Commission, February 12, 2019, No M.8964

Subject: Case M.8964 - DELTA / AIR FRANCE-KLM / VIRGIN GROUP / VIRGIN ATLANTIC

Commission decision pursuant to Article 6(1)(b) of Council Regulation No 139/2004 (1) and Article 57 of the Agreement on the European Economic Area (2)

Dear Sir or Madam,

(1)    On 08 January 2019, the European Commission received notification of a proposed concentration pursuant to Article 4 of the Merger Regulation by which Air France-KLM S.A. (“AFKL”, France), Delta Air Lines, Inc. (“Delta”, the United States of America) and Virgin Group (the British Virgin Islands) acquire within the meaning of Article 3(1)(b) and 3(4) of the Merger Regulation joint control of the whole of Virgin Atlantic Limited (“VAL”, the United Kingdom) by way of purchase of shares (“the Transaction”). (3) AFKL, Delta, Virgin Group and VAL are designated hereinafter as the “Parties”.

1. THE PARTIES

(2)    AFKL is a French company and is the holding company of Société Air France (“AF”), the French national carrier airline, and Koninklijke Luchtvaart Maatschappij N.V. (“KL”), the Dutch national carrier airline. (4) AFKL provides passenger air transport services, cargo air transport services and maintenance, repair and overhaul services (“MRO”). In 2017, AFKL carried 99 million passengers to 314 destinations in 116 countries with a fleet of 537 aircraft in operation.

(3)    Delta is an international airline with headquarters in the United States. Delta provides passenger air transport services and cargo air transport services, using belly space on its passenger aircraft, as well as MRO services. Delta carries approximately 180 million passengers a year to more than 300 destinations in  over 50 countries with a fleet of more than 800 aircraft.

(4)    Virgin Group is the holding company of a group of companies, including VAL, active in a wide range of products and services worldwide. The Virgin Group holding company, Virgin Group Holdings Limited, is wholly-owned by Sir Richard Branson.

(5)    VAL is the holding company of Virgin Atlantic Airways Limited (“Virgin Atlantic”) and Virgin Holidays Limited (“Virgin Holidays”). Virgin Atlantic is an international airline with headquarters in the United Kingdom (“UK”), providing passenger air transport services and cargo air transport services, using belly space on its passenger aircraft, as well as MRO services. Virgin Atlantic flies to 34 destinations worldwide. Virgin Holidays is a UK tour operator which supplies and distributes package holidays and ancillary products.

Pre-existing corporate and commercial relationships between the parties

(6)    Delta is a shareholder of both AFKL and VAL, and the Parties have existing alliance and joint venture relationships:

-   AF and Delta were founding members of the SkyTeam alliance in 2001. Since 2004 (shortly after the merger of AF and KL), KL has also been a full member of the SkyTeam alliance.

-   AFKL and Delta are also members of SkyTeam Cargo, a global airline cargo alliance, started in 2000.

-   Delta, AF, KL and Alitalia Compagnia Aerea Italiana S.P.A. (“Alitalia”) are parties to the TATL Joint Venture, a metal neutral joint venture arrangement covering various routes between Europe and North America, and other geographic regions. (5)

-   Delta and Virgin Atlantic are parties to the Delta-Virgin Atlantic Joint Venture, a metal neutral joint venture arrangement covering non-stop routes between the UK and North America. (6)

-   Delta, AF, KL, Alitalia and Virgin Atlantic are also parties to a coordination agreement dated 8 April 2013 (the “5-Way Coordination Agreement”), pursuant to which these parties established a framework for coordination of commercial activities covering city pair routings between the UK and North America. (7)

2. THE CONCENTRATION

(7)    The Transaction concerns the proposed acquisition by AFKL of a 31% joint- controlling interest in VAL. In addition, the Parties intend to implement a (new) transatlantic joint venture agreement (“JVA”) (8) between AFKL, AF, KL, Delta  and Virgin Atlantic and the Enhanced Strategic Cooperation  Agreement (“ESCA”) described in the ESCA memorandum of understanding (“ESCA MOU”) (9)  between Virgin Atlantic, AFKL, AF and KL.

AFKL entering as a new shareholder

(8)    Pre-Transaction, Virgin Group and Delta respectively own 51% and 49% of the shares in VAL and have joint control over VAL. (10)

(9)    Pursuant to a Sale and Purchase Agreement (“SPA”)  (11), executed on 15 May 2018, AFKL will acquire a 31% equity interest in VAL from Virgin Group.

(10)   Post-Transaction, AFKL will therefore own  31%,  Virgin  Group  20%  and  Delta 49% of the shares in VAL.

Implementation of a transatlantic joint venture agreement and the ESCA MOU

(11)   Pursuant to the JVA between AFKL, AF, KL, Delta and Virgin Atlantic, the Parties intend to establish a new transatlantic joint venture agreement   combining (i) [details of cooperation arrangement] (12)  and (ii) the metal neutral joint  venture arrangement between Delta and Virgin Atlantic (the “Delta-Virgin Atlantic Joint Venture” (13)). This expanded joint venture will cover the Parties [details on potential parties to new transatlantic joint venture agreement] (14) [further details on the joint venture agreement].

(12)   In addition to the JVA, Virgin Atlantic, AFKL, AF and KL have concluded an ESCA MOU which regulates how AFKL and Virgin Atlantic intend to organise and implement bilateral cooperation on certain routes between Europe (including the UK) and the rest of the world (excluding North America). (15)

Scope of this Decision

(13)   This Decision only assesses the impact on competition resulting from the acquisition by AFKL of a joint-controlling interest in VAL. The Commission  does not assess the impact on competition resulting from the pre-existing relationships between AFKL and Delta as these have been assessed in the Commission’s Decision in case AT.39964 – Air France/KLM/Alitalia/Delta. (16) In that respect, the Parties have notified the Transaction based on the assumption  that the TATL Joint Venture and Delta-Virgin Atlantic Joint Venture are in place. However, this decision is without prejudice to the application of Article 101 TFEU to any existing or envisaged agreement between the Parties, including without limitation the TATL Joint Venture, the DL-VS Joint Venture, the JVA and the 5-Way Coordination Agreement.

