this route. The average frequency for the charter companies is based on DAA data.

Table 81: Weekly flights between Dublin and Palma

<TABLE>

Source: Annex 7.3 (b) as updated on 22 August 2012 and Annex "Question 1.6" to DAA's response to the Commission's request for information of 20 September 2012 (1381)

(1473) In the IATA summer season 2011 Ryanair operated 6 weekly frequencies, while Aer Lingus operated 4 weekly frequencies. Three charter companies each operated on average 1 weekly frequency on the route at issue, while another charter company operated on average 4 weekly frequencies.

(1474) In the IATA winter season 2011/12 neither Ryanair, nor Aer Lingus had activities on this route in the benchmark month (January). There was no charter company active during the winter season.

(1475) In the IATA summer season 2012 Ryanair increased its weekly frequencies to 8 on the Dublin-Palma route, while Aer Lingus increased its weekly frequencies to 5. The weekly frequencies for charter companies in the IATA summer season 2012 amounted to a total of 6 (namely, two charter companies each operate on average 1 weekly frequency on the route at issue, while a third charter company operates on average 4 weekly frequencies).

- Market shares - capacity

(1476) Table 82 indicates the market shares of the Parties and the charter companies active on this route on the basis of seat capacity in the IATA summer season 2011 and in the IATA winter season 2011/12. As Ryanair did not provide data in this respect for charter companies, the market shares were computed on data made available by DAA.

Table 82: Market shares for seat capacity

Summer season 2011

Carrier / Market shares

Aer Lingus / [40-50%]

Ryanair / [40-50%]

Combined / [90-100%]

Charters / [5-10%]

Source: DAA's response to the Commission's request for information of 26 July 2012

(1477) As regards IATA summer season 2012, Ryanair had a market share of [50-60%] and Aer Lingus a market share of [30-40%]. Charter services represented a market share of between [5-10%] of the seat capacity (1382).

(1478) The very high combined market share of the Parties on this route is itself evidence of a dominant market position.

- Market shares - passengers

(1479) Table 83 indicates the market shares of the Parties and of the charter companies active on the route on the basis of passengers carried in IATA summer season 2011 and in IATA winter season 2011/12.

Table 83: Market shares for air services

Summer season 2011

Carrier / Market shares

Aer Lingus / [40-50%]

Ryanair / [50-60%]

Combined / [90-100%]

Charters / [5-10%]

Source: DAA's response to the Commission's request for information of 26 July 2012

(1480) As regards the IATA summer season 2012, Ryanair would have a market share of [50-60%] and Aer Lingus of [30-40%]. The charter companies carried between [5- 10%] of the passengers travelling on this route (1383).

(1481) Such a high combined market share of the Parties on this route is in itself evidence of a dominant market position.

- Conclusion

(1482) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Dublin-Palma route.

(1483) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that charter companies would be able to constrain the merged entity's market behavior sufficiently, especially with regard to fare setting, on the Dublin-Palma route.

8.9.6.11. Cork-Barcelona

- Route characteristics

(1484) On the Cork-Barcelona route, [...]* passengers travelled in IATA summer season 2011 while [...]* passengers travelled in IATA winter season 2011/12 (1384).

(1485) According to Ryanair (1385), the nature of the Cork-Barcelona route is a mix of business and leisure.

(1486) [...]* of Ryanair's passengers on this route originate at the Irish end of the route (1386).

- Parties' operations

(1487) Ryanair is the only airline that operates during the winter season. However, both Parties operate on the route with non-stop services in the summer season. Aer Lingus will exit the route in IATA winter season 2012/13 but will operate on this route in the summer season as from the next IATA summer season 2013. The Cork- Barcelona route is therefore an actual overlap route in the summer season.

(1488) Ryanair and Aer Lingus overlap on a city-to-city basis. Ryanair operates from Cork to Girona (GRO) and Reus (REU). Aer Lingus operates from Cork to Barcelona (BCN). Aer Lingus started to operate on this route in March 2004 (with a short interruption in IATA winter season 2012/2013), while Ryanair has been operating on it since summer 2010 to REU and just started to operate to GRO in summer 2012 (1387). The market investigation confirmed that the three airports are substitutable for flights ex-Cork (1388).

- Frequencies

(1489) Table 84 indicates the weekly flights operated by the Parties on this route. The average weekly frequency for the charter airlines is based on DAA data.

Table 84: Weekly flights between Cork and Barcelona

<TABLE>

Annex 7.3 (b) as updated on 22 August 2012 and Annex "Question 1.6 Additional info" to DAA's response to the Commission's request for information of 22 October 2012 (1389)

(1490) In the IATA summer season 2011, Ryanair operated 4 weekly frequencies to REU and Aer Lingus operated 3 weekly frequencies to BCN. On the Cork-Barcelona route, the Parties would therefore have combined 7 weekly frequencies for the IATA summer season 2011.

(1491) 1 weekly frequency was operated by a charter company in the IATA summer season 2011.

(1492) The frequencies for the IATA summer season 2012 for Ryanair and Aer Lingus combined increased to 9 weekly frequencies as Ryanair added 2 frequencies out of GRO, whereas the charter company accounted for only 1 weekly frequency.

- Market shares - capacity

(1493) Table 85 indicates the market shares of the Parties on the basis of seat capacity in IATA summer season 2011.As Ryanair did not provide data in this respect for charter companies, the market shares were computed on the basis of data made available by DAA.

Table 85: Market shares for seat capacity

Summer season 2011

Carrier / Market shares

Air Lingus / [50-60%]

Ryanair / [30-40%]

Combined / [90-100%]

Charters / [0-5%]

Source: DAA's response to the Commission's request for information of 26 July 2012

(1494) As regards the IATA summer season 2012, Ryanair had a market share of [50-60%] and Aer Lingus a market share of [40-50%]. The charter company represented a market share of between [0-5%] (1390).

(1495) The very high combined market share on this route is itself evidence of a dominant market position.

- Market shares - passengers

(1496) Table 86 indicates the market shares of the Parties and of the charter companies on the route at issue on the basis of passengers carried in the IATA summer season 2011.

Table 86: Market shares for air services

Summer season 2011

Carrier / Market shares

Air Lingus / [50-60%]

Ryanair / [30-40%]

Combined / [90-100%]

Charters / [0-5%]

Source: DAA's response to the Commission's request for information of 26 July 2012

(1497) As regards the IATA summer season 2012, Ryanair had a market share of [50-60%] and Aer Lingus of [30-40%]. The charter company represented a market share of between [0-5 %] (1391).

(1498) Such a high combined market share of the Parties on this route is in itself evidence of a dominant market position.

- Conclusion

(1499) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Cork-Barcelona route.

(1500) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that charter airlines would be able to constrain the merged entity's market behavior sufficiently, especially with regard to fare setting, on the Cork-Barcelona route.

8.9.6.12. Cork-Lanzarote

- Route characteristics

(1501) On the Cork-Lanzarote route, [...]* passengers travelled in IATA summer season 2011 while [...]* passengers travelled in IATA winter season 2011/12 (1392).

(1502) According to Ryanair (1393), the Cork-Lanzarote route is a leisure route.

(1503) [...]* of Ryanair's passengers on this route originate in Ireland (1394).

- Parties' operations

(1504) Both Parties operate on the route with non-stop services. Ryanair and Aer Lingus overlap on airport-to-airport basis. Both Ryanair and Aer Lingus operate from Cork to Lanzarote (ACE).

(1505) Aer Lingus has been operating on this route since October 2006, while Ryanair since June 2006 (1395).

- Frequencies

(1506) Table 87 indicates the weekly flights operated by the Parties on this route. The average frequency for the charter companies is based on DAA data.

Table 87: Weekly flights between Cork and Lanzarote

<TABLE>

Source: Annex 7.3 (b) as updated on 22 August 2012 and Annex "Question 1.6" to DAA's response to the Commission's request for information of 20 September 2012 (1396)

(1507) In the IATA summer season 2011, Ryanair operated 1 weekly frequency, while Aer Lingus operated 3 weekly frequencies. In the IATA winter season 2011/2012, Ryanair operated 1 weekly frequency, while Aer Lingus operated 3 weekly frequencies. On the Cork-Lanzarote route, the parties would therefore have had combined 4 weekly frequencies for the IATA summer season 2011 and IATA winter season 2011/2012.

(1508) Two charter companies each operated 1 weekly frequency in the IATA summer season 2011, while in the IATA winter season 2011/2012 only one charter company was active on the route and operated 1 weekly frequency.

(1509) The weekly frequencies for the summer season 2012 for Ryanair and Aer Lingus remained unchanged with 4 combined weekly frequencies, whereas the planned frequencies for the only charter company active on this route amounted to 1 weekly frequency.

- Market shares - capacity

(1510) Table 88 indicates the market shares of the Parties and of the charter companies active on the route on the basis of seat capacity in the IATA summer season 2011 and in the IATA winter season 2011/12. As Ryanair did not provide data in this respect for charter companies, the market shares were computed on data made available by DAA.

Table 88: Market shares for seat capacity

Summer season 2011

Carrier / Market shares

Aer Lingus / [60-70%]

Ryanair / [20-30%]

Combined / [90-100%]

All charters / [5-10%]

Winter season 2011/2012

Carrier / Market shares

Aer Lingus / [60-70%]

Ryanair / [20-30%]

Combined / [90-100%]

All charters / [0-5%]

Source: DAA's response to the Commission's request for information of 26 July 2012

(1511) As regards the IATA summer season 2012, Ryanair had a market share of [20-30%] and Aer Lingus a market share of [60-70%]. The charter companies represented market share of [0-5%] of the market (1397).

(1512) The very high combined market share on this route is itself evidence of a dominant market position.

- Market shares - passengers

(1513) Table 89 indicates the market shares of the Parties and of the charter companies active on the route on the basis of passengers carried in the IATA summer season 2011 and in the IATA winter season 2011/12.

Table 89: Market shares for air services

Summer season 2011

Carrier / Market shares

Aer Lingus / [60-70%]

Ryanair / [20-30%]

Combined / [80-90%]

All charters / [10-20%]

Winter season 2011/2012

Carrier / Market shares

Aer Lingus / [60-70%]

Ryanair / [20-30%]

Combined / [90-100%]

All charters / [0-5%]

Source: DAA's response to the Commission's request for information of 26 July 2012

(1514) As regards the IATA summer season 2012, Ryanair had a market share of [20-30%] and Aer Lingus of [60-70%]. The charter companies represented a market share of [5-10%] of the market (1398).

(1515) Such high combined share on this route is in itself evidence of a dominant market position.

- Conclusion

(1516) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Cork-Lanzarote route.

(1517) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that charter companies would be able to constrain the merged entity's market behavior sufficiently, especially with regard to fare setting, on the Cork-Lanzarote route.

8.9.6.13. Cork-Malaga

- Route characteristics

(1518) On the Cork-Malaga route, [...]* passengers travelled in IATA summer season 2011 while [...]* passengers travelled in IATA winter season 2011/12 (1399).

(1519) According to Ryanair (1400), the Cork-Malaga route is a leisure route.

(1520) [...]* of Ryanair's passengers on this route originate in Ireland (1401).

- Parties' operations

(1521) Both Parties operate on the route with non-stop services. Ryanair and Aer Lingus overlap on airport-to-airport basis. Both Ryanair and Aer Lingus operate from Cork to Malaga (AGP).

(1522) Aer Lingus has been operating on this route since 2004, while Ryanair since June 2010 (1402).

- Frequencies

(1523) Table 90 indicates the weekly flights operated by the Parties on this route. The average frequency for the charter companies is based on DAA data.

Table 90: Weekly flights between Cork and Malaga

<TABLE>

Source: Annex 7.3 (b) as updated on 22 August 2012 and Annex "Question 1.6" to DAA's response to the Commission's request for information of 20 September 2012 (1403)

(1524) In the IATA summer season 2011, Ryanair operated 4 weekly frequencies, while Aer Lingus operated 7 weekly frequencies. In the IATA winter season 2011/2012, Ryanair operated 2 weekly frequencies, while Aer Lingus operated 3 weekly frequencies. On the Cork-Malaga route, the Parties combined would therefore have had 11 and 5 weekly frequencies for the IATA summer season 2011 and the IATA winter season 2011/2012 respectively.

(1525) The charter company operated 1 weekly frequency in the IATA summer season 2011. No charter operations appear to have taken place during the winter season 2011/2012.

(1526) The weekly frequencies for the summer season 2012 for Ryanair amounted to 4 and for Aer Lingus to 7 weekly frequencies, whereas the frequencies for the only charter company active on the route amounted to 1 weekly frequency.

- Market shares - capacity

(1527) Table 91 indicates the market shares of the Parties and of the charter companies active on the route on the basis of seat capacity in the IATA summer season 2011 and in the IATA winter season 2011/12. As Ryanair did not provide data in this respect for charter companies, the market shares were computed on data made available by DAA.

Table 91: Market shares for seat capacity

Summer season 2011

Carrier / Market shares

Aer Lingus / [60-70%]

Ryanair / [30-40%]

Combined / [90-100%]

All charters / [0-5%]

Winter season 2011/2012

Carrier / Market shares

Aer Lingus / [50-60%]

Ryanair / [40-50%]

Combined / 100%

All charters / 0

Source: DAA's response to the Commission's request for information of 26 July 2012

(1528) As regards the IATA summer season 2012, Ryanair had a market share of [30-40%] and Aer Lingus a market share of [60-70%]. The charter companies represented a market share of [0-5%] of the market (1404).

(1529) It follows that post-Transaction, the Parties would have a quasi-monopoly on the route in the summer season and a monopoly in the winter season.

- Market shares - passengers

(1530) Table 92 indicates the market shares of the Parties and of the charter companies active on the route on the basis of passengers carried in the IATA summer season 2011 and in the IATA winter season 2011/12.

Table 92: Market shares for air services

Summer season 2011

Carrier / Market shares

Aer Lingus / [60-70%]

Ryanair / [30-40%]

Combined / [90-100%]

All charters / [0-5%]

Winter season 2011/2012

Carrier / Market shares

Aer Lingus / [50-60%]

Ryanair / [40-50%]

Combined / 100%

All charters / 0

Source: DAA's response to the Commission's request for information of 26 July 2012

(1531) As regards the IATA summer season 2012, Ryanair had a market share of [30-40%] and Aer Lingus of [60-70%]. The charter companies represented a market share [0- 5%] of the market (1405).

(1532) It follows that post-Transaction the Parties would have a quasi-monopoly on the route in the summer season and a monopoly in the winter season.

- Conclusion

(1533) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Cork-Malaga route.

(1534) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that charter companies would be able to constrain the merged entity's market behavior sufficiently, especially with regard to fare setting, on the Cork-Malaga route.

8.9.6.14. Cork-Palma - Route characteristics

(1535) On the Cork-Palma route, [...]* passengers travelled in IATA summer season 2011 there are no passengers in IATA winter season 2011/12 (1406).

(1536) According to Ryanair (1407), the Cork-Palma route is a leisure route and it has a seasonal pattern, as the operations of the Parties on the Cork-Palma route overlap only in the summer season.

(1537) [...]* of Ryanair's passengers on this route originate in Ireland (1408).

- Parties' operations

(1538) Both Parties operate on the route with non-stop services. Ryanair and Aer Lingus overlap on airport-to-airport basis. Both Ryanair and Aer Lingus operate from Cork to Palma (PMI).

(1539) Aer Lingus has been operating on this route since March 2011, while Ryanair started operating in March 2012 (1409). Therefore, the Parties' operations on the Cork-Palma route started overlapping in the summer season 2012.

- Frequencies

(1540) Table 93 indicates the weekly flights operated by the Parties on this route. The average frequency for the charter companies is based on DAA data.

Table 93: Weekly flights between Cork and Palma

<TABLE>

Source: Annex 7.3 (b) as updated on 22 August 2012 and Annex "Question 1.6" to DAA's response to the Commission's request for information of 20 September 2012 (1410)

(1541) In IATA summer season 2012, Ryanair and Aer Lingus both operated 2 weekly frequencies. On the Cork-Palma route, the merged entity therefore had 4 weekly frequencies in IATA summer season 2012.

(1542) The only charter company active on the route operated 2 weekly frequencies in IATA summer season 2012.

- Market shares - capacity

(1543) Table 94 indicates the market shares of the Parties and of the charter companies active on the route on the basis of seat capacity in IATA summer season 2012. As Ryanair did not provide data in this respect for charter companies, the market shares were computed on the basis of data made available by DAA.

Table 94: Market shares for seat capacity

Summer season 2012

Carrier / Market shares

Aer Lingus / [30-40%]

Ryanair / [50-60%]

Combined / [90-100%]

All charters / [5-10%]

Source: DAA's response to the Commission's request for information of 3 December 2012

(1544) Such high combined share on this route is in itself evidence of a dominant market position.

- Market shares - passengers

(1545) Table 95 indicates the market shares of the Parties and of the charter companies active on the route on the basis of passengers carried in IATA summer season 2012.

Table 95: Market shares for air services

Summer season 2012

Carrier / Market shares

Aer Lingus / [30-40%]

Ryanair / [50-60%]

Combined / [90-100%]

All charters / [5-10%]

Source: DAA's response to the Commission's request for information of 3 December 2012

(1546) The Transaction would give the merged entity a very high combined market share of above 90% in terms of estimated point-to-point passengers carried on the route in the 2012 summer season.

(1547) Such a high combined share on this route is of itself evidence of a dominant market position.

- Conclusion

(1548) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position in relation to the Cork-Palma route.

(1549) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on the route. Customers' travelling options therefore would be substantially reduced and it is unlikely that charter companies would be able to constrain the merged entity's market behavior sufficiently, especially with regard to fare setting, on the Cork-Palma route.

8.9.6.15. Conclusion on the routes where charter companies operate

(1550) Therefore, the Commission concludes that the Transaction is likely to significantly impede effective competition in particular as a result of the creation of a dominant position on the following 11 routes, where charter companies operate: Dublin- Barcelona, Dublin-Faro, Dublin-Gran Canaria, Dublin-Ibiza, Dublin-Lanzarote, Dublin-Malaga, Dublin-Palma, Cork-Barcelona, Cork-Lanzarote; Cork-Malaga and Cork Palma.

(1551) Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre- Transaction on these routes. Customers' travelling options therefore would be substantially reduced and it is unlikely that charter companies would be able to constrain the merged entity's market behavior sufficiently, especially with regard to fare setting, on these routes.

8.10. Potential competition

8.10.1. Analytical framework

(1552) Competition on the routes currently operated by both Ryanair and Aer Lingus ex Dublin, Shannon, and Cork, where both carriers have bases, cannot be regarded in isolation. Such an isolated analysis would imply that the respective product markets are entirely independent from each other. Potential competition is an important parameter of competition amongst Ryanair and Aer Lingus due to their bases at Irish airports and as evidenced by past experience of entry events.

(1553) Both carriers have the necessary flexibility to shift and add routes from their existing bases at these airports in reaction to changes in the competitive structure of the different routes operated from their bases. The analysis can therefore not only be static, that is to say, focusing only on the overlap routes. The analysis must also be dynamic looking at to what extent the disappearance of one carrier's very close, if not closest, competitor might eliminate potential competition that would have constrained Ryanair and Aer Lingus in the absence of the Transaction.

(1554) In this context the decision opening the proceedings raised serious doubts regarding potential competition issues on 24 routes (1411).

(1555) According to the Horizontal Merger Guidelines, "a merger with a potential competitor can generate horizontal anti-competitive effects, whether coordinated or non-coordinated, if the potential competitor significantly constrains the behaviour of the firms active in the market. This is the case if the potential competitor possesses assets that could easily be used to enter the market without incurring significant sunk costs." (1412)

(1556) Paragraph 60 of the Horizontal Merger Guidelines sets two conditions for establishing loss of potential competition:"[f]irst, the potential competitor must already exert a significant constraining influence or there must be a significant likelihood that it would grow into an effective competitive force. ... Second, there must not be a sufficient number of other potential competitors, which could maintain sufficient competitive pressure after the merger."

(1557) In previous cases, the Commission found that an airline active on a route could be constrained not only by its actual competitors on this route but also by potential competitors (1413).

8.10.2. Ryanair's arguments

(1558) Ryanair is of the view that the Transaction is not likely to significantly impede effective competition by eliminating the most credible potential entrant on 9 Ryanair routes and 11 Aer Lingus routes identified in the Statement of Objections. Ryanair argues that the Commission vastly overstates the alleged loss of potential competition as it fails to adequately take into account that Ryanair and Aer Lingus are differentiated competitors and these differences are particularly acute as regards nonoverlap routes, that Ryanair has already decided not to enter the non-overlap routes currently served by Aer Lingus and that Aer Lingus is unlikely to enter the nonoverlap routes currently served by Ryanair as it has no aircraft orders and no demonstrable intention to expand in such markets (1414).

(1559) In a more route-specific approach, Ryanair has put forward the following two types of arguments to dismiss the serious doubts expressed in the decision opening the proceedings (1415):

- First, [operations on routes]*.

- Secondly, [thinness of routes]*.

(1560) Ryanair also considers that Aer Lingus is not a potential competitor on Ryanair's non-overlap routes because of the following (1416):

- Aer Lingus is cutting back its short-haul capacity and traffic, which has declined from 9.5 to 8.4 MPPA over the past three years;

- on a number of overlap routes in recent years, Aer Lingus has essened its seat capacity from direct competition with Ryanair and replaced it with smaller Aer Arann turbo-prop aircraft (1417);

- Aer Lingus short-haul fares have been consistently double those of Ryanair, which makes it extremely difficult for Aer Lingus to successfully enter any Ryanair non-overlap route.

(1561) Finally, in the response to the Statement of Objections, Ryanair argues that the Commission inexplicably and without evidence rules out Wizz Air as a likely entrant on a number of routes mainly to Central and Eastern Europe, despite the fact that Wizz Air already operates to and from Cork. If the Commission believes that Wizz Air is not a credible entrant on these routes, then the Commission cannot at the same time claim that Ryanair and Aer Lingus are likely entrants on these non-overlap routes (1418).

8.10.3. Assessment

(1562) There is a significant number of routes to and from Irish airports where the activities of the Parties do not currently overlap.

(1563) Ryanair flies on 27 routes out of Dublin, on 8 routes out of Cork, on 6 routes out of Shannon and on 11 routes out of Knock, where Aer Lingus does not operate any services (1419).

(1564) Aer Lingus operates on 20 routes out of Dublin and on 8 routes out of Cork, (1420) where Ryanair does not operate any services (1421).

8.10.3.1. First limb of the test set out at paragraph 60 of the Horizontal Merger Guidelines

(1565) The first condition set out at paragraph 60 of the Horizontal Merger Guidelines for establishing loss of potential competition is that the potential competitor must already exert a significant constraining influence or there must be a significant likelihood that it would grow into an effective competitive force.

(1566) When one of the Parties is not active on a route which is operated by the other Party, it is in a position to enter this route relatively easily, should it find it profitable. Indeed, there is convincing evidence of the following:

- that both Ryanair and Aer Lingus have a significant presence at Irish airports, with bases at Dublin, Cork and Shannon airports, giving them the necessary flexibility to shift and add routes in reaction to changes in the competitive structure of the different routes operated from their bases.

- that entry on a new route by Ryanair or Aer Lingus, by means of the redeployment of an aircraft situated at their bases in Dublin, Cork and Shannon would not involve significant sunk costs. Due, inter alia, to base advantages and a strong city presence, no significant sunk costs would have to be incurred by the Party entering the new route (1422).

- that considering the concrete entry behaviour of Ryanair in relation to Aer Lingus' routes and the size of Ryanair's operations at its bases in Ireland, Aer Lingus is aware that Ryanair would be able to move capacity to other Aer Lingus' routes if prices were to increase on the latter routes, and vice-versa.

- that both Ryanair and Aer Lingus have a deep knowledge of the air transport service markets from and to Ireland.

(1567) Also, there is a pattern of dynamic entry and capacity reallocation in routes out of Dublin.

(1568) Since 2007, entry patterns have demonstrated that Aer Lingus and Ryanair frequently enter on each other's routes.

(1569) Indeed, Aer Lingus and Ryanair now compete on a greater number of routes compared to the 2007 Decision. The number of overlap routes has increased significantly, from 35 in 2007 to 46 (1423) in 2012. Furthermore, on almost 84% of these routes (63% in 2007) there are no other scheduled airlines active. These facts, in themselves, are already clear evidence that Aer Lingus and Ryanair are each other's most credible potential competitor.

(1570) Entry and exit events on the short-haul routes ex-Dublin have been described in Section 8.5.2. 15 of the routes operated by Aer Lingus ex Dublin in summer 2012 have had new entries by scheduled carriers since 2007 (1424). Of these 15 routes, 10 (67%) have been entered into by Ryanair only.

(1571) In total, since 2007, Ryanair entered on 9 routes out of Dublin, 1 out of Shannon, and 7 out of Cork where at least one carrier other than Aer Lingus operated (1425).

(1572) Moreover, 8 of the short-haul routes operated by Ryanair ex Dublin in summer 2012 have had new entries by scheduled carriers since 2007 (1426). Of these 8 routes, 3 (37%) have been entered into by Aer Lingus/Aer Arann only.

(1573) Several studies have tried to quantify the effect of potential competition. For example, Kwoka and Shumilkina (2010) have shown that, following the merger of US Air and Piedmont, "prices rose by 5-6% on routes that one carrier served and the other was a potential entrant." (1427) Also, Goolsbee and Syverson (2005) argue that "We find incumbents cut fares significantly when threatened by Southwest's entry. Over half of Southwest's total impact on incumbent fares occurs before Southwest starts flying. These cuts are only on threatened routes, not those out of non- Southewest competing airports." (1428)

(1574) Moreover, the 2007 Decision (1429) concluded that the notified merger was likely to significantly impede effective competition as a result of the creation or strengthening of a dominant position by the elimination of a credible potential entrant on 5 Aer Lingus' routes to and from Dublin (Dubrovnik, Naples, Nice, Palma de Mallorca and Santiago de Compostela) (1430). On 2 of these routes (Dublin-Nice and Dublin-Palma), there is now actual competition between Aer Lingus and Ryanair.

(1575) The 2007 Decision concluded also that the notified merger was likely to significantly impede effective competition as a result of the creation or strengthening of a dominant position by the elimination of a credible potential entrant on 10 of Aer Lingus' routes to and from Cork (Tenerife, Lanzarote, Faro, Malaga, Alicante, Nice, Madrid, Berlin, Rome, and Barcelona) (1431). On 6 of these routes (from Cork to Tenerife, Lanzarote, Malaga, Alicante, Faro and Barcelona), there is now actual competition between Aer Lingus and Ryanair.

(1576) The 2010 Judgement also highlighted that "the comparison made by the Commission does not take account of the influence on Aer Lingus's fares of Ryanair's presence as a potential competitor on routes out of Dublin (section 7.6). On those routes, it is in fact likely that Aer Lingus would charge lower fares than it would charge if Ryanair did not have a base at Dublin airport." (1432) (emphasis added)

(1577) Finally, the presence of Ryanair and Aer Lingus at Irish airports appears to be an important stimulant for the development of new routes, as both carriers are willing to benefit from a first-mover advantage, in particular in the context of a complex economic environment.

(1578) Therefore, the Commission considers that each Party already exerts a significant constraining influence on the other Party, or there is a significant likelihood that one Party would grow into an effective competitive force at least on some routes operated by the other Party.

- Ryanair routes of potential competition

(1579) The Commission focused initially on the following routes: (i) routes on which currently only Aer Lingus (including Aer Arann) operates and not Ryanair operates,

(ii) routes on which there are no other competing airlines offering services on the route (iii) routes on which Ryanair either has a base at the non-Irish end of the route or has a presence at that end with services provided from other Ryanair bases. Applying these criteria, 13 routes were initially identified (1433): Dublin-Bordeaux, Dublin-Hamburg, Dublin-Lyon, Dublin-Perpignan, Dublin-Santiago de Compostela, Dublin-Stuttgart, Dublin-Bologna, Dublin-Düsseldorf (1434), Cork-Paris, Cork-Munich, Cork-Nice, Cork-Brussels, and Cork-Gran Canaria (1435).

(1580) Further relevant convincing evidence has been gathered through the market investigation.

(1581) Firstly, while Ryanair has shifted significant capacity away from the Irish markets in recent years (1436), Ryanair describes its route selection policy [...]* (1437). It explains that no decisions are made on the choice of new routes until negotiations in relation to costs and facilities are completed with airports. There are numerous examples of significant shifts in Ryanair's choice of routes served over recent years (1438).

(1582) Furthermore, Ryanair recently made an offer to the Irish Government to grow traffic at Irish airports by 5 million passengers over 4 years (1439). Such expansion would mean an increase in excess of [...]*% in Ryanair's current traffic at Dublin, Cork and Shannon airports (1440). [...]* (1441).

(1583) The Commission acknowledges that the threat of entry is not necessarily equally significant on all the routes where only one of the Parties is present. Therefore, taking a prudent and "conservative" approach, the Commission has identified some routes on which the likelihood of entry is high. The Commission is of the view that Ryanair could enter on the following routes, (i) currently only Aer Lingus, or Aer Arann, operates (and not Ryanair), (ii) routes on which there are no other competing airlines offering services (airport pair) and (iii) routes on which Ryanair either has a base or at least a significant presence at the non-Irish end of the route.

(1584) Secondly, regarding the Aer Lingus non-overlap routes, Ryanair refers to routes (airport pairs) to major airports on which Ryanair would have chosen not to operate and has no intention to operate (1442). Of the relevant routes listed in the decision opening the proceedings, the following 6 routes would fall into this category: Dublin-Hamburg, Dublin-Lyon, Dublin-Stuttgart, Cork-Paris (Charles de Gaulle), Cork-Munich, and Cork-Brussels (Zaventem).

(1585) The Commission considers that Ryanair's arguments regarding major airports are irrelevant. The Commission is of the view that Ryanair would not enter these routes through major airports (Paris Charles de Gaulle, Munich, Brussels) but through the secondary airports to which Ryanair already operates, and which are substitutable airports (respectively Beauvais, Memmingen and Charleroi).

(1586) The Statement of Objections referred also to the Dublin-Hamburg (Lübeck) and Dublin-Stuttgart (Baden-Baden) routes. However, in a cautious approach in relation to airport substitutability, the Commission does not need to conclude whether a significant impediment to effective competition exists on these two routes as the overall conclusion of the present Decision would not be affected.

(1587) In addition, the Statement of Objections referred to the Cork-Brussels (Charleroi) routes. However, while no operations took place in summer 2011, there were less than 20 000 passengers in the IATA summer 2012 season. As explained below in recital 1595, this route falls below the threshold of 20,000 passengers.

(1588) Table 96 summarizes the situation on the remaining 3 relevant routes:

Table 96: 3 routes to major airports

Routes / Base or presence (1443) at non-Irish end / Passengers, summer 2011 and summer 2012

Dublin-Lyon / Not significant presence / [20 000 - 50 000]

Cork-Munich / Memmingen: 17 routes, 59 weekly frequencies / [20 000 - 50 000]

Cork-Paris / Beauvais: 44 routes, 228 weekly frequencies / [50 000 - 100 000]

(1589) Thirdly, [thinness of routes]* (1444).

(1590) [thinness of routes]*:

Table 97: passenger numbers on routes regarded as too thin by Ryanair

[thinness of routes]*

(1591) The routes in Table 97 have significant seasonality patterns, with the large majority of their passengers flying in the summer season (except for the Dublin-Agadir route).

(1592) However, the passenger numbers in Table 97 compare in terms of magnitude with passenger numbers on several routes operated by Ryanair and Aer Lingus in competition (for instance Dublin-Bilbao/Santander, Dublin-Marseille, Cork-Alicante, Cork-Tenerife, Dublin-Ibiza, and Cork-Lanzarote, in the summer season).

(1593) Moreover, the passenger numbers in Table 97 compare in terms of magnitude also with passenger numbers on several routes operated by Ryanair in competition with a third carrier. For example, on 2 August 2012 Ryanair announced its intention to enter the market with five new routes from Cork to Gdansk, Krakow, Warsaw, Wroclaw and Vilnius, which were served by Wizz Air. Most of these routes had less than 20 000 passengers in summer 2011 and less than 10 000 passengers in winter 2011- 2012. Ryanair has also been operating on thinner routes, such as Cork-Bordeaux, where it carried [10 000 - 20 000] passengers in the summer seasons 2010 and 2011.

(1594) In addition, several past entries by Ryanair (irrespective of whether these routes had been operated by third carriers or not) have taken place on routes with passenger numbers of 20 000 in the summer season or less.

(1595) It is therefore relevant, in a conservative approach, to apply a fourth criterion in order to define routes on which the likelihood of entry by Ryanair is high (1445): the routes should have had at least 20 000 passengers in the summer season 2011 and in the summer season 2012 (1446). In the case of low-cost airlines which often operate low frequency services, the 20 000 passengers threshold seems appropriate for the purposes of this Transaction (1447). Moreover, considering the summer season, when generally more passengers travel on the relevant routes, it is a cautious approach.

(1596) It appears from Table 97 that the only routes that pass the 20 000 passengers threshold and on which Ryanair also has a base or a significant presence at the other end are the Dublin-Bologna (base) (1448) and Dublin-Bordeaux (significant presence with 8 routes, 30 weekly frequencies) routes (1449). These two routes had [20 000 - 50 000] passengers in both summer 2011 and summer 2012 seasons.

(1597) Fourthly, in its response to the Commission's request of September 20, 2012, [operations on routes]*. However, the Commission checked the thickness of these routes and whether Ryanair would have a base or significant commercial presence at the non-Irish end of the routes. The Cork-Birmingham and Shannon-Birmingham routes meet both criteria, as Ryanair has a base at Birmingham. However, only the Cork-Birmingham route had [20 000 - 50 000] passengers in both summer 2011 and summer 2012 seasons.

(1598) In conclusion, for the purpose of this Decision, the Commission is taking a cautious and conservative approach to identify those concrete routes where it is most likely that Ryanair's entry has a significant constraining influence.

(1599) In particular in view of the opportunistic entry policy of Ryanair, and focusing on the following routes: (i) routes where currently only Aer Lingus or Aer Arann operate (and not Ryanair), (ii) routes where there are no other competing airlines offering services on the route , (iii) routes where Ryanair either has a base at the non-Irish end of the route or has a significant presence at that end with services provided from other Ryanair routes, (iv) routes where more than 20 000 passengers were carried in the summer 2011 and summer 2012 seasons on the routes,. The Commission considers in a cautious approach that Ryanair already exerts a significant constraining influence on Aer Lingus and Aer Arann, or there is a significant likelihood that Ryanair would grow into an effective competitive force on the following 5 routes: Dublin-Bologna, Dublin-Bordeaux, Cork-Paris (Beauvais), Cork-Munich (Memmingen), and Cork-Birmingham.

- Aer Lingus routes of potential competition

(1600) The Commission initially identified 11 specific routes (1450), (i) on which currently only Ryanair operates (and not Aer Lingus), (ii) on which there are no other competing airlines offering services on the route and (iii) on which the route was an overlap route in the 2007 Decision (iv) or on which Aer Lingus serves the airport or airports which are substitutable for the airport served by Ryanair from other Irish airports. The routes that were overlap routes in 2007 and are currently operated by Ryanair without competitors are: Dublin-Krakow (1451), Dublin-Newcastle and Dublin-Sevilla. The routes where Aer Lingus flies from other Irish airports and on which Ryanair without competitors are: Cork-Barcelona (1452), Cork-Bordeaux, Cork-Carcassonne, Cork-Gran Canaria, Cork-Milan, Cork-Fuerteventura, Shannon-Malaga, Shannon-Tenerife.

(1601) Contrary to Ryanair's assertions, Aer Lingus has not really cut back its short-haul capacity and flights. The number of short-haul seats operated by Aer Lingus has remained relatively constant over the period; increasing slightly from 11.0 million in 2007 to a high of 12.6 million in 2009 before decreasing to 11.9 million in 2012. This would represent an increase of 0.9 million seats between 2007 and 2012 (1453).

(1602) Further relevant evidence has been gathered through the second phase market investigation.

(1603) Aer Lingus entered in recent years on the following Ryanair routes (airport pairs): Cork-London Gatwick (2010) (1454), and Dublin-Stockholm (2012). Aer Arann entered on the following Ryanair routes (airport pairs) under the franchise agreement: Dublin-Bristol (2011), and Knock-Birmingham (2012). These events show that Aer Lingus and Aer Arann are able to enter on Ryanair's routes (airport pairs) and in fact enter on such routes where Ryanair is operating. This has been done despite the fact that Aer Lingus (and Aer Arann) fares tend to be higher than Ryanair's fares.

(1604) Aer Lingus considers that the Commission has been too conservative in its initial approach. In particular, the criterion limiting the routes of concern to routes where no other carrier offers services would be too restrictive. However, the same criterion was also used in the 2007 Decision and the Commission considers this cautious approach still to be warranted in this Decision.

(1605) Nevertheless, Aer Lingus highlights that other initial criteria were overly restrictive and contradicted by Aer Lingus' entry patterns since 2007 (1455). In particular, Aer Lingus entered the Dublin-Stockholm route without flying to Stockholm from any other airport. Furthermore, the Dublin-Stockholm route was not an overlap route in the 2007 Decision.

(1606) While Ryanair claims that Aer Lingus has no aircraft orders (1456), the Commission is of the view that the existence of aircraft orders is not a necessary requirement for a carrier to expand on specific routes (1457). Indeed, considering the base advantages Aer Lingus enjoys, opening new routes would not necessarily require the acquisition of new aircraft. Aer Lingus could easily redeploy aircraft and reduce frequencies on less profitable routes to enter new routes. Besides, Aer Arann is reported to have aircraft on order, which according to the Commission, could be used for instance on the relatively thick Dublin-Newcastle route.

(1607) The thickness of the routes would also be a relevant factor in the assessment of the Commission. The Commission considers that the same threshold of 20 000 passengers in the summer 2011 and summer 2012 seasons used for Ryanair should apply (1458). Aer Lingus and Aer Arann operate routes of this order of passenger numbers (for instance Dublin-Bilbao/Santander, Dublin-Marseille, Dublin- Warsaw/Modlin, Cork-Alicante, Cork-Tenerife, Dublin-Ibiza, and Cork-Lanzarote, in the summer season).

Table 98: routes thickness

Routes / Passenger summer 2011 and summer 2012

Dublin-Bydgoszcz / [20 000 - 50 000]

Dublin-Eindhoven / [50 000 - 100 000]

Dublin-Katowice / [20 000 - 50 000]

Dublin-Lodz / [20 000 - 50 000]

Dublin-Malta / [20 000 - 50 000]

Dublin-Nantes / [20 000 - 50 000]

Dublin-Newcastle (1459) / [50 000 - 100 000]

Dublin-Pisa / [20 000 - 50 000]

Dublin-Poznan / [20 000 - 50 000]

Dublin-Rzeszow / [20 000 - 50 000]

(1608) Moreover, applying to Aer Lingus a similar criterion to the one applied to Ryanair in terms of having a base or a significant presence at the non-Irish end of the route would not be appropriate as Aer Lingus and Ryanair operate under different models in this regard. It is in particular obvious that Aer Lingus operates most of its flights to destinations where it does not have a base or presence due to flights other than ex Dublin.

(1609) However, the Commission also takes into account the fact that Aer Lingus has entered onto Ryanair's operated routes less frequently than the opposite.

(1610) As a conclusion, the Commission is taking a cautious and conservative approach to identifying those concrete routes where it is most likely that Aer Lingus' entry has a significant constraining influence. In particular, focusing on (i) routes where currently only Ryanair operate (and not Aer Lingus and Aer Arann), (ii) routes where there are no other competing carriers offering services on the route, (iii) routes where the thickness of the route compares with other routes operated by Aer Lingus and Aer Arann. Considering entry patterns of Aer Lingus and Aer Arann, the Commission considers in a very cautious approach that Aer Lingus and Aer Arann already exert a significant constraining influence on Ryanair, or there is a significant likelihood that Aer Lingus and Aer Arann would grow into an effective competitive force on the Dublin-Newcastle route. The Commission has reached this conclusion also taking into account the business model of Aer Arann (that is to say, serving typically provincial United Kingdom routes).

(1611) The Commission does not need to conclude whether significant impediment to effective competition exists on the routes from Dublin to Bydgoszcz, Eindhoven, Katowice, Lodz, Malta, Nantes, Pisa, Poznan, and Rzeszowas, as the conclusion of the present Decision would not be affected.

8.10.3.2. Second limb of the test set out at paragraph 60 of the Horizontal Merger Guidelines

(1612) As regards the second limb of the test set out at paragraph 60 of the Horizontal Merger Guidelines, it is necessary to demonstrate that entry by other competing airlines on these routes is less likely and that they would not provide sufficient competitive pressure on the merged entity.

(1613) The market investigation has not identified any other potential competitor, whose entry would be timely, likely and sufficient, so that sufficient competitive pressure on the merged entity would exist after the Transaction, at least on the 6 routes listed at recitals 1599 and 1610.

(1614) Ryanair has claimed that "Ryanair and Aer Lingus are differentiated competitiors and these differences are particularly acute as regards non-overlap routes" (1460) without providing any further explanations. The Commission, on the contrary considers that the Parties are generally each other's closest competitor. In particular, when compared to other carriers, Aer Lingus and Ryanair are the only carriers with bases at Dublin and Cork airports, enjoying all the related base advantages (see Section 8.4.3). Both Ryanair and Aer Lingus can relatively easily adjust their schedule to react to new route opportunities as they enjoy scale effects, have early morning slots available and enjoy considerable brand awareness at least on the Irish side of the route. A carrier with a base on the other side of the route may still pose such potential competitive constraint, however, the strength of such effect is more likely to be proncounced between closest competitors and when the carriers have the same base.

(1615) The only other independent scheduled carrier with a base at Dublin airport is CityJet. Considering market circumstances, and its different business model and limited scope of activities in Ireland, CityJet cannot be considered to be a likely entrant who could provide sufficient competitive pressure post-Transaction on Ryanair's or Aer Lingus' routes listed at recitals 1599 and 1610.

(1616) No other independent carrier has a base at Cork airport, which would make entry likely on these routes.

(1617) Ryanair considers in particular that "the Commission inexplicably and without evidence rules out Wizz Air as a likely entrant on a number of routes mainly to Central and Eastern Europe, despite the fact that Wizz Air already operates to/from Cork." (1461)

(1618) Wizz Air operates indeed several flights from Cork airport but has no base in Ireland. Its business model focuses on operating from bases in central and eastern European countries. No change is contemplated by Wizz Air in this regard (1462). None of the 6 routes identified at recitals 1599 and 1610 relates to central and eastern European countries.

(1619) It is also noted that Wizz Air recently decided to exit several routes ex Cork, after Ryanair entered on the routes in the summer 2012 (see section 8.5.2.6).

(1620) Moreover, Wizz Air made clear that "also Wizz Air does not currently plan to enter routes ex-Ireland that serves Western European airports" and that "while Wizz Air is constantly evaluating routes, it does not have any immediate action plan for entering any other Irish route in the short/medium term." (1463)

(1621) Specifically, Wizz Air considers that, absent commitments, entry at Dublin airport "would be difficult, because of the competitive environment (i.e. Ryanair's strength and capacity in Dublin), the level of airport charges, scarcity of slots as well as due to the economic downturn." (1464)

(1622) To conclude, the Commission considers that Wizz Air would not be a likely entrant on any of the routes mentioned in recitals 1599 and 1610 after the Transaction, absent commitments.

(1623) Moreover, no similar pattern of entry to the one of Ryanair, Aer Lingus and Aer Arann on ex-Dublin and Cork routes can be observed from scheduled carriers with a base at the destination airport on any of the routes mentioned at recitals 1599 and 1610.

(1624) Charter operators, even those with a base at Irish airports or at the destination airports, cannot be considered to be likely entrants who could provide sufficient competitive pressure post-Transaction on any of the routes mentioned at recitals 1599 and 1610, in particular in view of their business models. This makes them at best very distant competitors of both Ryanair and Aer Lingus.

(1625) Finally, Ryanair's aggressive reaction to entry by increasing capacity and frequencies on the routes concerned would also act as a barrier to entry for potential competitors. Moreover, Ryanair has been very profitable in recent years, has deep pockets, and would therefore be able to sustain a prolonged period of very low prices.

(1626) To conclude, there are not a sufficient number of other potential competitors, which could maintain sufficient competitive pressure post-Transaction on the routes mentioned at recitals 1599 and 1610.

8.10.4. Conclusion on potential competition

(1627) As a consequence of the dynamic pattern of entry in competition with each other and the very limited impact of entry by other carriers on the Parties, Ryanair and Aer Lingus exert a potential competitive constraint on each other.