Joint control

(14)   The Board of Directors of VAL will consist of [details about composition of the Board of Directors]. (17) [Details about appointment of the Chairman]. (18) [Further details about appointment of the Chairman]. [Details about appointment of executive director]. (19)

(15)   [Details about voting mechanism] (20); [further details about voting mechanism]. (21)

(16)   The approval of strategic business plans, annual budgets, borrowings beyond the current annual budget, capital expenditure beyond the current annual budget, and any material contract not authorised by the current annual budget (each a “Special Matter”), will be subject to a special approval process, [details about voting mechanism and veto rights]. (22) [Further details about voting mechanism and veto rights].

(17)   In addition, each of Delta, AFKL and Virgin Group has a veto right over [details about veto rights]. (23)

(18)   Besides, for the sake of completeness, the Shareholders Agreement provides that, upon implementation of the JVA, which is conditional upon all necessary regulatory approvals having been obtained, (24)  [details about veto rights]. (25)

(19)   According to paragraphs 62 and 69 of the Consolidated Jurisdictional Notice, the power to co-determine the structure of the senior management, such as the members of the board and the power to block actions which determine the strategic commercial behaviour of an undertaking, usually confers upon the  holder the power to exercise decisive influence on the commercial policy of an undertaking.

(20)   As a result, post-Transaction, Delta, AFKL and Virgin Group will jointly control VAL and its subsidiaries (notably Virgin Atlantic and Virgin Holidays). (26)

Full-functionality

(21)   VAL has sufficient own staff, financial resources and dedicated management for its operation and for the management of its portfolio and business interests. Furthermore, VAL will not be limited to exercising a specific function for its parents as it will continue to have a market presence. It also does not have significant sale or purchase relationships with its parents. Finally, VAL is intended to operate on a lasting basis. Therefore, VAL is a full-functional joint venture.

3. EU DIMENSION

(22)   The undertakings concerned have a combined aggregate worldwide turnover of more than EUR 5 000 million (27) (AFKL: EUR 25 781 million, Delta: EUR 36 499 million, Virgin Group: EUR […] and VAL: EUR 3 038 million). Each of them  has an EU-wide turnover in excess of EUR 250 million (AFKL: EUR […],   Delta: EUR: […], Virgin Group: EUR […] and VAL: EUR […]), and they do not achieve more than two-thirds of their aggregate EU-wide turnover within one and the same Member State. (28)  The notified operation therefore has an EU dimension.

4. RELEVANT PRODUCT AND GEOGRAPHIC MARKETS

(23)   The Parties' activities overlap in (i) passenger air transport services (29), (ii) cargo  air transport services and (iii) maintenance, repair and overhaul (“MRO”) services.

4.1. Identification of the relevant markets in respect of air transport services of passengers

4.1.1. Relevance of the O&D approach

(24)   In respect of air transport services of passengers, the Commission has, in its prior decision practice related to air transport, defined the relevant markets for scheduled passenger air transport services on the basis of two approaches: (i) the "point of origin/point of destination" ("O&D") city-pair approach, where  the target was an active air carrier; (30) and (ii) the "airport-by-airport" approach, when the target included an important slot portfolio. (31)

(25)   Under the O&D approach, every combination of an airport or city of origin to an airport or city of destination is defined as a distinct market. 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. The effects of a transaction on competition are thus assessed for each O&D separately.

4.1.2. Relevance of the airport-by-airport approach

(26)   Under the airport-by-airport approach, every airport (or substitutable airports) is defined as a distinct market. Such a market definition has notably been adopted to assess the risks of foreclosure entailed by the concentration of slots at certain airports in the hands of a single undertaking. (32) Under this approach, the effects of a transaction on competition are thus assessed for all O&Ds taken together to or from an airport (or substitutable airports).

4.1.2.1. Introduction

Slots as an input for air transport services

(27)   By virtue of the Slot Regulation, (33) slots, i.e. the permission to land and take-off at a specific date and time at congested airports, are essential for airlines' operations. Indeed, only air carriers holding slots are entitled to get access to the airport infrastructure services delivered by airport managers and, consequently, to  operate routes to or from those airports.

(28)   The Commission has, in its prior decision practice, highlighted that the lack of access to slots constitutes a significant barrier to entry or expansion at Europe's busiest airports, such as London Heathrow airport. (34)

(29)   The Commission has also insisted, in the framework of its airport policy, that "slots are a rare resource" and "access to such resources is of crucial importance for the provision of air transport services and for the maintenance of effective competition." (35)

(30)   In addition, the Slot Regulation recalls that, with the increase of air traffic, there  is a continuously growing demand for capacity at congested airports. (36) Therefore, the lack of available slots has become a prominent feature of the EU airline industry and is expected to become an even more critical issue for air carriers in the near future.

Rules for the allocation of slots

(31)   In the context of the imbalance between demand and supply of airport capacity, the Slot Regulation defines the rules for the allocation of slots at EU airports. It aims to ensure that, where airport capacity is scarce, the latter is used in the fullest and most efficient way and slots are distributed in an equitable, non- discriminatory and transparent way.

(32)   Under the Slot Regulation, the general principle regarding slot allocation is that  an air carrier having operated its particular slots for at least 80% during the summer or winter scheduling period is entitled to the same slots in the equivalent scheduling period of the following year (the "grandfather rights"). (37) Conversely, slots which are not sufficiently used by air carriers (below 80%) are reallocated to other air carriers (the "use it or lose it" rule).

(33)   The Slot Regulation also provides for the setting up of "pools" containing newly- created time slots, unused slots and slots which have been given up by a carrier or have otherwise become available (e.g. via the “use it or lose it” rule). 50% of the slots from the slot pool shall first be offered to new entrants. The other 50% of the slots from the slot pool shall be placed at the disposal of other applicant airlines (incumbent airlines). If applications by new entrants amount to less than 50% of the capacity made available through slots from the pool, this remaining capacity shall also be placed at the other applicants' disposal. (38)

(34)   Under the Slot Regulation, slots cannot be traded. They may however be exchanged or transferred between airlines in certain specified circumstances and subject to the explicit confirmation from the coordinator under the Slot Regulation. (39)

4.1.2.2. Parties’ views

(35)   The Parties do not object to the airport-by-airport approach but submit that the Transaction does not make any material change to the position of any of the Parties at any EEA airport. (40)

4.1.2.3. Commission’s assessment

(36)   According to the Explanatory Memorandum for the Commission Proposal for a Regulation of the European Parliament and of the Council on common rules for the allocation of slots at European Union airports, (41) "the emergence of a strong competitor at a given airport requires it to build up a sustainable slot portfolio  to allow it to compete effectively with the dominant carrier (usually the “home” carrier)."

(37)   In this context, in a number of prior decisions related to transactions entailing the transfer of slots at certain airports, the Commission has considered the effects of the transaction on the operation of passenger air transport services at a given airport in terms of the slot portfolio held by a carrier at the airport, without distinguishing between the specific routes served to or from that airport. (42) Under this approach, the Commission assesses how the transaction strengthens the merged entity's position at certain airports and the potential effects thereof on the merged entity’s ability and incentive to foreclose other air carriers from accessing the relevant airports. Foreclosing access to airports may in turn foreclose those other air carriers from operating routes from/to the relevant airports. (43)

(38)   AFKL’s major hubs are Amsterdam and Paris-Charles de Gaulle airports, (44) where Delta holds a negligible amount of slots. (45) Delta’s slots at these airports are all used for existing TATL Joint Venture services. Virgin Atlantic does not hold any slot (and historic rights thereto) at Amsterdam or Paris-Charles de Gaulle airports, nor at any EEA airport outside the United Kingdom. (46) Therefore, the Transaction will not have any impact on any of the Parties’ slot portfolio at any EEA airport outside the United Kingdom. (47)

(39)   In the United Kingdom, Delta/Virgin Atlantic hold slots (and historic rights thereto) at London Heathrow, London Gatwick and Manchester airports in both IATA Seasons. AFKL also holds slots at London Heathrow and Manchester airports in both IATA Seasons. The Parties’ slot portfolios thus overlap at London Heathrow and Manchester airports. (48)

(40)   Therefore, in view of the above, the Commission considers it appropriate, for the purpose of this Decision, to apply the analytical framework designed to address the risk of foreclosure from access to airport infrastructure services potentially resulting from the acquisition of joint control over Virgin Atlantic, at airports where the slot  portfolio  of  Virgin  Atlantic/Delta  and  AFKL  overlap  in  Winter 2018/2019 and Summer 2019 IATA Seasons, namely London Heathrow and Manchester airports.

4.1.3.   Relevance of the market for air transport services to tour operators

(41)    Carriers, both charter and scheduled airlines, may sell seats (or entire flights) to tour operators, which then integrate the flights into package holidays or resell seats only to end customers.

(42)    In prior decisions, the Commission has 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. (49) Indeed, from a demand-side perspective, tour operators have different requirements from those of individual passengers (for example, purchase of large seat packages in advance from the start of the season, negotiation of rebates, taking into account passengers’ needs in terms of flight times). In Ryanair/Aer Lingus III, the Commission stated that  “competitive conditions in this market [the supply of seats to tour operators] 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).” (50)

4.1.3.1. Parties’ views

(43)    The Parties stated that the distinction between retail and wholesale supply of airline seats to tour operators is not meaningful for the competitive assessment  of specific routes in the case at hand. (51)

4.1.3.2. Commission’s assessment

(44)    The Commission considers that, for the purpose of assessing the horizontal effects of the Transaction, the supply of airline seats to tour operators only constitutes a meaningful market on routes where either AFKL or Virgin (52) are active to a significant extent. (53)  Indeed, in the absence of any (material)  overlap, the market for the supply of airline seats to tour operators cannot be considered as meaningful for the purpose of the Transaction. (54)

(45)    Therefore, for the purpose of the Transaction, a competitive assessment of the effects of the Transaction on the market for the supply of airline seats to tour operators could be meaningful only on the London – Shanghai route. (55) Nevertheless, the Commission notes that competitive features on the London – Shanghai route are similar as regards both the supply of scheduled air transport services to end customers and the supply of airline seats to tour operators. (56) In particular, the Commission notes that the main competitors of the Parties in the supply of air transport services to end customers are also active on the supply of seats to tour operators. For instance, the established carrier China Eastern operated a daily direct flight on the London – Shanghai route in Summer 2017 and six weekly frequencies in Winter 2017/2018 IATA Season. (57) While the Parties do not have access to third party data for the market for the supply of seats to tour operators, they estimate that their market shares (58) and  the  increment brought about by the Transaction are moderate under any plausible market definition, as in the market for supply of air transport services to end customers. (59)

(46)    In view of the moderate market shares and increments and the presence of established carriers offering direct flights on the markets for the supply of airline seats to (i) tour operators and (ii) end customers, the Commission considers that, for the purpose of this Decision, it is not necessary to further assess the effects  of the Transaction on the market for the supply of airline seats to tour operators with regard to the London – Shanghai route.

(47)    In any event, the Transaction is unlikely to raise serious doubts as to its compatibility with the internal market with respect to the supply of airline seats to tour operators, under any plausible market definition.

4.1.4.   Conclusion on the relevant market in respect of passenger air transport services

(48)    In the present case, the target is an active air carrier and the Parties’ slot holding overlap at London Heathrow and Manchester airports. In its prior decisional practice, the Commission noted that London Heathrow is an airport where barriers to entry are high due to the scarcity of slots. (60) Therefore, the relevant market for scheduled passenger air transport services will be defined under both the O&D approach and the airport-by-airport approach.

(49)    For the purpose of this Decision, the Commission considers that it is not necessary to further assess the effects of the Transaction on the market for the supply of airline seats to tour operators.