(1628) The Commission therefore concludes that the Transaction is likely to significantly impede effective competition, in particular as a result of the creation or strengthening of a dominant position, by eliminating the most credible potential entrant on the following routes on which Aer Lingus or Ryanair respectively exert a significant constraining influence and on which there is no sufficient number of other potential competitor which could maintain sufficient competitive pressure on the merged entity post-Transaction:

Table 99: Ryanair routes of potential competition

1. Dublin-Bologna

2. Dublin-Bordeaux

3. Cork-Birmingham

4. Cork-Munich (Memmingen)

5 Cork-Paris (Beauvais)

Table 100: Aer Lingus routes of potential competition

1. Dublin - Newcastle

8.11. Effects on consumers

(1629) As noted by the Commission in the 2007 Decision, "a merger leading to high market shares is particularly likely to lead to price increases when customers of the Merging Parties are likely to have difficulties or to be unable to switch to other suppliers, because there are few or no alternative suppliers. Such customers are, according to the [Horizontal Merger] Guidelines, 'particularly vulnerable to price increases'." (1465)

(1630) The Transaction would significantly reduce customers' ability to switch between different suppliers, since it would not only reduce the number of alternative airlines to whom they could turn in case of a price increase post-merger, but would remove entirely the possibilities to switch for customers entirely on many routes. Airline customers would have no countervailing buyer power against the merged entity to offset the anti-competitive effects of the merger, in particular because sales to larger corporate customers or to tour operators would be limited. Customers would be particularly vulnerable to price increases by the merged entity.

8.12. Conclusion on the competitive effects of the Transaction

(1631) The very large market shares achieved by the merged entity are in themselves, and save in exceptional circumstances, evidence of the existence of a dominant position. (1466)

(1632) Moreover, on the basis of all available evidence, the Commission concludes that Aer Lingus and Ryanair appear to be very close competitors, if not the closest competitors, on each of the overlap routes and they enjoy significant base advantages at Dublin airport in particular. In addition, the main barriers to entry identified in the 2007 Decision have not changed to a significant extent. If anything, barriers to entry have increased. While terminal congestion at Dublin airport has diminished, the Commission is of the view that the availability of slots for early morning frequencies would still be a relevant issue in Dublin and at certain congested airports. Although airport congestion or slot availability is not an issue at Cork and Shannon, there are significant barriers to entry identified due in particular to the Parties' strong position at these airports.

(1633) In markets in which Ryanair and Aer Lingus would reach a large combined market share or even would create a monopoly, a remote likelihood of entry cannot be sufficient to dismiss the Commission's competition concerns. The Commission considers that sufficient and timely entry appears unlikely after the Transaction. The market investigation has not identified, in the absence of commitments, post- Transaction entry or expansion projects of competitors on the routes to and from Ireland, which would be likely, timely and sufficient in order to provide competitive constraints offsetting the effects of the Transaction. The absence of material entry or capacity increases by competitors in recent years, irrespective of the cause, suggests that the prospects for entry or expansion by other carriers are highly uncertain.

(1634) Moreover, considering the identified barriers to entry and the absence of material pattern of entry by competitors on short haul routes to and from Ireland, the Commission is also of the view that a mere threat of entry or expansion would not sufficiently constrain the Parties post-Transaction. (1467)

(1635) In view of the foregoing, the Commission concludes that the Transaction is likely to significantly impede effective competition, in particular as a result of the creation or strengthening of a dominant position, on the 46 routes listed in Table 101. Moreover, the Transaction is also likely to result in the elimination of the very close competitive relationship between Ryanair and Aer Lingus and thus to eliminate the important competitive constraint that both carriers exert upon each other pre-Transaction on the routes on which the merged entity would operate alongside other scheduled carriers or charter carriers.

Table 101: Routes with likely SIEC

1. Dublin-Alicante/Murcia 2. Dublin-Palma

3. Dublin-Berlin 4. Dublin-Lanzarote

5. Dublin-Bilbao/ Santander 6. Dublin-Gran Canaria

7. Dublin-Birmingham/East Midlands 8. Dublin-Ibiza

9. Dublin-Brussels/Charleroi 10. Dublin-Munich/Memmingen

11. Dublin-Budapest 12. Dublin-CDG/Beauvais/Orly

13. Dublin-Edinburgh/ Glasgow 14. Dublin-Stockholm Arlanda/Skavsta

15. Dublin-Fuerteventura 16. Dublin-London

17. Dublin-Glasgow/Prestwick 18. Dublin-Madrid

19. Dublin-Manchester/ Liverpool/Leeds 20. Cork-Barcelona/Reus/Gerona

21. Dublin-Marseille 22. Cork-London

23. Dublin-Milan 24. Cork-Lanzarote

25. Dublin-Nice 26. Cork-Malaga

27. Dublin-Rome Fiumicino/Ciampino 28. Cork-Palma

29. Dublin-Tenerife 30. Cork-Alicante/Murcia

31. Dublin-Toulouse/ Carcassonne 32. Cork-Faro

33. Dublin-Venice/Treviso 34. Cork-Manchester/Liverpool

35. Dublin-Vienna/ Bratislava 36. Cork-Tenerife

37. Dublin-Warsaw/Modlin 38. Shannon-Manchester/ Liverpool

39. Dublin-Malaga 40. Shannon-London

41. Dublin-Faro 42. Knock-London

43. Dublin- Barcelona/Reus/Gerona 44. Knock-Birmingham/East Midlands

45. Dublin-Frankfurt/Hahn 46. Dublin-Bristol/Cardiff/ Exeter

(1636) Moreover, the Commission concludes that the Transaction is likely to significantly impede effective competition, in particular as a result of the creation or strengthening of a dominant position, by eliminating the most credible potential entrant on the routes listed in Table 102 and Table 103:

Table 102: Ryanair routes of potential competition

1. Dublin-Bologna

2. Dublin-Bordeaux

3. Cork-Birmingham

4. Cork-Munich (Memmingen)

5 Cork-Paris (Beauvais)

Table 103: Aer Lingus routes of potential competition

1. Dublin - Newcastle

9. EFFICIENCIES

9.1. The principles

(1637) According to the Merger Regulation (1468) and the Merger Horizontal Guidelines (1469), it is possible that efficiencies brought about by a merger counteract the effects on competition and in particular the potential harm to consumers that it might otherwise have. Parties to a concentration may thus detail the efficiency gains generated by the concentration that are likely to enhance the ability and the incentive of the merged entity to act pro-competitively for the benefit of consumers. (1470) Efficiency claims need to be verifiable. The parties have to demonstrate that such efficiencies are likely to be passed on to consumers and could not have been achieved to a similar extent by means that are less anticompetitive than the proposed concentration (the so-called "merger specificity").

(1638) The three conditions - verifiability, merger specificity and consumer benefit - are cumulative.

9.2. Ryanair's claims

(1639) Ryanair claims that the Transaction brings about substantial efficiencies, which would benefit all customers (1471) as Ryanair applies its cost-cutting expertise to improve Aer Lingus' efficiency, lower its costs and air fares, and enhance its competitiveness against other airlines at primary airports while growing both its short-haul and long-haul traffic. Ryanair expects to generate synergies and savings in most cost categories, in particular staff costs, turnaround times, aircraft costs, fuel costs, maintenance costs, airport and handling costs, and distribution and other costs.

(1640) In the Form CO, Ryanair claims that it plans to reduce Aer Lingus' annual cost base (excluding fuel) by at least [...]*% or approximately [...]*. Assuming that these cost savings are achieved, Aer Lingus' unit operating cost (excluding fuel) per passenger would be reduced from EUR 100.08 to EUR [...]*.

(1641) Regarding the verifiability criterion, Ryanair claims in particular that "a direct comparison of the costs currently incurred by Ryanair as compared to those incurred by Aer Lingus shows that Ryanair's efficiency claims are tangible and achievable." (1472)

(1642) Regarding the merger-specificity criterion, Ryanair considers that Aer Lingus would not be able to achieve a significant reduction in its cost base independently of the Transaction. Ryanair is of the view that Aer Lingus' "Greenfield" cost reduction program is not capable of generating more than an additional EUR 13 million in cost savings in 2012, which contrasts with Ryanair's conservative estimate of the EUR 50 million per annum reduction in Aer Lingus' costs post-Transaction. Furthermore, a significant proportion of the efficiencies projected by Ryanair are derived from economies of scale, which are not available to Aer Lingus (1473).

(1643) Regarding the consumer benefit criterion, Ryanair considers that efficiencies would not be offset against reductions in elements which are beneficial to the consumer, such as the quality of service or airport location. The efficiencies would derive from the size and scale advantages already enjoyed by Ryanair. In addition, Ryanair plans to increase the volume of Aer Lingus' passengers from 9.5 million to 14.5 million passengers per annum in the next 5 years. This will enable Ryanair to generate even greater cost savings and to reduce overall costs per passenger. Finally, after the Transaction, Aer Lingus will, as a result of this reduction in costs, be able to compete more aggressively with mega carriers such as US Airways and American Airlines on Ireland-United States of America routes. It will also be in a position to expand its long-haul offering by, for example, flying to the west coast of the United States of America, and creating more routes between the United States of America and Europe, whilst providing lower fares to consumers across the board (1474).

9.3. Aer Lingus' position

(1644) Aer Lingus has developed a cost reduction programme called Greenfield. As at 31 December 2011, Aer Lingus had achieved total Greenfield cost savings with an annual value of EUR 84 million. The majority of savings achieved during 2011 relate to non-staff items and include initiatives relating to sourcing of catering stock, fuel efficiency savings and ground-handling process improvements (1475).

(1645) Moreover, Aer Lingus considers that Greenfield remained on track in the first half of 2012. Aer Lingus expects to achieve cost savings in 2012 with an annual value of at least EUR 97 million in the context of Greenfield (1476). Implementation of this programme would continue beyond 2012 (1477).

9.4. Commission's assessment

(1646) Efficiencies to be acceptable as a countervailing factor under the Merger Regulation must be verifiable, likely to be passed on to consumers and merger specific to the extent that no other practicable less-anticompetitive alternatives exist to achieve the same benefits. (1478) The three conditions - verifiability, merger specificity and consumer benefit - are cumulative.

(1647) Most of the information allowing the Commission to assess efficiencies is solely in the possession of the parties to the Transaction. It is therefore 'incumbent upon the notifying parties to provide in due time the relevant information' (1479).

(1648) Ryanair has not provided relevant information allowing the Commission to assess whether the claimed efficiencies meet the cumulative criteria set out in the Horizontal Merger Guidelines.

(1649) The more precise and convincing the efficiency claims of the Parties are, the better the Commission can evaluate the claims. In that regard point 86 of the Horizontal Merger Guidelines states that, where reasonably possible, efficiencies and the resulting benefit to consumers should therefore be 'quantified' and that, when the necessary data are not available to allow for a precise quantitative analysis, it must be possible to foresee a clearly identifiable positive impact on consumers, not a marginal one.

(1650) In this regard, the Commission has not found verifiable evidence, beyond Ryanair's mere statements, that the latter can reduce Aer Lingus' costs without offsetting reductions in other elements beneficial to consumers, such as quality of service or airport location. For instance, claimed efficiencies would derive from airport and handling costs, where Aer Lingus' airport and handling cost per passenger would be [...]*, while Ryanair's would be [...]* (1480). However, Ryanair also explains that Aer Lingus would continue to operate to primary airports post-Transaction (1481). Therefore, it is unclear to what extent the difference in cost between the two carriers could be translated into efficiencies as there would be no volume effects linked to the Transaction when Ryanair operates from or to another airport than Aer Lingus.

(1651) Other efficiencies would derive from maintenance costs, for which Aer Lingus' maintenance costs would be [...]* per passenger, while Ryanair's would be [...]* (1482) Ryanair does not explain how it would engender material efficiencies in this regard while Aer Lingus fleet is comprised of different aircraft than those used by Ryanair (Airbus vs Boeing).

(1652) Further efficiencies would stem from turnaround times. Ryanair's turnaround time between landing and take-off is usually 25 minutes, while Aer Lingus' turnaround times on short-haul flights is significantly longer. Ryanair does not provide detailed information on how efficiencies would be created. The Commission is of the view that several elements may have an influence on boarding time, such as the way passengers are boarding and unboarding, or the congestion degree of the airports served. The Commission considers that Ryanair's very rapid average turnover time stems, inter alia, from the way passenger boarding and seating are managed and from the fact that Ryanair often operates at secondary airports which are not congested or at least less congested than primary airports. Since Aer Lingus operates at main airports, a faster boarding time is less likely to be achieved. Cost reductions would then either derive from a possible lessening in service quality for passengers, and would relate to a change in the airport mix served by Aer Lingus, which would also affect service quality.

(1653) Ryanair also notes in the response to the Statement of Objections that "a significant proportion of the efficiencies projected by Ryanair are derived from economies of scale." (1483) However, the magnitude of this proportion is not described.

(1654) In its response to the Statement of Objections, Ryanair refers to the ability of Aer Lingus post-Transaction to compete more aggressively on routes to the United States of America (1484). This assertion is not substantiated (1485).

(1655) Generally, the Commission is of the view that data provided by Ryanair are suffering from a significant lack of precision.

(1656) Moreover, looking at figures provided by Ryanair and Aer Lingus, the "net" efficiencies made of the insufficiently detailed EUR 50 million claimed by Ryanair minus the efficiencies which would stem from Aer Lingus' "Greenfield" project, which is not merger-specific, would represent, according to Ryanair's own estimates around EUR 37 million. According to Aer Lingus's figures, the "Greenfield" project achieved cost savings with an annual value of EUR 84 million and was expected to generate at least EUR 97 million for 2012. According to Aer Lingus preliminary results for 2012, the annualised cost savings target was exceeded by 7% (€104 million in annualised savings compared to targeted €97 million).

(1657) Furthermore, it is also "for the parties to provide in due time all the relevant information necessary to demonstrate that there are no less anti-competitive, realistic and attainable alternatives than the notified concentration". (1486) In order to establish merger specificity the Commission would only consider alternatives that are reasonably practical in the business situation of the parties, having regard to established practice in the air transport sector.

(1658) Ryanair satisfies itself with claiming that a significant proportion of the [...]* efficiencies projected by Ryanair would be derived from economies of scale, which are not available to Aer Lingus (1487). The Commission does not contest that Ryanair is significantly bigger in size than Aer Lingus'. However, it is not apparent that mere size would lead to merger specific efficiencies, for instance, in staff costs, which could also be achieved by Aer Lingus independently of the Transaction.

(1659) Moreover, the Horizontal Merger Guidelines state that "the incentive on the part of the merged entity to pass efficiency gains on to consumers is often related to the existence of competitive pressure from the remaining firms in the market and from potential entry". It is further indicated that: "It is highly unlikely that a merger leading to a market position approaching that of a monopoly, or leading to a similar level of market power, can be declared compatible with the common market on the ground that efficiency gains would be sufficient to counteract its potential anticompetitive effects" (1488).

(1660) Post-Transaction, as explained in detail in the present Decision, the merged entity would enjoy very significant market power on a high number of routes ex Ireland. Assuming profit maximisation behaviour by the merged entity (1489), and considering the fact that the claimed efficiencies would apply essentially to Aer Lingus, it is highly unlikely that the theoretical price-reducing effect of the efficiencies would be sufficient to reverse the price increase resulting from the Transaction. Even taking Ryanair's own expected cost savings of around 5% into account, this level of scost savings is extremely unlikely to be sufficient to counteract the likely increase in prices that would result from the merger.

(1661) The reference made by Ryanair that passing on efficiencies to consumers through lower fares would be the only way Ryanair can deliver on its offer commitment to grow Aer Lingus' traffic from 9.5 million to 14 million passengers over a five-year period, does not change the Commission's assessment. First, this offer was rejected by the Irish Government (1490). Secondly, as for any firm, Ryanair's objective would still be that of (at least long run) maximising profit.

9.5. Conclusion

(1662) The three conditions pertaining to efficiencies - verifiability, merger specificity and consumer benefit - are not met.

(1663) Given the extremely high combined market shares, often approaching monopoly levels, and the absence of timely, sufficient, and likely entry on the routes on which competition problems have been identified, it appears that any sufficient pass-on of alleged efficiencies to consumers would not likely take place.

(1664) Therefore, Ryanair has not provided sufficient evidence to demonstrate that any efficiency claims would materialise and counteract the competitive harm likely to arise from the Transaction.

10. COMMITMENTS

(1665) On 17 October 2012, Ryanair submitted its first set of Commitments aimed at addressing the serious competition issues identified by the Commission in the decision opening the proceedings (1491). These commitments were not market tested for the reasons stated in Section 10.2.1.

(1666) In order to address the competition concerns identified in the Statement of Objections, Ryanair submitted the Commitments of 7 December 2012.

(1667) The Commitments of 7 December 2012 were the subject of the first market test launched on 11 December 2012, by which the Commission sought to gather the views of competitors, customers, airport operators and associations on the effectiveness of the commitments submitted and their ability to restore competition on the markets where competition concerns were identified. In addition, the Commission sent further specific questions (1492), in particular to the identified potential upfront buyers ("UFB") (1493) in order to understand the UFBs' intentions towards the divestments proposed by the Commitments of 7 December 2012.

(1668) A State-of-Play meeting was held on 8 January 2013, at which Ryanair was informed of the results of the first market test and the Commission's preliminary assessment of the 7 December Commitments.

(1669) In order address the concerns expressed by the Commission, Ryanair submitted the Commitments of 15 January 2013. On 21 January 2013 (1494), the Commission launched the second market test (1495).

(1670) On 28 January 2013, a further State-of-Play meeting was held in which Ryanair was informed about the results of the second market test and the Commission's preliminary assessment of the Commitments of 15 January 2013.

(1671) Ryanair subsequently submitted the final Commitments dated 1 February 2013 (1496). On 4 February 2013, the Commission launched the final market test (1497). A final State-of-Play meeting was held with Ryanair on 12 February 2013.

10.1. The analytical framework for the assessment of proposed remedies

10.1.1. Overview

(1672) When a concentration raises competition concerns because it could significantly impede effective competition, in particular as a result of the creation or strengthening of a dominant position, the parties may seek to modify the concentration in order to resolve the competition concerns and thereby gain clearance of their merger. The commitments have to eliminate the competition concerns entirely and have to be comprehensive and effective in all respects. In assessing whether proposed commitments are likely to eliminate the competition concerns identified, the Commission will consider all relevant factors including inter alia the type, scale and scope of the proposed commitments, judged by reference to the structure and particular characteristics of the market in which the competition concerns arise, including the position of the parties and other participants on the market. (1498)

(1673) The Commission has to be able to conclude with the requisite degree of certainty that it will be possible to implement the commitments and that it is likely that the new commercial structures resulting from them will be sufficiently workable and lasting to ensure that the significant impediment to effective competition will not materialize (1499).

(1674) In the 2010 Judgment (1500), the Court clarified that "it must be held in that regard that commitments proposed by one of the parties to a merger will meet that condition only in so far as the Commission is able to conclude, with certainty, that it will be possible to implement them and that the remedies resulting from them will be sufficiently workable and lasting to ensure that the creation or strengthening of a dominant position, or the impairment of effective competition, which the commitments are intended to prevent, will not be likely to materialise in the relatively near future."

(1675) Ryanair's last two sets of Commitments were submitted well after 21 December 2013, which corresponded to day 65 of the procedure. As regards commitments which are submitted out of time, the Court held in the 2010 Judgment that "it is apparent from point 43 of the Notice on remedies acceptable under the former merger regulation and under Commission Regulation (EC) No 447/98 (OJ 2001 C 68, p. 3) ('the notice on remedies'), whose approach was adopted by the Commission in the present case as regards the merger regulation and Regulation No 802/2004, that the parties to a notified concentration may have such commitments taken into account subject to two cumulative conditions, namely, first, that those commitments clearly and without the need for further investigation resolve the competition concerns previously identified and, secondly, that there is sufficient time to consult the Member States on those commitments..." (1501) In that regard, it should be noted that the Commission exceptionally decided to examine the last two sets of Commitments submitted by Ryanair despite the fact that these Commitments did not represent only limited modifications of the previous commitments and that additional market tests had to be conducted.

(1676) Commitments must be put forward by the Parties and the Commission has the power to accept only such commitments as are capable of rendering the Transaction compatible with the internal market.

(1677) Ryanair's Commitments amount in substance to fix-it-first type of remedy. Indeed, Ryanair committed to enter into legally binding agreements with Flybe and IAG prior to the Commission's Decision, conditional on the Commission's acceptance of Flybe and IAG as suitable purchasers, the adoption of the Decision, and the closing of the Transaction (1502).

(1678) A fix-it first remedy implies that "the Parties identify and enter into a legally binding agreement with a buyer outlining the essentials of the purchase during the Commission's procedure" and "such agreements are normally conditional on the final Commission decision accepting the remedy in question" (1503).

(1679) The Commission welcomes fix-it-first remedies in particular where the identity of the purchaser is crucial for the effectiveness of the proposed remedy because the purchaser needs to have specific characteristics in order for the remedy to solve the competition concerns. If the parties choose to enter into a binding agreement with a suitable purchaser during the procedure by way of a fix-it-first solution, the Commission can in those circumstances conclude with certainty that the commitments will be implemented with a sale to a suitable purchaser. This can resolve the Commission's concerns particularly in cases where, given the circumstances, only very few potential purchasers can be considered suitable (1504).

(1680) A clearance decision which is combined with a fix-it-first remedy is appropriate in cases in which there is a sufficient degree of likelihood that the entrant would eliminate all identified competition concerns and that the remedy would be workable and implementable (1505).

(1681) Lastly the Commission notes that the Commitments of 17 October 2012, of 7 December 2012, and of 15 January 2013, as well as the final Commitments cannot be characterized as up-front buyer remedies as that requires inter alia a clear and precise clause which actually makes the implementation of the Transaction conditional on the implementation of the corresponding commitment. Such a clause does not exist in any of the Commitments proposed by Ryanair.

10.1.2. Elimination of all identified competition concerns

(1682) Before analysing the successive Commitments offered by Ryanair, it should first be emphasised that the Transaction comprises the merger of two Irish scheduled carriers based at Dublin, Cork, and Shannon airports, having by far the largest bases at these airports, and de facto covering the large majority of short haul routes ex Ireland.

(1683) As explained extensively elsewhere in this Decision, in order to be able to exert a sufficient competitive constraint on the merged entity so as to counteract a potential price increase post-Transaction, a new entrant would in particular have to be based at the Irish ends of the routes, have a strong brand to be able to compete with those of the merged entity, and be in a position to withstand in the medium term the competitive pressure of the merged entity.

(1684) In order to be effective, any proposed commitments would have to address these issues.

(1685) Regarding the exact form of the commitments, the General Court has pointed out that, while the Merger Regulation lays down the principle whereby the Commission may authorize a merger only if the proposed commitments address the competition concerns identified, it "leaves it a wide discretion as to the form which the commitments in question may take." (1506)

(1686) If "divestitures are the benchmark for other remedies in terms of effectiveness and efficiency", the Commission "may accept other types of commitments but only in circumstances where the other remedy proposed is at least equivalent in its effects to a divestiture" (1507). Often, a sufficient reduction of entry barriers is not achieved by individual measures, but by a package comprising a combination of divestiture remedies and access commitments or a commitments package with the overall aim of facilitating entry of competitors. If the commitments offered actually make entry by sufficient competitors timely and likely, they can be considered to have a similar effect on competition in the market as a divestiture. If it cannot be concluded that the lowering of the entry barriers by the proposed commitments is likely to lead to the entry of new competitors, such as to eliminate any significant impediment to effective competition in the market, the Commission will reject such remedy package. The notice on remedies points out specifically that, in air transport mergers, a mere reduction of barriers to entry by a commitment of the parties to offer slots on specific airports may not always be sufficient to ensure the entry of new competitors on those routes where competition problems arise and to render the remedy equivalent in its effects to a divestiture (1508).

(1687) According to paragraph 68 of the Horizontal Merger Guidelines, "for entry to be sufficient competitive constraint on the merging parties, it must be shown to be likely, timely and sufficient to deter or defeat any potential anti-competitive effects of the merger." (1509)

(1688) The effectiveness of the Commitments of 17 October 2012, of 7 December 2012 and of 15 January 2013, as well as of the final Commitments, and the results of the market tests have been analysed notably with a view to determining whether they create a prospect of entry that is likely, timely and sufficient to exert competitive constraint on the merged entity so as to offset the likely anticompetitive effects specifically established in this Decision (1510).

10.1.3. The remedy package must be workable and implementable within a short period of time

(1689) The Commitments have to eliminate the competition concerns entirely and have to be comprehensive and effective from all points of view. Furthermore, commitments must be capable of being implemented effectively within a short period of time as the conditions of competition on the market will not be maintained until the commitments have been fulfilled (1511).

(1690) The Commission has therefore investigated whether all the commitments submitted by Ryanair would be fully workable and implementable so that it could be concluded with the requisite degree of certainty that it would be possible to implement them and that the new commercial structures resulting from them would be sufficiently workable and lasting to ensure that the significant impediment to effective competition would not materialise in the near future.

(1691) In addition, in the case of fix-it first remedies, a legally binding agreement with one or more suitable UFB must be concluded during the Commission procedure, in order for the Commission's final decision to be able to determine whether the transfer of the divested business to the identified UFBs on the terms of the proposed divestiture will remove the competition concerns (1512).

(1692) It should be mentioned that while Ryanair referred to numerous UFBs at various stages of the investigation, the vast majority of the alleged UFBs eventually declined the offers made by Ryanair (this concerns easyJet, Etihad, Norwegian, Virgin Atlantic, Air France/City Jet and Wizz Air (1513)).

10.1.4. Commission precedents

(1693) Ryanair argued that all its proposed Commitments significantly exceeded the scope of prior commitments made in the context of airline mergers (1514).

(1694) The Commission points out that whether a remedy is suitable to eliminate the competition concerns identified can only be examined on a case-by-case basis (1515). The mere fact that certain remedies have been accepted in previous airline cases cannot be a decisive factor and justify accepting such remedies in the present case. The Commission has to assess the proposed commitments taking into account the specificities of the transaction under review.

(1695) The correctness of a case-by-case approach was acknowledged by the Court in the 2010 Judgment (1516), in the context of the remedies proposed in that case, in essence the release of slots: "...Unlike previous mergers in the passenger air transport sector (such as those which were at issue in Air France/KLM and Lufthansa/Swiss), the Commission could not be satisfied in the present case that mere slots would ensure access to a route. This is not a transaction involving active operators which have a home airport in different countries. Ryanair and Aer Lingus operate from the same airport, Dublin Airport, where they have significant advantages which could not easily be countered by competitors."

(1696) Furthermore, the Court found in ARD v Commission that "a comparison with other merger cases can be relevant only if it is established that they raise the same competition problems and concern markets with the same characteristics and where conditions have not changed." (1517) This reasoning also applies to the analysis of commitments.

(1697) The Commission notes that Ryanair's analysis of previous commitments made in the context of airline mergers ignores both the 2007 Decision involving Ryanair and Aer Lingus and the Decision in case COMP/M.5830 Olympic/Aegean Airlines (1518), where both Olympic and Aegean operated from the same airport. In the latter case, the proposed remedies (essentially in the form of slot releases) were dismissed inter alia as they were unlikely to induce a credible new player to create a base at Athens airport and exert a credible competitive constraint on the affected routes.

10.2. Description of the Commitments of 17 October 2012, 7 December 2012 and 15 January 2013

(1698) As noted above, Ryanair submitted four remedies packages to the Commission; on 17 October 2012, 7 December 2012, 15 January 2013 and lastly on 1 February 2013, as described in the Sections that follow.

10.2.1. Commitments of 17 October 2012

10.2.1.1. Description

(1699) According to Ryanair, the Commitments of 17 October 2012 identified multiple UFBs (1519) that would have opened new bases at Dublin and London Heathrow to guarantee continuing competition on all relevant routes by agreeing to take over some or all of the overlapping frequencies (and in some cases more) and/or to enter into block-space agreements ("BSAs") (1520) on thinner or seasonal routes beginning at the next practicable IATA season changeover.

(1700) Ryanair considered that it had offered a package of remedies based on tried and tested principles. In Ryanair's view, the Commitments of 17 October 2012 were argued to be unprecedented in their scale and scope (1521).

(1701) The Commitments of 17 October 2012 addressed 39 routes. Ryanair considered that 5 routes required no remedies given the strong competition that exists from one or more major Union carriers currently operating on the route from one of their major base airports (1522) and where (in 3 of the 5 cases) the airports were, in Ryanair's view, clearly not substitutable (1523). Furthermore, Ryanair explained that no remedies were required on 3 other routes (1524) operated by Ryanair and Aer Arann, because Aer Arann is an independently owned and managed airline that is not part of the Aer Lingus Group, so it would not be appropriate to treat these as overlap routes. In addition, on each of these routes, Ryanair claimed that no overlap arose in any event because Ryanair and Aer Arann fly to non-substitutable airports.

10.2.1.2. No market test was warranted before the Statement of Objections

(1702) Pursuant to Article 10(2) of the Merger Regulation, the Commission has to take a clearance decision as soon as the serious doubts referred to in Article 6(1)(c) of the Merger Regulation are removed as a result of commitments submitted by the parties. This rule applies to commitments proposed in phase II-proceedings before the Commission issues a statement of objections. The Commission may decide not to carry out a market test if it is clear from the information already at its disposal that the proposed remedies cannot be accepted.

(1703) Ryanair presented the Commitments of 17 October 2012 as being radical and Transformational (1525).

(1704) However, the Commitments of 17 October 2012 did not cover all routes where the decision opening the proceedings raised serious doubts. In particular 8 overlap routes were outside the scope of these Commitments. Moreover, the Commitments of 17 October 2012 did not cover all routes on which the SO preliminarily concluded that the Transaction would significantly impede effective competition (1526).

(1705) Therefore, it was clear that no market test was warranted for the Commitments of 17 October 2012.

10.2.2. Commitments of 7 December 2012

10.2.2.1. Description

(1706) Ryanair submitted revised commitments on 7 December 2012. According to Ryanair, the main aspects of these remedies (1527) were as follows:

(1707) IAG would take over Aer Lingus' entire existing Dublin-London Heathrow, Cork- London Heathrow and Shannon-London Heathrow schedule and operate aircraft bases at Dublin, Heathrow, Cork, and Shannon, with 6 aircraft in total (2 at Dublin, 1 at Cork, 1 at Shannon and 2 at Heathrow), in order to serve these 3 routes.

(1708) In addition, Ryanair committed to lease, or to cause Aer Lingus to lease, the Dublin- Heathrow Slots, the Cork-Heathrow Slots and the Shannon-Heathrow Slots to IAG for 6 IATA seasons. At the end of the 6 IATA seasons, IAG could elect either (i) to acquire these slots from Aer Lingus at their current open market value or (ii) to extend the leases for these Slots for another 4 IATA seasons. If IAG elected to extend the leases, it would have the right to acquire these Slots from Aer Lingus at their current open market value at the end of the additional 4 IATA seasons. In the event that Ryanair was prevented from extending the lease or from selling the slots to IAG as a consequence of the application of Article 10 of Aer Lingus' Articles of Association, Ryanair would enable IAG to continue to operate these Slots, either by means of a form of joint operation or through similar means such as wet-leasing of aircraft, until such time as it was possible for Ryanair to sell the slots to IAG.

(1709) Flybe would take over Aer Lingus' existing flights on 20 key overlap city pairs out of Dublin airport, Cork and Shannon (1528), establish and operate a 6 aircraft base in Dublin and enter into BSAs on 9 thin and/or seasonal city pairs from Dublin and Cork airports (1529).

(1710) Both IAG and Flybe would enter into BSAs on Aer Lingus' flights on 14 thin and/or seasonal city pairs, mainly charter/leisure routes, from Dublin, Cork and Knock Airports (1530).

(1711) In addition, Ryanair committed to preserve the Aer Lingus/Aer Arann franchise agreement on existing terms on all franchise routes with the exception of the 6 routes on which Aer Arann (operating under the Aer Lingus Regional brand) and Ryanair both operate and which, post-Transaction, would be operated by Flybe. The 6 routes are Dublin-Edinburgh, Dublin-Glasgow, Dublin-Bristol, Cork-Manchester, Shannon- Manchester and Knock-Birmingham.

(1712) Regarding the 23 overlap routes on which Ryanair would enter into BSAs with IAG and/or Flybe to guarantee continued competition, Ryanair also proposed the divestiture of slots to any applicant, including IAG and Flybe, which would plan to begin operating with its own frequencies on those routes (1531).

(1713) BSAs would also be offered on all 20 potential competition routes identified by the Commission, for 2 IATA seasons, on request by any applicant, the request to be made during the 6 IATA seasons or at any time thereafter (1532). Slot divestments, up to the number of frequencies currently operated by Ryanair / Aer Lingus, would be offered for an indefinite period of time to enable any applicant to enter any of the 20 routes with their own aircraft.

(1714) In addition, Ryanair committed not to enter the 3 Heathrow routes which IAG would operate with its own aircraft for 6 (or 10 in case of extension (1533)) IATA seasons, not to increase the remaining Ryanair or Aer Lingus frequencies on any route where Flybe entered with its own aircraft for 6 IATA seasons, and not to increase Aer Lingus' frequencies on routes subject to BSAs (while the BSAs were in force), unless otherwise requested or agreed by the UFB (1534). At the request of IAG, Flybe, or any applicant that acquired slots to operate on BSA routes, or Aer Arann in respect of Dublin-Edinburgh, Dublin-Glasgow, Dublin-Bristol, Cork-Manchester, Shannon- Manchester and Knock-Birmingham routes, Ryanair would allow such party to be hosted in Aer Lingus' Frequent Flyer Programme.

(1715) Ryanair stated that "its unprecedented remedies package will entail the withdrawal of up to 12 of Aer Lingus' 36 short-haul aircraft (33% of its short-haul fleet) from 23 city pairs" (1535) and the transfer of more than 4 million of Aer Lingus' 8.8 million short-haul passengers per annum (1536) to two substantial competitor airlines (IAG and Flybe)" (1537).

10.2.2.2. The Commission's assessment

(1716) On the basis of all the information available to the Commission and the evidence collected during the first market investigation, the Commission concluded that the Commitments of 7 December 2012 would not be likely to eliminate the competition concerns identified.

(1717) As regards the commitments regarding the Dublin-London Heathrow, Cork-London Heathrow and Shannon-London Heathrow routes, the Commission's investigation revealed in particular that it did not appear with the requisite degree of certainty that a permanent transfer of Heathrow slots to IAG would take place, in particular because of a protection mechanism enshrined in Article 10 of Aer Lingus' Articles of Association (see Section10.5.1).

(1718) Furthermore, the Commission found that the Commitments of 7 December 2012 did not contain any specific provision related to a commitment of IAG to buy the slots at the end of the lease period (1538), and no possibility existed that other interested carriers could apply for these slots. Hence it could not be excluded that at the end of 6 (or 10) IATA seasons, Ryanair/Aer Lingus would take the Heathrow slots back. Lastly, the Commitments of 7 December 2012, by identifying IAG as the proposed entrant on the three routes, were also likely to be insufficient in scope considering the likely reallocation to other (long haul) routes of at least some of the slots currently operated by IAG on the Dublin-London Heathrow route. Therefore the likely dominant position of the merged entity would have been maintained (1539).

(1719) Concerning the Flybe base commitment, the Commission considered that the competitive constraint exercised by Flybe from its 6 aircraft base in Dublin would not be sufficient to defeat the anti-competitive effects of the Transaction. Flybe would not be likely to maintain and develop its Dublin base, and operate all the 20 routes exerting sufficient constraint to defeat the anticompetitive effects of the Transaction. There were mixed answers to the first market test questionnaire (1540) with several respondents supporting the view that Flybe would be a sufficient constraint on the merged entity (1541), whereas equally credible other respondents (1542) held the opposite view. The Commission has concluded that Flybe is a distant competitor to the merged entity, with a rather different business model and different fleet composition, focusing on business passengers, and with very limited brand recognition outside the United Kingdom. Furthermore, Flybe is significantly smaller than Aer Lingus (in terms of revenues) and its Cost of Available Seat-Kilometer ("CASK") (1543), based on operating regional jets, are much higher than those of Ryanair and Aer Lingus, which operate B737 and A320 aircraft (1544). In addition, Flybe's financial situation does not appear to be strong (1545); while Aer Lingus has been profitable over recent years, and has significant cash reserves; so the Commission considered it unlikely that Flybe would be in a position to provide the medium term financial support to make sufficiently certain that the Dublin base would be maintained after the Commitment period (1546). Therefore, the Commission considered that Flybe would not constrain to a sufficient extent the merged entity on the 20 routes at stake.

(1720) Therefore, the Commission considered that the Commitments of 7 December 2012 did not eliminate the significant impediment to effective competition as far as the 20 Flybe Routes were concerned.

(1721) Furthermore, while the BSA and slot release measures might to some limited extent have facilitated entry on such routes, the Commission found that it was unlikely that these measures would make entry likely, at least on a significant number of routes such as [...]* as these routes are beyond the range accessible by Flybe's aircraft. Furthermore, [...]* and [...]* would not be profitable given Flybe's business model and the too long distances between Dublin and these two destinations.

(1722) Moreover, even if entry were likely, entry would remain insufficient on many routes. For instance, the UFBs' market shares obtained through BSA would remain generally low. Furthermore, the Commission found that BSAs were not likely to allow a carrier to establish a brand presence and independent track record, and to compete on all of the many parameters of competition in which differentiated carriers compete, including, for example, to freely set capacity (e.g. number of seats offered, aircraft type), to decide on flight times, training and incentives of crew members, quality of service, food provision, designated seating, seat pitch offered to passengers, onaircraft interiors, cost management, etc. Price competition would be at best very limited.

(1723) Therefore, considering the BSAs, the slot releases, and the related measures like the frequency limitation or freeze, the Commission concluded that Flybe, or other carriers, would not be likely to be able to effectively enter and build a presence on all the routes in the short-term and would not constrain the merged entity to a sufficient extent. Moreover, considering for instance its business model, IAG would not likely enter on (most of) these routes. Therefore, these measures would not be likely to lead to effective competition on a lasting basis.

(1724) In addition, as regards the termination of the Aer Arann franchise agreement on the 6 relevant routes as set out in the Commitments of 7 December 2012, it did not appear with the requisite degree of certainty that Ryanair would be in a position post- Transaction to implement the Commitment and modify the scope of the franchise agreement in a timely manner because of Aer Arann's consent rights (1547).

(1725) Lastly, no binding agreement with the UFBs was sent to the Commission in relation to the Commitments of 7 December 2012, despite their fix-it-first nature.

(1726) Based on all the available evidence, including the results of the first market test, the Commission considered that the Commitments of 7 December 2012 would not lower barriers to entry sufficiently to lead to the entry of new competitors able to exert a sufficient competitive constraint on the merged entity. In addition, the Commission could not conclude, with the requisite degree of certainty, that the Commitments of 7 December 2012 could be implemented and that they would be sufficiently workable and lasting to ensure that the significant impediment of effective competition which those Commitments were intended to remove would not be likely to materialise in the relatively near future.

(1727) On the basis of the assessment of the Commitments of 7 December 2012, the Commission therefore concluded that those Commitments were unable to remedy the significant impediments to effective competition identified, and, thus, could not render the proposed concentration compatible with the internal market.

10.2.3. The Commitments of 15 January 2013

10.2.3.1. Description

(1728) Ryanair submitted a revised set of remedies on 15 January 2013.

(1729) According to Ryanair (1548), the Commitments of 15 January 2013 eliminated all competition concerns identified in the SO. The Commitments of 15 January 2013 retained core elements of the Commitments of 7 December 2012, including the involvement of 2 UFBs (IAG and Flybe), but incorporated structural changes that, according to Ryanair, addressed the issues raised by the Commission regarding the Commitments of 7 December 2012. Ryanair considered that this remedies package was based on clear-cut structural divestitures and would result in the establishment of independent competitors with their own operations on all the overlapping routes.

(1730) Ryanair would enter into legally binding agreements with the approved UFBs prior to the adoption of the final Decision, those agreements being conditional only on the Commission's approval and the closing of the Transaction (1549). The Commitments of 15 January 2013 would therefore be "fix-it-first" commitments. Ryanair insisted that these Commitments would provide the Commission with far more certainty about their effective implementation than was provided in the context of any previous airline merger case (1550).

(1731) According to Ryanair, the main features of the Commitments of 15 January 2013 were as follows: - Divestiture of Aer Lingus' operations to Flybe on 43 overlap routes (1551)

(1732) Eliminating the BSAs offered under previous versions of its Commitments, Ryanair committed to divest a stand-alone business for operating Aer Lingus' actual operations on 43 (1552) (out of the 46) overlap routes to become a wholly-owned subsidiary of Flybe, possibly named "Flybe Ireland". The package would include the divestiture of up to 12 aircraft from Aer Lingus (1553), together with the divestiture of airport slots, and the transfer of personnel (for example pilots, cabin crew, and engineers), relevant contracts, forward bookings, a share of Aer Lingus existing capital, and any other elements necessary to operate the 43 city pairs via this standalone entity. Flybe would use the slots exclusively to operate the agreed schedule (1554) on each of the 43 routes during the first 6 IATA seasons. In case of misuse of the slots, such as cessation of services or reallocation of the slots to other routes, Flybe would need to hand back the slots to Ryanair.

(1733) The remedy also foresaw that Flybe would base the divested aircraft at Dublin airport for 6 IATA seasons.

(1734) The selling price for the divested business would be EUR 1 million. In addition, Ryanair undertook to sufficiently capitalise the divested business "with reference to historical performance of the 43 Flybe Routes and Aer Lingus' current capitalisation" (1555) to maintain its current estimated annual pre-tax profitability in the first 3-5 year trading period, and thus its long-term viability.

- Divestiture to IAG of slots on 3 overlap routes to London Heathrow (1556)

(1735) Subject to applicable law (in particular, Aer Lingus' Articles of Association), Ryanair committed to lease slots (at origin and destination) to IAG in order to enable it to operate the agreed schedule (1557) on the Dublin-Heathrow, Cork-Heathrow and Shannon-Heathrow routes. IAG would acquire slots to operate 6 daily frequencies on the Dublin-London Heathrow route, as well as the entirety of Aer Lingus' current schedule on the Cork-London Heathrow route (28 weekly frequencies) and Shannon- London Heathrow route (21 weekly frequencies). The merged entity would retain slots and continue to operate 47 weekly frequencies on the Dublin-London Heathrow route in competition with IAG and other carriers. During the first 6 IATA seasons (which, upon request by IAG, could be prolonged by another 4 IATA Seasons), IAG could only use the slots on the three above mentioned routes. After these 6 or 10 IATA seasons, IAG would acquire the slots from Aer Lingus against payment of an agreed price. In case of misuse, IAG would hand back the slots to Ryanair.

(1736) In the event that Article 10 of Aer Lingus' Articles of Association prevented Ryanair from leasing or selling the slots to IAG, Ryanair would enable IAG to operate these Manchester/Liverpool, Cork-Palma, Cork-Tenerife, Knock-Birmingham/East Midlands, Knock-London and Shannon-Manchester/Liverpool. slots, either by means of a form of joint operation or through similar means such as wet-leasing of aircraft, until such time as it was possible for Ryanair to sell these slots to IAG (1558).

- Aer Arann Franchise (1559)

(1737) Under the Commitments of 15 January 2013, Ryanair would have caused Aer Lingus to serve on Aer Arann, a notice of termination of the Aer Arann Franchise, as far as it related to any Flybe Routes, as soon as practicable following Closing, and in accordance with the terms of the Aer Arann Franchise (1560).

(1738) In addition to the commitment to modify the scope of Aer Arann's franchise agreement in relation to the Flybe Routes, Ryanair committed to facilitate the transfer of the Aer Arann Franchise to Flybe, so as to enable Aer Arann to operate under Flybe's franchise on any of the Flybe Routes, if requested by Flybe.

10.2.3.2. The Commission's assessment

(1739) On the basis of all the information available to the Commission and the evidence collected during the market investigation, the Commission concluded that the Commitments of 15 January 2013 would not be likely to eliminate all the competition concerns and did not remove the uncertainties as regards their implementation. The main reasons are explained below.

(1740) As regards the full divestiture of Aer Lingus' operations to Flybe on 43 overlap Routes (1561), the Commitments of 15 January 2013 were insufficiently clear with respect to the requirement for Flybe to establish and maintain a base with at least 12 aircraft in Dublin; moreover it seemed that Flybe would not establish a base at Cork, while the presence of a base has been identified as a significant element for a competitor to be able to exert sufficient competition constraint on the merged entity (see Sections 8.4.3 and 8.5.2.3). The Commission's investigation showed that the scope of the Commitments of 15 January 2013 was unlikely to be sufficient, considering in particular the insufficient number of divested frequencies on some routes (also taking into account some rounding effects), the fact that the completion penalty would only be payable by Flybe should Flybe not operate 80% of its daily frequencies on any of the Flybe routes (1562), and the risks attached to the size of the fleet.

(1741) In addition, based on the market investigation, and again despite mixed views expressed by respondents to the second market test questionnaire (1563), the Commission could not conclude with the requisite degree of certainty that Flybe would be able to constrain the merged entity sufficiently post-transaction, mainly due to its different business model, the lack of brand awareness in particular in Ireland, the route composition of the divested business, and its financial situation. Moreover, several respondents to the second market test (1564) expressed doubts as to the ability and the intention of Flybe to operate the divested business after 6 IATA seasons. Therefore, the uncertainty as to Flybe's commitment to operate the divested business for a sufficient period of time cast considerable doubts on its suitability as a purchaser of the divested business.

(1742) Moreover, several market participants raised concerns that the divested business would not be a viable business, because of the nature of the routes operated by Flybe (for example, the relatively high number of seasonal routes), the lack of brand awareness of Flybe and the fact that it would not benefit from the long haul operations of Aer Lingus (1565).

(1743) The profitability, and hence the viability, of the divested business was still unclear.

(1744) It is also noted that the Commitments of 15 January 2013 provided, without further elaboration, that "The Divestment business will be sufficiently capitalised, with reference to historical performance of the 43 Flybe Routes and Aer Lingus' current capitalisation, to maintain its current estimated pre-tax profitability of [...]* per annum in the first 3-5 year trading period, and thus its long-term viability." For a fix-it first type of remedy, such a clause would be unacceptable because of its lack of clarity. In addition, Ryanair's commitment to capitalise the divested business raised serious concerns as it could not be excluded at that stage that the financial support, to be provided by Ryanair in favour of Flybe, the concrete form of which was not described in the Commitments of 15 January 2013, would constitute a sponsored entry of Flybe by Ryanair, creating unacceptable continuing financial links between Ryanair and Flybe (1566).

(1745) As regards the divestiture to IAG of slots on 3 overlap routes to London Heathrow, the Commission concluded that the remedy could not be implemented as a consequence of restrictions on the disposal of slots at Heathrow in Aer Lingus' Articles of Association (see Section 10.5.1) (1567).

(1746) The Commission also found that the "joint operation" or "wet lease" approach, which was put forward in a very abstract and unclear manner by Ryanair, was not an appropriate alternative (1568). The Commission points out that it is for the Parties to provide all relevant explanations so as to satisfy the Commission of the precise content of the proposed commitments and it is not for the Commission to second guess the Parties' intentions in this regard.

(1747) It was only in a legal Opinion of 14 January 2013 that these concepts were explained by Ryanair. A "joint operation" was defined as a concept in the IATA Worldwide Slot Guidelines, involving "slots held by one airline being used by another airline". Under such joint operation, "the original slot holder retains historic precedence, not the operator of the slots. The slot holder is responsible for initial submissions and typically retains control of the slots until the Slot Return Deadline" and "[a]t the end of the shared operation, the slots involved in a shared operation remain allocated to the original slot holder".

(1748) Under the European Union Slot Regulation (1569) the concept of "joint operation" is caught under Article 10(8) (1570). Air carriers involved in such joint or shared operations must advise the slot coordinator of the detail of such operations prior to the beginning of such operations.

(1749) Ryanair's legal Opinion of 14 January 2013 further explained the concept of a "wetlease" as "an agreement between air carriers pursuant to which the aircraft is operated under the AOC [air operator certificate] of the lessor". It further explained that "under a wet lease arrangement an aircraft is operated by the lessee for the benefit of the lessor who essentially remains responsible for the state and maintenance of the aircraft i.e. the lessor retains effective control of the flight. The presumption, therefore, is that the lessor is the aircraft operator and that the flight plan will contain the ICAO designator of the lessor/owner or the registration marking of the aircraft."

(1750) The slot coordinator at London Heathrow confirmed to the Commission that in order to qualify as a joint operation "it would not be sufficient to have a mere codeshare arrangement but the airlines involved must demonstrate a substantive cooperation in providing the services jointly. This would typically involve a risk sharing mechanism or alike, for example a block seat arrangement" (1571) ACL further explained that "An important criterion for a cooperation to qualify as a "joint operation" would also be that tickets are actively marketed and sold on both airline's websites. ACL's role is to assess the operation in order to ensure that there is no fake arrangement, put in place just to facilitate the use of the slots at London Heathrow but that there is to the contrary a genuine commercial cooperation behind it." (1572)

(1751) Therefore, the Commission considered it questionable that a joint operation or a wet leasing arrangement could be a credible alternative to a Heathrow slot disposal. Such joint operation would have been likely to have created links between Ryanair and IAG(BA) which would not have been acceptable in the context of a remedy proposal which was meant to resolve an identified competition issue by creating a competitive constraint on the merged entity not by creating further links between the merged entity and its competitors. A commitment which would have maintained a structural long term relationship between Ryanair and IAG, such as a joint operation or a wet lease agreement, could not have been acceptable on competition grounds (1573).

(1752) Furthermore, the Commission also could not exclude that such "joint operation" or "wet-lease solution" would also have fallen under the provisions of Article 10 of Aer Lingus' Articles of Association. Indeed, as explained in detail in Section 10.5.1, the concept of "disposal transaction" is a broad one and in any event includes all measures "to otherwise dispose of or encumber" Aer Lingus' slots. In this scenario, the uncertainties as to the workability of such commitments are described in detail in Section 10.5.1 below (1574).

(1753) Lastly, the Commission notes that it could not rely on a commitment by Ryanair to resolve the restrictions which arose as a result of Aer Lingus' Articles of Association and to enable a sale of the Heathrow slots by using "best endeavours". Such commitment is too vague and uncertain and does not satisfy the Commission of its workability. As the issue here relates to the workability and enforceability on a lasting basis of the Commitments as regards the three routes to London Heathrow, the Commission was not satisfied that the Commitments of 15 January 2013 would have ensured that the significant impediment to effective competition (which the commitments were intended to prevent), would not have been likely to materialize (1575). There was a significant risk that the implementation of the Commitments of 15 January 2013 would have been delayed (1576).

(1754) Therefore, the Commission considers that the solution proposed by Ryanair to solve the issue of its (lack of) ability to dispose of the Heathrow slots is not sufficiently clear. Such solution appears hence unsuitable to resolve the Commission's concerns.

(1755) As regards the commitment regarding Aer Arann's franchise agreement, the concerns identified with respect to the Commitments of 7 December 2012 were still not alleviated. The Commission could not be satisfied to the requisite degree of certainty that Ryanair would be in a position post-Transaction to implement the commitment and modify the scope of the franchise agreement in a timely manner (1577).

(1756) Based on all the available evidence, including the results of the second market test, the Commission considered that the Commitments of 15 January 2013 would not guarantee that Flybe would be able to exert a sufficient competitive constraint on the merged entity. In addition, the Commission could not conclude, with the requisite degree of certainty that it would be possible to implement the Commitments of 15 January 2013 in particular as regards the slot divestiture to the benefit of IAG at Heathrow (even if implementable, the contemplated slot divestiture would not have been sufficient to avoid the creation of a dominant position of the merged entity on the Dublin-London route through the Transaction) and the change in the scope of Aer Arann's franchise agreement. The Commitments of 15 January 2013 appeared therefore not to be workable and lasting so as to ensure that the impediments of effective competition which those Commitments were intended to remove would not likely materialise in the relatively near future.

(1757) Lastly, no binding agreement with IAG and Flybe was sent to the Commission in relation to the Commitments of 15 January 2013, despite their fix-it-first nature.

(1758) The Commission considered that it was not possible to determine with the requisite degree of certainty that the Commitments of 15 January 2013 would and could have been fully implemented. It was therefore concluded that those Commitments were not able to remove the identified significant impediment to effective competition, and, thus, could not render the proposed concentration compatible with the internal market.

10.3. The final Commitments (1578)

10.3.1. General description by Ryanair

(1759) According to Ryanair (1579), the final Commitments eliminate all competition concerns identified in the Statement of Objections. This revised proposal retains the core elements of the Commitments of 15 January 2013, including the unprecedented involvement of two UFBs, Flybe and IAG/BA, and the divestiture of aircraft, slots, and other assets required to establish competing services on all 46 overlap routes identified by the Commission.

(1760) Ryanair is of the view that the final Commitments are qualitatively and quantitatively superior to the remedies accepted in any prior airline merger conditionally approved by the Commission over the past 20 years, which have involved offers of slot divestitures to unidentified potential entrants that in many instances have never materialised. Unlike in its previous proposals, which featured Heads of Terms negotiated with UFBs, Ryanair has now entered into binding agreements with both IAG (the "IAG Agreement") and Flybe (the "Flybe Agreement", the "Divestiture Agreements"). The Divestiture Agreements lead to clear-cut structural solutions that comprehensively remedy all overlaps.

(1761) Furthermore, Ryanair considers that the business sold to Flybe represents about onethird of Aer Lingus' shorthaul capacity (1580). It will carry some 4 million passengers per annum (close to half of Aer Lingus' short haul traffic), and is expected to generate annual revenues of approximately [...]*, with an annual pre-tax profit of [...]*, based on current performance of Aer Lingus' short-haul operations. Together with Flybe Ireland, Flybe will be a larger airline post Transaction than Aer Lingus is today.

10.3.2. Divestiture of the business regarding 43 routes to Flybe

(1762) Ryanair commits that it will divest to Flybe a stand-alone business comprising virtually all of Aer Lingus' present operations on 43 overlap routes (the "Flybe Routes").

(1763) Ryanair undertakes to transfer to Flybe a new and stand-alone company (the Divestment Business - possibly to be called Flybe Ireland) to which it will transfer a number of assets (as listed in Annex II to the final Commitments). The assets will include inter alia cash of EUR 100 million (1581), a lease for at least 9 Airbus A320 aircraft, airport slots (at origin and destination), a royalty-free, non-sub-licensable, non-exclusive and non-transferable license to the Aer Lingus trademark for 3 years, personnel, contracts, real estate etc. The transfer of these assets to the Divestment Business is referred to as the "Hive-Down".

(1764) For determining the assets and liabilities of the Divestment Business, Ryanair will, in consultation with Flybe, develop a business plan for the Divestment Business that will project a certain amount of annual pre-tax profits agreed upon between Ryanair and Flybe in the first year. Ryanair will ensure that the business plan is appropriate to meet the profitability target by identifying and transferring sufficient assets to the Divestment Business. If Ryanair is not able to structure the Divestment Business in a way that is shown by the business plan to meet this profitability target, Ryanair shall provide Flybe with additional cash for the Divestment Business in order to ensure the viability of the latter. The profit projection is however not a guarantee of profitability. Ryanair will, during the post-closing process, identify the assets to be included in the business and begin steps that may be required to restructure the business to ensure that it will reach the profitability targets in the business plan. In the event that Ryanair is not able to complete any restructuring required to justify a profit projection of EUR 20 million prior to the completion of the Hive-Down (1582), Ryanair will provide additional cash to assist Flybe in completing the restructuring process.

(1765) Flybe shall operate on the 43 Flybe Routes the schedule set out in Annex V of the final Commitments (the "Flybe Routes Schedule") during 6 IATA seasons (the "Minimum Period"), subject to its right to 'churn' routes. Flybe will acquire the slots on the 43 routes and may use them to operate the frequencies. 6 of the 43 Flybe Routes are currently operated by Aer Arann. With respect to these routes, Flybe shall be able to satisfy its obligation to operate on these routes by assuming the Aer Arann franchise agreement or by entering into a new franchise agreement with Aer Arann.

(1766) Regarding the churn provision, provided that the aggregate weekly frequencies per season scheduled by the Divestment Business on all routes originating and/or terminating in Ireland is at least equal to the aggregate weekly frequencies per season as set out in Annex V of the final Commitments, Flybe is only required to schedule flights on 90% of the Flybe Routes in the third and fourth IATA Season of the Minimum Period and 85% of the Flybe Routes in the fifth and sixth IATA Season of the Minimum Period.

(1767) Flybe will base the aircraft necessary to operate the Flybe routes schedule at Dublin Airport, provided that at least one aircraft is based operationally at Cork airport.

(1768) The Divestment Business will include leases for 9 Airbus A320 aircraft. Flybe will commit in addition to lease 5 Embraer 195 jet aircraft for operating the Flybe Routes Schedule (1583). Flybe will have the right to replace the Airbus A320's leased to the Divestment business with its own Embraer 195 jet aircraft. Where Flybe operates aircraft other than Airbus A320 aircraft, the number of weekly frequencies that Flybe is required to schedule shall be increased so that the total scheduled weekly seat capacity at least matches that which would prevail if Airbus A320 aircraft were used (1584).

(1769) To provide Flybe with additional publicity and brand awareness, the following measures are proposed in the final Commitments: - Trade Mark Licence: Flybe will be given a royalty-free, non-exclusive, non-sub-licenceable and non-transferable right to use the "Aer Lingus" trademark for the purposes of operating the Divestment Business for a period of three years. Other than during a short initial transitional period of 30 days, the Aer Lingus trademark can be used only in conjunction with Flybe's own brand name "Flybe", for example "Aer Lingus by Flybe" or "Flybe Aer Lingus". - Publicity for the Divestment Business on the Aer Lingus website: during the first 3 IATA seasons, the welcome screen of Aer Lingus' website will be divided vertically into two halves bearing the trade names and logos of Aer Lingus and the Divestment Business respectively. The welcome screen will be hyperlinked so that clicking on the Divestment Business' side of the screen directs the user to the Divestment Business' booking website and clicking on Aer Lingus' side of the screen directs the user to the booking website of the Aer Lingus group. After the first 3 IATA season, a banner will be put at the top of the Aer Lingus website stating "Visit the [Flybe Aer Lingus OR Aer Lingus by Flybe OR similar] for more routes and fares" and providing a web link to the Divestment Business' website.

(1770) In case of misuse of the slots, if Flybe operates less than 85% of the frequencies on any of the Flybe Routes (other than those for which it has used its 'churn right'), Flybe shall pay, without prejudice to any remedy available to Ryanair, a certain amount into a blocked account. The magnitude of the penalty will be determined in function of the revenues achieved on the route (during the previous two complete IATA seasons) and will vary between the IATA seasons (1585). Such amounts would be used, as decided by the Monitoring Trustee, to encourage competition and/or new entry on the 43 routes. In addition, if Flybe fails to operate 60% of the frequencies on any of the Flybe Routes for 2 consecutive IATA seasons, Ryanair would use its reasonable best efforts to identify another competitor that will enter the route and fly for at least 2 IATA Seasons to at least compensate for the shortfall in frequencies operated by Flybe and the Monitoring Trustee may use penalty payments to facilitate such new entry.

10.3.3. Slot divestitures to IAG/BA

(1771) During 6 IATA seasons, Ryanair undertakes that IAG/BA would operate on the Dublin-London route, the Cork-London and the Shannon-London route, using Airbus A 319 or equivalent aircraft, using its own slots in combination with those divested by Ryanair as explained below.

(1772) IAG and Ryanair may enter into a Gatwick Lease Agreement or a Heathrow - Gatwick Transfer Agreement, depending on whether and when the Heathrow Transfer Condition is satisfied. The Heathrow Transfer Condition consists in the determination by both Ryanair and IAG that giving effect to the Heathrow - Gatwick Transfer Agreement would not violate Article 10 of the Aer Lingus Articles of Association (or other successor provision) or any other applicable law or regulation.

(1773) If the Gatwick Lease Agreement is in effect, IAG would operate on the following routes and frequencies: - 68 weekly frequencies on the Dublin-London Gatwick route, 13 weekly frequencies on the Cork-London Gatwick route and 7 weekly frequencies on the Shannon-London Gatwick route using airport slot pairs owned by Ryanair and - an additional 2 weekly frequencies on the Dublin-Gatwick Route, 1 weekly frequency on the Cork-Gatwick Route and 7 weekly frequencies on the Shannon-Gatwick Route using slots currently owned or to be acquired by IAG.

(1774) If the Gatwick Lease Agreement is in effect and the Heathrow Slot Condition is satisfied, IAG may exercise its right to terminate the Gatwick Lease Agreement and give effect to the Heathrow - Gatwick Transfer Agreement (the IAG Call Option).

(1775) Furthermore, according to the final Commitments, if the Gatwick Lease Agreement is in effect, IAG would also operate 7 weekly frequencies on the Dublin-Heathrow route using slots that IAG would lease from Ryanair once the existing Aer Lingus short-term lease over these slots with a third party (the identity of which is currently unknown to Ryanair) expires. However a revised IAG Agreement (1586) provides that IAG and Ryanair would enter into a "First Heathrow Lease Agreement" for 7 Heathrow slots and IAG should have the option to lease further Heathrow slots under a "Second Heathrow Lease Agreement" which are not restricted in their use (1587).

(1776) Under the final Commitments, if the Heathrow - Gatwick Transfer Agreement is in effect, IAG would transfer its Gatwick slots (as identified above) and 7 additional Gatwick slots to operate on Dublin - Gatwick in exchange for slots currently owned by Aer Lingus at London Heathrow slots for IAG to operate: - 70 weekly frequencies on the Dublin Heathrow route, - 14 weekly frequencies on the Cork - Heathrow route, and - 14 weekly frequencies on the Shannon-Heathrow route. The revised IAG Agreement no longer includes the obligation to give the 7 additional Gatwick slots to Ryanair (1588).

(1777) The IAG Agreement also provides that Ryanair shall pay a so called "Break Fee" to IAG of [...]* if by the final day of the Minimum Period the Heathrow Transfer Condition is not satisfied and within [...]* IAG serves a written notice on Ryanair requiring such payment. On payment of the Break Fee the Gatwick Lease Agreement, and if it is in effect, the First Heathrow Lease Agreement and the Second Heathrow Lease Agreement shall terminate and IAG will transfer the Ryanair Gatwick Slots back to Ryanair. However, if IAG does not serve the notice within [...]*, the Gatwick Lease Agreement and the First Heathrow Lease Agreement and Second Heathrow Lease Agreement will continue in effect for a total term of no more than [...]* from the first IATA changeover date (Paragraph 5.4 of the IAG Agreement). If such termination occurs beyond the [...]*, the Ryanair Gatwick Slots will be transferred to British Airways.

(1778) The final Commitments provide for a mechanism in case of misuse by IAG of the IAG Routes Slots.

10.3.4. Other aspects

10.3.4.1. Frequent Flyer programme

(1779) At the request of IAG, Flybe, or Aer Arann, Ryanair shall allow such party to be hosted in Aer Lingus' Frequent Flyer Programme for any of the routes operated by Flybe, IAG or Aer Arann.

10.3.4.2. Potential competition

(1780) During the first 6 IATA seasons and at all times thereafter, Ryanair commits to transfer, or to cause Aer Lingus to transfer, slots for the Potential Competition Routes to any EU-licensed airline which so applies (an "Applicant"), to enable the relevant Applicant to operate frequencies with its own aircraft on any of the Potential Competition Routes, up to a maximum number of Frequencies as identified on a route-by-route basis in Annex VII. Potential Competition Routes are listed in Annex VII as being Dublin-Bologna, Dublin-Bordeaux, Cork-Birmingham, Cork-Munich, Cork-Paris and Dublin-Newcastle.

10.3.4.3. London-Ireland routes

(1781) Pursuant to a slot transfer procedure, and during the first 6 IATA seasons, Ryanair commits to transfer, or to cause Aer Lingus to transfer, to any Applicant, sufficient slots to operate frequencies with its own aircraft on the Dublin-London, Cork- London and/or Shannon-London routes, provided that the number of slots transferred does not exceed the route overlap difference (the lower of the Aer Lingus or Ryanair frequencies) on the relevant route.

10.3.4.4. Monitoring Trustee and Dispute Resolution

(1782) The final Commitments also set up a mechanism for a Monitoring Trustee and provide for a dispute resolution mechanism (which will be addressed in Section 10.4.3 and 10.5.4 below).

10.4. Assessment of the divestiture to Flybe

10.4.1. Conceptual framework

(1783) The Commission notes that, despite their submission so late in the investigation, it decided exceptionally to examine the final Commitments notwithstanding that they did not clearly resolve the competition concerns previously identified and therefore required a further market test (1589).

(1784) Under paragraph 23 of the notice on remedies, the divested activities must consist of a viable business that, if operated by a suitable purchaser, can compete effectively with the merged entity on a lasting basis. Furthermore, according to paragraph 24 of the notice on remedies, in proposing a viable business for divestiture, it is necessary to take into account the uncertainties and risks related to the transfer of a business to a new owner as these risks may limit the competitive impact of the divested business and may lead to a market situation where the competition concerns would not necessarily be eliminated.

(1785) Under paragraphs 25 and 26 of the notice on remedies, the business has to include all the assets which contribute to its current operation or which are necessary to ensure its viability and competitiveness and all currently employed personnel or at least the personnel necessary to ensure the business' viability and competitiveness. To this extent personnel and assets which are currently shared between the business to be divested and other businesses of the parties, but which contribute to the operation of the business or which are necessary to ensure its viability and competitiveness, also have to be included.

(1786) Therefore, the divested business has to contain the personnel providing essential functions for the business such as, for instance, group R&D and information technology staff even where such personnel are currently employed by another business unit of the parties - at least in a sufficient proportion to meet the on-going needs of the divested business. In the same way shared assets have to be included even if those assets are owned by or allocated to another business unit.

(1787) A viable business is a business that can operate on a stand-alone basis, independently of the merging parties as regards the supply of input materials or other forms of cooperation other than during a transitory period (paragraph 32 of the notice on remedies). In case the Commitments require a carve-out of a business, the Commission will only be able to accept such commitment if it can be certain that, at least at the time when the business is transferred to the purchaser, a viable business on a stand-alone basis will be divested and the risks for the viability and competitiveness caused by the carve-out will thereby be reduced to a minimum (paragraphs 35-36 of the notice on remedies). Also in case the divestiture consists of a combination of certain assets which did not form a uniform and viable business in the past, the resulting business should be viable and competitive (paragraph 37 of the notice on remedies).

(1788) Under paragraph 30 of the notice on remedies, the business to be divested has to be viable as such. However, if during the procedure a sales and purchase agreement with a specific purchaser is already concluded, its resources can be taken into account in the assessment of the viability of the divested business.

(1789) As set out in paragraph 47 of the notice on remedies, the intended effect of a divestiture will only be achieved if and once the business is transferred to a suitable purchaser in whose hands it will become an active competitive force in the market.

(1790) Paragraph 48 of the notice on remedies sets out the standard purchaser requirements, which may have to be supplemented on a case-by-case basis. The standard purchaser requirements are the following: (1590) - the purchaser is independent of and unconnected to the parties - the purchaser must possess the financial resources, proven relevant expertise and have the incentive and ability to maintain and develop the divested business as a viable and active competitive force in competition with the parties and other competitors, and - the acquisition of the business by a proposed purchaser must neither be likely to create new competition problems nor give rise to a risk that the implementation of the commitments will be delayed. Therefore, the proposed purchaser must reasonably be expected to obtain all necessary approvals from the relevant regulatory authorities for the acquisition of the business to be divested.

(1791) Considering that the final Commitments are presented as a fix-it-first type of remedy and the identity of the buyer is known, the Commission will assess the viability of the Divestment Business together with the suitability of the purchaser in Section 10.4.2 below.

10.4.2. Viability of the Divestment Business and suitability of Flybe as a buyer

(1792) Ryanair argues that the Divestment Business (for ease of reference "Flybe Ireland") will have all the assets required for it to be a strong, credible, and independent carrier including aircraft, sufficient capital, slots, personnel, contracts to operate on the routes, real estate and a license pursuant to which Flybe Ireland obtains a royalty- free, nonexclusive and non-transferable right to use the "Aer Lingus" trademark for a term of three years following closing (1591). Ryanair considers that Flybe has all capabilities necessary to compete with the merged entity on ex-Ireland routes (1592).

(1793) The Commission will first examine in Section 10.4.2.1 whether the Divestment Business will have the necessary assets in order to be a viable and competitive business. The Commission will then examine the suitability of Flybe as a purchaser in the light of paragraph 48 of the notice on remedies. Section 10.4.2.2 will describe Flybe's business model, Sections 10.4.2.3, 10.4.2.4 and 10.4.2.5 will discuss respectively Flybe's independence from the Parties, its past experience (in starting/taking over significant operations, in the Irish market as well as in competing with Ryanair), and its financial resources. The Commission will then assess the ability of Flybe to maintain and develop Flybe Ireland as a viable and active competitive force in Section 10.4.2.6 and its incentives to do so in Section 10.4.2.7.

10.4.2.1. Assets and liabilities of the Divestment Business

(1794) The responses to the final market test showed that respondents were largely positive about whether the list of assets to be transferred to Flybe Ireland (enumerated in Annex II of the final Commitments) would be sufficient for Flybe to take over Aer Lingus' operations on the 43 Flybe Routes and to continue these operations in a viable manner for at least 6 IATA Seasons (1593). Some respondents however raised concerns (1594).

(1795) The Commission notes that the business that will be transferred to Flybe only comprises short-haul activities while within Aer Lingus these activities are part of a global network, encompassing also long-haul activities. Ryanair argues that Aer Lingus's long-haul connectivity and code-share agreements are immaterial to the 43 routes in this respect (1595). Flybe however has admitted that it is "sceptical about the real profitability of Aer Lingus in the short haul business and it would need to see Aer Lingus' figures first, although it believes that there is capacity to reduce Aer Lingus's costs" (1596). The profitability of Flybe Ireland as a stand-alone business is examined below, in particular in Section 10.4.2.6.

(1796) The Commission further notes that Flybe Ireland business will not include the divestment of the Aer Lingus trademark, but a 3-year non-exclusive licence for Flybe to use the trademark in combination with its own trademark. The issue of brand awareness is further assessed in Section 10.4.2.6.

(1797) The Commission finally notes that there is no exact definition of the assets and liabilities that will be transferred to Flybe Ireland, other than the rather general list which is set out in Annex II to the final Commitments and the limited information given in the Form RM (1597). For example, Ryanair is very vague about the type and number of staff to be transferred and simply refers to "Personnel based in the Republic of Ireland (pilots, cabin crews and engineers) necessary to operate the Flybe Routes as currently operated by Aer Lingus" (1598). Ryanair is also vague about the contracts and real estate to be transferred and refers to "Contracts (receivables from forward ticket sales, airport and handling agreements, corporate account agreements) necessary to operate the Flybe routes as currently operated by Aer Lingus" (1599) and to "Real estate in the Republic of Ireland to necessary, taken to operate the Flybe routes as currently operated by Aer Lingus" (1600).

(1798) Ryanair indicates that the identification of the assets to be included in the Flybe Ireland business will be done during the post-closing process (1601).

(1799) Ryanair itself acknowledges that there are various uncertainties regarding the exact assets to be transferred. For example, it acknowledges that "Given the unsolicited nature of the Transaction, [it] has no knowledge of the existing corporate agreements that Aer Lingus has entered into" (1602) and that third party consents might be required to transfer relevant rights under contracts and that where such consent would not be obtained by the time of completion of the Hive-Down, it shall procure that the relevant member of Ryanair Group holds such rights for the benefit of Flybe Ireland (1603).

(1800) Under paragraph 34 of the notice on remedies, "[...] In cases involving a hostile bid, a commitment to divest activities of the target company may, in such circumstances of limited information available to the notifying parties about the business to be divested, increase the risk that this business might not, after a divestiture, result in a viable competitor which could effectively compete in the market on a lasting basis [...]"

(1801) In the absence of specifically defined assets to be divested as part of the Flybe Ireland business, the Commission finds that there is insufficient evidence that would allow the Commission to conclude that the Divestment Business would possess the necessary assets in order to be a viable and competitive business.

10.4.2.2. Flybe's business model

(1802) Flybe is a large regional airline group based in Exeter, Devon, England, operating scheduled, short-haul services on a point-to-point and hub basis. The airline employs around 3400 people, flies on over 200 routes and manages a fleet of around 100 aircraft (regional jets and turboprops).

(1803) Flybe operates a business model which combines low-frills features (such as low-fare options, high levels of internet sales, on-board sales including purchased food and beverages) with full-service characteristics (such as pre-assigned seating, business schedules, IATA membership and interlining).

(1804) In the United Kingdom, Flybe's largest bases are Birmingham airport, Belfast City airport, Southampton airport, and Manchester airport, with a total of 14 crew and aircraft bases across the United Kingdom.

(1805) In the financial year ending 31 March 2012, Flybe achieved GBP 678.8 million revenue under management1604 (up 14% on 2011). It transported 7.6 million passengers under management (up 6% on 2011) (1605).

(1806) As a result of the final Commitments, Flybe would take over several hundred employees (1606) and add 43 routes to its portfolio, to be operated with 9 Airbuses (the lease of which will be part of the Divestment Business) and 5 regional jets (for which Flybe undertook to procure that these would be leased by the Divestment Business (1607)). This is expected to raise Flybe's revenues by about EUR 300 million.

(1807) The final Commitments thus imply a significant increase in the scope of Flybe's business.

(1808) According to Ryanair, the business model of Flybe is well suited to compete aggressively with the merged entity for short-haul passengers (1608).

(1809) The responses to the final market test show mixed views in this regard (1609). A number of respondents consider that the operations on the routes match with Flybe's business model. Such respondents indicate for example that Flybe is an experienced regional airline (1610), already having a pan-European operation (1611) and whose model has been exported successfully to Nordic countries (1612). They also indicate that Flybe has a strong service offering and can target both business and leisure passengers (1613). A majority of respondents however consider that the operations on the 43 routes or at least part thereof would not match Flybe's business model. Respondents express doubts in particular with respect to routes with longer sector length such as routes to the Canaries (1614), while the provincial routes in the United Kingdom would fit in Flybe's portfolio depending on the type of aircraft used. Also, while Flybe specialises in targeting high yielding business traffic (1615), the 43 routes would include more nontime sensitive routes compared to the current Flybe operations. Several respondents also refer to the fact that operations with A320 aircraft would be new for Flybe (1616) and that Flybe would struggle to develop a connecting model to rival what Aer Lingus has today (1617).

(1810) When asked whether the operations on the 43 routes would match with its business model, Flybe distinguishes three types of routes, namely Business/Leisure, Mid Distance Leisure and Distant Leisure. According to Flybe, the Business Leisure routes (representing 75% of the divested frequencies) fit its core business model extremely well. The Mid Distance Leisure routes (representing 20% of the divested frequencies) are an important part of Flybe's network, but not currently at its core. The Distant Leisure routes (representing 5% of the divested frequencies) are outside of Flybe's existing business model, essentially as they are beyond the economic range of its aircraft (1618).

(1811) The Commission notes that Flybe has no experience in operating Airbus A320 aircraft. It has limited experience in operating with Boeing 737 aircraft, which are somewhat comparable in size to the Airbus A320 aircraft. However, as acknowledged by Jim French, Flybe's Chairman and Chief Executive Officer, that was only during an 18-months project that took place 8 years ago in a totally different context. Also Mr French indicated that " [...] an Airbus has got two wings, a couple of engines, two people sit upfront, a lot people are behind and it runs on fuel; an 80-seat jet does the same. The complexity is not running the jet, the aircraft, or the size of the aircraft - the complexity is understanding the market they operate in." (1619) (emphasis added).

(1812) While Ryanair undertakes to transfer personnel (1620) to the extent necessary to Flybe Ireland, due to the hostile nature of the Transaction, Ryanair is unable to provide the identities of such personnel (1621). Uncertainties exist whether Ryanair would be in a position to deliver upon this commitment and whether sufficient personnel with the required knowledge to manage a business operating with Airbus A320 aircraft would indeed be transferred to Flybe Ireland especially considering Aer Lingus' management resistance to Ryanair's bid. It is thus not unlikely that Flybe would face challenges in terms of operations and schedules at least in an initial period and would need to go through a learning process as regards the Airbus A320 aircraft operations.

(1813) In light of the above, the Commission considers that the operations of routes connecting Ireland to the United Kingdom would fit in Flybe's current business model, depending on the aircraft used. Other routes, especially leisure routes with longer sector lengths, would not fit in Flybe's current business model. Moreover, the Commission notes that Flybe, due to its business model, does not have experience with operating with Airbus A320 aircraft and therefore considers that it would thus face a challenge in understanding the market in which it would operate.

10.4.2.3. Flybe's independence from the Parties

(1814) Ryanair argues that Flybe is independent of, and unconnected to, both Ryanair and Aer Lingus. It indicates that there are no cross-shareholdings or cross-directorships between Flybe on the one hand, and Ryanair and Aer Lingus on the other. There are also no other commercial relationships - for example, alliances, common frequent flyer programmes, shared facilities at any airports, etc. - between Flybe, Ryanair and Aer Lingus (1622).

(1815) The Commission notes that Flybe and Ryanair have agreed that Ryanair would prepare a one year business plan for Flybe Ireland (its supposed competitor) in which it estimates projected revenues, costs and profits (in consultation with Flybe) (1623). In this respect, it is also noted that in the event that the one year business plan to be prepared by Ryanair for Flybe Ireland did not project a yearly EUR 20 million profit under the stipulated presumptions, then Ryanair would provide additional cash to Flybe Ireland to compensate for a reduced-profitability forecast (1624).

(1816) In its Form RM, Ryanair states its commitment to "ensuring that Flybe Ireland's cost base is structured in such a way to enable it to achieve a profit of at least €20 million in its first year of operation [...]" (1625). Pursuant to paragraph 10 of the final Commitments, "during the post-Closing process, Ryanair will identify the assets to be included in the Divestment Business and begin steps that may be required to restructure the business to ensure that it would reach the profitability targets in the business plan" (1626).

(1817) Ryanair indicates that it intends to structure a cost base for Flybe Ireland that is driven by its own cost management skills, together with input from analysis of carriers such as easyJet and Flybe. Ryanair is confident that it can provide a cost base that will provide Flybe Ireland with margins at least as good as Aer Lingus' margins (1627).

(1818) Flybe expects Ryanair to deliver to Flybe Ireland a cost base that should allow the latter to achieve EUR 20 million in pre-tax profits. It indicates in this respect that Ryanair has to undertake the transition and reorganisation of the Divestment Business to bring it down to the shape, size and cost structure before Flybe takes over Flybe Ireland (1628). Flybe itself does not propose any concrete actions as to how the cost base should or could be reduced, but leaves it largely up to Ryanair to achieve a lower cost base.

(1819) The Commission acknowledges that the arrangement whereby Ryanair prepares a one-year business plan and could proceed to the (re-)structuring of the cost base of the Divestment Business does not lead to a lasting relationship between the merged entity and the Divestment Business.

(1820) However, the Commission considers that the viability of the Divestment Business to be transferred to Flybe would largely depend on the ability and incentive of Ryanair to provide a profitable business. Uncertainties about Ryanair's ability to reduce sufficiently the cost base are discussed below in Section 10.4.2.6.

(1821) The Commission therefore concludes that the arrangement whereby Ryanair prepares a one year business plan for Flybe Ireland, and the arrangement whereby Ryanair could proceed to the (re-)structuring of the cost base of the Divestment Business, even though not leading to a lasting relationship between the merged entity and the Divestment Business, does not seem reconcilable with the concept of independent competitors. Even if strictly speaking the business plan is not binding upon Flybe, it is clear that Flybe would rely to a large extent on Ryanair to undertake the transition and reorganisation of the Divestment Business to bring it down to the shape, size and cost structure. Furthermore, as explained in Section 10.4.3.2, there are no appropriate ring-fencing measures in place.

10.4.2.4. Flybe's experience

- Experience in starting up/taking over operations

(1822) Ryanair and Flybe argue that Flybe is well placed to take over the Divestment Business inter alia due to its past experience in starting up or taking over operations.

(1823) Ryanair indicates that Flybe has significant experience in the airline industry. According to Ryanair, after the acquisition of the Divestment Business, Flybe will be larger than Aer Lingus is today in terms of total aircraft, routes served, and passengers flown (1629). Ryanair further states that Flybe is active across a number of markets, offering "low-frills" options in addition to catering for business passengers and operating across the spectrum of Aer Lingus' and Ryanair's business models (1630). Ryanair argues that Flybe, given its past experience, is well accustomed to establishing regional bases, developing brand awareness in new markets and operating new routes profitably (1631). Ryanair refers to Flybe's success with respect to Flybe Finland and the fact that it has grown 14% in 2011, despite weak economic conditions throughout Europe (1632).

(1824) Flybe acquired a stake in Flybe Finland (Flybe Finland is a subsidiary of Flybe Nordic, a joint venture with Finnair) in August 2011 as a loss making business. Flybe argues that the Flybe Finland business has been restructured and refocused, and although Flybe Finland is still loss making, Flybe plans substantial growth and performance. It further indicates that the business is on track to meet its original planned financial performance and that it expects 15% return on investment and forecasts profitable operations in the next financial year commencing March 2013. (1633)

(1825) The Commission notes that Flybe's joint venture with Finnair remains loss making and that Flybe has admitted that its expansion beyond the United Kingdom market has caused operational losses: as set out in Flybe's annual report "The key driver of the overall Group loss arises from the expansion into Europe with Flybe Europe generating a £3.7m loss, being its share of loss from joint venture operations in Finland together with related management costs. (1634)"

(1826) The Commission further notes that the majority of the routes operated by Flybe Finland are contract flying (namely 38 routes out of a total of 58 routes). On the contract flying routes, Flybe Finland operates (under a form of long term wet lease agreement similar to the regional carrier model in the United States) on behalf of Finnair; the commercial risk is borne by Finnair who sells the tickets. Flybe has no codeshare agreements on these contract flights. (1635) Compared to contract flying which is profitable as Flybe does not bear the commercial risk of these operations (1636), Flybe Finland makes losses on the routes which it operates itself and is expected to continue to do so in the financial year 2013/14. The own commercial operations of Flybe Finland have thus not yet proven to be profitable. Flybe acknowledges this and explains that the losses generated by routes that are its own commercial operations arise in particular due to the Finish market not being very responsive to stimulation (low prices) (1637). Moreover, the Commission notes that Flybe Finland is a subsidiary of Flybe Nordic, a joint venture with Finnair. This is a very different situation to the divestiture proposed in the context of the final Commitments, which would be the divestment of a stand-alone business, which would be operated exclusively by the Flybe Group.

(1827) In addition, Flybe acquired British Airways' regional business in the United Kingdom "BA Connect" in March 2007. BA Connect essentially operated regional flights within the United Kingdom and routes to continental Europe, primarily business routes. BA Connect comprised 39 aircraft, 1,700 staff and GBP 350 million revenue (in terms of revenues the BA Connect services were approximately equal to Flybe's revenues in 2007, prior to the restructuring of the business post acquisition). Flybe indicates that it received GBP 109 million cash from BA to fund the integration into Flybe, and that it fully integrated BA Connect within 12 months (with restructuring costs amounting to around GBP 96 million). Flybe indicates that while the BA Connect business generated GBP 40 million per annum losses at acquisition, the business made substantial profits at the end of the first 12 months' ownership (1638).

(1828) An independent analyst considered that "(...) the format of the deal [the proposed Commitments] appears to be a reminiscent of Flybe's acquisition of BA Connect operations a few years ago. Namely the business comes with a dowry to fund the necessary restructuring costs in order to improve profitability". (1639)

(1829) The Commission acknowledges that there may be a number of similarities between the acquisition of BA Connect and the acquisition of Flybe Ireland; for instance both transactions would involve a dowry payment (1640). The Commission nevertheless notes that the acquisition of BA Connect took place about 6 years ago, in a year which was described in Flybe's own words as "a year in which we became one of Europe's largest and most profitable regional airlines" (1641). Such a context is hardly comparable with today's context, where Flybe has just announced a major restructuring programme, and the current overall economic climate, in both the United Kingdom and Ireland is much more difficult compared to 2007. The network acquired in the context of the BA Connect acquisition also seemed much closer to Flybe's then existing business (focused on regional passengers in the United Kingdom), than the Divestment Business would today be for Flybe (1642). Finally, while in the BA Connect acquisition, Flybe implemented the restructuring itself, here Flybe relies on the business plan (which in turn relies on certain presumptions) for at least initial reassurance that the cost base of Flybe Ireland will be sufficiently low to justify the profit projection of EUR 20 million, and given the context of a hostile bid, Flybe only has very limited information on and knowledge of the business that it proposes to take over. Also the scope for cost efficiencies is not the same in the two cases, considering in particular the significant cost cutting measures Aer Lingus has already implemented since 2010.

(1830) Therefore, the Commission concludes that these previous experiences are rather different from the current operation and do not provide sufficient relevant evidence as to Flybe's ability to profitably restructure and operate Flybe Ireland as a sufficient competitive force against the merged entity in the medium term.

- Experience in the Irish market

(1831) Flybe only has a limited past experience in operating routes to and from Ireland. Currently, Flybe is present with its own operations on only two routes ex-Dublin (1643). On both routes Flybe is not serving the same airport on the non-Irish side of the route as Ryanair/Aer Lingus.

(1832) Furthermore, Flybe itself stated that it "does not have a very strong brand in Ireland, as it only has a presence on two routes ex-Dublin, albeit for many years" (1644). The brand issue is further discussed below in Section 10.4.2.6.

(1833) Flybe thus has only very limited experience in the Irish market which does not provide sufficient evidence as to Flybe's ability to profitably restructure/operate Flybe Ireland as a sufficient competitor force against the merged entity in the medium term.

- Experience in competing with Ryanair (1645)

(1834) To demonstrate that it would be able to compete with Ryanair post-Transaction, Flybe refers to the example of competition on routes from Belfast City. According to Flybe, Ryanair came into Belfast City for a couple of years, competed with Flybe on a number of routes, and then left the airport. It further indicates that it competed with Ryanair on some of the routes into Dublin from the South Coast of the United Kingdom (1646).

(1835) Flybe further provides a list of routes on which it currently competes with Ryanair. (1647) According to the information received from Flybe, on about half of these routes Flybe's market share increased between Winter 2009/10 and Winter 2011/2012 while Ryanair's market share decreased. On about the other half of these routes Ryanair's market share increased while Flybe's market share decreased. It thus appears that Flybe manages to successfully withstand competition from Ryanair at least on some of these routes. In addition, Flybe operates on one of the 43 overlap routes subject to the Decision, namely Dublin - Bristol/Cardiff/Exeter. On this route, Flybe has a very limited market share, below 10%.

(1836) Unlike Aer Lingus which faces competition from Ryanair on a significant portion of the routes it operates, Flybe competes with Ryanair on only a limited number of the routes it operates.

(1837) The Commission notes moreover that on most of these routes, Flybe and Ryanair do not operate from the same base, while Flybe Ireland would be operating on most of the routes in competition with the merged entity from the same base, namely in Dublin.

(1838) On balance, the Commission considers that Flybe, unlike Aer Lingus, has only limited experience and track record in competing with Ryanair.

(1839) Flybe acknowledges that it cannot match Ryanair on price given that Ryanair has a very low cost base. However, it believes that by driving down the cost base and having a price premium differential that is not too great, it can creep into the market against Ryanair (1648).

(1840) With respect to the Divestment Business, Flybe indicates that it would aim at bringing a different model to the market which would combine low cost through the operation of Airbus aircraft but also offering interesting frequencies (1649).

- Conclusion on Flybe's experience

(1841) Given that Flybe's previous experiences in operating in new markets are rather different from the proposed acquisition of Flybe Ireland, given that Flybe only has limited experience in the Irish market, and given that Flybe, unlike Aer Lingus, has only limited experience and track record in competing with Ryanair, the Commission considers that these elements do not provide sufficient evidence to support the conclusion that Flybe has the proven relevant experience to maintain and develop Flybe Ireland as a viable and active competitive force in competition with the merged entity. 10.4.2.5. Flybe's financial resources (1842) Flybe's1650 group revenues, profits (or losses) were the following in the financial years ending on 31 March (1651): in GBP million, including JV

<TABLE>

(1843) The increase in turnover in 2012 compared to 2011 is mainly explained by the revenues of the Joint Venture with Finnair, which represented GBP 63 million of out the total turnover of GBP 679 million.

(1844) After several profitable years, Flybe reported losses before tax for the past two financial years ending at 31 March 2011 and at 31 March 2012. Some independent analysts refer to Flybe's disappointing results (1652).

(1845) In the six months to 30 September 2012, Flybe incurred a loss before tax of GBP 1.3 million. According to Flybe, analysts' consensus is that Flybe will generate a loss before tax of between GBP 15 million and GBP 20 million for the year to 31 March 2013 (1653).