4.2. Definition of the relevant markets

4.2.1. O&D approach - Air transport services of passengers

4.2.1.1. Distinction between groups of passengers

The Parties’ views

(50)    The Parties note that in previous cases the Commission has considered that a distinction may be drawn between (i) time sensitive (“TS” or premium) passengers and (ii) non-time sensitive (“NTS” or non-premium) passengers, while ultimately leaving the market definition open in many instances. The Parties do not object to this distinction. Consistent with the Commission’s previous approach, the Parties have considered that (i) restricted economy class passengers were NTS passengers and (ii) all first class, business class and premium economy class passengers, as well as economy class passengers with the most flexible economy tickets were TS passengers. (61)

Commission’s assessment

(51)    The Commission has in its decisional practice (mostly concerning network carriers) considered distinguishing, for a given O&D route, between (i) time sensitive ("TS" or premium) 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 (ii) non-time sensitive ("NTS" or non- premium)  passengers  who  travel  predominantly  for  leisure  purposes,  do not require flexibility with their booking and are more price-sensitive than the first category. (62)

(52)    In its recent decisional practice relating to long-haul flights, the Commission has considered that it may be relevant to distinguish between TS and NTS passengers, although ultimately leaving the question open. (63)

(53)    The outcome of the market investigation is inconclusive, as to whether the distinction between TS and NTS passengers is relevant for the purpose of the Transaction. (64) While some competitors, travel agents and corporate customers indicated that the distinction was relevant for the assessment of the Transaction, other respondents to the market investigation considered that the distinction between TS and NTS passengers has become increasingly blurred, as some business travellers increasingly travel on economy fares, whereas leisure travellers might choose a “premium” cabin to secure a seat on a high-demand flight. (65)

Conclusion

(54)    In light of the above, the Commission concludes that the question whether it is necessary to distinguish between TS and NTS passengers can be left open, as  the Transaction would not raise serious doubts as to its compatibility with the internal market under any plausible market definition. The Commission will consider the relevance of the distinction between TS and NTS passengers in the route-by-route analysis where relevant.

4.2.1.2. Distinction between direct and indirect flights (66)

The Parties’ views

(55)    The Parties submit that on the overlap routes passengers consider indirect flights as an equivalent alternative to direct flights. In that regard, the Parties’ booking data show that some passengers do choose one-stop flights, although those indirect flights are more than 150 minutes longer in duration than non-stop flights. (67) Therefore, the Parties consider that one-stop flights may impose a competitive constraint on non-stop flights and that the extent to which one-stop flights are substitutable for non-stop flights is apparent from the booking data  for a certain route.. The Parties have submitted market shares for all non- stop/one-stop and one-stop/one-stop overlap routes.

Commission’s assessment

(56)    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 or by way of an "indirect" flight on the same O&D pair via an intermediate destination. (68)

(57)    The level of substitutability of indirect flights to 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. (69)

(58)    The Transaction does not give rise to any short/medium-haul overlap routes, because Virgin Atlantic does not operate any short/medium-haul services. (70) Therefore, the question of the substitutability of direct and indirect flights is  only needs to be addressed for long-haul services.

(59)    In prior decisions, the Commission has considered that with respect to long-haul routes (more than 6 hours flight duration), indirect flights constitute a competitive alternative to direct services under certain conditions (for example  if they are marketed as connecting flights on the O&D pair in the computer reservation systems). (71)

(60)    A majority of competitors having expressed their views confirmed that one-stop services could constitute competitive alternatives to non-stop services as identified above. (72) The majority of travel agencies stated that, on each of the relevant routes for the purpose of this decision, they sell tickets to both direct and indirect flights to their customers. The majority of corporate customers having expressed a view stated that they buy tickets for both direct and   indirect flights on the affected direct/indirect overlap routes. (73) Travel agents and corporate agents having expressed a view specifically mentioned that the following criteria can lead passengers to choose one-stop flights over non-stop flights: the price, the total travel duration (including stopover times), the schedules and the number of frequencies. (74)

Conclusion

(61)    The Transaction does not give rise to direct/direct overlap routes. (75) If direct and indirect flights were to be considered as two separate markets, the direct/indirect overlap routes would not be affected overlap routes as a result of the Transaction. For the purpose of this Decision, the Commission will consider that direct and indirect flights are part of the same market and assess the effects of the Transaction on direct/indirect overlap routes. Nevertheless, the conclusion  on whether or not direct and indirect flights belong to the same market can be  left open as the Transaction would not raise serious doubts as to its compatibility with the internal market under either of these alternative market definitions.

4.2.1.3. Distinction between charter flights and scheduled flights

The Parties’ views

(62)    The Parties submit that they face significant competition  from  charter companies selling so-called “dry seats” (seats only without other services) on a number of routes, especially in respect of sunshine/leisure destinations in the Caribbean (such as Cuba and Jamaica) served by Virgin Atlantic. (76)

(63)   In the Parties’ views, charter flights offering “dry seats” and scheduled flights are substitutable and are therefore part of the same relevant product market. (77)

Commission’s assessment

(64)    The Commission has previously considered that 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). (78)

(65)    The Commission has also considered that charter companies do not traditionally sell tickets directly to passengers. They sell seats on their aircraft to tour operators, which include 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 sell to some extent "dry seats" directly to passengers. (79)

(66)    The Commission has previously held 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 passenger transport services. (80) However, as regards charter carriers selling dry seats and scheduled point-to- point air passenger transport services, the Commission has left open whether they are part of the same relevant product market and has considered the relevance of dry seats offered by charter companies in its competitive assessment. (81)

(67)    The Parties submit that charter carriers are not represented in the market data they submitted. (82) The Commission notes that the market data submitted by the Parties includes on some routes the airlines Thomas Cook/Condor and TUI, which are integrated into a tour operating company.

Conclusion

(68)    In light of the above, the Commission considers that the question of whether dry seats are part of the same market as scheduled air transport services can be left open, as the Transaction would not raise serious doubts as to its compatibility with the internal market under any plausible market definition. The Commission will consider the relevance of the dry seats sold by charter airlines in the route- by-route analysis where relevant.