(1846) Moreover, Flybe's net assets at 30 September 2012 amounted to GBP 85.4 million, with total cash of GBP 59.1 million (1654).

(1847) Figure 5 below shows that Flybe's cash on balance sheet appears low (versus peers). This would suggest a limited ability to withstand shocks, which is important given the concerns on the viability of the Divestment Business as explained in this Section 10.4.2 and the risk of aggressive retaliation by Ryanair as explained in section 8.5.2.6. Figure 5: Flybe cash position versus its peers (1655)

<GRAPHIC>

Source: Company reports, Factset as at 2 January 2013.

(1848) In 2013, Flybe announced a major restructuring plan consisting of a two year roadmap to return the business to profitability (1656). Flybe explained that the restructuring is designed to make the business leaner but without compromising customer service which is one of the core principles of Flybe's business proposition (1657). The turnaround plan includes risk-reduction actions, and projects that breakeven per seat would not be restored before the year 2013/2014.

(1849) Flybe argues that, three years ago, Aer Lingus lost substantial sums and has had to radically reduce its costs and staff numbers, prior to returning to profitability (1658). Flybe is confident that it will be able to bring its existing business back to profitability in the coming 1 to 2 years (1659).

(1850) The Commission does not take a view on whether or not Flybe will bring its existing business back to profitability in the coming 1 to 2 years. The Commission nevertheless must examine whether Flybe possesses the necessary financial resources to maintain and develop the Divestment Business as a viable and active competitive force in competition with the merged entity on a lasting basis.

(1851) Flybe and Ryanair argue that the Divestment Business will be a well-capitalised business. They indicate that the Divestment Business will be capitalised in such a way as to compensate for potential trading losses and the investment required to strengthen the Flybe brand in Dublin and at a number of destinations (1660). They indicate that such a capital buffer should allow the Divestment Business to become a strong and viable competitor to the Merged Entity.

(1852) The Commission notes firstly that the fact that Flybe requires such significant capitalisation by Ryanair of the Divestment Business could be considered as an indicator that it does not itself have the required financial resources. This is corroborated by Flybe's above described financial position and recently announced restructuring.

(1853) At the same, time, the Commission notes that Aer Lingus has a significant level of cash reserves and that the amount of capital that will be injected into the Divestment Business by Ryanair would represent only a small proportion of the cash reserves that Aer Lingus currently has.

(1854) Secondly, the Commission notes that the (indicative) EUR 100 million cash is not restricted to the Divestment Business and could be used, for instance, for Flybe's main business in the United Kingdom. As set out above, the cash position of Flybe is rather weak compared to its peers in the industry, and Flybe is expected to continue to make losses on its own operations (in particular, breakeven per seat would not be restored before the year 2013/2014). The cash which would be injected into the Divestment Business would be available as from the day on which Flybe would take over the Divestment Business. Flybe's board has noted in this respect that the EUR 100 million cash would be unrestricted as to its use and the same fact was also acknowledged by Flybe during a conference call with investors, indicating that while the EUR 100 million is very much earmarked for Flybe Ireland because of its role as a revenue buffer contingency, the amount is not legally ring-fenced1661. In a communication to Flybe staff of 6 February 2013, Jim French, CEO of Flybe, also stated that "If this transaction is completed then Flybe Group plc will be in a much stronger position".

(1855) The capital which would be injected into the Divestment Business could thus represent a way for Flybe to have access to new funds without having to go to the shareholders or seek external financing in the market. It therefore cannot be excluded that at least part of the EUR 100 million that would be part of the Divestment Business would be transferred and used by Flybe for purposes other than restructuring and developing Flybe Ireland, especially if Flybe finds that the Divestment Business is not viable.

(1856) Even in the worst case scenario assessed by Flybe's Board, [...]* (1662).

(1857) Thirdly, Flybe would also not have the ability to provide further financing that may be necessary to withstand shocks (for instance in case of aggressive competition by Ryanair, during or after the Commitment period), or make additional investment to improve the likelihood of the Divestment Business being viable.

(1858) Lastly, while Ryanair claims that Aer Lingus is a "subscale peripheral carrier that is not consistently profitable and cannot grow or compete with the much larger carriers in Europe", Flybe appears as an airline that is in a financially more precarious position with limited ability to undertake new or risky activities without extra external financing, compared to Aer Lingus, who demonstrates profitability and strong cash reserves. One of the proposed rationales of Ryanair's bid is that Aer Lingus would not be in a position to compete with the "giant" carriers (1663). Following Ryanair's reasoning, this raises the question of how an entrant like Flybe which is not part of any of these giant groups and who is in a less favourable financial position than Aer Lingus would be a suitable purchaser of the Divestment Business.

(1859) In light of the above, the Commission concludes that Flybe does not possess the financial resources to maintain and develop Flybe Ireland as a viable and active competitive force in competition with the merged entity in the medium term.

10.4.2.6. Flybe's ability to compete with the merged entity post-Transaction

(1860) The market investigation showed mixed views with respect to the question whether, taking into account all proposed commitments, Flybe, through Flybe Ireland, would be able to constrain the merged entity so that the competition concerns identified by the Commission would be eliminated (1664).

- Brand recognition

(1861) In Section 8.5.2.4, the Commission analysed the entry barriers in terms of establishing a strong brand on the routes subject to this Decision. A number of competitors, including Flybe, have identified "brand recognition" as a significant barrier to entry on routes subject to this Decision, and marketing and promotion costs required to upgrade the brand would be significant. According to Flybe "it's important to promote the brand and new services when developing new routes (...)" Flybe also indicated that while both Ryanair and Aer Lingus enjoy a strong brand on the routes subject to this Decision, its own brand was considered as a "medium" brand by passengers on the routes subject to this Decision. The main reason for this was that it only has a presence on two routes ex-Dublin and that only moderate levels of marketing have been undertaken in Ireland. Flybe has estimated that it would take two years to upgrade its brand to a strong brand. At the same time, Flybe considered that Ryanair's and Aer Lingus' brand presence would not as such be a barrier for entry into operations ex-Dublin as it has a strong presence at the other end of the route, namely in the market in the United Kingdom.

(1862) Ryanair argues in the same manner that Flybe has strong brand recognition in the United Kingdom and throughout Europe, including in Ireland as it already operates a number of destinations from Ireland (1665).

(1863) The Commission acknowledges that a strong brand at destination may be of use when competing with the merged entity on certain routes to and from Ireland. For the reasons set out in Section 8.5.2.4 of this Decision, the Commission nevertheless considers that the benefits brought about by such brand recognition outside Ireland are limited compared to the advantages that Ryanair and Aer Lingus enjoy thanks to their strong brand in Ireland. Therefore, the Commission considers that it would be important for Flybe Ireland to have a strong brand in Ireland.

(1864) Moreover, while Flybe already operates to 15 of the 32 destinations outside Ireland, it only has operations with more than 100 weekly frequencies to 6 destinations (namely Birmingham, Manchester, Edinburgh, London Gatwick, Glasgow, and Paris-CDG). Post transaction only Dublin would be added to the list (on all other destinations, Flybe's maximum operations would not be more than 3 daily frequencies) (1666).

(1865) Ryanair argues that it offers commitments to address any of the Commission's concerns with respect to the Flybe brand. It will offer a licence to Flybe Ireland pursuant to which it obtains a royalty-free, non-exclusive, non-sublicenceable and non-transferable right to use the Aer Lingus trademark for a term of three years following closing. In addition, to provide Flybe with additional publicity and brand awareness, for a period of 3 IATA seasons post-closing, Ryanair will ensure that the welcome screen of the Aer Lingus website will be vertically divided into two halves, bearing the trade names and logos of Aer Lingus and Flybe Ireland, together with an agreed explanatory text. On the Flybe Ireland side of Aer Lingus' website welcome screen, there will be a hyperlink that directs customers to the Flybe Ireland booking website and on the Aer Lingus side there will be a link to the Aer Lingus booking website. Moreover, for the 3 IATA seasons thereafter, there will be a "banner" on the Aer Lingus website with a hyperlink to the booking website of Flybe Ireland.

(1866) Respondents to the market investigation gave mixed views with respect to the question whether the licence to use the Aer Lingus trademark for the Flybe Ireland operations is sufficient to enable Flybe Ireland to compete effectively with the merged entity (1667). Respondents to the market investigation also show mixed views with respect to the question whether the use by Flybe Ireland of the Aer Lingus trademark in combination with its own trademark would help Flybe to develop the brand awareness of its own Flybe / Flybe Ireland brand in Ireland (1668).

(1867) The Commission acknowledges that the temporary combined use of the strong Aer Lingus trademark may help Flybe Ireland to penetrate the routes and generate revenues in the very short term by gaining customers. The Commission nevertheless notes that the licence would be limited to the trademark alone and would not extend to the wider branding and image of Aer Lingus, which would remain with the merged entity. The Commission moreover notes that the merged entity would continue to use the Aer Lingus brand, even on some of the overlap routes. The Commission considers that the parallel use of the Aer Lingus brand by the merged entity and of the Aer Lingus trademark by Flybe Ireland (in combination with the Flybe brand) is quite likely to confuse customers, a concern which was also raised by several respondents to the market investigation (1669). It would create uncertainty in the mind of customers as to whether Flybe operates separately from the merged entity. Even worse, it could even lead customers to believe that Flybe is not an alternative independent option from the merged entity and thus dissuade them from investigating Flybe's offering.

(1868) The Commission further considers that such a branding strategy could be at the expense of Flybe creating a strong level of awareness of its own brand.

(1869) In this context, the Commission notes that the trademark licence is granted for the same term as the Minimum Period, which means that the moment at which Flybe would lose the right to use the Aer Lingus trademark would coincide with the moment at which it would be free to exit the routes without incurring a penalty. This could increase the incentive for Flybe to exit the routes at the end of the Minimum Period.

(1870) The responses to the final market test also show mixed and divergent views about the extent to which the website publicity measures would enable the migration or carrythrough of the Aer Lingus customer base to Flybe Ireland (1670).

(1871) As set out in Section 8.5.2, the Irish market is characterised by a high level of online sales and Ryanair and Aer Lingus both realised most of their bookings in Ireland from the internet. The airline's website is thus an extremely important tool for attracting customers. The website publicity measures as foreseen in the final Commitments could help to some extent to carry through part of the visitors of Aer Lingus' website to Flybe Ireland's website during the transitional period. However, it is not clear to what extent such transferred customer would actually go on to book on Flybe's website. This would depend on a number of factors such as frequencies and prices offered and the customer's willingness to travel with an airline other than (the original) Aer Lingus. Furthermore, the Commission notes that such website publicity measures are highly unusual between competitors and might create confusion in the minds of customers.

(1872) Moreover, Flybe indicates that it would require that the website be operational from day one. The reason for this is that during the summer, whilst Ryanair might start to reorganise and reengineer the Divestment Business, Flybe would want to see commercial sales coming in, so that when it took over the business (expected by Flybe to take place in October 2013), it would do so off the back of a strong commercial performance during the intervening period (1671). The final Commitments however only foresee website publicity measures after the Hive Down.

(1873) Therefore it is uncertain that the website publicity measures would significantly contribute to a quick migration of Aer Lingus passengers to Flybe Ireland.

(1874) On balance, the Commission considers that the trademark arrangement and the Aer Lingus website publicity measures would be likely to create confusion in the minds of customers, as it would be likely that a significant proportion of customers would not perceive Flybe Ireland as a separate entity from and competing with the merged entity. Therefore, the Commission considers that the proposed commitments consisting of a temporary licence and website publicity measures leave doubts as to whether Flybe Ireland would succeed in establishing itself as an active competitor in the market.

(1875) Flybe itself indicates that in order to be recognised as independent by its customers, it has to "[follow] through on its planned level of marketing expenditure and activities" (1672). Flybe explains in this respect that it plans to put substantially more money into advertising and promotion in the start-up phase of Flybe Ireland. It hopes that, by feeding off the substantial presence it has in Great Britain and in Northern Ireland, the Flybe brand would become very well known in Ireland and it could transition through to Flybe being the predominant brand within a three year timescale. (1673)

(1876) The Commission acknowledges in this respect that the cash that would be injected into Flybe Ireland (1674) could help Flybe to invest in the development of its brand and website. The Commission nevertheless also notes that it is not excluded that Flybe would use the cash for its main business rather than for investing in marketing in Ireland. Flybe's incentive to compete on the 43 Flybe routes is further developed in Section 10.4.2.7.

(1877) If Flybe did not follow through on its announced level of marketing expenditure and activities, it would remain dependent on the Aer Lingus brand and on referrals through the Aer Lingus website.

(1878) In view of the above, the Commission considers that, despite the proposed trademark licence, website publicity measures, and capital injection in Flybe Ireland, it is uncertain whether Flybe would be in a position to establish a sufficiently strong brand, in particular as regards passengers originating from Ireland, which would allow it to effectively constrain the merged entity so as to remove the competition concerns identified by the Commission.

- Proposed base operations

(1879) Pursuant to Clause 14 of the final Commitments, Flybe would base the aircraft necessary to operate the Flybe Routes Schedule at Dublin airport, provided that at least one aircraft is based operationally at Cork airport.

(1880) According to Ryanair (1675), Flybe would base thirteen of the aircraft used to operate the Flybe Routes in Dublin Airport and one in Cork Airport. It considers that a oneaircraft operation at Cork Airport would constitute a base in the special circumstance where aircraft are overnighted all year-round. Where this occurs, the airline will - according to Ryanair - accrue all of the benefits of having a small base (such as the ability to offer early-morning flights, operational flexibility, the commitment to servicing a particular route and the ability to build brand recognition). Ryanair further claims that there are no economies of scale or scope as a function of the number of aircraft based at a particular location and argues that many Union airlines successfully operate bases with 5 aircraft or less.

(1881) In the Flybe Agreement, Flybe undertakes that "all aircraft applied by [Flybe] for operating the [43 Flybe Routes] are operationally based at Dublin airport, except for one such aircraft, which shall be operationally based at Cork."

(1882) The Commission notes that it is unclear what exactly is meant by "operationally based". Moreover, no base is foreseen in the final Commitments for Shannon airport. Flybe would only be taking over the Shannon-Manchester route. However, not having at least one aircraft at Shannon airport might lead to a decrease in the service level as Flybe would not be able to offer early morning departure flights from Shannon to Manchester. Aer Lingus currently offers early morning flights to Manchester on some days of the week.

(1883) In view of the above, the Commission considers that Flybe Ireland's ability to effectively constrain the merged entity so as to remove the competition concerns identified by the Commission is likely to be affected to some extent by the proposed limited base operations.

- Divested frequencies and capacity

(1884) Ryanair proposes that Flybe Ireland would serve the 43 Flybe Routes with at least the equivalent of the lower of Aer Lingus' or Ryanair's frequencies on the route. The frequencies that Flybe will take over and operate are referred to as the "Flybe Routes Schedule".

(1885) Ryanair provided an overview of current (that is to say, pre-Transaction) and divested frequencies per route for the summer and winter seasons. Ryanair explains that the pre-Transaction frequencies have been adjusted to reflect jet equivalent capacity for the Aer Arann frequencies (operated with turboprop aircraft with about 66 to 68 seats), in order "to reflect the fact that the UFBs will be operating jet frequencies" (1676). The Commission nevertheless notes that when calculating the jet equivalents of the frequencies offered with smaller Aer Arann aircraft, there are several routes on which rounding effects would likely lead to capacity reduction (1677).

(1886) Ryanair also explains that frequency data are based on a week in September 2012 (for the summer season 2012) and on a week in November 2012 (for the winter season 2012/13). Ryanair considers these weeks to be representative of the average weekly frequencies per season as these weeks allegedly correspond to the normal level of operation during a season (that is to say outside the peak July/August/Christmas periods where capacity is often moved around between routes) (1678).

(1887) The Commission nevertheless considers that such approach fails to capture the higher frequencies and the resulting capacities that are offered in peak periods (1679). For example Aer Lingus operated the Dublin-Palma route with five weekly frequencies between 23 June and 15 September in Summer 2012 (1680).The final Commitments however foresee only 4 weekly frequencies to be divested. Ryanair indicates that in peak season, capacity is often moved around between routes. However, Flybe Ireland would only have a limited number of aircraft and, moreover, it would be required to operate the Flybe Routes Schedule with these aircraft. This means that de facto Flybe Ireland would have only limited operational flexibility to move aircraft around between routes.

(1888) Furthermore, Flybe Ireland would not have the ability to deploy aircraft with a higher seat capacity in these peak periods such as Aer Lingus currently does (for example, Aer Lingus operates an A330 aircraft with at least 267 passengers (up to 322) on the Dublin-Malaga route during the summer season (1681)).

(1889) This means that Flybe Ireland would not be able, to the same extent as the merged entity would be, to deploy additional frequencies on the routes in peak seasons or to deploy aircraft with higher seat capacity on the routes in order to meet increased demand. The competition constraint exercised by Flybe Ireland on the merged entity in this regard would thus be limited.

(1890) Moreover, the final Commitments foresee that Flybe will be required to pay a penalty in case it does not operate at least 85% of these frequencies (1682). This de facto gives the possibility to Flybe not to operate up to 15% of the frequencies on any given route, without breaching the terms of the final Commitments and would imply that de facto the increment in terms of frequencies resulting from the Transaction would not be fully covered in case Flybe makes use of this facility.

(1891) It is finally noted that the divested frequencies are based on the assumption that Flybe Ireland flies with Airbus A320 aircraft. Flybe Ireland is given the right to replace the Airbus A320 aircraft with smaller aircraft (such as Flybe's regional jets) provided that it then increases frequencies so that the offered capacity remains the same (1683). However, there is no clear mechanism in the final Commitments to guarantee that the slots required for such increased frequencies would be provided to Flybe Ireland (1684). Moreover, a rounding provision is also offered in this regard, implying an additional risk of capacity reduction.

(1892) Therefore, the Commission considers that Flybe Ireland's ability to effectively constrain the merged entity so as to remove the competition concerns identified by the Commission is likely to be affected to a material extent by the proposed number of divested frequencies and implied capacities.

- Doubts as to whether Flybe Ireland could operate profitably on the 43 Flybe Routes

(1893) The Commission considers that an assessment of the viability of the Divestment Business is important in deciding whether the Flybe Commitments are workable. A complication in the case at hand is that due to the nature of the Transaction (it consists of a hostile bid by Ryanair) Flybe and Ryanair do not have access to the underlying data of Aer Lingus that would allow them to establish a credible business plan. Such limited availability of information in the context of a hostile public bid and the risk attached thereto is recognised in the notice on remedies. Paragraph 34 of the notice on remedies states that in cases involving a hostile bid, a commitment to divest activities of the target company may, in such circumstances of limited information available to the notifying parties about the business to be divested, increase the risk that this business might not, after a divestiture, result in a viable competitor which could effectively compete in the market on lasting basis. The notice on remedies also states that it may therefore be more appropriate to propose to divest activities of the acquiring company in such scenarios.

(1894) The fact that Ryanair, in order to find an upfront buyer, was required to give a guarantee for the cost base and a EUR 100 million buffer for covering shortfalls in revenues can be considered as a first indicator that the Divestment Business would not be really viable from a profitability point of view.

(1895) The Commission nevertheless made an in-depth assessment of the likelihood that Flybe Ireland's operations on the 43 Flybe Routes would be profitable. To this end, the Commission has not limited itself to information received from Ryanair, but it has also requested information it considered relevant from other parties, such as Aer Lingus and Flybe. The Commission examined all available information at its disposal, including route profitability data from Aer Lingus but also Flybe Ireland's balance sheet and Profit and Loss account prepared by Ryanair, the high level modelling prepared by Flybe and the terms in the final Commitments with respect to the Divestment Business Asset Plan.

(1896) The following analysis discusses in particular the revenue and costs assumptions underlying these plans.

- Aer Lingus's route profitability on the 43 routes

(1897) [...]* (1685), Aer Lingus' operations on the Divestment Business routes are [...]*.

(1898) Ryanair (1686) argues that the figures submitted by Aer Lingus should be treated with extreme scepticism given that Aer Lingus would be a biased party. It claims that the data are unsubstantiated and inconsistent with more detailed data in the case file. Ryanair further argues that the 43 Flybe Routes represent a balanced package of Aer Lingus' main short-haul routes [...]*. It provides a number of examples of ways in which Aer Lingus could have manipulated data [...]*. It also argued that its own model which is further discussed below in this Section, [...]*Ryanair has estimated this figure at EUR 350 million and conservatively assumed that Flybe Ireland would achieve just slightly over EUR 300 million in the first year. It states that there is no reason why [...]*.

(1899) The Commission notes that the allegations made by Ryanair are vague and imprecise. For example, Ryanair does not explain how the data submitted by Aer Lingus would be inconsistent with more detailed data in the case file, nor does it explain to which data it refers.

(1900) Furthermore, Ryanair considers that the allocation of costs by Aer Lingus would be a key element in any assessment of Flybe Ireland (1687). [...]*. In addition, while Ryanair now contests the Aer Lingus data, in the Flybe Agreement, Ryanair has agreed with Flybe that, for the preparation of the one year business plan, "the Aer Lingus Group's accounting policies, principles, determinations and judgements will be applied" (1688). Moreover, Aer Lingus confirmed that its costs are allocated using cost drivers (1689), for example aircraft ownership costs are allocated by elapsed time. Most costs are not allocated with regard to passenger numbers.

(1901) Ryanair also mentions that the accounting of connecting passenger revenues and costs is complex (1690). However, the allocation of costs in Aer Lingus' accounts is not related to whether passengers are connecting or non-connecting (1691). Regarding Ryanair's comment that Aer Lingus' profitability analysis is impacted by long haul operations, the Commission considers in addition that the Divestment Business will not benefit from this, at least not to the same extent as Aer Lingus.

(1902) The Commission finally notes that the turnover figure of EUR [...]* million represents the full frequencies operated by Aer Lingus, while the Divestment Business will typically operate on the increment resulting from the Transaction (that is to say the lower of Ryanair or Aer Lingus frequencies). This leads to a [...]* reduction in the Divestment Business turnover versus Aer Lingus' current turnover. Moreover, a further reduction of turnover can be reasonably expected because of lower load factors, lower fares, as a consequence of brand issues, possibly lower frequencies, etc., as explained in recitals 1905 ff below. Therefore, there is no reason to think that Ryanair's revenue hypotheses are particularly conservative.

(1903) On balance, the Commission considers that there is no reason to believe that the Aer Lingus figures were manipulated and do not accurately represent Aer Lingus' results on the routes.

(1904) Furthermore, the Commission notes that the Divestment Business is an artificiallycreated business, to accommodate the Commission's competition concerns. Currently Aer Lingus serves the 43 routes as part of an overall network, including long haul but also many other short haul routes ex-Irish airports. Aer Arann's operations that would be part of the Divestment Business also have relatively significant levels of connectivity to Aer Lingus' network (1692) (see however Section 10.6, where the Commission concludes that the commitments related to Aer Arann are not likely to be implemented in a timely manner).

- Flybe Ireland's balance sheet and Profit and Loss account prepared by Ryanair

(1905) Ryanair has made an estimated balance sheet and profit and loss account for Flybe Ireland (1693). In its model, Ryanair estimates passenger numbers for the 43 routes, using estimates of Aer Lingus' capacities and of Aer Lingus' load factors for the summer season 2012 and winter season 2012/13. This leads to an estimate by Ryanair of a total of around 4.32 million passengers for the two seasons together. Using a flown average fare of EUR 108.12 (1694), as estimated by Ryanair, the revenues generated on these routes would be EUR 467.5 million.

(1906) Ryanair also takes into account the fact that Flybe Ireland, under the Commitments, would not operate the full frequency schedule of Aer Lingus. Applying the ratio of the divested frequencies over Aer Lingus' current frequencies on these routes (1695), Ryanair's estimates indicate a number of around 3.47 million passengers for the two seasons. The total revenue of such reduced scale of the operations on the 43 routes would amount to around EUR 375 million (that is to say, in Ryanair's model, the reduced number of frequencies alone would lead to a reduction of revenues of more than EUR 90 million for the Divestment Business).

(1907) Then, Ryanair assumes that the Aer Lingus' load factor as initially estimated would be reduced by 5 percentage points to "allow for FLYBE branding". The total number of passengers for the 43 routes is consequently further adjusted down to 3.24 million for the two seasons together. This passenger figure is used by Ryanair to estimate the Divestment Business' turnover (1696), using the same flown average fare of EUR 108.12 (1697). Total revenues of the Divestment Business would thus amount to EUR 350.5 million under the above assumptions.

(1908) Ryanair applies a further discount factor of 15% to the yield per passenger (1698), still to "allow for FLYBE branding", reaching revenues of EUR 304.8 million for the Divestment Business. The operating expenses are estimated at EUR 286.5 million, leading to an operating profit before interest and tax of EUR 19.5 million (that is to say, a 6% operating profit margin).

(1909) According to Aer Lingus, Ryanair's model does not stand up to scrutiny (1699). As a general comment, Aer Lingus argues that Flybe Ireland would not benefit from the economy of scale advantages currently enjoyed by Aer Lingus in its key bases such as Dublin. Furthermore, according to Aer Lingus, the estimated operating costs set out in the Ryanair model are significantly understated and the projected profit before tax is overstated by a corresponding magnitude. Aer Lingus also questions the assumptions on which the Ryanair model is based.

(1910) [...]* (1700).[...]*. Therefore, while insufficient details were made available to the Commission by Ryanair, the Commission understands that very significant cost reductions are envisaged in these draft financial statements - [...]*. However, Ryanair does not explain in any clearly substantiated manner the sources of such cost reductions.

(1911) Indeed, the Commission notes that the balance sheet and profit and loss accounts as prepared by Ryanair are vague and rely on broad assumptions which are not verifiable (1701). This is not surprising given that Ryanair does not have access to the relevant Aer Lingus' information and has to rely to a large extent on less accurate publicly available information for making its model. Furthermore, Ryanair's estimates are sometimes even internally inconsistent. (1702)

(1912) In addition, the Commission has doubts about the relevance of a number of assumptions in the model. In particular, the Commission notes that the model is based on the assumption that passenger numbers would decrease in proportion to the reduction in frequencies operated by Flybe Ireland (in comparison to the frequencies operated by Aer Lingus pre-Transaction) and any reduction to the number of passengers would be due to the "brand effect"; such an assumption appears optimistic as the decrease could be more than proportionate (1703). Furthermore, the passenger load factor used by Ryanair has been set at Aer Lingus' published average short haul load factor of 76% (79% in Summer 2012 and 68% in Winter 2012/2013), reduced by 5%. Ryanair does not explain how Flybe Ireland would be able to achieve the projected load factors1704. Even if Flybe operated its current smaller aircraft on (part of) the routes, it is unclear whether Flybe Ireland would indeed be able to achieve such relatively high load factors. A difference in load factor has a significant impact on profitability. Moreover, aggregate figures of Aer Lingus are taken into account by Ryanair, whereas route specific fares (and load factors) would have been relevant information for more accurate estimates. Also, by basing the model on aggregate figures, it is very likely that there is an upward bias in the yield estimates for the routes that are part of the Divestment Business, as the Commission expects, (as is corroborated by the quantitative analysis), that in routes where Ryanair is also present, Aer Lingus fares are, ceteris paribus, lower. Besides, the fare reduction of 15% taken into account by Ryanair to "allow for FLYBE branding" is not supported by sufficient evidence as to its adequacy. A difference in average fare has also a significant impact on profitability.

(1913) Moreover, as set out in Section 10.6, the commitments related to Aer Arann are not likely to be implemented in a timely manner as far as the franchise agreement is concerned (see Section 10.6). Therefore, data related to Aer Arann should be discounted from the analyses (1705).

(1914) Furthermore, Flybe Ireland would operate with 12 Airbus A320 aircraft, while the final Commitments and the Flybe Agreement foresee that Flybe Ireland would have 9 A320s and 5 E195s aircraft. Operations with regional jet aircraft are less cost efficient and lead to a higher cost per seat. The fact that the model is prepared for a different fleet of aircraft (with different operating costs attached to it) raises significant uncertainties on its reliability and accuracy. Further examples are provided in the analysis made in the sub-Section "Cost base" below.

(1915) By way of conclusion, the Commission considers that Flybe Ireland's balance sheet and the Profit and Loss accounts prepared by Ryanair are based on a number of broad assumptions, which are not supported by convincing evidence and which offer an unrealistically high estimate of the profitability of the Divestment Business. Therefore, the Commission does not consider that such financial statements can shed much light on the viability prospects of the Divestment Business.

- High level modelling prepared by Flybe (1706)

(1916) Flybe has undertaken some own high level modelling on the bundle of the 43 routes, showing a profitability of EUR 21 million. The assumption used by Flybe is that operations would be as efficient as those of [...]* (as an existing [...]* operator), taking into consideration some "prudency".

(1917) However, the Commission notes that there is no explanation of why Flybe Ireland could be as efficient as [...]*, even including some prudency, in particular given that Flybe has no experience with [...]* aircraft operations and would not have the same scale of operations.

(1918) Furthermore, as another example of unsupported estimates, the 6 routes operated by Aer Arann in Flybe's high level modelling would achieve a profit of more than EUR [...]* million, thus accounting for much of the total EUR 21 million profit estimated by the model. The profit margin on revenues for these 6 routes would amount to around [...]*%, which is a highly improbable ratio, considering for instance Aer Arann's profitability (1707). Moreover, as set out in Section 10.6, the commitments related to Aer Arann are not likely to be implemented in a timely manner as far as the franchise agreement is concerned (see Section 10.6). Therefore, data related to Aer Arann should be discounted from the analyses. (1708)

(1919) The Commission considers that the Flybe's modelling does not demonstrate a sufficient level of reliability. The Commission notes in this respect that Flybe's board acknowledges that it does not have reliable data. (1709)

- Divestment Business Asset Plan

(1920) The final Commitments foresee that Ryanair will, in consultation with Flybe, develop a one year business plan for Flybe Ireland.

(1921) According to paragraph 3 of Annex I of the final Commitments, the Divestment Business Asset Plan delivered to Flybe under clause 2(c) will be prepared on the basis that, applying the following presumptions:

(a) subject to (b) below, the costs of the Divestment Business shall be the predicted costs for the operation by the Target Company of the Divestment Business for the 12 months following completion of the Hive-Down (the "Predicted Costs");

(b) notwithstanding the provisions of (a) above, the Predicted Costs shall be calculated on an operationally efficient basis assuming that the Divestment Business were operated by a reasonably prudent operator and on the basis that the Divestment Business is to be operated using 12 operating leased Airbus A320 aircraft;

(c) the Aer Lingus Group's accounting policies, principles, determinations and judgments will be applied; and

(d) it will be assumed that the Divestment Business earns the same revenue as it would have earned in the [...]* prior to the completion of the Hive-Down had it flown the Divestment Business Routes,

((a) to (d) together, the "Presumptions"), the projected pre-tax profits of the Divestment Business will be not less than €20 million in the first twelve months from the date of completion of the Hive-Down."

(1922) It should be clarified that this business plan is not binding on Flybe except insofar as it determines the amount of the cash contribution to be paid to Flybe for taking over the Divestment Business. This business plan will be prepared on the basis that, applying the above presumptions, the projected pre-tax profits of the Divestment Business will be not less than EUR 20 million in the 12 months following the transfer to Flybe (1710). In the event that the business plan does not project EUR 20 million in pre-tax profits, there is an agreed adjustment mechanism as set out in Annex I to the final Commitments (1711).

(1923) It is unclear how and whether the profitability target would be met. Ryanair simply refers to "steps that may be required to restructure the business" (1712) without any further details regarding this.

(1924) In particular, it appears doubtful that such projected pre-tax profits can be achieved even within the framework of the Divestment Business Asset Plan (even if optimistic but binding presumptions on revenues are accepted) given Aer Lingus current performance on the routes and given that Aer Lingus group's accounting policies and principles will be applied. The only improvement on the performance could come from Aer Lingus not being a "reasonably prudent operator"; however, the Commission considers that a company that has withstood competition from Ryanair is likely to be a "reasonably prudent operator".

(1925) Relying on a competitor to reduce one's cost base, even if there is a penalty involved, does not appear credible. The Commission is of the view that the additional cash to be provided to Flybe in the likely case where the projected pre-tax profits of the Divestment Business are less than EUR 20 million, offers only a short term palliative which does not ensure the long term viability of the Divestment Business. It should also be noted that, despite the additional cash to be paid by Ryanair, Flybe's incentive to operate the Divestment Business would be negatively affected. Indeed, a worse performing Divestment Business would be less profitable and considering that the closure cost would remain the same, Flybe's incentive to continue to operate the Divestment Business is likely to decrease.

- Cost base

(1926) The Commission notes that Ryanair did not provide any substantial and convincing evidence to demonstrate how Flybe Ireland's cost base would actually be reduced. For example in a letter of 11 February 2013, it was merely indicated that "[...] Ryanair intends to structure a cast base for Flybe Ireland that is driven by its own cost management skills, together with input from analysis of carriers such as easyJet and Flybe( ...) By leveraging its own skills in managing airports, labour, aircraft ownership, etc., Ryanair is confident that it can provide a cost base that will provide Flybe Ireland with margins at least as good as the 5% overall margin that Aer Lingus achieves [...]" (1713).

(1927) Ryanair announced efficiencies related to the acquisition of the entirety of Aer Lingus activities (see Section 9). While the Commission already concluded that the criterion of verifiability of the alleged efficiencies was not met, Ryanair claimed that a significant proportion of [...]*efficiencies projected by Ryanair would be derived from economies of scale, which would not be available to Aer Lingus (1714). However, the economies of scale which would be enjoyed by Flybe Ireland would be likely to be very small, if indeed there were any such economies to be achieved, considering the fact that Flybe Ireland represents only about one-third of Aer Lingus' short haul capacity (1715). Other efficiencies, in relation to maintenance or distribution costs, if any, also appear to be unlikely, and in any event very limited.

(1928) For airport and handling costs, Flybe has larger operations than Aer Lingus at several airports (for example Edinburgh, Glasgow, Manchester, Birmingham); these would be likely to yield savings. However, Aer Lingus currently has much larger operations at Dublin airport than Flybe Ireland would have, where corresponding savings would likely be lost for Flybe Ireland as regards all routes ex Dublin and Cork (while the focus of the Divestment Business is on routes ex Dublin and Cork). Moreover, Aer Lingus convincingly explains that airport charges estimated by Ryanair (1716) had been substantially underestimated. Indeed, while Ryanair uses a broad number based on Aer Lingus' average cost, Aer Lingus handles a substantial proportion of its ground operations itself, including passenger handling services at Dublin airport. These operations are not reported as airport charges by Aer Lingus but as staff costs. In turn, staff costs estimated by Ryanair do not appear to include self-handling costs (Ryanair assumes 6 crews made of 6 persons for each of the 12 aircraft). Moreover, Flybe Ireland would need additional fixed structure, as it is not likely that Flybe could provide such support functions without significant additional staff when around 40% turnover would be added to its current operations. This does not appear to be included in Ryanair's modelling.

(1929) The Commission notes that Aer Lingus already has a very efficient cost base, in particular following actions undertaken in the context of its Greenfield cost cutting plan. The Aer Lingus' 2012 Preliminary results announced on 6 February 2013 show that Aer Lingus has exceeded its annualised cost savings target by 7.5% (EUR 104 million in annualised savings compared to targeted EUR 97 million). In addition, the Commission is of the view that the generally lean cost structure of Aer Lingus is further demonstrated by its ability to sustainably compete, and gain market shares, against Ryanair on its large portfolio of routes.

(1930) [...]*.

(1931) The Commission also emphasises that while some further cost cutting may be possible, what is of relevance is "efficient" cost cutting, i.e. cost cutting that does not affect the willingness of consumers to pay for the services provided. If one assumes generally a lower cost structure the implications on the revenue should also be analysed.

(1932) Flybe expects Ryanair to deliver to Flybe Ireland a cost base that should allow the latter to achieve EUR 20 million in pre-tax profits. Flybe indicates in this respect that Ryanair has to undertake the transition and reorganisation of the Divestment Business to bring it down to the shape, size and cost structure before Flybe takes over Flybe Ireland (1717).

(1933) Flybe itself does not propose any concrete actions as to how the cost base should or could be reduced, but leaves it largely up to Ryanair to achieve a lower cost base.

(1934) As appears from the minutes of the Flybe board meeting held on 1 February 2013, Ryanair has not provided supporting evidence to Flybe for its view that a profit before tax of EUR 20 million would be achievable by Flybe Ireland. Acknowledging the absence of reliable data, save for that which is publicly available and the terms of the proposed agreement with Ryanair, Flybe's directors considered at their meeting of 1 February 2013 various scenarios against a so called "base case" in which Flybe Ireland would deliver revenues and costs in line with expectations (namely revenues of EUR [...]* million and costs of EUR [...]* million, leading to a profit before tax of EUR 20 million) (1718)

(1935) Flybe's directors considered the most likely scenario to be that the costs would be [...]* in the "base case" by EUR [...]* million per annum in all three first years, without however explaining the source of these [...]* costs (1719).

(1936) The Commission also notes that Flybe in practice will use 5 regional jets, whose CASK is higher than CASK of Airbus A 320/ Boeing B737 aircraft, as used by Aer Lingus and Ryanair respectively (1720). Therefore, the Commission considers that the real cost base of Flybe would likely be higher than the one used in these modelling exercises.

(1937) The Commission having reviewed all the evidence collected in the investigation concludes that it is unlikely that the cost predictions of Ryanair and Flybe for the Flybe Ireland business would materialise. Ryanair and Flybe do not explain the source of such cost reductions and do not discuss whether and to what extent - even if achieved - such cost reductions would also affect revenues.

- Estimated revenues

(1938) Flybe acknowledges that it would probably take time to get the revenues back to the Aer Lingus level (1721). At Flybe's board meeting of 1 February 2013, the most likely scenario was considered to be [...]* (1722).

(1939) The Commission considers that Flybe Ireland would face a number of disadvantages as compared to Aer Lingus with respect to the revenues it would achieve.

(1940) Firstly, Aer Lingus enjoys a well-established brand, in particular in Ireland. Ryanair considers that the Divestment Business would be able to set [...]*% lower fares than those Aer Lingus currently charges, and would have [...]*% lower load factor, at least initially. As set out above, despite the proposed trademark licence, website publicity measures, and capital injection in Flybe Ireland, it is uncertain whether Flybe would be in a position to establish a sufficiently strong brand, in particular as regards passengers originating from Ireland.

(1941) Secondly, Flybe Ireland would not be able to capture intra-company connecting traffic as Aer Lingus/Aer Arann currently do. Also, it is unclear whether it would succeed in achieving a similar level of connectivity with third parties as Aer Lingus currently does thanks to its open network structure and numerous early morning frequencies (which Flybe Ireland could likely not match) (1723).

(1942) Thirdly, Flybe acknowledges that it cannot compete with Ryanair on prices (1724) and argues that it would offer a different type of product with high frequencies. However, often (and except for some routes where it would operate with regional jets, in which case it would face a higher cost per seat), the number of frequencies offered by Flybe Ireland would be at most equal to or less than the frequencies offered by the merged entity. It can also not be excluded that Flybe Ireland would not obtain the necessary slots to offer a business day-return on routes where Aer Lingus today does so and on which it would continue to operate alongside Flybe post-Transaction. Annex II of the final Commitments, 3rd indent provides in this respect that "where possible the relevant Aer Lingus Group slots will be allocated so as to enable the Divestment Business to fly a business day-return, as long as the remaining Aer Lingus schedule remains reasonably attractive". (emphasis added)

(1943) In light of all the above, the Commission concludes that Flybe is unlikely to maintain the same level of revenues as Aer Lingus currently has in the Divestment Business even in the medium term.

- Conclusion

(1944) Taking into account also the conclusion on the cost structure, and the current profitability of Aer Lingus in the relevant routes, the Commission thus considers it unlikely that Flybe Ireland would achieve sufficient revenues and such cost base to achieve the targeted profits before tax of EUR 20 million.

(1945) More importantly, the Commission considers it unlikely that Flybe Ireland's operations on the 43 Flybe Routes, considered altogether, could be profitable at all.

10.4.2.7. Flybe's incentive to continue to operate on a lasting basis on the 43 routes or on the Irish market in general

(1946) The responses to the final market test show that a majority of respondents either considers it unlikely that Flybe would continue to operate on the 43 Flybe Routes beyond the compulsory period of 6 IATA seasons or indicates that this would depend on the profitability of the operations (1725).

(1947) Flybe's board considered a number of cash and profit scenarios (1726). These cash and profit scenarios were based on a number of assumptions, including that projected revenues for the reference year would be in the amount of EUR [...]* million. The envisaged outcomes summarised at the end of each scenario are based on keeping the business open, or closing it; no restructuring scenarios are considered in this high level risk analysis.

(1948) In a majority of the various sensitivity scenarios, Flybe Ireland would be [...]* (1727).

(1949) In the "base case", Flybe Ireland would deliver revenues and costs [...]*. In this "base case", the Flybe Ireland business would generate revenues of EUR [...]* million and would incur costs of EUR [...]* million. In the worst case scenario, revenues would be down by [...]*% and costs EUR [...]* million higher than expected. In such scenario, the business would be [...]*.

(1950) Flybe's board minutes indicate that "(...) [The] Directors had a good understanding of the range of potential results of entering into the proposed transaction and viewed this as a long term commitment to the Irish market." (1728)

(1951) However, the Commission finds several elements that do not corroborate the stated long term commitment of Flybe to the Irish market.

(1952) First, the purchase price of EUR 1 million, compared to the initial capital provided by Ryanair to the Divestment Business, is not indicative of a medium-long term commitment by Flybe to the Irish business given that Flybe clearly indicated that it is not prepared to take any financial risk with respect to the Flybe Ireland acquisition (1729). Because of its recently announced restructuring program and the low cash available to Flybe, it seems highly unlikely that Flybe would inject any additional cash in the Flybe Ireland business in the likely case the EUR 100 million would be fully utilised, in particular if it transpired that the Divestment Business was not a viable business. This is corroborated by the described sensitivity scenario analysis, whereby [...]*.

(1953) Secondly, the Commission further notes that the penalty mechanism foreseen in the final Commitments does not constitute a real deterrent for Flybe not to close routes. Flybe itself indicates that, even though it is its intention to stay on the routes, "(...) if push come to shove we would be able to terminate routes at a penalty that we think, whilst a deterrent, is not an overriding deterrent - let's put it that way" (1730). It also indicated that while that is not the aspiration, "(...) legally [it] could withdraw from anything at any time and pay penalties" (1731). On the contrary the Commission notes that according to Flybe's figures, closure costs would remain [...]*. Therefore, the Commission concludes that Flybe takes essentially no commercial risk in acquiring the Divestment Business. Coupled with the significant concerns on the viability of the Divestment Business, the risk of aggressive retaliation by Ryanair (see Section 8.5.2.6) and the limited cash resources available to Flybe, this raises very significant concerns over Flybe's incentive to continue operating the Divestment Business on a lasting basis.

(1954) Thirdly, the Flybe Board's expectations are for outcomes [...]*.

(1955) In the "Sensitivity F" scenario, revenue would [...]* while the costs would be EUR[...]*million [...]* by the "base case". In such scenario, [...]*. The Flybe Board thereby notes that [...]*. Considering Flybe's aversion to taking any commercial risk in relation to this transaction, the Commission considers it likely that Flybe would close the Divestment Business even before the end of year 3 so as not to incur the [...]*.

(1956) In the "Sensitivity G" scenario, revenue would [...]* as compared to the reference year. Costs would again be EUR [...]* million [...]*. In this scenario, Flybe Ireland would be [...]*. This scenario depends however on costs being lower from the outset than projected by the "base case", but as noted above, the Commission considers it unlikely that even the "base case" costs would materialise, far less that they could immediately be reduced by a further EUR [...]* million.

(1957) Therefore, even Flybe's Board [...]*. This would be the case in one of the two scenarios that it considers most likely and the other scenario depends on an unsubstantiated projection of cost reductions.

(1958) Fourthly, and as already noted above, the EUR 100 million capital is not restricted or ring-fenced to Flybe Ireland and could be used e.g. for Flybe's main businesses in the United Kingdom. Considering the cash position of Flybe Ireland and the prospect of continued losses [...]*, this might be an incentive for Flybe to exit Flybe Routes (in part or entirely) before the EUR 100 million capital was exhausted.