4.2.1.4. Airport substitutability

Analytical framework

(69)    When defining the relevant O&D markets for passenger air transport services, the Commission has previously found that flights to or from airports with sufficiently overlapping catchment areas can be considered as substitutes in the eyes of passengers (particularly if the airports serve the same main city). In  order to correctly capture the competitive constraint that flights to or from two different airports exert on each other, a detailed analysis taking into consideration the specific characteristics of the relevant airports is necessary. (83)

(70)    The evidence used to characterise airport substitutability includes inter alia a comparison of actual distances and travelling times to the indicative benchmark of 100 km/1 hour driving time, (84) the outcome of the market investigation (views of the competitors and other market participants), and the competitors’ practices in terms of monitoring of competition.

(71)    In the present case, taking account of the relevant routes where the Parties’ activities overlap, the question of airport substitutability arises for the routes to or from the following cities: Shanghai, Lyon, Paris, Düsseldorf, Milan, London, Manchester, New York, Orlando and San Francisco

(72)    Nevertheless, for the purpose of this Decision, the question of airport substitutability is not relevant for Shanghai, Lyon, Düsseldorf, Manchester, Orlando and San Francisco.  (85)

Assessment of airport substitutability

a) Paris

(73)    Paris has two main airports, namely Charles de Gaulle (CDG) and Orly (ORY).

(74)    In its prior decision practice, the Commission has considered ORY and CDG as substitutable airports, but ultimately left the question open. (86)

(75)    For the purposes of the O&D assessment of the Transaction, the question of airport substitutability is relevant for the following routes: Paris – San-Francisco and Paris – Boston.

(76)    On the Paris – San Francisco route, AFKL/Delta operates direct services and markets an indirect service to/from CDG. VAL only operates indirect services to/from CDG while Frenchbee operates a direct service to/from ORY. (87)

(77)    On the Paris – Boston route, AFKL/Delta operates direct services and markets an indirect service to/from CDG. VAL only operates indirect services to/from CDG. (88)

(78)    The travel distances and times between Charles de Gaulle and Orly airports and the centre of Paris are summarised below:

(79)   Both competitors and corporate customers and travel agents having expressed a view gave mixed replies as to whether passengers consider CDG and ORY as substitutable. (89) The majority of competitors having expressed a view do monitor air carriers flying on the Paris – San Francisco and Paris - Boston routes to/from CDG or ORY. (90) The outcome of the market investigation is therefore inconclusive.

(80)   In any case, for the purpose of this Decision, the question of whether the relevant market consists of Charles de Gaulle, or of Charles de Gaulle and Orly, can be  left open, as the Transaction would not raise serious doubts as to its compatibility with the internal market under either plausible market definition.

b)  Milan

(81)   Milan has three airports, namely Malpensa (MXP), Linate (LIN) and Bergamo (BGY). In light of the fact that mostly domestic and short/medium-haul flights are operated from Bergamo, the latter will therefore not be included in the airport substitutability analysis.

(82)   In its prior decision practice, the Commission has considered Malpensa  and Linate as substitutable, but the market definition was ultimately left open. (91)

(83)   For the purposes of the O&D assessment of the Transaction, the question of airport substitutability is relevant for the routes: Milan – Los Angeles and Milan – San Francisco.

(84)   With respect to both the Milan – Los Angeles and Milan – San Francisco routes, the Parties operate indirect services to/from MXP and to/from LIN. (92)

(85)   The travel distances and times between Linate and Malpensa airports and the centre of Milan are summarised below:

(86)   Both competitors and corporate customers and travel agents having expressed a view gave mixed replies as to whether passengers consider LXP and LIN airports as substitutable. (93) The majority of competitors having expressed a view do monitor air carriers flying on the Milan – Los Angeles and Milan – San Francisco routes to/from MXP or LIN. (94) The outcome of the market investigation is therefore inconclusive.

(87)   In any case, for the purpose of this Decision, the question of whether the relevant market consists of Malpensa, or of Malpensa and Linate, can be left open, as the Transaction would not raise serious doubts as to its compatibility with the internal market under either plausible market definition.

c)  London

(88)   London has six main airports, namely Heathrow (LHR), Gatwick (LGW), City (LCY), Stansted (STN), Luton (LTN) and Southend (SEN). (95)

(89)   In its prior decisional practice relating to long-haul services to/from London, the Commission examined the effects of the transaction on a market comprising flights to and from Heathrow only and on a wider market comprising flights to and from LHR, LGW and LCY, but the market definition was ultimately left open. (96)

(90)   The Parties consider that there is no reason to deviate from this practice. (97)

(91)   For the purposes of the O&D assessment of the Transaction, the question of airport substitutability is relevant for the following routes: London – Seattle, London – Atlanta, London – Minneapolis, London – Detroit, London – Portland, London – Lagos, London – Shanghai, London – Havana and London – Montego Bay.

(92)   On the London – Seattle route, VAL/Delta operates direct and indirect services to/from LHR. AFKL operates indirect services to/from LHR via AMS or CDG. VAL/Delta also operates indirect services to/from LGW while Norwegian operates direct services to/from LGW. (98)

(93)   On the London - Atlanta route, VAL/Delta operates direct and indirect services to/from LHR. AFKL operates indirect service to/from LHR via AMS or CDG. VAL/Delta also operates an indirect service to/from LGW. (99)

(94)   On the London – Minneapolis route, VAL/Delta operates direct and indirect services to/from LHR. AFKL operates indirect services to/from LHR via AMS or CDG. (100)

(95)   On the London – Detroit route, VAL/Delta operates direct and indirect services to/from LHR. AFKL operates indirect services to/from LHR via AMS or CDG. (101)

(96)   On the London – Portland route, VAL/Delta operates direct and indirect services to/from LHR. AFKL operates indirect services to/from LHR via AMS or CDG. (102)

(97)   On the London – Lagos route, VAL operates direct and indirect services to/from LHR. AFKL operates indirect services to/from LHR and to/from LCY. (103)