(1959) The Commission notes that the sensitivity analysis made by Flybe's Board is a high level risk analysis that is based on keeping the Divestment Business open or closing it, and that [...]* (1732). The Commission cannot foresee the outcome of a possible restructuring as it does not have any concrete business plan or scenario from Flybe. However, given Flybe's current financial situation and its risk adversity, such a restructuring would be likely to have a significant impact on the number of routes on which Flybe would continue to operate. Such a restructuring would vitiate the effectiveness of the remedy as a means to remove the identified significant impediments to effective competition, and, thus, could not render the proposed concentration compatible with the internal market.

(1960) Therefore, the Commission concludes that Flybe would not have a sufficient incentive to continue to operate at least a substantial part of the 43 Flybe Routes on a lasting basis.

10.4.2.8. Conclusion on the viability of the Divestment Business and suitability of Flybe as a purchaser

(1961) The Commission therefore considers that Flybe is not a suitable purchaser in whose hands the Divestment Business would become an active competitive force in the market.

(1962) The Commission further considers that the remedy is insufficient to remove the identified significant impediments to effective competition, and, thus, could not render the Transaction compatible with the internal market.

10.4.3. Further relevant aspects as regards the final Commitments

10.4.3.1. Uncertainties as to the timely implementation of the final Commitments

(1963) The notice on remedies provides at paragraph 95 that commitments require safeguards to ensure their effective and timely implementation. Furthermore, the notice on remedies also provides at paragraphs 97 and 100 that the divestiture has to be completed within a fixed time period agreed between the parties and the Commission and that for fix-it-first solutions for which in general, a binding agreement with the purchaser will already be entered into during the procedure, after the Commission's decision, only a further period for the closing has to be foreseen. Such period is according to the notice on remedies generally of 3 months.

(1964) The final Commitments provide that, under the Flybe Agreement, Ryanair shall sell and Flybe shall acquire the Divestment Business "as soon as practicable after the successful completion by Ryanair of the Offer." The Flybe Agreement for its part foresees at clause 4.2. that as of the date upon which the Offer becomes effective, Ryanair shall "as soon as reasonably practicable" (a) procure the incorporation of the Divestment Business, (b) in connection with Flybe develop a one-year business plan for the Divestment Business and (c) deliver to Flybe such business plan. The Flybe Agreement further foresees certain additional steps which could possibly further delay the acquisition of the Divestment Business by Flybe such as an adjustment mechanism if Flybe believes that the business plan does not meet the agreed profitability targets (clause 4.5 of the Flybe Agreement) or the involvement of an independent expert should there be such disputes (clause 4.7 of the Flybe Agreement). While there are indeed fixed timeframes for the adjustment and the resolve of the independent expert, should a revised business plan be necessary, it will be delivered by Ryanair to Flybe "as soon as reasonably practicable" (clause 4.8 of the Flybe Agreement). Once all these hurdles have been passed, Ryanair will use its "reasonable efforts to complete the Hive-Down" (clause 4.12 of the Flybe Agreement).

(1965) As foreseen in the notice on remedies, in the a fix-it-first type of remedy the implementation of commitments is expected to happen in a much swifter timeframe than the implementation of other types of remedies because the identity of the buyer is known and an agreement has already been reached. However, the proposed Commitments set-up a framework in which there are no time limits for Ryanair to sell the Divestment Business to Flybe beside "as soon as reasonably practicable" by using "reasonable efforts".

(1966) In the light of the above, the Commission concludes that the final Commitments do not sufficiently guarantee that they would be implemented in a timely manner and at the very lease create uncertainties as to when the Divestment Business would be acquired by Flybe and competition restored on the market.

10.4.3.2. Involvement of Ryanair post-Transaction

(1967) Paragraph 65 of the final Commitments provides that, as set out in the Flybe Agreement, from the date the Offer becomes effective (Closing) until the sale and purchase of the Divestment Business pursuant to the Flybe Agreement, Ryanair shall preserve the economic viability, marketability and competitiveness of the Divestment Business in accordance with good business practice.

(1968) Under paragraph 66 of the final Commitments, as soon as practicable following the date the Offer becomes effective (Closing), Ryanair will carve-out the Divestment Business form Aer Lingus and transfer it into a separate standalone company in a manner that will allow and enable this company to operate as a separate independent entity. Following this process and until the sale and purchase of the Divestment Business pursuant to the Flybe Agreement, Ryanair will keep the Divestment Business separate from the part of Aer Lingus which Ryanair is retaining and will ensure that Key Personnel of the Divestment Business - including the Hold Separate Manager - have no involvement in any business retained by Ryanair.

(1969) Furthermore, until the date the Offer becomes effective (Closing), Ryanair shall assist the Monitoring Trustee in ensuring that the Divestment Business is managed as a distinct and saleable entity separate from the business retained by Ryanair. In addition, Ryanair shall appoint a Hold Separate Manager who shall be responsible for the management of the Divestment Business under the supervision of the Monitoring Trustee. The Hold Separate Manager shall manage the Divestment Business independently and in the best interest of the business with a view of ensuring its continued economic viability, marketability and competitiveness and its independence from the business retained by Ryanair and Aer Lingus.

(1970) However, as appears from clause 4.4 of the Flybe Agreement, the Hold Separate Manager would be appointed "no later than the date on which the Hive-Down is completed". As explained above, under the Flybe Agreement, Ryanair shall use reasonable efforts to complete the Hive-Down but the timing of this event (which constitutes the trigger for the timetable for Completion of the Flybe Agreement) is subject only to an obligation to use "reasonable efforts".There is high uncertainty as to when precisely the Hold Separate Manager would be appointed and how the final Commitments guarantee that the Divestment Business is managed independently under the supervision of the Monitoring Trustee.

(1971) Furthermore, even if the Hold Separate Manager were to be appointed right after the Offer is completed, Ryanair's commitment towards Flybe that it would itself develop a business plan of the Divestment Business which would ensure that a certain level of profitability is forecasted, casts serious doubts as to the question whether such activities could be qualified as management of the "day-to-day business" of the Divestment Business.

(1972) Lastly, contrary to the Commission's standard commitments model text, the proposed Commitments do not provide for any ring-fencing mechanism which would ensure that Ryanair does not obtain any business secrets, know-how, commercial information, or any other information of a confidential or proprietary nature relating to the Divestment Business. Quite to the contrary, as Ryanair itself committed towards Flybe to develop the business plan of the Divestment Business, there is a serious risk that confidential information of the Divestment Business would be made available to Ryanair.

(1973) In the light of the above, the Commission considers that the final Commitments do not sufficiently guarantee that the Divestment Business would be held separate from Ryanair sufficiently promptly after Ryanair taking control over Aer Lingus.

10.4.3.3. Complex and not sufficiently clear-cut commitments

(1974) The final Commitments are complex and are not sufficiently clear-cut.

(1975) The Commission has in its previous decisional practice (1733) accepted commitments which provided that the agreements which were concluded to implement the commitments would provide for a fast-track dispute resolution procedure.

(1976) The final Commitments provide in paragraph 88 that "In the event that an Applicant has reason to believe that Ryanair is failing to comply with the requirements of the Commitments vis-à-vis a third party, the fast track dispute resolution procedure described in [that] Section will apply". Considering the definition of "Applicants" in the final Commitments, it would appear that the fast-track dispute resolution clause is not applicable to any disputes which may arise with the identified entrants Flybe and IAG.

(1977) Indeed the Flybe Agreement provides in Paragraph 26.2 that "Each Party agrees that any proceeding, suit or action out of or in connection with this Agreement shall be subject to the non-exclusive jurisdiction of the English courts to which the Parties hereby submit" and "Each Party waives (and agrees not to raise) any objection on the ground of forum non conveniens or on any other ground to the taking of Proceedings in any court in accordance with the provisions of this clause".

(1978) Ryanair for its part commits in Clause 2 of the final Commitments to enforce its contractual rights under the Flybe Agreement "In the event of a breach by Flybe ... of the terms of the Flybe Agreement ..." to ensure full compliance by Flybe with the terms of the Flybe Agreement.

(1979) The rationale for including the same fast-track dispute resolution mechanism in the Commitments and the respective agreements is that the Commission has to be persuaded that - should disputes arise - there is a working mechanism to ensure that these are resolved within a strict and tight timeframe. Commitments have to be implemented in a timely manner to ensure that the significant impediment to effective competition does not materialise.

(1980) In light of the complexity of the final Commitments, the settlement of disputes arising from the implementation of the Flybe Agreement through litigation in English Courts instead of a faster way of settlement raises doubts as to whether the final Commitments would be implemented in a timely manner.

10.4.4. Conclusion on the divestiture to Flybe

(1981) The Commission considers that the final Commitments do not eliminate the competition concerns entirely and are not comprehensive and effective from all points of view. Furthermore, the Commission is not able to conclude with the requisite degree of certainty that the new commercial structures resulting from them will be sufficiently workable and lasting to ensure that the significant impediment to effective competition will not materialise. Moreover, the Commission cannot clearly determine that the final Commitments, once implemented, fully and unambiguously resolve the competition concerns identified in this Decision.

10.5. Slot divestitures to IAG/BA on the three London routes (Dublin, Cork and Shannon)

(1982) The Commission recalls that, exceptionally, it decided to examine the final Commitments although these did not clearly resolve the competition concerns previously identified and therefore required a further market test (1734).

(1983) According to the notice on remedies, whilst divestiture is the remedy preferred by the Commission, the Commission also may accept others. There may be situations where divestiture of a business is impossible. However, divestitures are the benchmark for other remedies in terms of effectiveness and efficiency (1735). In such circumstances, the Commission has to determine whether or not other types of remedy may have sufficient effect on the market to restore effective competition (paragraphs 14 and 61 of the notice on remedies).

(1984) Slot commitments of the type proposed by Ryanair on the three routes between London and Dublin, Cork and Shannon respectively, are based on the fact that the availability of slots at Gatwick or Heathrow is a significant entry barrier (1736).

(1985) The final Commitments for the three London to Ireland routes are without prejudice to the question whether and when the so called Heathrow Transfer Condition is satisfied.

(1986) In the Sections that follow, the Commission will first assess the question whether the Heathrow Transfer Condition is likely to be satisfied (see Section 10.5.1 below). The Commission will then show that the merged entity is unlikely to be sufficiently constrained during and after the Minimum Period (Sections 10.5.2 and 10.5.2) Lastly, the Commission will set out other concerns in relation to proposed Commitments on the three London-Ireland routes (Section 10.5.4) and draw its conclusions in Section 10.5.5.

10.5.1. Transfer of Heathrow Slots unlikely

10.5.1.1. Introduction

(1987) Under the notice on remedies, the requisite degree of certainty concerning the implementation of proposed commitments may in particular be affected by risks regarding the transfer of a business to be divested, such as third party rights in relation to the business.

(1988) The Commission made clear on several occasions (1737) that the ability of Ryanair to dispose of Aer Lingus' Heathrow slots post-Transaction would raise complex issues and lead to uncertainties as to the implementation of such commitments, particularly because of the restrictions in Article 10 of Aer Lingus' Articles of Association ("Article 10") and other relevant legal provisions.

(1989) It is incumbent on the Parties to remove any uncertainties as to the implementation of the remedy when submitting it to the Commission1738. However Ryanair asserted that obstacles to delivering the disposal of Heathrow slots were "unlikely" and insisted in all its proposals that the disposal of Aer Lingus' slots at Heathrow did not raise any legal issues and could be implemented. However, Ryanair's letter to the Commission of 15 January 2013, accompanying the Commitments of 15 January 2013, simply stated that Ryanair's professed ability to overcome Article 10 "is addressed in a legal opinion from Martin Hayden SC and Ross Aylward BL" provided to the Commission (the "Ryanair opinion").

(1990) The question of Ryanair's ability to secure the disposal of the Heathrow slots has indeed been the subject of extensive legal submissions by both Aer Lingus and Ryanair, which have concentrated on (i) the interpretation of Article 10 and (ii) whether minority shareholders in Aer Lingus could invoke other legal provisions to prevent the disposal of the Relevant Slots, even if they were unable to invoke Article 10 to do this.

(1991) In order to assess the validity of the arguments made in the Opinions submitted by Aer Lingus and Ryanair, the Commission has obtained its own legal opinion from an independent legal counsel Rossa Fanning BL ("the Fanning Opinion") (1739). The Commission's doubts as to the deliverability of a commitment involving Heathrow slots are shared by the Fanning Opinion, which concludes (at paragraph 75) that "Aer Lingus has a strong basis for contending that there is a domestic law impediment to a LHR Commitment ... arising out of Article 10".

(1992) In the final Commitments, Ryanair proposes to set-up a mechanism whereby IAG would lease Gatwick slots to operate on the three London routes until the Heathrow Transfer Condition is satisfied and IAG can take over Aer Lingus' Heathrow slots.

(1993) Therefore, the final Commitments still include the prospect of Aer Lingus' Heathrow slots being disposed of.

10.5.1.2. Description of the Heathrow Transfer Condition

(1994) Under the final Commitments, Ryanair commits that IAG(BA) would provide services on the Dublin-London, Cork-London and Shannon-London routes either from Heathrow airport or from Gatwick airport. An entry by IAG(BA) at Heathrow or Gatwick and the type of agreement which Ryanair would enter into with IAG(BA) depends on whether and when the so called "Heathrow Transfer Condition" is satisfied.

(1995) The final Commitments define the Heathrow Transfer Condition as "the determination by each of Ryanair and IAG each acting in its reasonable discretion that giving effect to the Heathrow Gatwick Tranfer Agreement would not violate Article 10 of Aer Lingus' Articles of Association (or other successor provision) or any applicable law or regulation".

(1996) Under paragraph 25 of the final Commitments, if two months before the IAG Commencement Date (1740) the Heathrow Transfer Condition has not been satisfied, IAG(BA) and Ryanair would enter into the Gatwick Lease Agreement. However, paragraph 24 of the final Commitments provides that, during the term of the Gatwick Lease Agreement, Ryanair will use its best efforts to achieve satisfaction of the Heathrow Transfer Condition (including without limitation the exercise of any rights it may hold as a shareholder in Aer Lingus and/or as a result of being able to control the board of directors of Aer Lingus and/or seeking the agreement of the relevant Irish governmental authorities and/or seeking declaratory judgment from a court of competent jurisdiction.

(1997) Under paragraph 26 of the final Commitments, if the Heathrow Transfer Condition were to be satisfied prior to two months before the IAG Commencement Date (1741), IAG(BA) and Ryanair would enter into the Heathrow - Gatwick Transfer Agreement. Paragraphs 28-29 of the final Commitments also provide that if the Heathrow Transfer Condition were satisfied at any point prior to the termination of the Gatwick Lease Agreement and IAG exercised its Call Option, the Heathrow - Gatwick Transfer Agreement would also come into effect and the Gatwick Lease Agreement and the Heathrow Lease Agreement (if in effect) would terminate (1742).

(1998) The Commission considers that the final Commitments do not state in a sufficiently clear manner the precise meaning of the Heathrow Transfer Condition. Considering in particular the previous existing disputes and diverging views on the question of the interpretation of Article 10 (and other legal provisions) it is likely that Ryanair would itself "determine acting in its reasonable discretion" that there are no issues involving Article 10 or other applicable laws or regulations, while minority shareholders of Aer Lingus would disagree. Furthermore, what is actually in issue for the Commission's assessment is not Ryanair's determination in its "reasonable discretion" but whether Ryanair could in reality transfer the Heathrow slots.

(1999) The Sections which follow provide an assessment of the likelihood that the Heathrow Transfer Condition would be satisfied.

10.5.1.3. Irish Government unlikely to sell its 25.11% stake in Aer Lingus

(2000) As the applicability of Article 10 is very likely to depend on the question whether the Irish Government would sell its shares in Aer Lingus to Ryanair, it is relevant to first assess the question of what is the most likely course of action by the Irish Government before entering into the mechanics of Article 10 which will be described in Section 10.5.1.4).

(2001) Ryanair argued that it is highly unlikely that the Irish Government's 25.11% shareholding in Aer Lingus will be retained post-Transaction as the Irish Government has decided to sell its stake in connection to the commitments made to the Troika of the European Union, the European Central Bank and the International Monetary Fund.

(2002) While it is true that a decision in principle has been taken by the Irish Government to sell its shareholding in Aer Lingus, the timing of a possible sale and any conditions for a sale have not yet been agreed (1743). However, the Irish Government issued a public statement on 18 December 2012 (1744) and on 12 February 2013 (1745) whereby it confirmed that while it remains committed to the sale of the 25.11% stake in Aer Lingus at the right time under the right conditions, it is not prepared to support any offer that would significantly undermine connectivity or competitiveness for Ireland (1746). The Irish Government reiterated its position not to sell its shareholding in Aer Lingus to Ryanair in response to the final market test: "This revised remedies package [of 1 February 2013] does not satisfy the Irish Government's concerns about connectivity, competition and employment in the Irish market. The Government has recently indicated that it is not prepared to support any offer that would significantly impact connectivity and competition in the Irish market." (1747)

(2003) Even if Ryanair's Offer succeeds, it is likely that the position of the Irish Government would remain the same, as the Irish Government would still be keen on maintaining Aer Lingus' operations from London Heathrow for connectivity reasons. The Irish Government reiterated that "While it retains its shareholding in Aer Lingus, the Government will invoke its rights in the procedure in Article 10 of the Aer Lingus' Articles of Association to protect access to Heathrow services from Ireland which remains very important for Ireland's connectivity..." (1748) Therefore, Ryanair's projection that, post Transaction, the Irish Government would eventually accept the situation and allow Ryanair to sell the slots to IAG is simply not sufficiently certain to materialize (1749).

(2004) Furthermore, a sale by the Irish Government of its shares in Aer Lingus to a third party also appears unlikely within the medium term. The Irish Government explained that "while the Government has made a decision in principle to sell its shareholding, the timing of a possible sale and any conditions for a sale have not yet been considered. The context in which the Government has made this decision in principle to sell its shareholding is its agreement with the EU/ECB/IMF Troika to reach a target of €3 billion from a programme of State asset disposals. The Government identified a number of assets for disposal under the programme. The potential value of the Government's shareholding in Aer Lingus is small relative to some of the other assets identified and the Government is not committed to a set timetable in relation to the sale of its shareholding. It has indicated that the shareholding will be sold at an appropriate time but only when market conditions are favourable and at an acceptable price to be agreed by Government." (1750)

(2005) While Ryanair has an obligation under the final Commitments and the IAG Agreement to "use its best efforts" to achieve satisfaction of the Heathrow Transfer Condition, it appears that, as long as the Irish Government maintains its shareholding in Aer Lingus, Ryanair would be unlikely to be in a position to dispose of Aer Lingus' Heathrow slots. As mentioned above, the Irish Government clarified that while it retains its shareholding in Aer Lingus, it would invoke its rights under Article 10.

(2006) Therefore, it appears likely that the Irish Government would not sell its stake in Aer Lingus to Ryanair. Even if the Irish Government were to sell its shareholding in Aer Lingus to a third party, such sale would not be timely.

10.5.1.4. Article 10 of Aer Lingus' Articles of Association

(2007) In brief, Article 10 of Aer Lingus' Articles of Association ("Article 10") prevents a "Disposal Transaction" in relation to Aer Lingus' Heathrow slots without first giving minority shareholders the opportunity to call an Extraordinary General Meeting ("EGM") of Aer Lingus's shareholders (1751). A "Disposal Transaction" is defined very broadly as any transaction pursuant to which slots would be sold, transferred, disposed of, leased, surrendered, mortgaged or otherwise alienated or encumbered (1752).It thus appears that the exchange of Aer Lingus' Heathrow slots as proposed by Ryanair in the final Commitments would have to be notified to the shareholders holding more than 10% of the share capital in Aer Lingus (1753).

(2008) Under Article 10(c), shareholders holding at least 20% of Aer Lingus' issued share capital can, within 28 days of receipt of such notification, request that an EGM is convened to consider such proposed Heathrow slot disposal transaction. If no such request for an EGM is made, the disposal of the Heathrow slots may be concluded within 12 months of the expiry of the 28 day period without the need for shareholder approval (1754).

(2009) The Irish Government is the only other Aer Lingus shareholder which has alone more than 20% of Aer Lingus' shares (1755), as it has decided not to sell its 25.11% stake in Aer Lingus to Ryanair as explained above in Section 10.5.1.3. Therefore the Irish Government is the only shareholder that could make such a request for an EGM in a timely manner.

(2010) Pursuant to article 10(f), the Irish Minister for Finance would exercise its rights to convene an EGM in accordance with the recommendations of the Minister for Transport.

(2011) The Minister for Transport has set out the criteria by reference to which he intends to make recommendations to the Minister for Finance as to whether to convene an EGM (1756). As a general principle and by assessing each case individually, the Minister for Transport will only recommend to the Minister for Finance to convene an EGM if the disposal would result in the slots at Heathrow falling below the critical level for ensuring connectivity to and from Ireland (1757).

(2012) The final Commitments cover an exchange of most of Aer Lingus' slots at London Heathrow for slots at Gatwick (1758) which is significantly less suitable as a connecting airport. As explained in Section 10.5.1.3 above, the Irish Government has publicly stated that it would not sell its shareholding in Aer Lingus to Ryanair and further expressed its concerns as regards the connectivity issue. The Commission is therefore of the view that the disposal of Heathrow slots as proposed in the final Commitments would raise sufficient issues for the Minister for Transport to recommend to the Minister for Finance to convene an EGM

(2013) Indeed, the Irish Department of Transport, Tourism and Sport confirmed that while it retains its shareholding in Aer Lingus, "the Government will invoke its rights in the procedure in Article 10 of the Aer Lingus' Articles of Association to protect access to Heathrow services from Ireland which remains very important for Ireland's connectivity." (1759)

(2014) The Commission is therefore satisfied that the Irish Government would request that an EGM is convened to decide on the disposal of the Heathrow slots under the final Commitments and that it would vote against such proposal.

(2015) At the EGM, the resolution to dispose of LHR slots would have to be carried by a "Specified Percentage", which is defined as 100% minus the shares held by the Irish Government, as a percentage of the total issued shares, plus 5%. A vote in favour of a resolution to dispose of Heathrow slots would have to be carried by the Specified Percentage of the votes cast.

(2016) At present, the Irish Government holds 25.1%, so a blocking vote would be 25.1% + 5% = 30.1% of votes cast, plus one vote. Given the Irish Government's current shareholding therefore, Article 10 implies that, to approve the disposal of Heathrow slots at the EGM, a resolution must be passed by 69.89% (1760) of the votes cast by the shareholders attending and voting at the EGM (1761). Thus, if there were a 100% turnout, the Irish Government would require support from holders of another 5% of shares, but if the turnout were less than 100%, the Irish Government's 25.1% shareholding might be sufficient to constitute more than the Specified Percentage of votes actually cast.

(2017) In Ryanair's view, the Irish Government's ability to block a Heathrow slot disposal would depend on the number of Aer Lingus shares held by Ryanair after a successful bid and the number of Aer Lingus' shareholders that attended an EGM to approve (or block) such disposal.

(2018) Ryanair explained that, if a request for an EGM was to be made, taking a hypothetical attendance figure of 71.98% (the turnout for voting on the resolution on which most of Aer Lingus shareholders voted on at the last Aer Lingus' Annual General Meeting) and if Ryanair's bid was successful and it held for example, 51% of Aer Lingus' share capital (1762), Ryanair would surpass the 69.89% approval threshold required to pass a Heathrow slot disposal.

(2019) The Commission considers that past attendance rates can hardly provide a reliable indicator of the likely attendance at an EGM post-Transaction, in particular as the issue at stake would be very likely to crystallise diverging views on the future of Aer Lingus and on more general transport issues, which may be seen as critical for Irish people.

(2020) More fundamentally, while Ryanair's claim is true in that, if taking the hypothetical attendance figure of 71.98%, a shareholding of 51% would give Ryanair 70.85% of the votes cast (1763), this scenario is only realistic if, before such a meeting, the Irish Government had sold - quod non - its shareholding in Aer Lingus to Ryanair. As explained above however, the more realistic scenario is that the Irish Government would not sell its shareholding to Ryanair in which case the attendance rate at the EGM (if it were to be attended by these two shareholders alone) would be 76.11% (1764). Therefore, the Irish Government would be in a position to block the disposal of the Heathrow slots (because with its 25.11% shareholding it would represent more than 30.12% of the vote cast at an EGM with an attendance rate of 76.11%).

(2021) Ryanair confirmed that (1765), considering the attendance rate at Aer Lingus' Annual General Meetings, the Government shareholding as a percentage of votes cast was higher than 30.1% in recent years. Therefore, on the basis of attendance rates at previous Annual General Meetings, the Irish Government would be in a position to prevent the disposal of Heathrow slots (as it would have the required 30.12% of votes cast needed to block a disposal of Heathrow slots). Moreover, if attendance at the EGM were lower than around 83%, the Irish Government with its 25.1% shareholding would be in a position, acting alone, to block the disposal of Heathrow slots (1766).

(2022) Ryanair further contended that, if other shareholders' attendance rate is assumed at only half of the average in previous years (i.e. assuming that attendance by shareholders other than Ryanair and the Irish Government represents approximately 8% of the total number of shares in issue), and Ryanair holding as little as 51% of Aer Lingus' shares, the EGM attendance rate would be approximately 84%. Ryanair therefore concluded that the Government's 25% shareholding would not be sufficient to block a Heathrow slot disposal (1767).

(2023) It is true that the Irish Government alone could not block the disposal of the Heathrow slots should the attendance rate be above 83% at an EGM. This means that, absent the sale by the Irish Government of its shareholding to Ryanair (which is the most likely scenario as explained above), attendance and voting at the EGM would have to be above 83% for Ryanair to possibly be in a position to secure the disposal of the Heathrow slots.

(2024) However, it is likely that at least some other shareholders (other than the Irish Government) would also not support Ryanair's proposal to dispose of the Heathrow slots.

(2025) The Commission is aware of shareholders who have decided not to sell their shareholding in Aer Lingus to Ryanair. For instance, the Irish Airlines (Pilots) Superannuation Scheme (the "Scheme") which holds around 2.25% of Aer Lingus' shares stated that "The Board discussed the current Ryanair bid on Aer Lingus at a Board meeting on 25 July 2012 and resolved not to sell its stake in Aer Lingus to Ryanair"1768. Likewise, the O'Brien group of companies, holding around 3.8% of shares in Aer Lingus, also confirmed that they would not sell their shares to Ryanair (1769).

(2026) It appears most likely that the shareholders who have decided not to sell their shareholding in Aer Lingus to Ryanair would not favour the acquisition of Aer Lingus and would be opposed to the Heathrow slot disposal1770. It also appears likely that such shareholders would be more interested in participating or being represented at the EGM than in the past.

(2027) Furthermore, the Commission is also aware that some shareholders such as the O'Brien Group of companies would vote against a proposal to transfer the London Heathrow slots to IAG (British Airways) if an EGM were convened pursuant to Article 10 of Aer Lingus' Articles of Association. In addition, other shareholders like Ethiad who has a share of 2.98% in Aer Lingus expressed their concerns to the Commission as regards a commitment involving the sale of Heathrow slots to IAG (1771). Such shareholders would likely vote against the disposal of Heathrow slots to IAG.

(2028) Therefore, the Commission considers that it is far more likely that the Irish Government (either alone or with other shareholders representing at least 30.12% of the shares of the votes cast) would be in a position to block a disposal of the Heathrow slots than it is for Ryanair to obtain acceptance of such disposal.

(2029) In any event, even if the Heathrow Transfer Condition were satisfied during the Minimum Period, the Commission considers that the merged entity would not be sufficiently constrained (see Section 10.5.2). In addition if the Heathrow Transfer Condition were satisfied after the end of the Minimum Period, the entry into force of the Heathrow Gatwick Transfer Agreement no longer depends on Ryanair but on IAG which would have to exercise its call option. Furthermore, even if IAG did exercise its call option, it would be under no obligation to use the Heathrow slots on routes to and from Ireland (contrary to IAG's obligation to operate the IAG Routes schedule during the Minimum Period from Gatwick) and the merged entity would not be sufficiently constrained post-Transaction (see Section 10.5.3).

(2030) It should also be pointed out that Ryanair also stated that Aer Lingus may alter its Articles of Association post-Transaction by passing a special resolution of its shareholders. The special resolution may be passed at an EGM, which can be called by the board of Aer Lingus for this specific purpose. A special resolution would need to be passed by at least 75% of the members of the company which are present and entitled to vote at a general meeting (1772).

(2031) Considering the Irish Government's share of 25.11%, Ryanair would probably not be in a position to obtain 75% of the votes cast at such an EGM, even if all other attending shareholders voted in favour of a change of Aer Lingus' Articles of Association. In any event such an amendment of the Articles of Association could not be secured in a timely manner.

(2032) Lastly, Aer Lingus also argued that a commitment by Ryanair to dispose of Aer Lingus'Heathrow slots infringes other legal provisions such as Section 60 of the Irish Companies Act 1963, which prohibits the provision of financial assistance by a company in relation to the purchase of its shares and might also lead to further corporate law breaches (oppression and disregard of the interests of minority shareholders, breach of duty by Aer Lingus' directors or unlawful distribution of Aer Lingus' assets) (1773).

(2033) Considering the other uncertainties as described above and below, the Commission does not find it necessary to take a position on these specific aspects of Irish corporate law. The Commission notes however that the Irish Government informed the Commission that it reserves all its rights to take legal action against a disposal of Heathrow slots (1774).

10.5.1.5. Conclusion

(2034) The Commission finds therefore that Ryanair has not demonstrated to a sufficient degree that it is able to remove the uncertainties as to the implementation of the remedy. Therefore, it does not appear to the Commission with the requisite degree of certainty that the Heathrow Transfer Condition would be satisfied during the Minimum Period and that Ryanair would be in a position to implement the final Commitments in a timely manner insofar as they relate to an exchange of Aer Lingus' Heathrow slots for the Gatwick slots. Therefore it is unlikely that that the Heathrow Gatwick Transfer Agreement would enter into force during the Minimum Period, or even after the Minimum Period.

(2035) If the Heathrow Transfer Condition were satisfied after the Minimum Period, the implementation of this aspect of the final Commitments would lie in the hands of a third party (IAG) and in fact might not have any impact on the three London to Ireland routes as there is no obligation on IAG to operate on these routes.

(2036) Even if the Heathrow Transfer Condition were fulfilled and the Heathrow Gatwick Transfer Agreement did enter into force, such Commitment would in any event not be sufficient as explained in Sections 10.5.2 and 10.5.3.

10.5.2. Merged entity unlikely to be constrained during the Minimum Period

10.5.2.1. Gatwick Lease Agreement

(2037) Under the final Commitments, if the Gatwick Lease Agreement is in effect, IAG would operate 68 weekly frequencies on the Dublin-Gatwick route, 13 weekly frequencies on the Cork-Gatwick route and 7 weekly frequencies on the Shannon- Gatwick route using airport slot pairs owned by the Ryanair Group (the "Ryanair Gatwick Slots") and an additional 2 weekly frequencies on the Dublin-Gatwick Route, 1 weekly frequencies on the Cork-Gatwick Route and 7 weekly frequencies on the Shannon-Gatwick Route using slots currently owned or to be acquired by IAG (the "IAG Gatwick Slots" and, together with the Ryanair Gatwick Slots, the "Gatwick Slots").

(2038) The majority of respondents to the final market test consider that IAG(BA) would constrain the merged entity post-Transaction on the three routes to London Gatwick (1775).

(2039) Table 104 below provides an overview of the frequencies which are currently operated and the frequencies which would be operated post-Transaction by each of Ryanair, Aer Lingus, Aer Arann, IAG and Air France/Cityjet as applicable under the Gatwick Lease Agreement on the Dublin - London, Cork - London and Shannon - London routes: Table 104: Overview of weekly frequencies under the Gatwick Lease Agreement (1776) <TABLE> (2040) As can be seen from Table 104 above, the merged entity would enjoy a significant frequency advantage on the three routes.

(2041) On the Dublin-London route the merged entity would operate more frequencies (178) (1786) than the other two remaining competitors together (138), while on the Cork- London and Shannon-London routes it would operate about two to three times more frequencies than IAG(BA).

(2042) Furthermore, under paragraph 22 of the final Commitments, IAG would use the 132- seat Airbus A319 aircraft or equivalent to operate the IAG Routes Schedule (i.e. the frequencies listed above). Aer Lingus is currently using a combination of 174 seat Airbus A320 aircraft and 212 seat Airbus A321 aircraft to operate on these routes while Aer Arann is using 68 seat turboprops ATR42-500. Therefore, the final Commitments would lead to an overall decrease in capacity on Dublin-London and Cork-London routes (1787). However, on the Shannon-London route, there would be an increase in capacity because of an increase in the total number of frequencies (1788).

(2043) The merged entity would remain dominant in terms of capacity on these routes (1789). On the Dublin-London route, the Parties would maintain a significant share of capacity of above 60% while IAG and Air France/City Jet would have a market share of around 30% and 6% respectively (1790). On the Cork-London route, the merged entity would have a market share of around 80% and IAG(BA) would only have a market share of around 20%. On the Shannon-London route, the merged entity would have a market share of above 75% while IAG(BA) would have a market share of around 25% (1791).

(2044) The majority of respondents to the final market test pointed out that IAG(BA) has a brand which is well known in the Irish market and would be a strong and well established competitor (1792). Other respondents on the other hand emphasized that IAG(BA) has a business model which provides a higher quality & higher cost product to consumers (1793) compared to the product offered by Ryanair and Aer Lingus.

(2045) As explained in detail above in Section 8.9.5.3, IAG(BA) has a different business model than Ryanair and Aer Lingus. Unlike Ryanair and Aer Lingus, IAG (BA) is a full service network airline operating a hub-and-spoke model. It offers both economy and business class on the Dublin-London route and provides an integrated passenger product with additional services such as ticket flexibility, complementary luggage transport, access to airports lounges, frequent flyer programmes, board meals, seat allocation. IAG (British Airways) operates a large network of medium and long-haul services out of London Heathrow and uses short-haul routes including Dublin- London to feed its hub at London Heathrow.

(2046) In addition, while both Ryanair and Aer Lingus operate a true one-way pricing policy, IAG(BA) also sells return tickets and applies "fare fencing" (1794). Bmi used to operate a single cabin on this route (as do Ryanair and Aer Lingus); however, IAG(BA) has now introduced a Club Europe cabin (business class) in line with its other European routes (1795). IAG itself considers that both Ryanair and Aer Lingus "offer a practically identical service in terms of product (typical low cost product / service). Both companies follow the same business model [...]" (1796).

(2047) More importantly, being more focused on business passengers, IAG(BA) could most likely not offer "low-fares" services in competition to Ryanair and Aer Lingus that would substitute the disappearing competitive pressure from Aer Lingus to a sufficient extent. Although it cannot be excluded that IAG (BA) may offer "cheap flights" to point-to-point customers to fill its capacity, it cannot be reasonably expected that IAG(BA) would offer tickets at a price which is comparable to those of Ryanair or Aer Lingus (1797).

(2048) It is also notable that, while IAG(BA) has a strong brand recognition (1798) it has very limited activities on routes out of Ireland (i.e. it only operates on Dublin-London, while Dublin-Madrid is operated by Iberia). Both Ryanair and Aer Lingus have the strongest "city presence" in Dublin and the other Irish routes and therefore an advantage in winning Irish originating customers who still represent a significant share on the three London routes route (1799). Previous interaction between IAG(BA), Aer Lingus and Ryanair also shows that IAG(BA) was unsuccessful in competing with Aer Lingus and Ryanair on Dublin-London Gatwick (1800). Indeed, IAG(BA) exited this route in March 2009 upon entry of Aer Lingus on this route in October 2007.

(2049) Lastly, Ryanair explained in the Form RM at paragraph 77 that, "IAG will simply replace Aer Lingus on these routes and any associated schedules". However, significant uncertainties exist as to how this is going to materialise in practice. Nowhere do the final Commitments or the initial IAG Agreement specify which slots are being made available to IAG(BA) by Ryanair and which slots IAG(BA) contributes. The revised IAG Agreement contains a list of slots but in any event some slots would still have to be procured or retimed by Ryanair. Further uncertainties exist as to the precise schedule. IAG(BA) itself emphasized in relation to the Commitments of 7 December 2012 and of the Commitments of 15 January 2013 that "The "base" which IAG envisages [in the Commitments of 7 December 2012 and in the Commitments fo 15 January 2013] that one of its operating airlines (eg. BA) would set up at Dublin would reflect the operation needs of the LHR/DUB schedule which envisages two early morning departures from Dublin. It is only by having 2 aircraft based and overnighting in Dublin that this can be achieved [...]." (1801) In the final Commitments, there is no commitment by IAG(BA) to at least overnight additional aircraft at Dublin, Cork, or Shannon which might have an impact on the service level.

(2050) Therefore, the Commission cannot conclude with the requisite degree of certainty that, if the Gatwick Lease Agreement were in effect, IAG(BA) would constrain the merged entity to a sufficient degree post-Transaction and therefore that the significant impediment to effective competition would be removed.

10.5.2.2. Heathrow - Gatwick Transfer Agreement

(2051) Under the final Commitments, the Heathrow Gatwick Transfer Agreement is an agreement to transfer the IAG Gatwick slots identified above (namely 2 weekly slots for the Dublin-London route, 1 weekly slot for the Cork-London route and 7 weekly slots for the Shannon-London route) and 7 weekly additional Gatwick slots for the Dublin-London route (1802) to Ryanair in exchange for slots at Heathrow for IAG to operate 70 weekly frequencies on the Dublin-Heathrow route, 14 weekly frequencies on the Cork-Heathrow route and 14 weekly frequencies on the Shannon-Heathrow route and the additional slots for the 7 weekly frequencies to operate the Dublin- London Heathrow route.

(2052) Under paragraph 22 of the final Commitments, during the Minimum Period (6 IATA Seasons), IAG will cause British Airways to operate 70 weekly frequencies on the Dublin-Heathrow route, 14 weekly frequencies on the Cork-Heathrow route and 14 weekly frequencies on the Shannon-Heathrow route (the Heathrow Exchange Slots) (1803).

(2053) The majority of respondents to the final market test considers that IAG(BA) would constrain the merged entity post-Transaction if it were to operate on these routes to Ireland from London Heathrow (1804).

(2054) Ryanair explains in the Form RM that "in any event (even following the disposal of the Heathrow Slots), during the Minimum Period (6 IATA Seasons), the merged entity Ryanair/Aer Lingus will continue to operate 19 weekly frequencies on the Dublin-London Heathrow, 14 weekly frequencies on the Cork-London Heathrow and 7 weekly frequencies on the Shannon-London Heathrow route".

(2055) Assuming that Ryanair will maintain Aer Lingus' frequencies on the Dublin - London Gatwick route and its own frequencies on the other airports, the situation post-Transaction would be similar to the one under the Gatwick Lease Agreement described above.

(2056) In particular, post-Transaction, the merged entity would enjoy a significant frequency advantage as it would have 176 (1805) weekly frequencies compared to IAG and Air France/CityJet who would have 107 and 31 weekly frequencies respectively (1806). Furthermore, the seat capacity on this route would also overall decrease and the merged entity would maintain a significant share of capacity on this route (1807).

(2057) As regards the other two routes, on each of Cork - London and Shannon - London, Ryanair proposes to divest 14 frequencies, which is significantly below the overlap of 23 and 21 frequencies respectively1808. Furthermore, as already mentioned in Section 10.5.2.1, IAG (BA) would operate lower capacity aircraft on these routes, leading to an overall decrease in capacity on these routes and to the merged entity continuing to enjoy a dominant position on these routes.

(2058) Furthermore, and as explained already in Section 10.5.2.1, in terms of business model, IAG (BA) would not be likely to constrain the merged entity to a sufficient degree. IAG(BA), which focusses on business- and connecting passengers particularly for services at London Heathrow, could most likely not offer "low-fares" services in competition to Ryanair that would substitute for the disappearing competitive pressure from Aer Lingus to a sufficient extent (1809). (2059) The Commission notes that in its previous decisional practice (1810), it accepted remedies in which slots were made available to prospective new entrants who intended to begin or increase regular operations on the routes on which the Commission identified serious doubts or a significant impediment to effective competition. However, as regards the final Commitments, it cannot be excluded that IAG(BA) would actually replace some of its own current frequencies between Dublin and London Heathrow with the slots it would obtain under the final Commitments with the ultimate result that post-Transaction the total number of frequencies on the Dublin-London city pair would further decrease.

(2060) IAG(BA) confirmed in a request for information concerning the Commitments of 7 December 2012 that it [...]* (1811). Given its business model, it is likely that these frequencies (or a great majority of these) would be operated from London Heathrow. IAG(BA) is operating, on average, 7 frequencies per day between Dublin and London Heathrow in the current IATA Winter season 2012/13. While the Commission acknowledges that, post-Transaction, there would be an increase in IAG(BA)'s frequencies from 7 daily to around 10 daily, the final Commitments de facto lead to a situation in which IAG would probably only operate Aer Lingus' divested frequencies (70 weekly) and pull out its existing 7 frequencies per day from the Dublin-London route. IAG confirmed that: [...]* (1812) Overall this means that the Transaction would lead to a drop in the total number of carriers from four (Ryanair, Aer Lingus, IAG(BA) and Air France/City Jet) to three (the merged entity Ryanair/Aer Lingus, IAG(BA) and Air France/City Jet), and a drop in the total number of operated frequencies and capacities.

(2061) Therefore, the Commission cannot conclude with the requisite degree of certainty that, if the Heathrow Gatwick Transfer Agreement were in effect, IAG(BA) would constrain the merged entity to a sufficient degree post-Transaction such that the significant impediment to effective competition would not materialise.

10.5.2.3. Heathrow Lease Agreement

(2062) Under paragraph 27 of the final Commitments, if at any point prior to Closing, the Heathrow Lease Condition is satisfied, (i) Ryanair has to notify this fact to IAG within 30 business days, (ii) within 15 business days of notification, Ryanair and IAG procure that they would enter into the Heathrow Lease Agreement (1813) and (iii) the Heathrow Lease Agreement will come into effect the first IATA seasons following such notice (1814).

(2063) The Heathrow Lease Condition is defined in the final Commitments as the termination of the existing short term lease of airport slots at Heathrow by Aer Lingus to a third party.

(2064) The revised IAG Agreement however provides for the possibility that two Heathrow Lease Agreements are entered into and specific conditions apply to each agreement.

(2065) As explained in Section 10.5.1, there is a certain procedure under Article 10 as regards "Disposal Transactions" (1815) in relation to Aer Lingus' Heathrow slots. However, "Short Term Leases" are exempted from this procedure according to Article 10 (a)(i).

(2066) A Short Term Lease is defined as "a lease, temporary swap or temporary exchange (whether or not in return for consideration) of a Slot or a pair of Slots the term of which is no longer than 36 months provided that a series of such leases (whether in favour of the same or different parties) in respect of [the same] Slot or pairs of Slots shall not be regarded for purposes hereof as a Short Term Lease if in the aggregate they exceed 36 months over a 4 year period in respect of the relevant Slot. Provided further that if such a lease is to be put in place in relation to a Slot and there is already a Short Term Lease in place in relation to a Slot, that lease should not be considered to be a Short Term Lease. A proposal to renew a Short Term Lease which has lasted for 36 months (whether continuous or discontinuous) shall be treated as a Disposal Transaction and subject to the provisions hereof."

(2067) In such a case, Aer Lingus need not notify its shareholders of the Short Term Lease or go through the shareholder approval process at all (1816). However, as stipulated above, Aer Lingus cannot have more than one short term lease at the same time so that if Aer Lingus already has one Short Term Lease in place, any further lease of a Heathrow slot(s) (even if it would last no longer than 36 months) will not be considered as a Short Term Lease and will thus fall under the procedure set out in Article 10 (1817). Furthermore, a renewal of a Short Term Lease which has lasted for 36 months would also be treated as a Disposal Transaction (1818).

(2068) However, the text of the final Commitments and of the initial IAG Agreement did not make it clear that the intention is only to enter into a Short Term Lease within the meaning of Article 10. For example, the duration of such lease was not limited to a maximum of 36 months and hence such a Heathrow Lease Agreement would a priori not qualify as a "Short Term Lease" under Article 10. It therefore appeared likely that Ryanair would not be in a position to enter into a Heathrow Lease Agreement under the wording used in the initial IAG Agreement.

(2069) In any event, if Ryanair is proposing only a Short Term Lease, such an agreement would be exempted from the definition of a "Disposal Transaction" under Article 10 of Aer Lingus' Articles of Association only once the existing Short Term Lease were terminated and if it concerned a different slot or pair of slots. Aer Lingus explained that it currently has a Slot Exchange Agreement in place with British Airways in respect of a year-round daily slot pair which is due to terminate at the end of the winter 2012/13 winter season in respect of the winter slot and the 2014 summer season in respect of the summer slot (1819).