(98)   On the London – Shanghai route, VAL operates direct services to/from LHR. AFKL operates indirect services to/from LHR and to/from LCY. (104)

(99)   On the London – Havana route, VAL operates direct and indirect services to/from LGW. AFKL operates indirect services to/from LHR. (105)

(100)   On the London – Montego Bay route, VAL operates direct services to/from LGW while AFKL and VAL/Delta operate indirect services to/from LHR. (106)

(101)   The travel distances and times between Heathrow, Gatwick and City airports and the centre of London are summarised below:

(102)   Both competitors and corporate customers and travel agents having expressed a view gave mixed replies as to whether the relevant market consists of flights to/from LHR only, or to/from LHR, LGW and LCY taken together. (107) The majority of competitors having expressed a view do monitor air carriers flying to/from LHR, LGW or LCY on the routes listed in paragraph (91) above. (108) The outcome of the market investigation is therefore inconclusive.

(103)   In any case, for the purpose of this Decision, the question of whether the relevant market consists of flights to/from Heathrow only, or to/from Heathrow, Gatwick and City airports, can be left open, as the Transaction would not raise serious doubts as to its compatibility with the internal market under either plausible market definition.

d) New-York

(104)   The city of New York is served by three primary airports, namely John F. Kennedy (JFK), Newark Liberty (EWR) and La Guardia (LGA) airports. La Guardia predominantly serves North American destinations and there are  no direct flights to and from La Guardia relevant for the assessment of the Transaction. (109) La Guardia will therefore not be included in the airport substitutability analysis.

(105)   In its prior decisional practice relating to transatlantic flights to/from New York, the Commission examined the effects of the transaction on a market comprising flights to and from JFK only and a wider market comprising flights to and from JFK and EWR, but the market definition was ultimately left open. (110)

(106)   For the purposes of the O&D assessment of the Transaction, the question of airport substitutability is relevant for the following routes: Manchester – New York, Edinburgh – New York, Nice – New York, Dublin – New York, Lyon – New York, Marseille – New York and Hamburg – New York.

(107)   On the Manchester – New York route, VAL/Delta operates direct and indirect services to/from JFK and indirect services to/from EWR. AFKL operates indirect services to/from JFK while United Airlines, Lufthansa and Air Canada operate direct services to/from EWR. (111)

(108)   On the Edinburgh – New York route, VAL/Delta operates direct and indirect services to/from JFK. AFKL operates indirect services to/from JFK via AMS or CDG while United Airlines, Lufthansa and Air Canada operate direct services to/from EWR. (112)

(109)   On the Nice – New York route, AFKL/Delta operates direct and indirect services to/from JFK while VAL operates indirect services to/from JFK. The Parties also operate limited indirect services to/from EWR. (113)

(110)   On the Dublin – New York route, AFKL/Delta operates direct and indirect services to/from JFK. VAL operates indirect services to/from JFK and to/from EWR while United Airlines, Lufthansa, Air Canada and Aer Lingus operate direct services to/from EWR. (114)

(111)   On the Lyon – New York route, the Parties operate indirect services to/from JFK and to/from EWR. (115)

(112)   On the Marseille – New York route, the Parties operate indirect services to/from JFK and to/from EWR. (116)

(113)   On the Hamburg – New York route, the Parties operate indirect services to/from JFK and to/from EWR while United Airlines operates a direct service to/from EWR in summer only. (117)

(114)   The Parties consider JFK and EWR as substitutable.  (118)

(115)   The travel distances and times between these airports and the centre of New York are summarised below:

(116)   The outcome of the market investigation is inconclusive. While a majority of competitors having expressed a view consider that JFK and EWR airports are substitutable and monitor air carriers flying on the routes listed in paragraph (106) above to/from JFK or EWR, (119)  the corporate customers and travel agents gave mixed replies as to whether passengers consider JFK and EWR as substitutable airports. (120)

(117)   In any case, for the purpose of this Decision, the question of whether John F. Kennedy and Newark airports are substitutable can be left open, as the Transaction would not raise serious doubts as to its compatibility with the internal market under either plausible market definition.

e) Orlando

(118)   Orlando is served by two airports, namely Orlando International (MCO) and Orlando Sanford International (SFB).

(119)   In its prior decisional practice, the Commission found that MCO and SFB can be considered as substitutable from the point of view of passengers, at least with regards to NTS passengers, with regard to the routes London – Orlando and Manchester - Orlando. (121)

(120)   For the purpose of the O&D assessment of the Transaction, the question of substitutability of MCO and SFB is relevant for the following routes: Dublin – Orlando and Amsterdam – Orlando.

(121)   On the Dublin – Orlando route, the Parties operate indirect  services  to/from MCO, while Aer Lingus operates direct flights from MCO. (122) With respect to the Amsterdam – Orlando route, the Parties operate from MCO while TUI operated a direct service to/from SFB. (123)

(122)   The travel times and distances between these airports and the city centre of Orlando are summarised below:

(123)   The outcome of the market investigation is inconclusive. While a majority of competitors having expressed a view consider that Orlando and Sanford airports are substitutable and monitor competitors on the route Amsterdam – Orlando flying to/from MCO or SFB, (124) the corporate customers and travel agents gave mixed replies as to whether passengers consider Orlando and Sanford as substitutable. (125)

(124)   In any case, for the purpose of this Decision, the question of whether Orlando and Sanford airports are substitutable can be left open, as the Transaction would not raise serious doubts as to its compatibility with the internal market under either plausible market definition.

Conclusion

(125)   For the purpose of this Decision, it is not necessary to decide whether (i) Paris- Charles de Gaulle and Paris Orly airports, (ii) Milan Malpensa and Milan Linate airports, (iii) London Heathrow, Gatwick and City, (iv) John F. Kennedy and Newark, (v) Orlando and Sanford airports are substitutable as the Transaction would not raise serious doubts as to its compatibility with the internal market under any plausible market definition.