(2070) The revised IAG Agreement seeks to clarify these points and limits both leases to 36 months with clear reference to Article 10 of Aer Lingus' Articles of Association.

(2071) The Commission concludes nevertheless that, even if Ryanair and IAG were in a position to enter into the Heathrow Lease agreement (or the First Heathrow Lease Agreement), it would not have a material impact as the merged entity would still benefit from a frequency advantage, remain dominant in term of capacity on the route and face a limited competition constraint from IAG (BA) post -Transaction as explained in detail in Sections 10.5.2.1 and 10.5.2.2 above. The Second Heathrow Lease Agreement does not appear to have an impact on any of the divested routes.

10.5.2.4. No proper misuse clause

(2072) The Commission has in its previous practice accepted commitments which included misuse mechanisms, in order to guarantee that, if slots are misused, in particular in the sense that they are not used on the routes of concern, these slots are given back to the parties and made available to any other interested carrier.

(2073) However, the final Commitments do not provide for such a mechanism during the Minimum Period should IAG decide to stop operating on these routes. The merged entity therefore does not face the threat that if IAG(BA) were to decide to stop operating, another carrier would be given the possibility (and means) to enter on these routes and compete with the merged entity. As these slots would not be made available to other carriers, IAG(BA) also does not face the pressure to operate on these routes during the Minimum Period.

(2074) Under paragraphs 31 to 34 of the final Commitments, a situation of misuse would arise where IAG(BA) during the 6 IATA seasons fails to fulfill its obligation to operate on the three routes (with the frequencies and slots depending on which type of agreement is in place), if it transfers, assigns, swaps, loses etc the Heathrow Slots or the Gatwick Slots. The final Commitments further provide for a cure period. However, the remedy if the misuse is not cured is simply that Ryanair has the right to terminate the Gatwick Lease Agreement, the Heathrow Lease Agreement and if completion of the sale and purchase of the Heathrow Slots had occurred under the Heathrow Gatwick Transfer Agreement, IAG would hand back the Heathrow slots to Ryanair.

(2075) The final Commitments therefore do not include any obligation on Ryanair to make these slots available to other carriers nor do they specify what Ryanair would have to do to ensure that the significant impediment to effective competition on these routes would be removed.

10.5.2.5. No entrant identified

(2076) Paragraph 41 of the final Commitments provides that "During the Minimum Period Ryanair commits to transfer, or to cause Aer Lingus to transfer, to any Applicant, sufficient Slots to operate Frequencies with its own aircraft on the Dublin-London, Cork-London and/or Shannon-London routes, provided that the number of slots transferred does not exceed the Route Overlap Difference on the relevant route". (emphasis added).

(2077) "Applicant" is defined in the final Commitments as "any EU licensed airline which applies to acquire one or more of the Potential Competition Routes Slots from Ryanair in accordance with these Commitments" while "Potential Competition Routes Slots" are defined as the slots required by an Applicant to operate the Potential Competition Routes which are listed in Annex VII as being Dublin- Bologna, Dublin-Bordeaux, Cork-Birmingham, Cork-Munich, Cork-Paris and Dublin-Newcastle.

(2078) The procedure that would be followed is set out in paragraphs 43 to 49 of the final Commitments. Ryanair explains in the Form RM that it offers to divest to "any European airline", 29 weekly frequencies on the London - Dublin route, 9 weekly frequencies on the London - Cork route and 7 weekly frequencies on the London - Shannon route.

(2079) As is clear from the notice on remedies, commitments shall fully specify the substantive and implementing commitments entered into by the parties. This is manifestly not the case here. As it appears from the final Commitments, Ryanair only seems to commit to transfer or to cause Aer Lingus to transfer slots to an EUlicensed airline which applies to acquire one or more of the Potential Competition Routes Slots (1820). Ryanair however explains in the Form RM that these additional slots would be made available to "any European airline" (1821) and therefore there would seem to be no requirement for a carrier interested in the London slots to have also applied to acquire the Potential Competition Routes Slots. These types of inconsistencies point to the lack of clarity as to what is precisely being offered as a commitment by Ryanair. Furthermore, the final Commitments also do not specify at which London airport slots would be made available.

(2080) In any event, the likelihood that slots would actually be taken up by interested carriers is further reduced by the fact that slots are only available for 6 IATA seasons only. During these 6 IATA seasons, a new entrant would be operating in competition with the merged entity, IAG(BA) (who would be operating the divested frequencies) and Air France/Cityjet.

(2081) Lastly, the Commission has not identified any sufficient, likely and timely entry on these routes during its market investigation1822.

10.5.2.6. Conclusion

(2082) Therefore, the Commission is not able to conclude with the requisite degree of certainty that the new commercial structures resulting from the final Commitments are sufficiently workable and lasting to ensure that the significant impediment to effective competition on the three London routes does not materialise during the Minimum Period.

10.5.3. Merged entity unlikely to be constrained after the Minimum Period

10.5.3.1. No incentive for IAG to remain on the routes beyond the Minimum Period

(2083) The Commission has in its previous decisional practice1823 accepted slot remedies when it identified entry which was sufficient, likely and timely. Slots were made available and granted to applicants which showed a sufficient interest in operating on a specific route. To increase the likelihood of entry, slot commitments included the possibility to obtain grandfathering rights only after the slots were used on the routes of concern for a certain period (normally 6 IATA seasons). However obtaining grandfathering rights is subject to Commission approval.

(2084) In the present case, the final Commitments set-up a mechanism whereby IAG would cause British Airways to operate on the three London routes during the first six IATA seasons (the Minimum Period). However, significant uncertainties exist as to IAG's (BA's) commitment to stay on these routes after 6 IATA seasons. Although the final Commitments identify IAG(BA) as the entrant and allegedly guarantee its presence on the routes only for the first 6 IATA seasons, they make it rather unlikely that IAG(BA) would stay on the routes after the Minimum Period, considering IAG(BA)'s business model and its incentives.

(2085) According to Annex 1, clause 2(c) of the initial IAG Agreement, the termination date of the Gatwick Lease Agreement is the earlier of the following two dates (i) the date on which the Heathrow Lease Agreements Slots and the Heathrow Exchange Slots are transferred to British Airways pursuant to the Heathrow-Gatwick Transfer Agreement (meaning that the Heathrow Transfer Condition has been satisfied) or (ii) a date nominated by IAG subject to prior written notice to Ryanair which can however not occur before the end of the Minimum Period (6 IATA Seasons). As explained in detail in Section 10.5.1 above, the Commission considers that it is unlikely that the Heathrow Transfer Condition would be satisfied during the Minimum Period (1824). If, by the end of the Minimum Period, the Heathrow Transfer Condition is not satisfied, IAG can decide whether it wants to collect the break fee of EUR [...]* or instead maintains its lease of the Gatwick slots (1825). If IAG decides to collect the Break Fee, this leads to the termination of the Gatwick Lease Agreement and of the Heathrow Lease Agreement(s) if in effect.

(2086) IAG(BA) considered that "LON/DUB is the single largest O&D market in Europe and 24% bigger than AMS. There is significant transfer potential for BA here [...]*. Therefore, BA would be interested in serving this route [...]*. Similarly both Cork and Shannon are both routes with similar profiles to routes already on the BA network from LHR [...]*. These routes are predominantly p2p with less transfer potential, but both would be worth serving with access to the EI slots." (1826)

(2087) Several respondents to the Commission's final market test pointed to the interest of IAG(BA) to secure Heathrow slots rather than to operate on the Irish routes on a lasting basis (1827).

(2088) As appears from the final Commitments (and the history of the earlier versions of the Commitments), the IAG Agreement and the Form RM, IAG's interest lies in the conclusion of the Heathrow - Gatwick Transfer Agreement which would give IAG(BA) access to Aer Lingus' valuable Heathrow slots. There are a number of provisions to ensure that Ryanair uses its best efforts to enable that the Heathrow Transfer Condition is satisfied as soon as possible (1828). Indeed, while the Heathrow Gatwick Transfer Agreement would lead to a permanent transfer of the Heathrow slots to IAG(BA), the Gatwick Lease Agreement would only be an interim solution whereby IAG(BA) would lease the slots and have the option to decide later on whether it is interested in keeping the slots (and not necessarily for the Irish routes).

(2089) If (as seems most likely) the Heathrow Transfer Condition were not satisfied during the Minimum Period, the Gatwick Lease Agreement could come to an end after 6 IATA seasons. From that moment, as IAG(BA) would no longer have any obligation to operate on the three routes from Gatwick, IAG would consider its own commercial benefit and its own network in its decision as to whether or not to keep these routes from Gatwick or the Gatwick slots. Furthermore, as the final Commitments do not set out precisely which slots would be made available to IAG, uncertainties also exist as to value of these slots and the real interest of IAG in acquiring these slots at the end of the Minimum Period. IAG and Ryanair seem to have tried to address this issue in the revised IAG Agreement which does specify a number of slots, however, a considerable number of slots would still need to be "procured" or retimed by Ryanair.

(2090) Even if IAG(BA) did decide to maintain the lease of these slots and forego the Break Fee of [...]* (which appears to the contrary as an incentive on IAG(BA) to terminate the Gatwick Lease Agreement), there is no commitment and incentive for IAG(BA) to continue to use the Gatwick slots on the Irish routes. The Irish routes do not seem to have been of significant interest for IAG(BA) in the past. IAG(BA) used to operate on the London Gatwick - Dublin route but decided to exit from this route in March 2009 upon entry by Aer Lingus on London Gatwick - Dublin and only recently entered on the London Heathrow-Dublin by operating London Heathrow - Dublin continuing the operation of bmi.

(2091) If IAG(BA) were to be able to exchange its Gatwick slots with Aer Lingus' Heathrow slots under the Heathrow Gatwick Transfer Agreement, IAG's incentive to remain on the routes from Heathrow after the Minimum Period decreases even more.

(2092) Slots at London Heathrow, one of the most congested airports in Europe are extremely valuable, in particular to IAG(BA) which has its main operations at London Heathrow. If IAG(BA) were to obtain the scarce slots from Aer Lingus, it would be free to reallocate these slots to routes which it would find more profitable at the end of the Minimum Period. Indeed, there is a high likelihood that IAG would decide that these slots should be allocated to long-haul routes which are much more profitable than the short-haul routes.

(2093) While it can be expected that IAG(BA) would maintain a certain number of frequencies on the London Heathrow - Dublin route, it can be expected that overall it would decrease the number of frequencies.

(2094) Furthermore, as IAG(BA) considers that the Cork and Shannon routes have less transfer potential, it can reasonably be expected that IAG(BA) would significantly scale back the operations on these routes and reallocate the slots to routes with more transfer potential or to the more profitable long-haul routes. In addition, under the Commitments of 17 October 2012, IAG(BA) would not have entered with its own operations on Cork and Shannon but only through BSAs. IAG(BA) explained that, to start operating with its own aircraft on these two routes, "current market and demand conditions in the Irish market would have to improve [...]*." (1829) Without the prospect of obtaining the prized Heathrow slots, IAG(BA) would not seem to have been interested in these two routes. This casts serious doubts as to IAG(BA)'s actual commitment to these routes.

(2095) Therefore, the Commission does not consider that IAG(BA) has an incentive to stay on these three routes at the end of the Minimum Period, irrespective of whether the Gatwick Lease Agreement or the Heathrow Gatwick Transfer Agreement were in force. It is therefore most likely that IAG would exit these three routes from Gatwick and at least significantly scale back the operations on these three routes from Heathrow at the end of the Minimum Period.

10.5.3.2. No mechanism for entry at the end of the Minimum Period

(2096) As mentioned above in Section 10.5.3.1, grandfathering rights are normally included in slot remedies as a further incentive for new entrants to pick up the slots and start operating on the routes in competition with the merged entity. In the final Commitments, if IAG(BA) decides not to continue operating on these routes at the end of the Minimum Period, there is no mechanism in the final Commitments to ensure that another carrier could obtain the slots.

(2097) A difference between the Commission's previous practice in airline mergers and the final Commitments is in their treatment of grandfathering rights (i.e. the possibility of carriers to use the slots obtained in a remedies framework to operate on other routes than the ones for which the slots were initially allocated). If a carrier applies for slots to operate on a certain route, it means that it is interested in that particular route and it has to submit sufficient evidence to demonstrate its interest. Slots made available as part of a remedies package should therefore enable entry of a carrier on a particular route and help that carrier in exercising a competitive constraint against the merged entity. The Commission's practice has been that, once the slots have been used for a given timeframe (which is decided on a case-by-case basis but is normally 6 IATA seasons), on the basis of the information available, the Commission decides whether that carrier should obtain grandfathering rights. Furthermore, obtaining grandfathering rights does not mean that that carrier would actually exit the routes of concern.

(2098) However in the final Commitments, it is clear that the obligation on IAG(BA) is to stay on the route for only 6 IATA seasons. At the end of this period, the Commission has no influence over the fate of the Heathrow or Gatwick slots. As explained above, the Commission considers that it is likely that these slots (or at least a part of these) would be reallocated to other more profitable routes. Therefore, the final Commitments are not sufficiently workable and lasting to ensure that the impairment of effective competition does not materialise on the three London routes.

10.5.3.3. Conclusion

(2099) Therefore, the Commission is not able to conclude with the requisite degree of certainty that the new commercial structures resulting from them are sufficiently workable and lasting to ensure that the significant impediment to effective competition on the three London routes does not materialise after the Minimum Period.

10.5.4. Complex and not sufficiently clear-cut commitments regarding the three London routes

(2100) The final Commitments raise significant concerns as to their complexity. They set up a framework of conditions which would lead to one or the other agreement being entered into. Furthermore, there are inconsistencies between the text of the final Commitments and the IAG Agreement which were also pointed out by IAG (1830). For example paragraph 27 of the final Commitments refers to "Closing" which is defined as the date of completion of the Offer whereas the corresponding clause 4.4. of the initial IAG Agreement refers to the "Date of Completion" which is defined as the completion of the sale and purchase of the Heathrow Slots under the Heathrow Gatwick Transfer Agreement. Also, while the IAG Agreement also includes an obligation in clause 6.2 on Ryanair to operate the Heathrow slots in such a manner as to maintain their historic precedence, paragraph 23 of the final Commitments text only refers to an obligation on IAG to maintain the precedence of the Heathrow Slots and the Gatwick Slots.

(2101) Furthermore, in some places the final Commitments contain what are plainly mistakes and are not clear as to their precise meaning. For example in paragraph 27 the reference under point (ii) should be to "the Heathrow Lease Agreement" and not to the "Heathrow Gatwick Transfer Agreement". It is also not clear what is the definition of "Applicant" as in the final Commitments text this is referred to as an EU carrier which applied for Potential Competition Routes Slots, whereas the Form RM only refers to "an European carrier".

(2102) The revised IAG Agreement seeks to clarify some of the points. However, the Commission notes these revisions came at a very late stage in the procedure. These changes also create further inconsistencies between the text of the final Commitments and the text of the IAG agreement as revised.

(2103) Considering the complexity of the framework, such inconsistencies, mistakes and vague language, and a revised IAG agreement received at a very late stage, would be likely to lead to disputes at the implementation stage of the final Commitments therefore giving rise to serious uncertainties as to the likelihood of an efficient and timely implementation of the final Commitments.

(2104) The Commission therefore considers that the final Commitments are not sufficiently clear to be implemented in an efficient and timely manner post-Transaction.

(2105) Furthermore, as with the Flybe Agreement, the fast track dispute resolution mechanism in the Commitments does not apply to the IAG Agreement (See Section 10.4.3.3).

(2106) The complexity of the final Commitments, the inconsistencies between the final Commitments, the Form RM and the IAG Agreement and the settlement of disputes arising from the implementation of the IAG Agreement ultimately through litigation in English Courts raise doubts as to whether the final Commitments would be implemented in a timely manner.

10.5.5. Overall conclusion on the slot divestiture on the three London routes

(2107) The Commission considers that the final Commitments do not eliminate the competition concerns entirely and are not comprehensive and effective from all points of view. Furthermore, the Commission is not able to conclude with the requisite degree of certainty that it will be possible to implement them and that it will be likely that the new commercial structures resulting from them will be sufficiently workable and lasting to ensure that the significant impediment to effective competition will not materialise. Moreover, the Commission cannot clearly determine that the final Commitments, once implemented, fully and unambiguously resolve the competition concerns identified in this Decision.

10.6. Commitments relating to overlap routes where Aer Arann operates

10.6.1. Description of the commitments

(2108) According to the final Commitments (1831), Flybe would be able to satisfy its obligation to operate on the Flybe routes currently operated by Aer Arann (1832) by assuming the Aer Arann franchise or by entering into a new franchise agreement with Aer Arann providing for Aer Arann to operate the same number of frequencies on those routes and on terms no less favourable to Aer Arann than the Aer Arann franchise, in each case subject to agreement with Aer Arann. In that case, Ryanair will use reasonable endeavours to enable Flybe to assume or enter into such franchise agreement, including by terminating the Aer Arann franchise or assigning it to Flybe.

(2109) In addition, Ryanair commits to allow Aer Arann to be hosted, upon Aer Arann's request, in Aer Lingus' Frequent Flyer Programme (1833).

10.6.2. Ryanair's views

(2110) Despite the fact that Ryanair contests that its services overlap with those of Aer Arann, Ryanair commits to use reasonable endeavours to enable Flybe to assume or enter into such franchise agreement, including by terminating the Aer Arann franchise or assigning it to Flybe so that Aer Arann would become a franchisee of Flybe. According to Ryanair (1834), the final Commitments preserve the viability of Aer Arann as an independent competitive force in Ireland.

10.6.3. The Commission's assessment

10.6.3.1. The final Commitments are unclear

(2111) The final Commitments as regards the six routes currently operated by Aer Arann are unclear and ambiguous. While Ryanair acknowledges that assigning or terminating the franchise agreement are "in each case subject to agreement with Aer Arann", Ryanair commits only that it will "use reasonable endeavours" for this commitment to happen by "terminating the Aer Arann Franchise or assigning it to Flybe." (1835) The Flybe Agreement is also equally unclear in this respect as it states that: "With respect to the [routes] that fall within the Aer Arann Franchise, in connection with the Hive- Down, [Ryanair] undertakes to use reasonable endeavours to enable [Flybe Ireland] to operate on those routes by (at [Flybe's option] acceding to the Aer Arann Franchise or entering into an equivalent franchise agreement with Aer Arann..." (1836)

(2112) Such unclear drafting raises concerns as to whether Flybe actually commits to enter on these routes with its own operations or whether it only commits to taking over the Aer Arann franchise, should Aer Arann agree to it.

10.6.3.2. Ryanair cannot modify, terminate or assign the Aer Arann franchise agreement in a timely manner

(2113) The term of the franchise agreement between Aer Lingus and Aer Arann is 10 years. Subject to a one-off termination right in 2018, the franchise agreement would be in effect until 31 December 2022 (1837).

(2114) Furthermore, Ryanair would not be entitled to modify or terminate unilaterally the scope of the franchise agreement post-Transaction as any changes to or amendments of the franchise services are subject to mutual agreement of Aer Lingus and Aer Arann (1838). This is acknowledged by Ryanair itself when it states that assigning or terminating is "in each case subject to agreement with Aer Arann." (1839)

(2115) Specifically, Appendix A to the franchise agreement provides the list of routes which the franchise agreement covers and Aer Arann or Aer Lingus cannot unilaterally change/terminate them: "There can be no changes made to this Appendix A and/or the flights the subject matter of the Services set out herein unless such change is evidenced in writing and agreed by both Aer Lingus and Aer Arann in advance of such change." (1840)

(2116) Similarly, an assignment of the franchise agreement to Flybe also cannot occur without Aer Arann's consent. Indeed, Clause 12 of the franchise agreement provides that: "Neither party may assign or otherwise transfer all or any of its rights and/or obligations hereunder without the prior written consent of the other party (such consent not to be unreasonably withheld)." (1841)

(2117) Aer Arann has also confirmed to the Commission that the franchise agreement is non assignable and in any event that Flybe Ireland or Ryanair would not be a realistic substitute due to their different distribution model (1842).

(2118) Finally, Aer Arann has clearly indicated that it will not consent to any form of termination or the removal of routes as the franchise agreement is fundamental to its anticipated return on investment (1843). These routes include Aer Arann's most important services (1844).

(2119) In conclusion, the final Commitments are not likely to be implemented in a timely manner as far as the franchise agreement is concerned (either by way of termination or assignment). Moreover, the Commission cannot clearly determine that the final Commitments, once implemented, fully and unambiguously resolve the competition concerns identified in this Decision as regards the Aer Arann routes.

10.7. Commitments relating to potential competition routes

(2120) The Commission notes that it decided to exceptionally examine the final Commitments although these did not clearly resolve the competition concerns previously identified and there was a need for a further market test (1845).

10.7.1. Description of the commitments

(2121) According to the final Commitments (1846), during the first 6 IATA seasons and at all times thereafter, Ryanair commits to transfer, or to cause Aer Lingus to transfer, slots for the potential competition routes to any Applicant (1847), to enable the relevant Applicant to operate frequencies with its own aircraft on any of the potential competition routes, up to a maximum number of frequencies as identified on a routeby- route basis in Annex VII. Applicants shall comply with the procedure for slot divestiture to obtain the slots.

10.7.2. Ryanair's views

(2122) According to Ryanair and, notwithstanding their view that the Commission's concerns regarding non-overlap routes are not valid (1848), Ryanair's revised remedies package addresses these potential competition concerns in two ways (1849).

(2123) First, Flybe Ireland, as a strong competitor with base operations in Ireland, will be well placed to enter any routes that will be operated by the merged entity, including those identified by the Commission as giving rise to potential competition concerns. This remedy ensures that, at the very least, the pre-Transaction likelihood of entry on these routes is maintained.

(2124) Secondly, Ryanair offers slot divestitures to any third party on the 6 non-overlap routes with regard to which the Commission maintains its potential competition concerns.

(2125) Ryanair considers that Flybe, able to rely on its fleet of 97 aircraft (in addition to those added as part of the remedies package), aircraft leases or aircraft purchases, will be exceptionally well placed to launch services in competition with Ryanair/Aer Lingus. (1850) Flybe will continue to exert the same competitive constraint that Aer Lingus provides today on all 6 potential competition routes. (1851)

(2126) In addition, Ryanair considers that it addresses the Commission's remaining potential competition concerns on 6 non-overlap routes by offering slot divestitures to any third party. (1852) Ryanair will offer slot divestments on these routes for an indefinite period of time, enabling new entrants (including the Divestment Business) to operate up to the same number of frequencies as are currently operated by Aer Lingus or Ryanair. (1853)

10.7.3. The Commission's assessment

(2127) First, taking also into account the previous assessment of Flybe, the Commission considers that Flybe Ireland would not exert a sufficient competition constraint on the merged entity, at least on a significant number of the Divestment Business routes. Therefore, despite having a base at Dublin and one aircraft based operationally at Cork airport, Flybe is unlikely to be able to exert a sufficient competition constraint on the merged entity on the potential competition routes.

(2128) Secondly, the evidence collected in the market investigation indicates that no entry or expansion projects of competing carriers can be considered as timely, likely and sufficient enough to constitute a competitive constraint on the merged entity on the six routes of potential competition (1854).

(2129) Therefore, the Commission concludes that the proposed commitments with respect to the potential competition routes are not able to remove the identified significant impediment to effective competition, and, thus, could not render the proposed concentration compatible with the internal market.

10.8. Conclusion on the final Commitments

(2130) Based on all available evidence, including the results of the final market test, the Commission considers that the final Commitments would not likely lead to the entry of new competitors able to exert sufficient competitive constraint on the merged entity.

(2131) The final Commitments do not allow the Commission to conclude, with the requisite degree of certainty, that it would be possible to implement them in a timely manner and that they would be sufficiently workable and lasting to ensure that the impairment of effective competition which those Commitments is intended to remove would not be likely to materialise in the relatively near future.

(2132) The Commission cannot clearly determine that the final Commitments, once implemented, fully and unambiguously resolve the competition concerns identified in this Decision.

(2133) It is concluded that the final Commitments offered by Ryanair are not able to remedy the identified significant impediment to effective competition, and, thus, cannot render the Transaction compatible with the internal market.

11. CONCLUSION

(2134) The Commission concludes that the Transaction is likely to significantly impede effective competition in the internal market or in a substantial part thereof within the meaning of Article 2(3) of the Merger Regulation as a result of the creation of a dominant position of Ryanair and Aer Lingus on 46 routes from and to Dublin, Shannon, Cork and Knock.

(2135) For those markets where other scheduled carriers operate and where charter companies operate, the Transaction is likely to significantly impede effective competition in the internal market or in a substantial part thereof within the meaning of Article 2(3) of the Merger Regulation as a result of the elimination of the particularly close competitive relationship between Ryanair and Aer Lingus and thus of the important competitive constraints that both airlines exert on each other pre- Transaction.

(2136) In addition, the Transaction is likely to significantly impede effective competition in the internal market or in a substantial part thereof within the meaning of Article 2(3) of the Merger Regulation by the elimination of a credible potential entrant on 6 routes from and to Dublin and Cork.

(2137) The concentration must, therefore, be declared incompatible with the internal market and the EEA Agreement pursuant to Article 8(3) of the Merger Regulation and Article 57 of the EEA Agreement.

HAS ADOPTED THIS DECISION:

Article 1

The notified operation whereby Ryanair would acquire sole control of Aer Lingus within the meaning of Article 3(1)(b) of the Merger Regulation is hereby declared incompatible with the internal market and the EEA Agreement.

Article 2

This Decision is addressed to: Ryanair Holdings plc Ryanair Corporate Headquarters Dublin Airport Ireland - County Dublin Done at Brussels, 27.2.2013

Annex I - Price Correlation Analysis

(1) This Annex provides quantitative support for the conclusions reached by the Commission on market definition, and in particular concerning the substitutability of close-by airports.

(2) In the first section the Commission explains the relevance of the price correlation analysis for the present case and in the second the Commission discusses the limitations of the price correlation technique for the purposes of market definition. The third section explains the methodology followed by the Commission and discusses how this methodology addresses the general limitations of price correlation analysis. The results of the analysis are provided in the fourth section and the final section contains the conclusions of the Commission.

Explanation of price correlation and relevance in the present case

(3) The Commission has undertaken a price correlation analysis. (1)

(4) The correlation analysis was performed on the average monthly fares across a number of routes between Dublin, Cork and Shannon (as well as Knock and Belfast/Derry) on one hand and European destinations with potential horizontal overlaps on the other, for the purposes of market definition.

(5) Price correlation analysis measures the contemporaneous adjustments in the price changes. If the prices of the two airport pairs move perfectly in line with each other, the correlation coefficient is equal to one and if there is no relationship the correlation is equal to zero. The higher the correlation coefficient between two airport pairs the higher the degree of co-movement between them.

(6) Price correlation analysis is based on the idea that in the absence of common demand- or supply-side shocks the extent of co-movement in the prices in different airport pairs will provide information about the substitutability amongst them. This is because price competition results in an alignment of prices in the two airport pairs if they belong to the same relevant geographic market. The economic intuition is a simple arbitrage argument. For example for two airport pairs that were to belong to the same market, if an airport pair specific cost shock were to lead to a price rise in the airport pair i, then consumers would substitute their purchases away from this airport route and travel instead in airport pair j. As a result of the increase of the demand in airport pair j the price in this route should increase and so a co-movement (measured by a positive correlation) between the changes in the prices in the two airport pairs is expected.

(7) Hence, if two airport pairs are considered to be part of the same market then the price movements are expected to be correlated. However, absolute price convergence is not necessary for products or services to belong to the same geographical market.

Limitations

(8) The Commission acknowledges that price correlation analysis is not a perfect measure for the purposes of market definition and the results of this analysis should be interpreted with caution and in parallel with the qualitative evidence collected during the market investigation.

(9) Price correlation may lead to false positives (that is to say, correlation coefficients indicating a single market while the underlying series do not belong to the same market). A main caveat of correlation analysis is that common shocks (such as changes in the fuel price) can induce spurious correlation and therefore high correlations could be driven by factors which can cause a co-movement but are still unrelated to consumer and producer substitution. Similarly, two price series may be found to be correlated only because each of them has a trend (again, leading to spurious correlation). Correlation techniques are spurious in case they are applied to time series that are non-stationary (that is to say, a series in which when a shock occurs at a certain point in time has a persistent effect in the following periods while in a stationary series a shock does not have such a persistent effect). Also if the price series are serially correlated the resulting correlation between the two series will be affected by the relation of each of the price series with their past values and may lead to erroneous results (notably in case of non-stationary series). To address these concerns, correlation analysis is performed on the following: i) the first difference of the prices and ii) on series that are stationary and iii) by including lagged values of the revenues in the estimations.

(10) At the same time, price correlation may also lead to false negatives. This outcome may arise as price correlation only measures contemporaneous correlation (while price changes may take longer to materialise). It is unlikely though that such effect would be material for airline cases, especially in cases where correlation is measured at monthly average fares (that is to say, relatively long period).

(11) Another critique of correlation analysis is that there is no generally agreed level for the threshold which defines whether the series are sufficiently correlated to belong to the same geographical market. To assess whether the prices are sufficiently correlated, it is typical to use as a benchmark some other correlations for which one has strong (a priori) indications that they belong to the same market. As in the 2007 Decision, the Commission has taken as a first benchmark the average correlation on the routes where both Aer Lingus and Ryanair fly to the same airport, as for these routes there is a strong presumption that they belong to the same market. If the correlation lies above the benchmark this is an indication that the two products lie in the same relevant market. Therefore, the most useful source of information from this correlation analysis comes from a comparison of correlations among different pairs.

Methodology

(12) For the purpose of this Decision, correlation analysis has been carried out on the data provided by Ryanair and Aer Lingus on average monthly prices across a number of airport pairs between Dublin, Cork and Shannon (as well as Knock and Belfast/Derry) on one hand and European destinations with the horizontal overlaps as identified in the Decision opening the proceedings on the other. (2) These horizontal overlaps arise either because the Parties fly to the same airport or because they fly to different airports in the same geographic area. The purpose of the analysis is to test whether for the airport-pairs where the two companies do not fly in the same geographic area the correlation of prices does not differ from the instances where both Parties fly to the same airport.

(13) Given that the analysis is performed on a monthly basis, for a number of routes in which both Parties did not operate for a sufficient number of months simultaneously, there are not sufficient observations. For example, this is the case for the Dublin- Stockholm and Dublin-Warsaw/Warsaw Modlin routes.

(14) The Commission has followed a number of steps (similar to the methodology followed in 2007) to ensure that the limitations of the correlation analysis are addressed in the case at hand.

Stationarity tests

(15) Correlation analysis can be meaningfully performed only on stationary series (as discussed above). When instead the series are non-stationary, correlation analysis would result in spuriously high correlation (that is to say, false positive). When price series are non-stationary (of order one) correlation analysis can be performed on the first price difference of the price series, <Formula> .

(16) When performing the analysis on the first difference of the price series the correlation analysis shows that the price changes themselves, rather than the price levels, are potentially correlated, and therefore relative convergence is tested. A "high" correlation of price changes across products or services for different geographical areas still indicates that these prices co-vary and hence suggests products belong to the same geographical market.

(17) To statistically test whether the series are stationary or not the Commission has computed the Augmented Dickey-Fuller ("ADF") test for each price series to see whether each price series is non-stationary. Appendix A reports the finding of the individual raw price series but also of the residuals of the partial correlation analysis (for ease of reading the ADF results of only the raw series and of the FD-2 strategy described below are reported in Appendix A). (3)

Controlling for seasonality, common input costs (fuel price) and trend - Partial correlation

(18) As explained above, if common shocks exist they may cause a false positive and the products may appear more correlated than they really are and therefore the Commission would run the risk of drawing erroneous conclusions about the size of the relevant market.

(19) For stationary time series, a better reflection of whether there is a co-movement of prices is the partial correlation coefficient, not the ordinary correlation coefficient. The partial correlation coefficient differs from the ordinary correlation coefficients, as the former, but not the latter, has been purged of the influence of common factors.

(20) In the present case, the partial correlation coefficient is calculated by regressing each price series on time series of the price of fuel using Ordinary Least Squares ("OLS") and seasonal dummies and then measuring the correlation of the regression residuals.

(21) As both Ryanair and Aer Lingus monitor the Jet Kerosene Cargoes CIF NWE (4) this series was used as a proxy for common fuel cost. The Commission also corrects for seasonality by introducing quarterly dummies in the regression. This is done in order to account for possible seasonal variation in the data. It may have been relevant to include several other factors, but due to lack of such data an extended analysis is omitted in the present exposition. In general, the partial correlation coefficients should be preferred to the standard simple correlations.

(22) The residuals from the regression of the price series on the fuel price and the quarterly dummies are considered as a measure of prices purged of common factors. The partial correlation coefficient is the ordinary correlation coefficient between the residuals from two such regressions.

(23) Furthermore, in the Commission's analysis is the Statement of Objections some series in question have been identified as trend stationary or trend non-stationary. In the present analysis the Commission has also included a trend in the partial correlation regression to control for this trend, addressing a comment raised by Ryanair (see Ryanair's comments on Commission's analysis). (5)

Serial correlation

(24) The Q test shows that in some instances the residuals of the regression of the price series on fuel price and the quarterly dummies are serially correlated, that is to say, past value of the price series influence the current level. By including lagged values of the prices in the estimation (see partial regression in the correlation on price levels), the Commission addresses the serial correlation issue.

Steps in the analysis

(25) As in the 2007 Decision, the Commission followed two complementary approaches in the present case. Both approaches generate similar results and do not affect the conclusions.

I. Correlations on price levels

- Plot the candidate price series

- Run the ADF test of stationarity on individual price series (results reported in Appendix A)

Compute the raw price correlation coefficient (RAW in Appendix B)

- Regress the price series on the common influence variables (fuel costs and seasonal dummies) and a trend

- Run the ADF test of stationarity on the residuals of the above regression

- Correlate the residuals from the above regressions (SC in Appendix B)

- Regress the price series on the common influence variables (fuel costs and seasonal dummies), on the lag of prices and a trend

- Run the ADF test of stationarity on the residuals of the above regression

- Correlate the residuals from the above regressions (SCL in Appendix B)

II. Correlations on first differences

First strategy

- Take the first difference of the price series and the common costs

- Run the ADF test of stationarity on the first difference of the price series

- Correlate the first difference of the price series (RAWFD in Appendix B)

- Regress the first difference of the price series on the first difference of the common costs, seasonal dummies and a trend

- Run the ADF test of stationarity on the residuals from the regression above

- Correlate the residuals from the above regression (FD1 in Appendix B)

Second Strategy

- Regress the price series on the common influence variables (fuel costs and seasonal dummies) and a trend

- Take the first difference of the residuals from the regression above

- Run the AD test of stationarity on the residuals of the above regression

- Compute the correlation coefficient of the first differences of the residuals (results in column FD2 in Appendix B)

(26) The analysis was performed on the net fare revenues from Aer Lingus and the gross revenue of Ryanair (with the baggage fees included). (6)

Results

(27) Overall, it is concluded that the correlation analysis is consistent with the qualitative evidence collected during the market investigation. Also, the results were remarkably similar to the results obtained in the 2007 Decision. The scope of the analysis in the present case, however, is much larger than in the 2007 Decision as there are many more overlap routes during the period under consideration (that is to say, from November 2004 to July 2012). Many of the routes for which correlations could be measured are not actual overlaps, but were overlaps at some moment in time during this period.

(28) The correlation coefficients provided are all the estimates obtained from the data that the Commission collected from Ryanair and Aer Lingus. This relates to all related airport-pairs where sufficient time series data was available. However in a significant number of cases the number of observations is very small (typically cases with less than 20 observations). In these cases no large significance should be attached to the correlation coefficients.

(29) There are four types of price correlations. The first category is when both Aer Lingus and Ryanair serve the same airport. The second when Ryanair serves two airports within the same city pair (for example in London where Ryanair serves London Gatwick, London Stansted and London Luton). The third when Aer Lingus serves two airports within the same city pair (similarly in London Aer Lingus serves both London Heathrow and London Gatwick). The fourth category refers to situations where Ryanair and Aer Lingus serve different airports within the same city pair (for example Ryanair serves Bratislava airport while Aer Lingus serves Vienna airport).

(30) The results of the correlation analysis are provided in Appendix B, Table 1.(7) The first column indicates the city pair (the main city close to the airports that are in the same area). The two next columns refer to the two series whose correlation in reported in the last 6 columns (columns 5-10). In the first row of the table, "EIFAO" in column 2 and "FRFAO" in column 3 indicate that the correlations provided in columns 5-10 are the correlations between Aer Lingus' fares to Faro and Ryanairs' fares to Faro (from Belfast/Derry). The number of observations for the given pair is also provided in column 4. The correlations are provided in the same sequence as explained above. Therefore column 5 ("raw") shows the raw correlations of the two price series. Column 6 reports the partial correlation coefficients after controlling for seasonality and the fuel price ("sc") and column 7 the partial is the partial correlation taking also into account a lagged dependent variable ("scl"). The last three columns relate to the raw correlation of the first difference of the price series ("rawfd"), the correlation coefficients of the first difference of the price series after regressing on the first difference of fuel price and seasonal dummies ("fd1") and the partial correlations from the second strategy ("fd2") respectively.

(31) The results of the correlation analysis indicate that the correlations are relatively stable in the different specifications. This finding implies that the amount of correlation observed in the raw data is not explained by common variation in the fuel costs and seasonal dummies. In cases with very few observations, there are not sufficient degrees of freedom to consistently compute the correlations and therefore for some of these cases there is a significant difference in the correlations computed across specifications. There are some exceptions to this, especially in cases of very few observations. As mentioned above, the results obtained bear great similarity to the correlations obtained in the 2007 Decision - and this applies to both relatively high correlations as well as to low correlations. Notably, as in the 2007 Decision, the correlation between Ryanair's fares to Charleroi and Aer Lingus' fares to Brussels is very low (and in some specifications not significant). Routes with very high correlations between the fares of the two air carriers (that is to say, either type 1 or type 4 correlations) across all specifications include the Dublin-Krakow, Dublin- Faro, Dublin-Poznan and Dublin-Warsaw routes. (8)

(32) In what follows the Commission will refer to the correlations in the last column of Table 1 in Appendix B, namely the correlation in first differences controlling for common costs, seasonality and trend. The Commission follows this approach as for this specification there are the smallest number of non-stationary fares (the fares for this last specification are non-stationary for only a few series as shown in Appendix A, column 5).

(33) Following the same approach as in 2007 the Commission uses as a first benchmark the average correlation on the routes where both Aer Lingus and Ryanair fly to the same airport. This benchmark has been useful for routes where no better benchmark could be used. A better benchmark has been used in cases where within one city pair both carriers fly to the same airport in addition to other airport(s), which arises for example for routes to London or Manchester, as explained below).

(34) There are 30 instances in which both carriers fly to the same airport. However, for Dublin-Fuerteventure, Dublin-Gatwick, Dublin-Bristol, Dublin-Nice and Dublin- Prague at least one of the series is non-stationary and therefore these pairs are dropped from the analysis. Among the remaining 26 such pairs, the correlation coefficients range from 0.29 in the Dublin-Liverpool route to 0.96 in the Dublin- Poznan route. The average correlation in these routes is around 0.79. (9) Most importantly the weighted average of the correlations (that is to say, each correlation coefficient is weighted by the revenues that were generated by both carriers in a given route) is 0.66. This weighted average is lower than the simple average as routes with significant traffic (such as Dublin-Birmingham and Dublin Manchester routes) have relatively low correlation. Therefore, one would expect that for thicker routes lower correlation threshold applies.

(35) Concerning routes in which Ryanair contests the market definition the Commission notes the following: in the case of Toulouse-Carcassonne, Bilbao-Santander, Barcelona-Reus, Barcelona-Girona, Alicante-Murcia, (10) Rome Fiumicino- Rome Ciampino, Milan Bergamo-Milan Linate, Milan Bergamo-Milan Malpensa, Vienna- Bratislava the correlations are higher than even the simple average (that is to say, higher than 0.79). The pair Venice-Treviso also has correlation above this threshold.

(36) In one other route, Paris Charles De Gaulle- Paris Beauvais, the correlation is between the weighted and simple average benchmarks of 0.66 and 0.79. As this is a relatively thick route it is useful to compare the computed correlation with the weighted average of the routes where both Parties fly to the same airport (as they are more likely to be similar to such thick routes).

(37) The Dublin- Glasgow/Glasgow Prestwick route is slightly below the 0.66 benchmark; but only marginally at 0.64. Finally, the Frankfurt-Frankfurt Hahn pair has a correlation of 0.59. As these figures are significant, and not significantly lower than the 0.66 benchmark the Commission considers that a conclusion can only be taken after taking into account other qualitative evidence.

(38) The analysis does not provide any significant information on the Stockholm airport's substitutability as there are only 5 observations in the given route.

(39) Similarly, for the Belfast-Derry routes (that is to say to Faro and to London) the Commission considers that the quantitative analysis does not provide any significant evidence to conclude either that Belfast and Derry belong to the same market or not, given the relatively small number of observations in the relevant routes. (11)

(40) As regards London, Manchester and Birmingham "within route" benchmarks can be obtained. Such benchmarks arise as in these cities both carriers fly to the same airport. However, for Dublin-London this is not possible as the series of Aer Lingus' fare to London Gatwick is non-stationary.

(41) Concerning Manchester, the relevant benchmark is relatively low as both Aer Lingus and Ryanair fly to both Liverpool and Manchester with relatively weak correlations of (non-significant) 0.29 and 0.43 respectively. The correlation between Ryanair's Liverpool fares and Aer Lingus' Manchester fares is slightly weaker, with a correlation of 0.34. At the same time, the within carrier correlation between these two airports is very high. Ryanair's Manchester and Ryanair's Liverpool fares have correlation of 0.92 and Aer Lingus fares' correlation in these two airports is 0.74. Therefore both air carriers appear to view the two airports as substitutable. If the two airports were not related one would not expect even Ryanair's/Aer Lingus' own prices to be highly correlated across the two different airports as even Ryanair/Aer Lingus would price independently in each of these airports. (12) Concerning Leeds Bradford, the evidence is not conclusive as amongst the Parties' fares the correlation is much weaker and non significant (-0.11 for Leeds-Liverpool and 0.17 for Leeds- Manchester). However the correlation of Ryanair's fares between Leeds Bradford and i) Manchester and ii) Liverpool is much higher, both at 0.74.

(42) Concerning Birmingham, the relevant benchmark is between Aer Lingus and Ryanair's fares in Birmingham airport, whose correlation is 0.45. Ryanair's own fares between Birmingham and East Midlands is very high at 0.86 implying that at least Ryanair is likely to consider these two airports as substitutable. The correlation between Aer Lingus fares in Birmingham and Ryanair's fares in East Midlands is lower at 0.35.

(43) Concerning London (ex-Cork) the quantitative evidence is not conclusive on the substitutability of all London airports. Both carriers fly to London Gatwick with correlation of 0.60, which is the within route benchmark. The correlation of fares of the two air carriers flying in different airport pairs are somewhat lower (for example between Aer Lingus' Gatwick fares and Ryanair's Stansted fares is 0.51 ). At the same time Ryanair's fares amongst different airport pairs, namely Gatwick and Stansted is very high (above 0.89) and therefore at least Ryanair regards these airports as substitutable. The correlation of Aer Lingus' London Gatwick and London Heathrow fares is also slightly higher than the benchmark in the London case (0.63 vs. benchmark of 0.60) High correlations amongst Ryanair's Gatwick and i) Luton and ii) Stansted fares are also seen ex-Shannon (0.85 and 0.9 respectively). Finally, ex-Knock the picture changes slightly as even the between air carrier's correlations are higher. The correlation between Aer Lingus London Gatwick and Ryanair's fares in Luton and Stansted are much higher (0.72 and 0.74 respectively for Knock). This lends support to the view that ex-Knock it is more likely that destination airports that may be further away would be considered substitutable.

(44) As in 2007 the data for the route Dublin-Brussels/Charleroi is odd as the Commission obtains a non-significant (and negative) correlation. This is also inconsistent with the results of the market investigation and the qualitative analysis on this airport pair.

(45) The Commission notes that as at least one of the series for Dublin-London, Dublin- Munich/Memmingen, and Dublin-Bristol/Cardiff (13) is non-stationary the correlations obtained for these pairs may be biased and the Commission cannot, therefore, comment on their magnitude.

(46) The Commission has also confirmed that the results carry forward when computing the correlation between the Ryanair fares and the Aer Lingus fares for nonconnecting passengers.

Ryanair's comments on Commission's analysis

(47) Ryanair raises several criticisms on the correlation analysis performed by the Commission.

(48) The first criticism is that the model lacks robust economic underpinning and neglects a range of relevant factors such as the nature of competition, the quality of service, the impact of different passenger types notably in terms of connecting vs. nonconnecting passengers, and the presence of competitors. According to Ryanair, as certain variables are shown to influence prices in the regression analysis (as shown in Annex III to this Decision), the Commission's correlation analysis should have included such variables (notably, the presence of other airlines). (14)

(49) The Commission has employed a standard procedure for undertaking correlation analysis. The Commission has recognised that the correlation analysis is not a flawless exercise, as discussed in detail in the Commission's analysis above, however, the Commission's procedure takes steps to address the shortcomings of the correlation technique.

(50) The rationale behind price correlation is to test how the fares of one company respond to the changes in fares of another company. To avoid the issue of spurious correlation the source of information that identifies the correlation coefficients should be provided by asymmetric shocks (that is to say, it should not be common shocks that drive the correlation). For this reason the Commission has controlled for common shocks in the partial correlation analysis. The Commission's analysis has controlled for common cost shocks (fuel price) and common demand shocks (seasonal dummies). In a correlation analysis asymmetric shocks should not be stripped from the analysis - contrary to a regression analysis that has a different scope (that is to say, to identify the relationship between two variables controlling for all possible other factors). For example, it is not obvious whether the entry of a third carrier would constitute a common shock to both companies, as it may affect the two carriers differently (as shown also in the regression analysis). To confirm the robustness of its correlation analysis against this criticism of Ryanair, the Commission has introduced in the regressions of the partial correlations the total frequencies offered by all airlines at the destination airport (a further control for common demand shocks that may influence Aer Lingus' and Ryanair's average fares). The results of the stationarity analysis are presented in column 6 of the Table in Appendix A and the correlations are presented in Appendix B, Table 2. The results remain largely unaffected. On the stationarity of the series, the only significant difference is that Aer Lingus' fares in London Gatwick become stationary and therefore an analysis of the London routes is meaningful. (15)

(51) On the correlation analysis, the average arithmetic benchmark is 0.76 and the weighted benchmark 0.63. (16) Overall, the results are similar. However the Commission notes that for Dublin- Glasgow/Glasgow Prestwick the correlation is above the weighted benchmark (0.64 vs 0.63) and the correlation in Frankfurt/Franfkurt Hahn is very close to the benchmark (0.61). Last, the correlation in Toulouse/Carcasonne is 0.73 and while slightly below the arithmetic average it is well above the weighted average.

(52) Concerning London (ex-Dublin), the quantitative evidence in this specification provides some evidence on the substitutability of London airports. On the Dublin- London route, both carriers fly to London-Gatwick with correlation of 0.37, which is the within route benchmark. The correlation between Aer Lingus' fares in Gatwick and Ryanair's Stansted fares is higher at 0.40. The other pair correlations are somewhat lower than 0.37 (for example, Aer Lingu's fares in Heathrow and Ryanair's in Gatwick is 0.3). At the same time Ryanair's fares amongst different airport pairs, namely Gatwick, Stansted and Luton, is very high (above 0.85) and therefore it appers that at least Ryanair regards these airports as substitutable. The correlation of Aer Lingus' London Gatwick and London Heathrow fares is also higher than the benchmark in the London case (0.48 vs. benchmark of 0.37).

(53) Furthermore, the Commission notes that Ryanair erroneously states that the Commission has failed to distinguish between passengers types, notably connecting vs. non-connecting. On the contrary, the Commission undertook such a robustness check to confirm the findings of its correlation analysis. (17) In particular, the Commission replicated its analysis considering the correlation between Aer Lingus' non-connecting passengers and Ryanair's passengers (which are considered nonconnecting). (18) Also, the Commission analysed the correlation between Ryanair's gross average fares and Aer Lingus' gross average fares. These results were available to Ryanair and Aer Lingus in the data room and the results confirmed the Commission's findings.

(54) Second, Ryanair argues that the correlation results presented in the SO are biased for several routes, unless the differenced residuals of both routes are stationary. They provide a table with the 29 series (corresponding to 19 routes) that are either trend stationary or non-stationary (with or without a trend). They argue that correlations of these series would be biased.

(55) The Commission agrees with Ryanair that correlation analysis should be performed on stationary series. An overview of the series that are not 'stationary' (that is to say, either trend stationary or (trend) non-stationary) indicates that the majority of these routes are trend-stationary. An obvious approach is then to include in the partial correlation a time trend that would ensure that the correlation measured is not driven by any such trend. By definition, the introduction of a time trend in the regression rules out the possibility that the residuals are trend-stationary. The Commission has included this in the main presentation of its results. As reported in Table 1 in Appendix B, the correlation figures of this robustness check are substantially similar to the ones of the Commission's previous analysis, and the vast majority of the fare series on which the Commission makes inferences are stationary, as shown in Appendix A. The Commission notes that this change has not affected its conclusions from the correlation analysis compared to the preliminary conclusions reached in the SO, with the exception of the route Dublin-Munich/Memmingen for which the results has to be considered inconclusive since one of the fare series is non-stationary. (19)

(56) Ryanair also, erroneously, claims that results are also likely to be biased as there is divergence of costs (fuel price) and fares, providing a specific route as an example. (20) However, the Commission notes that, contrary to Ryanair's finding the correlation between fares and fuel price is positive for the route in question. More generally, even if fuel prices and average fares do not move in the same direction all the time, this does not invalidate the analysis. The aim of the exercise is not to explain the movement of the average fare but to determine the co-movement of prices taking into account common factors.

(57) Third, Ryanair claims that some correlations are counterintuitive. For example, Ryanair notes that in some instances when both carriers fly to the same airport the fares correlation is relatively low while for some pairs such as Toulouse/Carcassonne the correlation is relatively high. The Commission notes that the results of the analysis are very similar to the findings in the 2007 Decision. It appears that not significant changes in market circumstances have taken place for the airport/city pairs which were also operated during the timeframe analysed in the 2007 Decision. Notably, for Toulouse/Carcassonne similar results were obtained in the 2007 analysis (price correlation of 0.95, albeit for a relatively small number of observations). The robustness of the results provides confidence on the reliability of the results. (21)

Conclusion

(58) The results of the analysis provide support for the conclusions reached by the Commission for the purposes of market definition. Overall, the results lend support for wide market definitions for most of the airport pairs while for Frankfurt/Frankfurt Hahn, Glasgow/Glasgow Prestwick a conclusion can only be taken after taking into account other qualitative evidence. In the case of London airports, Leeds Bradford, Bristol/Cardiff, Munich/Memmingen and Brussels/Charleroi the results are inconclusive.

Footnotes Annex I:

1 The price correlation analysis in the present case is in line with the methodology followed in the 2007 Decision as described in detail in that decision, Annex III.

2 In the following analysis, and the only exception to the potential city pairs as considered in the Decision opening proceedings, the Commission has not tested whether the fares of the Parties co-move in the pair Edinburgh- Glasgow.

3 The Commission has used the AIC criterion to select the optimal lags of the auxiliary regression. However in Toulouse/Carcassonne the optimal lags reduced the degrees of freedom of the auxiliary regression to below 20. Using an alternative approach to select the optimal number of lags, Q-test, a smaller number of optimal lags was selected. For this route the results presented refer to the procedure on the basis of the Q-test (shown in the Table of Appendix A in bold).

4 For Aer Lingus, see Aer Lingus' response to the Commissions' request of information of 3 October 2012.

5 Oxera, "Assessment of the Commission's analysis in the Annexes I, II, and III of the SO," page 3-5. The Commission notes that the results of the analysis including trend are largely similar to the ones without a trend.

6 A robustness check was performed on Aer Lingus total gross revenue and Ryanair sum of total gross/net revenue and total other revenues. These two metrics reflect for both companies the total revenue generated from its passengers on the given route (see also section on the description of data). A further robustness check was performed using only Aer Lingus non-connecting passengers. However any analysis with non-connecting passengers would have to be constraint to the time period post June 2008. The results overall remain robust.

7 The rows in bold indicate the first type of correlation i.e. when both Aer Lingus and Ryanair serve the same airport.

8 These routes are not necessarily actual overlaps.

9 The simple average of the highest and lowest correlations is around 0.63. The Commission has not considered in its computations airport pairs with less than 20 observations.

10 The pair Alicante-Murcia is not crucial for the purposes of the quantitative analysis as both Parties fly to Alicante from Dublin and Cork.

11 Even if there is a significant number of observations to London for one airport pair (London Heathrow and London Stansted), there is no within route benchmark (i.e. instance where both Parties fly to the same airport) as in the other routes to London ex-Dublin, Cork and Shannon. The correlation of this pair is not significantly different from the correlation of these two airports ex-Dublin, Cork or Shannon.

12 The Commission however notes that it is possible that certain common factors may induce a higher correlation amongst routes of the same air carrier.

13 The pair Bristol-Cardiff is not crucial for the purposes of the quantitative analysis as both Parties fly to Bristol from Dublin.

14 Oxera, "Assessment of the Commission's analysis in the Annexes I, II, and III of the SO," page 3.

15 For completeness the Commission also notes that Ryanair's Knock-Stansted series is now also nonstationary. However, Ryanair's Knock-Luton is stationary and therefore the analysis above holds.

16 These average as above relate to airport pairs where both series are stationary.

17 See for instance paragraph 45 of Annex I of the Statement of Objections.

18 See paragraph 45 of Annex I of the Statement of Objections.

19 For Dublin-London, the results are inconclusive as the series of Aer Lingus' fares in Gatwick is non stationary. However, under the specification including the destination frequencies the series becomes stationary as discussed above.

20 Oxera, "Assessment of the Commission's analysis in the Annexes I, II, and III of the SO," page 5; see also email correspondence between the Commission and the Notifying Party's economic consultants of 17 and 18 December 2012.

21 Also Ryanair makes an erroneous claim that it is unclear how the Commission has calculated the correlation coefficient for Liverpool as Aer Lingus does not fly to Liverpool airport (see Oxera, "Assessment of the Commission's analysis in Annexes I, II and III of the SO", footnote 10 ). The Commission notes that Aer Lingus currently does not fly to Liverpool but it did during the period of the analysis (and in particular during the November 2004 to June 2006 period).

Appendix A: Serial Correlation (Q test) and Stationarity (ADF test) Table

Table 1: Augmented Dickey Fuller tests of the underlying series

<TABLE 1>

Appendix B: Correlation Coefficients Table

Table 1: Correlation Coefficient Results with trend

<TABLE 1>

Table 2: Correlation Coefficient Results with trend and destination frequencies

<TABLE 2>

Annex II - Ryanair IPR analysis

Description of Ryanair's analysis

(1) Ryanair has submitted a study based on Illustrative Price Rises (Methodology for Map-Based Analysis), annexed in Ryanair's response to the decision opening the proceedings. This analysis, according to Ryanair relates to the substitutability of airports within O&D pairs, as defined by the Commission in the decision opening the proceedings.

(2) The analysis was performed on 5 city pairs (from Dublin to Paris, Frankfurt, Munich, Stockholm and Vienna) where Aer Lingus and Ryanair operate to different destination airports, and uses customer location data from Ryanair's booking system. The first goal of the analysis is to determine to what extent the catchment areas of Ryanair and Aer Lingus overlap in these city pairs on the basis of these booking data.

(3) The analysis only looks at passengers at the non-Irish side of the route (that is to say, includes bookings where the outbound leg started at the non-Irish end of the route as the aim of the analysis is to look at the catchment areas around the non-Irish airports).

(4) The analysis assumes that there is symmetry between Ryanair and Aer Lingus passenger responses and therefore Aer Lingus passengers are assumed to be equally willing to switch between airports as Ryanair's. In particular, the analysis assumes that every Ryanair passenger from within 100km of the Aer Lingus airport, who previously purchased a Ryanair ticket, would switch between airlines and airports whereas no-one outside 100km of the Aer Lingus airport is willing to do so.

(5) On the basis of the post code data extracted from Ryanair's reservation system and the above assumptions. Ryanair estimates the percentage of Ryanair's passengers that are within 100km of the Aer Lingus airport. This share is between [10-20]*% (in the case of FMM and MUC in Munich) and [50-60]*% (in the case of CDG and BVA in Paris). Based on these figures, Ryanair concludes that "the bulk of Ryanair passengers are located well away from the airport which Aer Lingus serves, with a low proportion of the Ryanair passengers located within 100km of the Aer Lingus airport" (1).

(6) Ryanair further assumes that in routes with a third competitor there is a need to divide up the individuals living within 100km of the Aer Lingus airport between the various airlines operating Dublin services. Ryanair performs this splitting on the basis of weekly frequencies offered by competitors in that airport. Therefore the estimated percentage of Ryanair passengers switching to Aer Lingus is between [10- 20]*% in Frankfurt and [30-40]*% in Vienna. These figures, according to Ryanair, reflect the "actual" diversion ratio between Ryanair and Aer Lingus for these airport pairs.

(7) Ryanair then calculates a critical diversion ratio which is required to generate an indicative price rise of [...]*% by making further assumptions on the Aer Lingus and Ryanair gross margins as well as the relative prices of Ryanair and Aer Lingus. The analysis uses average gross margins across all routes (for Ryanair between [...]*% and [...]*% for Aer Lingus and a price ratio of [...]* (with Aer Lingus being almost twice as expensive as Ryanair).

(8) The formula for critical diversion ratio, which assumes symmetric diversion and linear demand, yields a critical diversion ratio for Ryanair of approximately 30%. Therefore if the actual diversion between Ryanair and Aer Lingus is lower than this 30% threshold then a price increase for the hypothetical monopolist is not profitable and jointly they do not constitute a relevant market.

(9) Ryanair concludes that for example in the case of Frankfurt that "The estimated critical diversion ratio for Ryanair is [...]*%. However, Oxera estimates that only approximately [...]*% of all of Ryanair's HHA-Dublin passengers who would switch away from Ryanair following a price increase would actually switch to Aer Lingus. Since this number is smaller than the critical diversion ratio, it follows that the illustrative post-merger price rise by Ryanair would be lower than 5%. This analysis therefore indicates that Aer Lingus' services from FRA place little competitive pressure on Ryanair's services from HHA". (2)

Shortfalls of the analysis

(10) The Commission notes several shortcomings of the analysis. (3) The Commission considers that these shortcomings affect both the reliability of the analysis but also show that a different, more likely interpretation of the results exists under which the two airports in each relevant airport pair belong to the same market:

(a) The Commission notes that the assumptions made by Ryanair for the purpose of estimating the actual diversion ratio on the willingness of passengers to switch airport are arbitrary and do not necessarily reflect the true diversion patterns due to a small but significant non-transitory price increase ("SSNIP"). First, the assumption that everyone within 100km around the Aer Lingus airport will be willing to switch and the others will not is a very strong one that is likely to bias the result. Large proportions of Ryanair's passengers have a credit card address that is further than 100 km from Ryanair's airport. For example, for Paris the change in the proportions of Ryanair passengers located within 100km of the Ryanair airport compared to the Aer Lingus airport is very small [...]*. Therefore, this raises the issue of how the Ryanair passengers whose credit card address is outside 100km are treated, as these passengers represent a significant proportion of the passengers on most routes. Table 1 below shows that a significant proportion of passengers are located outside 100km from both Ryanair and Aer Lingus airports and that the proportion of Ryanair passengers within 100km of a Ryanair and of a Aer Lingus airport are very similar in several of these examples.Secondly, Ryanair's analysis excludes the Irish passengers for whom the distance between the Ryanair and Aer Lingus airports could be less important than the distance between the city center and the airports.

Table 1

City Pair / Ryanair passengers within 100km of Aer Lingus airport / Ryanair passengers within 100km of Ryanair airport

Vienna-Bratislava / [30-40]*% / [40-50]*%

Paris / [50-60]*% / [60-40]*%

Stockholm / [40-50]*% / [50-60]*%

Munich / [10-20]*% / [30-40]*%

Frankfurt / [20-30]*% / [50-60]*%

Source Ryanair

(b) The Commission also considers that it is arbitrary to allocate passengers across competitors on the routes where there is a competitor. Ryanair's analysis splits passengers in the Aer Lingus' airport on the basis of the number of frequencies of different competitors in that airport. First, the Commission notes that the actual point-to-point passengers flown should more closely reflect the market position of each competitor. Also, Aer Lingus would be the closest competitor of Ryanair on at least some of these routes (see for instance Section 8.9.5) and therefore a greater share of Ryanair's passengers is likely to switch to Aer Lingus than to the other competitors (operating in the Aer Lingus' airport) in the relevant routes. Properly accounting for these would lead to a higher actual diversion reflecting the greater constraint that Aer Lingus imposes on Ryanair. The different actual diversion ratios according to Ryanair are provided in Table 2 below.

Table 2

City Pair / Actual diversion to the Aer Lingus airport / Actual diversion to Aer Lingus (based on frequencies) / Actual diversion to Aer Lingus (based on passengers)

Vienna-Bratislava / [30-40]*% / [30-40]*% / [30-40]*%

Paris / [50-60]*% / [10-20]*% / [30-40]*%

Stockholm / [40-50]*% / [20-30]*% / [20-30]*%

Munich / [10-20]*% / [10-20]*% / [10-20]*%

Frankfurt / [20-30]*% / [10-20]*% / [10-20]*% (4)

Source Ryanair

(c) The Commission notes that the analysis provided by Ryanair computes critical diversion ratios (that would make a 5% price increase profitable for the Parties) for market definition instead of calculating the likely illustrative price increase from a merger. Using the "actual" diversion ratios between the Parties one can compute the illustrative price rises ("IPR") due to the Transaction (that Ryanair's analysis does not report). However, even a finding that the price effect is relatively small (less than 5%) could also indicate that the Parties' services do not constitute a relevant market on their own before the Transaction, other rivals also constrain them and the market definition should include these other rivals as well.

(d) The Commission using the assumptions of Ryanair on the diversion ratios, computed the predicted IPR and performed several robustness checks of the analysis. Ryanair's diversion ratios predict a price increase between 1% and 7% for Ryanair and a 1% to 4% price increase for Aer Lingus.

- However, the formula used by Aer Lingus (notably the linear demand assumption) limits the magnitude of the price increase. Under other typically used demand functions the magnitude of predicted price increases are usually larger (and the predicted critical diversion ratios are smaller). The Commission tested whether the results are robust assuming a different demand function for the purposes of computing the critical diversion ratio. The Commission notes that assuming a log-linear demand function the critical diversion ratio ranges from [...]*% to [...]*% for Aer Lingus and [...]*% to [...]*% for Ryanair. Sensitivity analysis for the log-linear demand function is provided in Appendix I of this Annex. More precisely, under the [...]*% margin assumption for Aer Lingus and the [...]*% margin assumption for Ryanair (the assumptions made in Ryanair's analysis), critical diversion under log-linear demand function falls to 20% for Ryanair and to 13% for Aer Lingus. Therefore comparing these benchmarks with the actual diversion to Aer Lingus, on all routes (in Munich and Frankfurt price increases would be profitable only at the Aer Lingus airport) the price effect would be greater than 5% (that is to say, greater than the 5% indicative price rise as set by Ryanair's analysis). The predicted price increase under the log-linear demand assumption ranges from 1% to 19 % for Ryanair and from 3% to 22% for Aer Lingus (depending on the airport). (5) The analysis for the illustrative price increases is provided in Appendix II of this Annex.

- Furthermore, the way margins are computed may well affect the robustness of any results. Typically in the airline industry computing gross margins is a very complex exercise. It is likely that the relevant margin in the setting of the IPR relates to the margins for the marginal units, which are likely to be significantly higher than the average margins in a given route. Ryanair's analysis has used gross margins of [...]*% for Aer Lingus and [...]*% for Ryanair. Increasing gross margins implies that the critical diversion ratio decreases. In particular assuming [...]*% gross margins for Aer Lingus, the critical diversion ratio falls (under loglinear specification) to around [...]*% and therefore the prediction of Ryanair's analysis would indicate that all airport pairs (including Munich and Frankfurt under the proper passenger-based allocation as shown in Table 2 above) significantly constrain each other and belong to the same market.

(e) Regarding the representativeness of the dataset, the Commission notes that the analysis is made only on a relatively very small number of observations. For the Vienna, Paris, Stockholm, Munich and Frankfurt route, respectively [...]* and [...]* observations were used in the analysis. This is contrasted with more than 13 000 passengers that flew on the route Dublin-Paris in March 2012 (while 13 000 passengers actually flew in March 2012 the number of bookings provided by Ryanair refer to bookings made in March 2012 and not to the actual flights in March 2012). Based on Ryanair's own data there were more than [...]* booked passengers on the respective routes in March 2012. Therefore the dataset used by Ryanair typically represents less than [...]* % of the actual number of total passengers on these routes in March 2012. Ryanair claims that all booking information has been included in the original dataset. Ryanair further explains that each booking may represent more than one passenger (a single booking may refer to a group of passengers) (6). However, in any case Ryanair acknowledges that at least the passengers booking from Ireland are not included in the dataset. Therefore, the Commission considers that there are concerns regarding the representativeness of the sample.

Ryanair's comments

(11) Ryanair has also commented on the Commission's views on the IPR analysis. While acknowledging that the IPR analysis' assumptions may bias the diversion estimates, Ryanair claims that the assumption on diversion does not necessarily lead to a downwards bias of the diversion estimates.

(12) Ryanair further suggests that the Commission has considered a 'worst case scenario' rather than a balanced assessment. It claims that the Commission has not recognised that a range of conservative assumptions have been made in the IPR analysis and that the Commission has applied in a cumulative fashion several layers of cautious assumptions to consider a worst-case scenario to compare against Oxera's original results.

(13) The Commission does not agree with Ryanair's views for the following reasons.

(14) First, Ryanair has not addressed the Commission's concerns that the assumptions made in the IPR analysis for the purposes of estimating the actual diversion ration are arbitrary. The Commission can have no basis for considering that this metric reflects the actual diversion. The actual diversion as computed by Ryanair appears to be an arbitrary "assumed" diversion. Ryanair argues that this assumption does not necessarily lead to downward bias. However, Ryanair has not replied to the question of why a significant proportion of passengers are located outside 100km from both Ryanair and Aer Lingus airports and that the proportion of Ryanair passengers within 100km of a Ryanair and of an Aer Lingus airport are very similar in several of the examples.

(15) Secondly, the Commission does not consider it has used a worst case scenario. The Commission has first raised very serious doubts regarding the reliability of the actual diversion ratios, as computed by Ryanair. Also, Ryanair has not addressed the Commission's point that it does not take into account the closeness of competition (Aer Lingus being the closest competitor of Ryanair) when it allocates passengers across competitors in the route (point b above). The Commission has further argued that the analysis is not well suited for market definition purposes (point c). A finding that the price effect is relatively small (less than 5%) is not informative on whether Ryanair would be in a position to increase fares for flights to its own airport (and therefore them being in the same market) and similarly a small price effect only indicates that the Parties' services may not constitute a relevant market on their own (but together with other rivals they may). Ryanair again does not comment on this. Last, Ryanair has remained silent on the fact that the IPR analysis is made only on the basis of a relatively small number of observations which does not include passengers originating ex-Ireland (point e).

(16) Notwithstanding all these fundamental flaws in Ryanair's analysis, the Commission has tested the robustness of the IPR analysis. The Commission also notes that Ryanair does not explain why any of the assumptions put forward by the Commission are more unrealistic than the one put forwards by Ryanair.

(17) The Commission also notes that the assumptions made by Ryanair are not conservative, for the following reasons:

- The Commission notes a stark contradiction in Ryanair's IPR analysis and on its reply to the SO. Ryanair has claimed that "Passengers who purchase flights on Ryanair typically do so because they are highly price sensitive (that is to say, will chose the cheapest ticket available), because Ryanair's flight schedule is more convenient, or because Ryanair flies to the most convenient airport for the actual travel origin/destination. These passengers will not switch to Aer Lingus in response to a 5-10% price increase by Ryanair because the schedule and destination will not change, and Ryanair will generally still be (by a wide margin) the cheapest available option (Ryanair's average short-haul fares are half those of Aer Lingus)". (7) In the IPR analysis, as Ryanair had no access to booking data of Aer Lingus, it has assumed that there is symmetry between Ryanair and Aer Lingus passenger responses (that is to say, they have assumed that Aer Lingus' passengers are equally willing to switch between airports as Ryanair's passengers). Ryanair considers this "to be a conservative assumption since Ryanair's passengers are generally likely to be more price-sensitive, and hence more willing to switch between airlines and airports than Aer Lingus' passengers" (8), that is to say, the exact opposite of what is claimed in the reply to the Statement of Objections. Besides, the Commission's regression analysis shows that both airlines significantly constrain each other.

- In routes with competitors, for the computation of actual diversion ratios the Commission cannot regard the allocation on the basis of frequencies rather than passengers as a conservative assumption. The underlying figures for actual diversion change significantly from [10-20]*% to [30-40]*% for Paris and from [10-20]*% to [10-20]*% for Frankfurt. Furthermore, Ryanair does not explain why a measure of closeness of competition is not taken into account. The Commission cannot consider the assumption made by Ryanair as a conservative assumption.

- Furthermore, Ryanair criticises the Commission for using a log-linear demand assumption on the demand function, claiming it generates unrealistic results with weak theoretical foundation for the IPR. Also, the Commission's robustness analysis has shown that under the log-linear specification the critical diversion ratios are much lower (and the expected price effects significantly higher). (9) The Commission agrees that some of the log-linear demand predictions may not hold for large price increases as it assumes a constant elasticity of demand which is not likely to hold for very large price increases. However, for the IPR analysis provided by Ryanair, the Commission does not agree that the linear approximation is necessarily a better specification than the log-linear. (10) A standard approach in the literature is to consider both specifications. (11)

Conclusions

(18) The analysis provided by Ryanair for the Map-Based analysis relies on several strong and arbitrary assumptions. First, the Commission does not consider that the assumptions made by Ryanair (notably the 100km benchmark and the assumptions on routes with competitors) provide a sufficiently meaningful comparator for measuring actual diversion ratios for the purposes of market definition and effects based analysis. Secondly, the results of the analysis are not robust to changes in the assumptions of the computation of the critical diversion ratio. Thirdly, the Commission cannot exclude that the sample provided is not a representative sample. The Commission does not consider that the results of this analysis provide any evidence that the relevant airports belong to different markets; on the contrary, even under the strong assumptions made by Ryanair in several robustness checks the results indicate the opposite.

Footnotes Annex II:

1 Annex 2(b) of Draft Form RM of 17 August 2012.

2 See Ryanair's observations on the decision opening the proceedings.

3 Ryanair's analysis itself acknowledges that it is based on some assumptions and the interpretation of the results would depend on these assumptions ("In reality, neither of these assumptions will be true in most cases [...]. Whether the analysis under- or over-represents the true number of switchers will depend upon the balance between these two groups"), Annex 2(b) of Draft Form RM of 17 August 2012.

4 The Commission notes that Ryanair has not properly distributed the [...]*% across competitors on the basis of market shares on passengers as following Ryanair data (Annex 7.3(b) as updated on 22 August 2011) in Summer 2011 and in Winter 2011/2012 Aer Lingus represented 42% and 59% of all passengers on the route (compared to 35% and 21% for Lufthansa respectively). Therefore the correct actual diversion as the basis of Ryanair's computations would be closer to 20% rather than [...]*% (16% for Summer 2011 and 21% for Winter 2011/2012).

5 The IPR methodology does not take into account the feedback effects between the other rivals (these factors typically increase the magnitude of the predicted prices increases) and does not attempt to account for efficiencies, which typically decreases the predicted price increases.

6 Ryanair explains that "each "booking" of a Ryanair flight typically corresponds to around 3 one-way flights (or sectors) (while only one billing address will be associated with each booking). The exact bookings/sectors ratio may vary from route to route. In addition, as you [the Commission] noted, the "bookings" observations used for the mapping analysis relate to flights from one end of the route only (i.e., the non-Irish end), which further understates "bookings" in relation to total passengers on a route", see Ryanair's reply to Commission's Request of Information of 26 October 2012.

7 See Ryanair's Response to Statement of Objections, page 6.

8 See "Methodology for Map-Based analysis", Annex 2(b) of Draft Form RM of 17 August 2012.

9 The Commission also notes that it has tested, as a robustness check, the results of the analysis when Aer Lingus' gross margin is [...]* instead [...]* points increase/decrease).

10 Ryanair also claims that the Commission has incorrectly assumed that the average Ryanair fares are as expensive as those of Aer Lingus' in the computation for the log-linear demand. The Commission first notes that Ryanair does not provide these calculations. More importantly, for example in Shapiro (2010), "Unilateral Effects Calculations", the computations of illustrative price changes under constant elasticity of demand are not dependent on the price ratio (contrary to the linear demand where the price ratio affects the IPR). Therefore the Commission does not consider correct Ryanair's claim (which in any case does not even change the figures substantially (from 20% to 14% for Ryanair and from 13% to 22% for Aer Lingus).

11 See Shapiro, C (2010) "Unilateral Effects Calculations" and Parker, D (2009)"Illustrative Price Rises from Mergers in Differentiated Products Markets", GCP. Ryanair provides no evidence that the own elasticity of demand changes for relatively small price changes in this particular case.

Appendix I: Sensitivity analysis of critical diversion analysis of Ryanair (under log-linear demand)

<TABLE>

Appendix II: Illustrative Price Rises (IPR) computed with Ryanair's assumptions on the diversion ratios, different assumptions on the value of the margins of Ryanair and Aer Lingus and with linear and log-linear demand curves.

[...]*

Annex III - Regression analysis.

Introduction

(1) The Commission has used regression analysis technique, as in the 2007 Decision, to test the competitive interaction between the Parties. Regression analysis is a statistical tool for understanding the relationship between two or more variables. Using this regression analysis the Commission has investigated whether the presence of one of the Parties has an impact on the fares of the other, to quantify the magnitude of such an effect and to help predict the average increase in fares post- Transaction (if any). The Commission in the 2007 Decision used different regression techniques but concluded that the fixed-effects analysis provides the most appropriate approach in the case at hand. The principal advantage of the fixed-effects analysis is that it takes into account important but unobserved (or unmeasured) influences on price that do not vary over time on any single route. Therefore the Commission in the present case has followed the same fixed-effects methodology.

(2) A regression analysis with route specific fixed-effects accounts for specificities on given routes. When an analysis does not control for such specificities the results of the analysis may be biased due to the so-called omitted variable bias (that affects alternative regression analysis techniques, namely cross-sectional regressions). By controlling for route specificities the omitted variable bias risk is mitigated because unobservable cost or demand factors, whose variation across routes would be likely to affect fares, are more likely not to vary over time on any single route. The Commission regards this methodology as the most suitable to assess the competitive constraint exerted by Ryanair on Aer Lingus.

Data Description

(3) The Commission has requested data from DAA, Ryanair and Aer Lingus.

DAA data

(4) The DAA has provided in its first submission (3 August 2012) a dataset containing information on all the routes out of Dublin, Cork and Shannon for the period between November 2004 to July 2012. The information provided was for each carrier (split between its scheduled and charter flights), month and route:

- Total monthly frequencies (movements);

- Total number of passengers, split between point-to-point passengers and connecting passengers;

- Total number of seats offered.

(5) A number of clarification emails were exchanged between the Commission and DAA with a view to understanding the nature of some observations (see for example email from the Commission to DAA of 09 October 2012). On 12 October 2012 DAA provided a final dataset.

Aer Lingus Data

(6) Aer Lingus submitted in various stages the reply to the data request sent by the Commission on 30 August ("Company data on the City Pair of Reference").

(7) Aer Lingus raised several issues on data availability (see Aer Lingus' email of 10 and 12 September). (1)

(8) Aer Lingus' main difficulty has been to provide disaggregated data as their data is only stored and available from June 2008 to the present day. This difficulty relates to data disaggregated on the basis of connecting vs. non-connecting passengers as well as for different fare types. Aer Lingus also explained that it does not have revenue data from Aer Arann operations.

(9) The dataset provided by Aer Lingus included data on short-haul routes out of Dublin, Cork and Shannon to non-Irish destinations. Aer Lingus provided aggregate data on 21 September 2012. The variables included the following:

- Number of passengers: which relates to the number of total flown passengers;

- Total capacity;

- Number of sectors (movements);

- Load factors;

- Gross revenue: which is the sum of the net revenue, recoverable charges (which includes airport taxes, change fees, FFP etc), administration and excess baggage fees and total other revenues;

- Net revenue; which includes the pure passenger revenue;

- Total administration and excess baggage fees: includes baggage fees, excess baggage fees, online booking fees and telesales fees;

- Recoverable charges: includes recoverable airport charges but also revenues from insurance/security charges;

- Total other revenues: includes change/cancellation fees, baggage fees, seat selection, retail proportion of Plus fare, inflight sales and commissions. Cargo and mail revenues are not included;

- Total fuel costs.

(10) Aer Lingus also provided data (passengers, net and gross revenues) on passenger origination (where sales made within the sales region of Ireland have been attributed as passengers originating at an Irish airport) on 26 September 2012 (for period June 2008 to July 2012). Gross revenue is allocated to Irish-originating passengers in the same proportions as applies to net revenues.

(11) Aer Lingus also provided information on the business class passengers on its flights before March 2005 (28 September 2012).

(12) Aer Lingus further submitted data (on passengers and net revenues) on connecting vs. non-connecting passengers but only as of June 2008 (see Aer Lingus' submission of 8 October 2012).

(13) Aer Lingus provided information on when Flex fares and Plus fares were introduced on a route-by-route basis in their submission of 8 October 2012 and has confirmed that following the phasing out of business class tickets in March 2005 there were not different fare types on Aer Lingus flights until the introduction of Flex fares and Plus fares. (2)

(14) The Commission also asked Aer Lingus data for the following routes (9 October 2012):

- Knock-Birmingham (including EMA);

- Knock-London (all airports);

- Belfast/Derry (both Belfast airports and Derry) - Faro;

- Belfast/Derry (both Belfast airports and Derry) - London (all airports).

(15) Aer Lingus provided these data on 16 October 2012, with the exception of Knock- Birmingham (as this is operated by Aer Arann).

(16) Another data request was sent to Aer Lingus on 30 August 2012 requesting data on airport distances as well as Schedule Reference Service data ("SRS data").

(17) The distance dataset included the following variables:

- Distance of an airport to the city center;

- Travelling time to city center.

(18) Aer Lingus also submitted SRS data (submitted on 11 September and a more extended dataset including more airports on 8 October) which provide information on airport use by carrier. The information included the number of outbound scheduled flights per week to destinations that either Aer Lingus or Ryanair (requested on 03 October 2012) had served (during the period November 2004-July 2012). This dataset has been used to assess the degree of presence of each of the competitors at the destination airport for each route/city pair.

(19) The Commission also requested (3 October 2012) from Aer Lingus the fuel price they most closely monitor. Aer Lingus has provided (8 October 2012) the time series:

- Jet Kerosene Cargoes CIF NWE for the EU (in $/MT); and

- Jet 54 USGC pipeline for the US (in $cent/gal). Ryanair Data

(20) Several data requests were sent to Ryanair.

(21) A data request ("Company data on the City Pair of Reference") was sent on 30 August. Ryanair provided data on all city-pair routes operated by Ryanair and departing from Dublin, Cork or Shannon during November 2004-July 2012. Data were requested for each regularly operated flight on a monthly basis.

(22) Ryanair submitted the first reply on September 11 (and revised versions on October 3 and October 8) and provided the following variables:

- Number of booked passengers: Ryanair confirmed that it cannot provide data on the flown number of passengers (see email from Ryanair of 11 October 2012);

- Total capacity;

- Number of sectors (movements);

- Load factors;

- Total (gross/net) revenue: this includes all revenues generated from the sale of the ticket and includes published airport charges, check-in fee, insurance levy, ETS levy and EU 261 levy, other miscellaneous revenues. It excludes government taxes as Ryanair acts as a tax collector;

- Total administration and excess baggage fees: these are calculated by reference to the total administration and excess baggage fees generated during the particular month, divided by the total number of passengers booked during the month (and then multiplied by the number of passengers on each route);

- Total other revenues: includes bag fees, administration and baggage fee, on board sales, and commissions;

- Total airport taxes: the charges that were paid by Ryanair to the airports;

- Total fuel costs: are calculated by taking the total fuel costs incurred during the month and dividing the total flight hours for the month (which is then multiplied by the total flight hours on each route).

(23) Ryanair also submitted data by passenger origination since April 2005 on 17 October 2012.

(24) The data request ("Airport data") was sent on 30 August 2012 and asked Ryanair to provide distance data between airports.

(25) Ryanair submitted further disaggregation of the baggage fee on 29 September 2012. This was only submitted for the United Kingdom routes. On 8 October 2012, Ryanair submitted data on all routes which includes:

- Total (gross/net) revenue: This excludes baggage fees.

- Total other revenue: This includes baggage fees, administration and excess baggage fees, on-board sales, car hire and hotel revenues, and other ancillary revenues. The revenue generated is divided by the total number of passengers during the month (then multiplied by the number of passengers who travelled on each route).

(26) Further clarification questions were sent to Ryanair on 8 October 2012 and 9 October 2012. The Commission has asked why Ryanair cannot provide data for actual flown passengers (rather than booked) and why they cannot provide data origination.

(27) The Commission also requested data on the fuel price time series that Ryanair monitors (at a monthly level), as well as data for the following routes:

- Knock-Birmingham (including EMA);

- Knock-London (all airports);

- Belfast/Derry (both Belfast airports and Derry) - Faro;

- Belfast/Derry (both Belfast airports and Derry) - London (all airports).

(28) Ryanair provided this data on 11 and 17 October 2012.

(29) A separate request was sent to Ryanair on airport data on 30 August 2012. Ryanair submitted data on September 11. Ryanair provided data regarding the following:

- Distance of an airport to the city center;

- Travelling time to city center;

- Closest Aer Lingus airport;

- Distance to the closest airport to which Aer Lingus operates (whenever the distance is below 150km).

(30) Ryanair computed these distance and travelling time data on the basis of Google Maps.

(31) The Commission has also requested from Ryanair (3 October 2012) the fuel price they most closely monitor. Ryanair has provided (8 October 2012) the

- Jet Kerosene Cargoes CIF NEW.

Econometric methodology

(32) The Commission have followed the empirical strategy employed in the 2007 Decision and considered a reduced-form specification. The basic idea is to estimate the effect of the entry or exit of one of the merging parties on the competitive behaviour of the fares of the other. In the Commission's analysis the dependent variable is the price of one merging party and the explanatory variable of interest is whether the other is present on the route (city of airport depending on the specification). Other variables are added to "control" for other possible systematic influences on fares, which refer to route characteristics that may affect demand or supply on that route.

(33) The Commission's (merged) panel data tracks the prices set by both Ryanair and Aer Lingus on individual routes over time. The data includes observations on the average price charged by each airline per month from November 2004 to July 2012 both on routes to the same airport and on routes to the same city but a different airport. The Commission has considered in its analysis the average monthly fares. The Commission computed the average net fares based on total net revenue (3) on a passenger basis. In the case of Ryanair the booked passengers and in the case of Aer Lingus the flown passenger numbers were available and used for the computations.

(34) Data in this format makes it possible to use an empirical strategy, the so-called fixedeffects regressions, to assess the extent to which the merging firms exert a competitive constraint on each other (holding constant other factors such as competition from other airlines).The fixed-effects regression analysis with panel data, exploits the variation in market structure at individual routes over time.

(35) The advantage of using a fixed-effect approach is that it is possible to control for important but unobserved or unmeasured influences on price that vary from route to route (or month to month as this approach controls for route specific effects as well as month specific effects). When important variables affecting price on different routes (or months) are unobserved and correlated with the explanatory variables included in the regression, the estimated coefficients can be subject to bias. This problem is often referred to as omitted variable bias. For example, prices may be higher in monopoly routes not because there is no competition but because on this particular route, demand is relatively low or costs are relatively high (for example when high entry barriers are correlated with high operation costs). The opposite is also true, prices on monopoly routes may be relatively low because demand is weak (or very elastic) on such routes.

(36) Fixed effect regressions exploit the variation in market structure on individual routes over time. This approach, introducing appropriate assumptions on the distribution of the error term, uses information on changes in the market structure within a route over time. For example, the entry of Ryanair on a route dominated by Aer Lingus may affect the latter's price (after controlling for observable changes in other variables such as entry by other rivals). Effectively the method compares the level of Aer Lingus prices on a route after Ryanair entered, with the level before Ryanair entered. This before-and-after comparison is done systematically for all routes where Aer Lingus operates and thereby generates the average effect of Ryanair's presence on Aer Lingus fares.

(37) This before-and-after comparison (for a given route) can mitigate the omitted variable bias. It is more likely that unobservable or non-measurable cost or demand factors affecting fares and varying across routes do not vary over time within a given route (such as the type of destination, the popularity of the route according to purpose of travel, customer awareness, destination airport characteristics, number of alternative airports at destination, safety considerations, total duration of travel, air traffic regulations at country of destination). Thus, the primary advantage of fixedeffects regressions comes where most unobservable or non-measurable factors affecting price are unlikely to vary much during the sample period.

(38) Fixed-effects regressions are suitable if there is sufficient time series variation in the data to permit precise estimates of the relationship between price and presence of a rival. When the time series variation in the data is limited, fixed effect estimates of the price effects of market structure may be very imprecise.

(39) The fixed-effect approach is suitable if there are sufficient entry and exit events to allow for identification of the relevant events. Notably, there has been consistent entry and exit of Ryanair on routes where Aer Lingus already operates and similarly, there have been a number of events of Aer Lingus' entry and exit on routes where Ryanair has been present. The entry and exit events are shown in Appendix A and B of this Annex. Entry and exit events of Ryanair provide useful information for the regression of Aer Lingus' fares and similarly entry and exit events of Aer Lingus provide the necessary variation in the data for Ryanair's regression.

(40) Fixed effect models also exacerbate the bias that may result from imprecisely measuring the explanatory variables. This is often referred to as "errors-in-variables" bias, which is the difficulty in detecting the influence of an explanatory variable when that variable is measured with error. Where errors in measuring the explanatory variable of interest are random this will attenuate the estimated magnitude of the coefficient estimate towards zero (that is any such errors would lead to an underestimate of the coefficient).

(41) Similarly to the 2007 Decision, four hypotheses were tested. If Ryanair were competitively constrained by the presence of Aer Lingus, it can be expected that Ryanair would have to offer lower fares on average when Aer Lingus is on the same airport or city pair. Conversely, if Ryanair imposes a competitive constraint on Aer Lingus one should expect that Aer Lingus fares are negatively affected by Ryanair's continued presence. This line of argument gives rise to the following testable hypothesis:

- Hypothesis I: Ryanair's (Aer Lingus) presence is associated with a statistically (4) and economically significant reduction in Aer Lingus' (Ryanair's) fares in the various short-haul routes where they overlap.

- Hypothesis II: Ryanair (Air Lingus) exerts a stronger competitive constraint on Aer Lingus' (Ryanair's) fares than any other actual or potential competitor does.

- Hypothesis III: The existence of an actual or potential competitor operating from a base at the destination airport on a route originating in Dublin has a limited impact on the Parties' prices.

- Hypothesis IV: The stronger the presence of Ryanair (Aer Lingus) on the route the more pronounced the effect on Aer Lingus (Ryanair) fares.

(42) It should be noted that the estimated effects of Ryanair on Aer Lingus prices (and of Aer Lingus on Ryanair's prices) are likely to be underestimated as the presence of Ryanair/Aer Lingus in Dublin exerts a potential competitive constraint on Aer Lingus/Ryanair. For example, on routes out of Dublin where Ryanair is the only carrier, it can be expected that it sets prices which are lower than what it would charge if Aer Lingus had no Dublin base (as there is the threat that Aer Lingus can enter such a route). Since the regression analysis considers only fares' overtime variations within each route and only captures price reductions subsequent to Aer Lingus' entry, this potential competition constraint does not show up in the empirical results.

(43) In summary, the results suggest (i) that the entry of Ryanair leads to a significant [...]* % lower Aer Lingus prices when considering airport-pairs and around [...]*% lower prices when considering city-pairs. The entry of Aer Lingus leads to around 3- [...]* % lower Ryanair prices when considering airport-pairs and around [...]*% lower prices when considering city-pairs (ii) that Ryanair (Air Lingus) exerts stronger constraint on Aer Lingus (Ryanair) prices than other competitors whose constraining effect is significantly lower or non-existent (iii) that bases of competitors at the destination airport have a small or even a positive effect on the Parties' fares and finally, according to the Commission's regressions, and (iv) that a 1 % increase in frequencies of Ryanair decreases the fare of Aer Lingus by approximately a significant [...]*%. A 1 % increase in frequencies of Aer Lingus decreases the fare of Ryanair by approximately a significant [...]* %.