4.2.2.   Airport-by-airport approach – Air transport of passengers

(126)   As explained in paragraph (40) above, the Commission considers it appropriate to apply the airport-by-airport approach for the purpose of this Decision. The Commission will assess the effect of the Transaction on the Parties’ slot holding  in both IATA Seasons at London Heathrow and Manchester airports in relation to the markets for air transport services of passengers to or from the relevant airports and the market for airport infrastructure services provided at the relevant airports.

(127)   The Commission will consider below the various possible delineations of these two relevant markets under the airport-by-airport approach.

4.2.2.1. Air transport services of passengers to or from the relevant airports

Relevant product market

(128)   In prior decisions, the Commission has not deemed it necessary to consider under the airport-by-airport approach, when all O&Ds to or from an airport are aggregated, the same distinctions as those considered when each O&D market is examined separately (e.g. time sensitive vs. non-time sensitive passengers, direct vs. indirect flights, charter flights vs. scheduled flights, wholesale vs. retail supply of airline seats). (126)  On the basis of the information in the file, the Commission considers that there are no grounds for it to deviate from this past practice for the purposes of this Decision.

Relevant geographic market

(129)   In prior decisions, the Commission has considered whether the relevant airports were substitutable with other airports in view of their overlapping catchment areas. (127)

(130)   In the present case, the substitutability from the point of view of passengers of

(i)   London Heathrow, Gatwick, City, Stansted, Luton and Southend airports    and

(ii)  Manchester, Liverpool and  Leeds  airports  has  been  considered  in  section 4.2.1.4. above. (128)

Conclusion

(131)   For the purpose of its airport-by-airport assessment of the Transaction in this Decision, the Commission will assess the competitive effects of the Transaction on the markets for the provision of passenger air transport services, encompassing all routes to or from an airport, or to or from substitutable airports.

(132)   For the purpose of its airport-by-airport assessment of the Transaction in this Decision,  the  question  of  whether  the  relevant  geographic  market  consists of (i)   flights to/from London Heathrow only or Heathrow, Gatwick and City airports and (ii) flights to/from Manchester only or Manchester, Liverpool and Leeds can be left open, as the Transaction would not raise serious doubts as to its compatibility with the internal market under either plausible market definition (see section 5.2. below).

4.2.2.2. Airport infrastructure services

(133)   For the purpose of providing passenger air transport services at congested  airports, airlines have to source infrastructure services at those airports. As indicated in paragraph (27) above, at congested airports, infrastructure capacity is managed through the allocation of slots, which enable air carriers to fly to and from the airports. Slots are therefore defined, from the point of view of airports,  as "a planning tool for rationing capacity at airports where demand for air travel exceeds the available runway and terminal capacity." (129) From the point of view  of airlines, the granting of a slot at an airport means that the airline may use the entire range of infrastructure necessary for the operation of a flight at a given time (runway, taxiway, stands and, for passenger flights, terminal infrastructure). This in turn enables the airlines to provide passenger air transport services to and from that airport.

(134)   As a consequence, through the Transaction and the combination of slot portfolios, the Parties together obtain a right of access to a higher share of airport infrastructure capacity. The Transaction therefore has an impact on (the demand- side of) the markets for airport infrastructure services at the relevant airports and also on the markets for passenger air transport to and from those airports.

Relevant product market

(135)   The Commission has, in its prior decisional practice, delineated a product market for the provision of airport infrastructure services to airlines, which includes the development, maintenance, use and provision of the runway facilities, taxiways and other airport infrastructure. (130)

(136)   The Commission has considered sub-dividing the market for airport infrastructure services on the basis of airline customers (i.e. charter operators, scheduled full service carriers and scheduled low cost carriers) and on the basis of the type of flights (i.e. short-haul and long-haul). (131)

(137)   However, in prior decisions relating to the transfer of slots at airports, the Commission has not considered it appropriate to further distinguish within the market for airport infrastructure services, considering that slot portfolios give access to all infrastructure services necessary to operate at the airport.

(138)   The Commission considers that there is no element in the file that would require deviating from the Commission's past practice for the purposes of this Decision.

Relevant geographic market

(139)   In its prior decisional practice, the Commission has, defined the geographic scope of the market for airport infrastructure services as the catchment area of  individual airports.

(140)   The Commission has also considered additional criteria relevant for assessing airport substitutability in relation to the market for airport infrastructure services, while acknowledging that the airlines' choice of airports ultimately depends on passengers' demand. In addition to the catchment area of a particular airport, the Commission has notably analysed the capacity constraints for slots and facilities, passenger volumes or the positioning of the airport (e.g. a niche airport serving high yield time-sensitive passengers or an airport serving mainly leisure, less time-sensitive passengers). (132)

(141)   The Commission has taken account of all the above-mentioned criteria when assessing the geographic scope of the airport infrastructure services markets relevant for the assessment of the effects of transfer of slots. (133)

Substitutability of London Heathrow, Gatwick, City, Stansted, Luton and Southend airports

(142)   VAL, AFKL and Delta each hold slots at Heathrow airport. Virgin Atlantic/Delta hold slot at Gatwick, while AFKL holds slot at City. (134) The Transaction therefore gives rise to an overlap between AFKL and Virgin Atlantic/Delta at Heathrow, and on a broader geographic scope comprising at least (i) LHR and (ii) LCY and/or LGW.

(143)   The question of the  catchment  area  of  London  airports  is  addressed  in  section 4.2.1.4. above. From the point of view of passengers, the relevant markets consists of flights to/from London Heathrow only or to/from Heathrow, Gatwick and City airports.

(144)   As regards capacity constraints, these three airports are coordinated (Level 3) airports.