Fixed-effects regression analysis

(44) The Commission considered the basic fixed-effect specification, as done in the 2007 Decision as the approach remains relevant for the present case. In this specification the route fixed-effects capture all factors affecting prices that differ across routes but do not vary over time. Thus such factors can be omitted from the regression specification without any risk that this might lead to omitted variable bias. In contrast, in the case of cross-section regressions if an important variable, say flight duration, was omitted but this variable was correlated with any of the explanatory variables this would lead to biased estimates (5).

(45) The Commission's empirical strategy focuses on the impact of Ryanair's "presence" and "strength of presence" on Aer Lingus' average net monthly fares and vice versa.

(46) The baseline fixed-effects regression is as follows: <FORMULA>

Where:

- The dependent variable is the average net monthly fares of first Aer Lingus and then Ryanair.

- ai is the route fixed effect (time invariant dummy variables =1 for the route and 0 otherwise). The ai dummy accounts for systematic but unobserved or nonmeasurable differences in costs or demand within that route.

- f(.) is a function of competitor variables. These are the explanatory variables of interest. In certain specifications these are the Ryanair and then Aer Lingus "presence" dummies. In other specifications these are the number of frequencies flown on the given route to indicate the "strength of presence." Similar variables included for flag and non-flag carriers, Cityjet and Aer Arann (before the introduction of the franchise agreement on a route-by-route basis. (6)) This information is typically provided from the DAA dataset, described above.

- Dt is a dummy for each time period (a month). The month dummies allows for controlling for cost shocks that affected all routes during the same time period.

- Xit is a vector of cost and demand controls added in certain specifications

(47) Similarly to the 2007 case, the Commission has constructed two data sets. The first is consistent with the city-pairs market definition adopted in the SO. The second tests the relationship for airport-pairs thereby implicitly assuming that each airport close to the destination city is a separate market. All specifications, if appropriate, are performed on both data sets and relevant differences are pointed out.

(48) The Commission has tested a number of alternative specifications that essentially differ in the competitor variables included. The Commission first tests the presence specification where a dummy variable for the other merging firm and for other rivals present on the route is included. Then the Commission tests the frequencies specification where the frequencies of the other merging firm and that of other rivals is included in logarithmic form (in addition to absence dummies). This specification allows measuring the sensitivity of each firm's fares to the strength of its various rivals on the route.

(49) The Commission first starts with the simplest possible specification (or baseline) and gradually include relevant controls, paying particular attention to the statistical robustness and economic significance of the explanatory variables of interest. All results are presented in tables below that allow for direct comparison of coefficients in alternative, slightly more general, specifications.

(50) The Commission employs "robust regression" and "weighted regression" techniques that control for the influence of outliers (essentially it drops some outliers and assigns less weight to others). Robust and weighted regressions assign weights to individual observations. In the robust regression, the weights are assigned based on a mathematical formula while in the weighted regression the Commission has weighted the observations on the basis of the net revenues generated for each observation, to control for the relative importance of the observations. We start by reporting the results from fixed-effects regressions on Aer Lingus price. We consistently find that Ryanair exerts a competitive constraint on Aer Lingus' prices. Next, we turn to describing our efforts in applying the fixed-effects procedure to test the influence of Aer Lingus on Ryanair prices, which indicated that Aer Lingus also constrains Ryanair, especially when they compete head-to-head on the same route.

Regressions with Aer Lingus' monthly average net fares as the dependent variable

Presence Specification

Airport-pairs database

(51) The presence specification investigates the effects of Ryanair entry on Aer Lingus'monthly average net fares controlling for the presence of other carriers and time and route specific fixed-effects. The regressions include dummy variables for the following: (i) the presence of Ryanair (r_FR), (ii) the presence of one or more flag carriers (r_flag_scheduled) and (iii) the presence of one or more non-flag carriers (r_nonflag_scheduled). Furthermore, in the 2007 Decision, Ryanair argued that besides Ryanair and Aer Lingus, other carriers, in particular Aer Arann (r_RE) and Cityjet (r_WX) have a base at Dublin airport. Allegedly, these carriers could exert a significant competitive constraint on the Parties (7). Dummy variables were introduced for these two carriers separately from that of other non-flag carriers. This allows the Commission to compare directly the effect of their presence both in absolute terms and relative to that of Ryanair. The regressions also includes Aer Lingus' log of capacity (ln_seats_air_lingus) as a scale factor. Furthemore, Aer Lingus' had two main restructuring events in the time period used for the estimation. The first dummy (r_business_EI) captures the phasing out of business type tickets on Aer Lingus routes that was completed in March 2005. The second dummy (r_newfareEI) captures the effect of Aer Lingus's introduction of flexible fares on a route-by-route basis, which started in November 2008. (8)

(52) The specification reported in Table 1 in columns (3) to (8), also includes a variable to control for the total frequencies offered by all carriers at the destination airport (ldest_frequency). This variable controls for the traffic at the destination airport and is therefore a good proxy for variations in demand at the airport-pair level (anticipated at least at the time that carriers set their schedule). These destinationrelated variables were constructed on the basis of the SRS data submitted by Aer Lingus. The specifications reported in columns (5) to (6) of Table 1 include a dummy to capture the effect of the joint presence of flag carriers and Ryanair. The specifications reported in columns (7) to (8) of Table 1 include dummies to capture the effect of the presence of at least one flag (base_route_flag) or at least one nonflag carrier (base_route_noflag) with a base at the destination airport. (9) The results of these regressions are presented in Table 1 below.

TABLE 1 -PRESENCE SPECIFICATION - BASELINE- AIRPORT PAIRS - AER LINGUS

[...]*

(53) The coefficients of the regressions are robust to many alternative specifications and estimation techniques. The coefficient of most interest is that of Ryanair's presence, "r_FR". It is between [...]* and [...]* and significant even at the 1% level. It means that over the period considered and across all routes, on average, Aer Lingus prices, depending on the specification, are [...]*% lower when Ryanair is also present on the route after controlling for Aer Lingus' own capacity and the presence of other competitors. These estimates are higher than the [...]*% that was found in the 2007 Decision in similar specifications.

(54) Similarly to the findings in the 2007 Decision, no other rival has a negative and robustly, statistically significant effect on Aer Lingus' fares. The flag carrier dummy is significant and positive in the robust regressions. However, it seems that this effect is mostly due to the joint presence of Ryanair and the flag carriers. This is shown by the results of the regression in columns (5) and (6) of Table 1. These specifications include a dummy that controls for this (by interacting the effect of Ryanair's presence with flag's presence, r_FR_flag). In these results the flag carrier effect becomes insignificant. Aer Arann's presence is associated with an increase in Aer Lingus prices, which appears counterintuitive. Since Ryanair's presence is associated with lower Aer Lingus' prices and Aer Arann tended to be present on routes where Ryanair was not, Aer Arann's presence would then be generally associated with higher Aer Lingus' prices. Similarly the results for the presence of CityJet indicate that its presence does not constrain the pricing of Aer Lingus (10).

(55) The r_newfareEI dummy is significant both statistically and economically. It indicates that after the introduction of flexible fares the average fare of Aer Lingus was about [...]*% lower than before this period. This result could possibly indicate the general reduction in the average fares that may have been associated with Aer Lingus' fares restructuring. Such a dummy corrects for the risk that the regression would assign such an impact to Ryanair's (or other competitors') presence.

(56) Results in columns (7) and (8) in Table 1 indicate that the presence of flag carrier or non-flag carrier with a base at destination has a positive and statistically significant effect (for flag carriers only in specification shown in column (8)) on Aer Lingus' prices. This finding indicates that the fares of Aer Lingus are not constrained when a carrier at the destination airports increases its overall activities.

(57) Finally the Commission notes that Aer Lingus' capacity has a significant and positive relationship with average fares (that is to say, on average, more capacity is planned on routes with higher expected demand, and therefore higher average fares).

(58) In all our regressions, the R-squared is high (above 0.8), which implies that the model fits adequately. This is largely due to the route and date fixed-effects. The route fixed-effects capture all time-invariant differences across the route, for example distance or thickness of the route that might affect prices. The date dummies capture unobserved sources of variation over time, including seasonal peaks in demand.

(59) Table 2 reports the results from adding a number of additional controls to the baseline specification. In this set of regressions, we introduce a dummy to control for the presence of charter flights (r_charter) and use a direct measure of total costs on the route (lfuel). Both variables turn out to be highly significant in all specifications but there is little variation in the coefficients of the presence variables. In particular, Ryanair's presence remains statistically highly significant and its negative impact is of similar magnitude (8-13%). Non-flag carriers do not appear to constrain Aer Lingus (their presence is associated if anything with higher prices for Aer Lingus, as in most specification the coefficient is positive or statistically insignificant). The impact of flag carriers is typically negative; however once their impact is interacted with Ryanair's presence (columns (3) and (4) of Table 1) flag carrier presence's impact is even positive (and non-significant). The coefficient of the charter presence variable also indicates that charters' presence does not impose any competitive constraint on Aer Lingus' pricing (and on the contrary across routes in which charter companies entered Aer Lingus has since then set higher fares).

(60) Again, the presence of a non-flag carrier based at the destination has a positive effect on Aer Lingus prices whereas the presence of a flag carrier with a base has a positive but not robust effect (significant only in the weighted regressions), as shown in columns (5) and (6)). It should be noted that this variable aggregates the effect of all non-flag carriers with a base. (11)

(61) The Commission has tested whether the results are robust for an alternative way to control for seasonality. As seasonality may differ between routes this could potentially bias the results. The specifications in columns (7) and (8) controls for route level seasonality by introducing route level month dummy variables. The results remain largely robust. The fuel and capacity variables lost most of their significance due to the extra controls, as expected, but Ryanair still have a statistically significant negative [...]*% effect on Aer Lingus' average fares. (12)

TABLE 2 -PRESENCE SPECIFICATION - ADDITIONAL CONTROLS - AIRPORT PAIRS - AER LINGUS

[...]*

City-pairs database

(62) The Commission tested the presence specification also in the city-pairs database. In this case, the tested hypothesis is that if Ryanair serves a different airport to a given city it does not compete at all with Aer Lingus. In this specification all variables are defined at the city pair level (that is to say, the c_ln_seats_air_lingus variable controls for the capacity of Aer Lingus on the given city pair including all airports and similarly the compc_flag_scheduled variable controls for all competing flag carriers on the given citypair). In specifications (7) and (8) the Ryanair effects were separated into airport (EI_FR) versus citypair overlaps (ryan_diff_route). Specifications (9) and (10) control for the potential route level seasonalities. The results of these regressions are presented in Table 3 below.

TABLE 3 -PRESENCE SPECIFICATION - CITYPAIRS - AER LINGUS

[...]*

(63) Ryanair continues to be the carrier that mostly constrains Aer Lingus. Its presence leads to approximately a [...]*% reduction in Aer Lingus' fares. The flag carrier dummy now is also significantly negative with values ranging from [...]*. However, as discussed above flag carriers have an effect on Aer Lingus prices only when they appear on the citypair together with Ryanair, as the inclusion of the comp_FR_flag dummy variable indicates in specifications (3) and (4). Therefore, it is likely that this effect captures the competitive constraint of Ryanair on Aer Lingus (especially in light of the result of Ryanair's presence coefficient which is not significantly affected by the introduction of the interaction dummy). Furthermore, these values should be interpreted jointly with the flag carrier base dummy at the destination citypair. In specifications (5) and (6) where the base dummies are included the base_compc_flag variable's values ranges from [...]*% to [...]*%. Therefore in these specifications the cumulative effect of the flag carriers is negligible. The non-flag carrier with a base at the destination variable also appears to have a significant but positive effect.

(64) Specifications (7) and (8) separate the effect of the head-to-head airport from the city-pair competitive effect of the presence of Ryanair. The estimates indicate that while Ryanair has a much higher effect on Aer Lingus prices when competing headto- head to the same airport ([...]*%), Ryanair also constrains Aer Lingus on the citypair overlaps where the effect is a significant [...]*% reduction in Aer Lingus average fares. However, the Commission notes that these figures should be interpreted with caution (in particular for instances where entry took place at a different airport within a given city pair) as they relate to a much smaller number of entry and exit events (see Appendix A). (13)

Frequency specification

(65) The frequency specification is intended to test the hypothesis that the stronger the presence of one of the merging parties the more pronounced the effect on fares of the other.

(66) The hypothesis is difficult to test because to measure the strength of a carrier on a route it is necessary to use some capacity variable, such as the number of frequencies on the route. However, any capacity variable is likely, to some extent, to be determined simultaneously with fares and the frequency levels of rival carriers (this is the so-called endogeneity problem which could bias the results of the regression analysis).

(67) The frequency specification includes variables for the logarithms of the total number of frequencies of (i) Ryanair (ln_cfreq_ryanair), (ii) the flag carriers (ln_cfreq_flag) and (iii) the non-flag carriers (ln_cfreq_nonflag). The Commission also includes dummy variables to control for the absence (14) of (i) Ryanair (c_absence_ryanair), (ii) the flag carriers (c_absence_flag) and (iii) the non-flag carriers (c_absence_nonflag). The regressions also includes Aer Lingus' log of capacity (ln_seats_air_lingus) or Aer Lingus' log of frequencies (ln_cfreq_aerlingus) as a scale factor. We include demand and cost variables (ldest_frequency, lfuel and lfuelperseat) to limit the endogeneity problem and dummy variables for the bases of flag and non-flag carriers.

(68) Specification (5) to (8) separates the airport (ln_rfreq_ryanair) versus citypair but different airport (ln_cdiff_freq_ryanair) frequencies of Ryanair. Finally, specification (7) and (8) introduce the destination city-pair flag (base_compc_flag) and non-flag (base_compc_noflag) dummies.

(69) The results of the baseline frequency specification are reported in the first column in Table 4 below.

TABLE 4 -FREQUENCY SPECIFICATION - CITYPAIRS - AER LINGUS

[...]*

(70) As anticipated the log frequencies of Ryanair has a significant and negative effect on Aer Lingus' prices. On average a 1% increase in Ryanair frequencies decreases Aer Lingus prices by [...]*%. This effect is robust to the inclusion in the regression of base destination dummies and demand controls.

(71) In this set of regressions the frequencies of flag carriers (ln_cfreq_flag) also have a significant negative [...]* effect on Aer Lingus' fares but smaller in magnitude than that of Ryanair (typically around half). Moreover, similarly to the presence specification, the presence of a flag carrier base at the destination dummy (base_compc_flag) also has a significant but positive effect on Aer Lingus' fares. The demand control (ldest_freq_total) is highly significant, as in the presence specification. Regressions with Ryanair's monthly average net fares as the dependent variable

(72) In the 2007 Decision the regressions on Ryanair's fare were given less weight due to the low number of entry or exit events of Aer Lingus on Ryanair's routes. Since then Aer Lingus has entered or exited from numerous routes where Ryanair operates so the Commission can meaningfully apply the fixed-effects method also on Ryanair's prices.

(73) The Commission has estimated similar regressions on Ryanair fares as on Aer Lingus fares.

TABLE 5 -PRESENCE SPECIFICATION - BASELINE- AIRPORT PAIRS - RYANAIR

[...]*

(74) As shown in Table 5 Aer Lingus' presence (r_EI) appears to have a significant negative [...]*effect on Ryanair's prices. Flag carriers, Aer Arran and City Jet do not appear to constrain Ryanair in any way (the coefficients on Aer Arann are even positive). Non-flag carriers, on the other hand, in the robust regression specifications appear to constrain Ryanair as well as Aer Lingus. This effect however is not present in the weighted regressions. This suggests that the non-flag carrier effect does not arise once the routes are weighted in terms of the revenues generated on each route. Also the presence of a non-flag base at the destination airport has a positive effect. Flag carriers do not appear to constrain Ryanair's average fares.

(75) To further investigate route level effects of Aer Lingus entries and exits on Ryanair average fares the Commission considered adding a number of additional controls to the baseline specification (reported in Table 6). These set of regressions introduce a dummy to control for the presence of charter flights (r_charter), use a direct measure of total costs on the route (lfuel) and include dummies to capture the effect of the presence of at least one flag (base_route_flag) and one non-flag carrier (base_route_noflag) with a base at the destination airport. All these control variables turn out to be highly significant in all specifications but there is little change in the magnitude of the parameters of interest.

(76) Finally, the specifications in columns (7) and (8) include controls for route level seasonality by introducing route level month dummy variables. Route level seasonality seems to matter a lot more in the Ryanair regression, possibly because Ryanair may target relatively more leisure passengers. The effect of Aer Lingus entry increases significantly from [...]*% in both the robust and the weighted regressions. The non-flag carriers' effect remains similar in the robust but becomes significant in the route level seasonally adjusted weighted regressions. Based on these results, nonflag carriers seem to have an effect on Ryanair prices but these effects are about half as large as an Aer Lingus entry on the airport pair.

TABLE 6 -PRESENCE SPECIFICATION - ADDITIONAL CONTROLS - AIRPORT PAIRS - RYANAIR

[...]*

(77) The first eight columns of the citypair level regressions (Table 7) show similar effects. The Aer Lingus dummy has a significant 3-4% negative effect in each specification. Regressions (7) and (8) suggest that most of this effect can be attributed to the route level competition and on the citypair level (aerlingus_diff_route) this negative effect is not statistically significant. However, the Commission notes that these figures should be interpreted with caution (in particular for instances where entry took place at a different airport within a given city pair) as they relate to a much smaller number of entry and exit events (see Appendix B). In the specifications with route specific seasonal dummies, column (9) and (10), Aer Lingus' impact on Ryanair's average fares is even more significant and rises to [...]*%.

(78) Non-flag carriers' presence at the citypair level appears to constrain Ryanair prices to a similar extent to Aer Lingus. However, non-flag carriers' presence has a lower impact than Aer Lingus in the specifications with route specific seasonal dummies (column (9) and (10)). Furthermore, the presence of a non-flag carrier base has a positive effect, counteracting the negative impact of the non-flag carrier presence.

(79) The flag carriers, Aer Arann, CityJet presence dummies are either non-significant or has positive effects. The only exception is CityJet's impact in the last two specifications (where, however, its impact is less than half the impact of Aer Lingus).

TABLE 7 -PRESENCE SPECIFICATION - CITYPAIRS - RYANAIR

[...]*

(80) In the frequency specifications (Table 8) Aer Lingus is the only carrier that has a negative and significant influence on Ryanair. A 1% increase in frequencies of Ryanair decreases Aer Lingus prices by around [...]*%. Interestingly, in the frequency specifications the increase in the frequencies of the non-flag carriers in fact significantly increases Ryanair average fares. Furthermore, the Commission notes that the coefficient of Aer Lingus' frequencies in different airports have significant negative effect on Ryanair's prices (one could interpret the results as indicating that it is not the mere presence of Aer Lingus in different airports within a city-pair that constrains Ryanair but the frequencies it operates). (15) Non-flag carriers' frequencies do not appear to constraint Ryanair's pricing and in some specifications their impact is positive (this is in contrast with the findings in the presence regression).

TABLE 8 -FREQUENCY SPECIFICATION - CITYPAIRS - RYANAIR

[...]*

The Parties' comments on the Commission's regressions analysis following the Statement of Objections

(81) This section reports and responds to the Parties' assessment of the Commission's analysis in Annex III of the SO (16). The economic experts of both Ryanair and Aer Lingus had separate access to the data and the Commission's computer codes in the data room and they have replicated, commented on and criticized the Commission's analysis and interpretation of the results.

The comments and analysis of Ryanair's experts

(82) According to Ryanair the Commission's regression analysis is not robust to different model specifications and does not support the view that Aer Lingus and Ryanair constrain each other. Ryanair raises three criticisms to the Commission's regression analysis.

(83) Firstly, the price impacts estimated by the Commission are much lower if the analysis is restricted to city-pairs where Ryanair and Aer Lingus fly to different airports (different airport analysis).

(84) Secondly, the Commission's regressions are based on a static model, contrary to the analysis in Annex I, and the results are not robust to dynamic model specifications.

(85) Thirdly, the Commission's specifications suffer from an omitted variable bias in that they fail to control appropriately for route specific demand conditions and/or fail to account for endogenously determined fares and frequencies.

(86) Ryanair concludes that the Commission's analysis "suffer from serious shortcomings." The Commission disagrees with this conclusion. In this section the Commission addresses the criticisms raised by Ryanair.

Different airports analysis

(87) Ryanair argues that "the Commission baseline model predicts that Aer Lingus' presence does not have a statistically significant impact on Ryanair's fares" on routes between city-pairs where Ryanair and Aer Lingus fly to different airports. Moreover, "the Commission's baseline model predicts that Ryanair's presence may not have a statistically significant impact on Aer Lingus' fares on routes between city pairs where Ryanair and Aer Lingus fly to different airports, as suggested by the results of the analysis in scenario (5) [...]". (17) Ryanair bases these arguments on model estimations where the estimation sample is restricted to those city pairs where the two carriers fly to different airports.

(88) The Commission disagrees with this statement and observes that Ryanair has misinterpreted their own findings, has applied the wrong methodology to the wrong data, and when the Commission applied their suggestions properly the effects are still significant.

(89) Firstly, the Commission does not consider reliable such estimates given the small sample size. In the Aer Lingus citypair regressions provided in above Table 3 the Commission noted that "Ryanair also constrains Aer Lingus on the city-pair overlaps where the effect is a significant [...]*% reduction in Aer Lingus average fares. However, the Commission notes that these figures should be interpreted with caution (in particular for instances where entry took place at different airport within a given city pair) as they relate to a much smaller number of entry/exit events (see Appendix A)." Furthermore in the Ryanair citypair regressions provided in above Table 7 the Commission observes that "Regressions (7) and (8) suggest that most of this effect [i.e. of Aer Lingus's entry on Ryanair prices] can be attributed to the route level competition and on the city-pair level (aerlingus_diff_route) this negative effect is not statistically significant. However, the Commission notes that these figures should be interpreted with caution (in particular for instances where entry took place at different airport within a given city pair) as they relate to a much smaller number of entry/exit events (see Appendix B)." In particular there are only seven entry and exit events in each of the regressions.

(90) Also, Ryanair suggested estimating the regressions only on those limited number of city-pairs where entry (of either Ryanair or Aer Lingus respectively) took place at different airports within the given city-pair. Due to the limited number of entry/exit events, the estimations are much less reliable. The results of Ryanair show that Aer Lingus had a non-significant [...]*% impact in most of the robust regressions and a non-significant [...]*% impact in the weighted regressions on Ryanair's prices. On the other hand, they have found that relative to the SO's estimates the impacts of Ryanair on Aer Lingus prices are much larger. In 6 specifications out of 7 reported in table 4.2 of Ryanair's submission (18) the impact is significant and between [...]*%. In the remaining specification, the "direct" effect is a non-significant [...]*% but the effect on routes where both Ryanair and a flag carrier are present increases to (significant) [...]*%. Given that the flag carriers' "direct" impacts are always nonsignificant (and in one specification significantly positive) this effect could be safely assigned to Ryanair's entry. Therefore, at least to what concerns the impact of Ryanair on Aer Lingus' fare Ryanair has largely misinterpreted its own findings. Interpreted correctly, these results show that Ryanair indeed has an economically and statistically significant effect of Aer Lingus' prices.

(91) Secondly, Ryanair has applied the wrong methodology to the wrong data set. As already explained, Ryanair has restricted the sample to city pairs where the Parties fly to different airports. Ryanair used the Commission's airport pairs specification with this sample. The Commission notes that there is a clear contradiction in doing so, as the Commission's airport pairs specification is meaningful only when both Parties fly to the same airport. For example, in the Aer Lingus' airport pairs regression the explanatory variable of interest (r_FR, a dummy variable equal to one when Ryanair is present on a route and zero otherwise) captures the effect of the presence of Ryanair on Aer Lingus' fares when both carriers fly to the same airport. If the sample is restricted to city pairs where the Parties fly to different airports, as the Notifying Party has done, r_FR will be constant and equal to zero in the whole subsample. Any meaningful coefficient estimate of r_FR will be therefore impossible. The Commission notes that Ryanair manages to estimate the coefficient by restricting the sample to the wrong data set. Ryanair has arbitrarily limited the sample to some city pairs which appears to reflect the instances where the Parties fly to the same city pair but not to the same airport today, that is, in the last observed time period of the sample. However, there are instances that within these selected pairs while currently the two air carriers do not fly to the same airport, they might have flown to the same airport in the past. For example, Ryanair restricts the sample to (among others) the city pair Dublin-Malpensa/Linate/Bergamo. Currently Ryanair flies to Bergamo, whereas Aer Lingus flies to Malpensa and Linate (i.e. they fly to different airports within the city pair). However Aer Lingus (as well as Ryanair) flew to Bergamo from April 2006 to October 2006, and this event is the only one taken into account by Ryanair's regression on this city pair. In short, Ryanair's methodology measures the impact that the presence of one carrier has on the other carrier's fares when they both fly to the same airport, unreasonably restricting the sample to some city pairs. Therefore, the Commission finds these regressions irrelevant and meaningless for the scope of this analysis.

(92) The Commission applied Ryanair's suggested model to the right data (that is to say, excluding the city pairs in which in at least one period Ryanair and Aer Lingus flew to the same airport). Table 9 below reports the Ryanair results and Table 10 reports the Aer Lingus results.

TABLE 9 -PRESENCE SPECIFICATION - CITYPAIRS-DIFFERENT AIRPORTS - RYANAIR

[...]*

TABLE 10 -PRESENCE SPECIFICATION - CITYPAIRS-DIFFERENT AIRPORTS - AER LINGUS

[...]*

(93) While the impact of Aer Lingus' entry on Ryanair's fares when the entry occurs in a different airport is significant only in the robust regressions, Ryanair's impact on Aer Lingus's fares is significant in all specification and in magnitude similar to what was reported in the baseline analysis above. All specifications confirm the findings that neither destination-based flag nor non-flag carriers exert a significant constraint on Aer Lingus. Table 9 indicates that non-flag carriers seem to constrain Ryanair by about the same magnitude as Are Lingus does.

Dynamic model

(94) Ryanair argues that "The static nature of the model may further bias the Commission's results. Failure to specify a dynamic model when the true data generating process exhibits intertemporal dependence is likely to bias the coefficient estimates ... downward." Furthermore, "if a dynamic model is applied to the data used by the Commission, the estimated constraint that the airlines pose on each other may be smaller compared to the Commission's static model and the constrain is statistically insignificant in some model specifications." (19)

(95) Ryanair proposed seven different dynamic models that included the lagged fares of Ryanair to control for the serial correlation. Ryanair reports that "for example, the estimated long-run impact of Ryanair's presence on Aer Lingus fares is [...]*% in model (2) in table 4.3, which is less that then [...]*% range estimated by the Commission. The long-run estimates obtained in the various model specifications in table 4.3 are between [...]*%." (20)

(96) The Commission believes that this criticism is inconsequential. The Commission acknowledges that just as in the analysis of the 2007 Decision serial correlation is present in Ryanair's prices. Serial correlation arises when the prices charged in the past could affect the level of prices today. As it was stated also in the 2007 Decision, serial correlation does generally not affect the consistency of the estimation results (provided that there is no lagged dependent variable included on the right hand side of the regression, as is the case in the Commission's regressions).

(97) In any case, the Commission has tested the specifications proposed by Ryanair and found that the results remain robust. The fact that the average fare charged by Ryanair on a certain route at a given point in time could to some extent depend on past values of the fare for that route has no effect on the coefficient for variables assessing competition, given the setting of the Commission model.

(98) Firstly, in four of Ryanair's specifications the long-run effects of Ryanair entry were significant and similar to those found by the Commission ([...]*% versus [...]*% found by the Commission). This clearly implies that irrespective of the estimation procedure the competitive constraint that Ryanair exerts on Aer Lingus is economically significant. Even in specification (4) of Table 11 where Ryanair's "direct" effect is non-significant, the effect on routes where both Ryanair and a flag carrier are present increases to a significant [...]*%. Since the same estimation finds significant positive effect for flag carriers this negative impact is very likely to be caused by Ryanair's entry.

(99) Secondly, in the two specifications provided by Ryanair where Ryanair's impact was not significantRyanair did not use the estimation command properly (21). The specifications where the Ryanair entry impact is not significant are implemented using the xtabond2 STATA command. However, Ryanair failed to specify the instruments properly. Ryanair has not specified the time and route dummies as included instruments. (22) Ryanair's experts responded that "even if it were the case that all variables are treated as endogenous (and therefore instrumented by their lags), this would not impede the consistency of the obtained results, but it would reduce the efficiency of the gmm estimator." (23) While this statement is true, the efficiency or the bias of the estimates can be rather substantial especially if the instruments are weak in the statistical sense. Moreover, it has to be noted that the implication of applying a less efficient statistical estimator can result in finding statistically non-significant price effects, even when there are significant such effects. (24) Hence, using an inefficient statistical estimator (as admitted by Rynair) is not a sound practice when one argues, as Rynair does, that the price effects are statistically non-significant. The Commission has re-estimated the model with the proper instrumental variables specifications and found that Ryanair's impact remained significant and negative (shown in table 11 below).

(100) To correct for the presence of serial correlation and the heteroskedasticity, another possible approach, also used in the 2007 Decision, is to use the estimation method proposed by Newey/West (implemented via the newey2 command in STATA). This correction allows the error term to be heteroskedastic and/or correlated within each route. The only limiting assumption imposed by this method is that the serial correlation decays relatively fast overtime within each route. This hypothesis is reasonable in the present case.

(101) Table 11 below presents the results for the Commission's implementation of the model suggested by Ryanair and the Newey correction model applied to the main specifications. The results reported are fully consistent with the results of the Commission's analysis.

(102) The first four specifications report estimation results of the specification used by Ryanair with proper instrumental variable definitions. The last 3 models include the specification used by the Commission in the 2007 Decision to control for serial correlation. The direct long run effect ranges between [...]*% and [...]*% in the specifications suggested by Ryanair. The effect is larger when flag carriers are also present on the route. In the specification used by the Commission in 2007, the direct effect ranges were between [...]*% and [...]*%. In each specification, the findings that neither destination-based flag nor non-flag carriers exert a constraint on Aer Lingus are confirmed.

TABLE 11 -DYNAMIC PRESENCE SPECIFICATION - AIRPORTS - AER LINGUS

[...]*

Risk of omitted variable bias and endogeneity

(103) Ryanair also raises concerns that "while the Commission recognises the importance of accounting for unobserved or unmeasured factors in the analysis (see para 13 of Annex III), its modelling approach only controls for unobserved or unmeasured factors that do not vary across time. Regional demand drivers which tend to vary over time - such as unemployment - are therefore not captured by the analysis." Furthermore, "the competitor variables used in the Commission's analysis (see para 57c) are assumed to be independent of other variables in the model, such as demand conditions. However, it is likely that this assumption is violated." Finally, "the capacity variable used in the Commission's analysis may also violate the independence assumption with respect to other variables in the model."

(104) Firstly, the Commission acknowledges, as it also did in the 2007 Decision, that all entry and capacity variables could indeed depend on other variables, such as demand conditions. However, the chosen estimation methodology mitigates this risk. Carriers generally have to determine the level of capacity well before they observe demand and hence the level of capacity is predetermined relative to demand. Furthermore, despite the potential methodological difficulties arising from such dependencies across variables, the coefficients of Ryanair' and Aer Lingus' presence are robust to the use of different estimation methodologies (including the instrumental variables estimations presented above in Table 11).

(105) Secondly, by including route and time fixed-effects the Commission controlled for unobserved route or time invariant characteristics that could correlate with entry and capacity variables. This correlation could bias the results if the fixed effects were not included. The parameters of interest are estimated from the changes in the relevant variable on a particular route, and not from the changes in the variable across routes or across time in general. In other words, unobserved or unmeasured route or time specific factors do not bias the estimation results.

(106) Furthermore, the Commission has estimated different specifications with different time dummies. The Commission's regressions include both route and time dummy variables to control for route and also time varying fixed effects. In particular the specifications with time varying fixed effects (specification (7) and (8) in Table 2 and Table 6) capture all time varying effects that are route specific, for example, regional demand drivers such as unemployment. The results remain robust.

Aer Lingus' comments

(107) Aer Lingus agrees with the Commission's focus on fixed-effects regression analysis, which "mirrors" the quantitative analysis Aer Lingus has also undertaken (see above). Aer Lingus makes a number of comments broadly in support of the conclusion reached by the Commission from its own fixed-effects regressions that Ryanair exercises a strong competitive constraint on Aer Lingus' prices. Aer Lingus examined the robustness of the Commission's findings adopting the Commission's methodology on different samples.

(108) Firstly, Aer Lingus restricts the sample considering a more recent period, from 2009 to 2012. In the airport pairs specification, Aer Lingus' fares decrease by 7-9% when Ryanair is present on the route. These results are confirmed in the city pairs specification, where Aer Lingus' fares decrease by 6-8% when Ryanair is present. Additionally, Aer Lingus finds that for the period 2009-2012, Aer Lingus' fares were not constrained by the presence of any carriers other than Ryanair.

(109) Secondly, Aer Lingus considers only Aer Lingus' "key routes", identified by either:

- Excluding the routes where Aer Lingus operated for less than 12 months;

- Considering only the routes where Aer Lingus operated for 4 or more years;

- Considering the routes where Aer Lingus operated for the whole period.

(110) In all subsamples, the presence of Ryanair is associated with a (statistically significant) drop in Aer Lingus' fares. In particular, Aer Lingus' fares fall by 10-15% when Ryanair is present in the same airport (airport pairs specification) and 9-13% when Ryanair is present in the same city pair (city pairs specification).

(111) Thirdly, Aer Lingus further examined the robustness of the Commission's findings by eliminating from the airlines presence variables what Aer Lingus considered "spurious entry and exit events". The results by adopting these modified presence variables are similar to the ones mentioned above.

(112) In conclusion, according to Aer Lingus the Commission thus finds clear evidence that Ryanair significantly constrains Aer Lingus' pricing behaviour. They also note that the results of the Commission's fixed-effect regressions are in line with those submitted by Aer Lingus. This pertains to both the magnitudes of the estimated coefficients, and the conclusion that the Commission reached.

(113) The Commission has carefully considered the comments and criticisms made by both Ryanair and Aer Lingus with regard to the Commission's regression analysis. Most of the criticisms were similar to those made in the 2007 Decision. The Commission's further analysis confirms that the findings that the presence of Ryanair has a statistically and economically significant effect on Aer Lingus' prices (and vice versa) are very robust. Conclusion

(114) The Commission's price regression analysis tests the following hypothesis:

- the presence of one of the Parties on the route is associated with a statistically and economically significant reduction in the fares of the other;

- the Parties exert on each other a stronger competitive constraint than any other existing competitor;

- the existence of an actual or potential competitor with a significant presence at the destination airport on a route originating in Dublin, Cork and Shannon has an impact on the Parties' prices;

- a stronger presence of one of the Parties (in terms of number of frequencies) has a more pronounced effect on the other's fares.

(115) The fixed-effects regressions with Aer Lingus's fares as the dependent variable show that Ryanair exerts a significant competitive constraint on Aer Lingus's fares. In particular the Commission's analysis finds the following results:

- First depending on the specification, Ryanair's presence is associated with Aer Lingus charging around 9-14 % lower prices when considering airport-pair and around 7 - 11% lower prices when considering city-pairs. This effect is economically and statistically significant in all tested regressions. This result is also robust, correcting for the presence of outliers, heteroskedasticity and serial correlation. It is also highly robust to the use of alternative specifications including alternative demand and supply controls. Notably, in practically all cases the control variables in the different regressions have the expected signs and are statistically significant. The explanatory power of the regression (that is to say, the fit of the model) is also high with R2 consistently around 80%. Ryanair also appears to impose a more significant constraint on Aer Lingus when it serves the same airport;

- Secondly, comparing the coefficients of Ryanair with that of flag-carriers and non-flag carriers, as well as Aer Arann and CityJet, Ryanair's presence or number of frequencies have a much stronger economic impact than that of any other type of carrier. This also holds true for charter carriers. In fact, in most cases the regressions indicate that the presence of other carriers has no economic or statistically significant effect on Aer Lingus fares. The presence of flag carriers has a negative effect on the average fares of Aer Lingus but at the same time this effect seems to be significant only if Ryanair is also present on the route and is typically significantly smaller than the effect of Ryanair;

- Thirdly, neither destination-based flag carriers nor destination-based flag carriers exert a significant constraint. Moreover a flag carrier base at the destination airport has a significant positive effect on Aer Lingus prices, limiting the effect of flag carriers. Thus it cannot be expected that the merged entity would be effectively constrained by flag or other non-flag carriers post- Transaction;

- Fourthly, measuring the strength of Ryanair's presence using the number of frequencies on the route as a proxy provides further confirmation that Ryanair constrains Aer Lingus. Depending on the specification, the price effect of a 1 % increase in frequencies of Ryanair on Aer Lingus prices is negative at around 5-6 %. This adds to the robustness of the results derived from the presence specifications.

(116) The fixed-effects regressions with Ryanair's prices as the dependent variable also support the same conclusion. Aer Lingus exerts a competitive constraint on Ryanair's prices.

- Firstly the presence of Aer Lingus is associated with Ryanair charging around 3-7 % lower prices when considering airport-pair and around 3-4 % lower prices when considering city-pairs. This effect is economically and statistically significant in all tested regressions. This result is also robust, correcting for the presence of outliers, heteroskedasticity and serial correlation. It is also highly robust to the use of alternative specifications including alternative demand and supply controls. The explanatory power of the regression is also high with R2 consistently above 80%. On the city pair analysis, Aer Lingus appears to impose a statistically significant constraint on Ryanair only when it serves the same airport (however there is a limited number of entry/exit events when they fly to different airports).

- Secondly, comparing the coefficients of Aer Lingus with that of flag-carriers and non-flag carriers, as well as Aer Arann and CityJet, Aer Lingus's presence or number of frequencies have stronger economic impact than that of any other type of carrier. In fact, in most cases the regressions indicate that the presence of flag-carriers, as well as Aer Arann and CityJet, has no economic or statistically significant effect on Ryanair fares. The presence of non-flag carriers has a significant negative effect on Ryanair's prices but this effect is not robust in the different specifications.

- Thirdly, neither destination-based flag nor non-flag carriers exert a constraint on Ryanair. Moreover a non-flag carrier base at the destination airport has a significant positive effect on Aer Lingus prices, limiting the effect of non-flag carriers. Thus it cannot be expected that the merged entity would be effectively constrained by flag or other non-flag carriers post-Transaction.

- Fourthly, measuring the strength of the presence of Aer Lingus using the number of frequencies on the route as a proxy provides further confirmation that Aer Lingus constrains Ryanair. The price effect of a 1 % increase in Aer Lingus' frequencies on Ryanair prices implied by the Commission's frequency regressions is around a 2 % decrease. This adds to the robustness of the results derived from the presence specifications. Also, the frequencies of non-flag carriers do not appear to constrain Ryanair's pricing.

(117) The Commission's regression analysis confirms and complements the conclusions set in the Decision on the basis of qualitative evidence regarding the closeness of competition between Ryanair and Aer Lingus (for routes from Dublin, Cork and Shannon). The results indicate that Aer Lingus prices are currently constrained by competition from Ryanair and vice versa.

(118) Furthermore, post-Transaction as predicted by standard ''non-coordinated effects" analysis, both carriers would internalise the effects of setting higher fares on each other. In particular, the merged entity would have the incentive to set higher fares for Aer Lingus as most of the customers lost would be captured by Ryanair (and similarly set higher prices for Ryanair as most of its customers would be captured by Aer Lingus).

Footnotes Annex III:

1 Aer Lingus' Comments to Request for Information 8.

2 Aer Lingus' response to the Commission's request of information of 8 October 2012. Aer Lingus has further provided data with a split between Flex and other fares for the Dublin, Cork and Shannon routes (25 October 2012).

3 For Ryanair this includes the published airport taxes, while for Aer Lingus this only refers to the net fare.

4 Statistical significance is a formal way of assessing whether observed associations are likely to be explained by chance alone.

5 If the unobserved variables are expected to be uncorrelated with the other explanatory variables an alternative approach known as "random-effects" is likely to be more efficient than the fixed-effects method.

6 Aer Arann franchise agreement with Aer Lingus, details were provided by Aer Arann, see Aer Arann's reply to Commission's request for information of 17 October 2012.

7 This refers to the activities of Aer Arann before the introduction of the franchise agreement. Once a franchise agreement is reached for each route then the Commission has treated Aer Arann's operations as part of Aer Lingus operations. See in Section 8.2 of this Decision, where the Commission considers in particular that Aer Arann is a competitor of Ryanair, but not of Aer Lingus.

8 See Aer Lingus' response to the Commissions' request of information of 8 October 2012.

9 A carrier is considered to have a base at the destination airport if it operates more than 100 flights a week from that airport to any other short-haul destination. This roughly corresponds to operating 2-3 planes with a turnover slightly above 3 return flights per day for every day of the week. The base variable however, is not necessarily able to capture separately the effect of a new base at the destination airport and it may also reflect partially some variation such as seasonality. Hence, the parameter estimates of the base variables should be interpreted with some caution.

10 The City Jet presence was adjusted relative to the DAA data. The dummy takes into account that Air France's routes ex-Dublin, except Paris-Dublin are marketed by CityJet.

11 Aer Lingus' capacity (ln_seats_air_lingus), now has a significant and negative coefficient since the lfuel variable controls for any possible endogeneity of the capacity variable. Given frequencies, higher supplied capacity results in lower average fares.

12 Charter's presence remains positive but is now smaller at 4-5%. Aer Arann's presence has no significant effect on Aer Lingus' average fares.

13 The rows in bold indicate a city pair in which the entry/exit event took place in different airports.

14 The absence dummies are needed since the logarithm of 0 is undefined and therefore these observations would be dropped.

15 Furthermore, in the frequency specification there is greater identification power as there is greater variation in the frequency variable of Aer Lingus compared to the entry/exit of Aer Lingus in these airports (different airports within the same city pair).

16 Oxera, "Assessment of the Commission's analysis in the Annexes I, II, and III of the SO,"

17 Oxera, "Assessment of the Commission's analysis in the Annexes I, II, and III of the SO,", page 8

18 Oxera, "Assessment of the Commission's analysis in the Annexes I, II, and III of the SO,"

19 Oxera, "Assessment of the Commission's analysis in the Annexes I, II, and III of the SO," page 10.

20 In its reply, Ryanair refers to the Kennedy adjustment. Namely, if b is the estimated coefficient on a dummy variable and V(b) is the estimated variance of b then: g = 100 (exp(b - V(b)/2) - 1) gives an estimate of the percentage impact of the dummy variable on the variable being explained. See for example Van Garderen, Kees Jan and Shah, Chandra, "Exact Interpretation of Dummy Variables in Semilogarithmic Equations". The Econometrics Journal, Vol. 5, pp. 149-159, 2002. This adjustment is necessary when the coefficient on a dummy variable suggests a relatively large proportionate change in the level of the dependent variable. When the coefficient is relatively small the difference between the approximate and the exact interpretation is minimal. In the regressions reported in this annex, the coefficients of interests are relatively small, mostly below 0.20, so the difference do not change the results significantly neither qualitatively nor quantitatively.

21 There was one specification where standard errors were not reported due to "nonsymmetric or highly singular variance matrix."

22 Roodman (2006) "How to Do xtabond2: An introduction to "Difference" and "System" GMM in Stata" explains clearly (p.39):"The most important thing to understand about the xtabond2 syntax is that unlike most Stata estimation commands, including xtabond, the variable list before the comma communicates no identification information. The first variable defines Y and the remaining ones define X. None of them say anything about Z even though X and Z may share columns. Designing the instrument matrix is the job of the ivstyle() and gmmstyle() options after the comma".

23 Email correspondence between the Commission and Notifying Party's economic consultants of 5 December 2012.

24 Efficiency is a measure of the precision of the estimator of a parameter. An estimator which is more efficient than another has a lower variance, that is, it has a smaller statistical uncertainty surrounding it. The more efficient estimators are preferred to the less efficient ones because they allow more precise, accurate statistical inference. When a test based on an efficient estimator shows that the parameter estimate is non-significant, this can be a strong indication that the true parameter value is indeed not different from zero. When, however, a test based on a less efficient estimator shows non-significance, there is a higher likelihood (compared to the case of the test based on the efficient estimator) that this result reflects only the statistical uncertainty of the estimator and not that the true parameter value is zero.

Appendix A: Entry/Exit Events of Ryanair

<TABLE>

Appendix B: Entry/Exit Events of Aer Lingus

<TABLE>