(145)   In 2018, 78 million passengers used Heathrow airport, (135) 46.1 million passengers travelled to/from Gatwick (136)  and 4.8 million passengers used City airport. (137)

(146)   As regards positioning, Heathrow is the “world’s busiest international airport” and a major international hub for air transport, with more than 81 airlines operating 204 destinations. (138) Gatwick is used by 50 airlines, which provide “a combination  of  full  service,  low-cost  and   charter   services”   and   fly   to  230 destinations. (139) While 56% of City airport’s passengers travel for business purpose, 44% are leisure passengers. In that regard, City airport “has developed an attractive mix of airlines and destinations with 11 airline customers […] serving 43 different destinations (11 UK, 34 European and 1 US destination)”. (140)

(147)   Considering that Heathrow, City and Gatwick have different positioning and strategy, the Commission concludes that, for the purpose of this Decision, the geographic scope of the market for the provision of airport infrastructure services to airlines is limited to Heathrow airport.  (141)

Substitutability of Manchester, Liverpool and Leeds-Bradford airports

(148)   Virgin Atlantic/Delta and AFKL hold slots at Manchester airport. The  Transaction therefore gives rise to an overlap between Virgin Atlantic/Delta and AFKL’s slot portfolios. […]. (142)

(149)   As regards catchment area, Manchester, Liverpool and Leeds airports are each all located within 100 km/1 hour driving time from Manchester city centre.

(150)   In previous decisions, the Commission considered that Manchester and Liverpool or Manchester, Liverpool and Leeds airports could belong to the same market in respect of certain short-haul routes. (143)

(151)   As regards capacity constraints, Manchester airport is a coordinated (Level 3) airport during both IATA Seasons, while Liverpool and Leeds airports are both schedules facilitated airports (144) (Level 2) (145) during both IATA Seasons.

(152)   As regards positioning and passenger volumes, Manchester serves in particular full service carriers. It is used by 70 airlines serving 220 destinations. The number of passengers at Manchester airport reached 59 million in 2018. Leeds and Liverpool are mainly used by low-cost carriers such as Ryanair and easyJet (at Liverpool  only). (146)    3.6  million  passengers  travelled  to/from  Leeds  airport,  (147) compared to 5.1 million to/from Liverpool airport.  (148)

(153)   Considering  that  (i)  only  Manchester   airport   is   a   coordinated   airport, and (ii)   Manchester airport has a different positioning and market strategy, the Commission concludes that, for the purpose of this Decision, the geographic  scope of the market for the provision of airport infrastructure services to airlines  is limited to Manchester airport.

Conclusion

(154)   For the purpose of this Decision, the Commission will assess the effects of the Transaction on the market for the provision of airport infrastructure services to airlines.

(155)   For the purpose of this Decision, the Commission considers that the geographic scope of the market for airport infrastructures services in London is London Heathrow airport and the geographic scope of airport infrastructure services in Manchester is limited to Manchester airport.

4.2.3. Air transport services of cargo

4.2.3.1. Relevant product market

(156)   In prior decisions, the Commission considered a market for air transport of cargo including all kinds of transported goods provided by all types of air cargo carriers, (149) without any further subdivision to be made according to the nature of the goods transported (for example, dangerous or perishable goods) or the type of air cargo carrier. (150)

(157)   In fact, the Commission has concluded that four types of air cargo  carriers, namely (i) cargo airlines with dedicated freighter planes; (ii) airlines with only belly space  cargo  capacity  on  passenger  flights;  (iii)  combination  airlines  (i.e. airlines  with  both  dedicated  freighter  airplanes  and  belly  space  cargo capacity); and (iv) integrators, compete with each other for business with the  same kinds of customers. (151)

(158)   Based on the Commission’s prior decisions, the O&D approach to market definition is not appropriate for air cargo transport services because cargo is (i) in principle less time-sensitive than passengers, and (ii) usually transported “behind” and “beyond” the points of origin and destination by trans-modal transport methods and thus can be routed via a higher number of stops than passengers. (152) Consequently, the Commission considers that a wider market for air transport of cargo exists as, unlike passengers, cargo can be transported with a higher number of stopovers and therefore any one-stop route is a substitute for any non-stop route. (153)

(159)   In addition, according to the Commission's precedents, cargo transport markets should be assessed on a unidirectional basis, due to differences in demand at each end of the route. (154)

(160)   The Parties agree with the Commission's decision-making practice. (155)

(161)   Therefore, in line with its prior decisional practice, the Commission will assess the effects of the Transaction on a broader market for air transport of cargo encompassing all types of air cargo carriers and including all kinds of transported goods on a unidirectional basis.

4.2.3.2.  Relevant geographic market

(162)   In prior decisions, the Commission defined the market in intra-European routes of air cargo transport as European-wide. (156) As regards intercontinental routes, the Commission established that catchment areas at each end of the route broadly correspond to continents where local infrastructure is adequate to allow for onward connections (for example, by road, train, or inland waterways, etc.), such as Europe and North America. As regards continents where local infrastructure  is less developed, the relevant catchment area has been considered the country of destination. (157)

(163)   The Parties agree with the Commission's previously established geographic definition.

(164)   Therefore, in line with its prior decisional practice, the Commission will assess the effects of the Transaction on a continent-to-continent and continent-to-country basis as the case may be. The Transaction gives rise to the following affected cargo flows: EEA - North America (USA & Canada), North America (USA & Canada) - EEA, Mexico - EEA, UK - North America (USA & Canada), North America (USA & Canada) - UK, Mexico - UK, UK - Mexico, EEA - Caribbean, Caribbean - EEA, EEA - Cuba, Cuba - EEA, EEA - South Africa, South Africa - EEA, EEA - Nigeria, Nigeria - EEA.

4.2.4. Maintenance, repair and overhaul (“MRO”) services

4.2.4.1. Relevant product market

(165)   In prior decisions, the Commission distinguished four separate segments within the MRO market based on the part of the aircraft to be serviced and the level of service required, namely (i) line maintenance (minor checks carried out on  aircraft and performed at the different airports), (ii) heavy maintenance (comprehensive inspection and overhaul of the aircraft, for which the aircraft is taken out of service), (iii) engine maintenance, and (iv) components maintenance (inspection, repair and overhaul of specific aircraft components). (158) The Commission also considered but ultimately left the question open, whether a distinction between commercial and business aviation is appropriate. (159) It moreover noted that line maintenance and heavy maintenance can be further subdivided according to nature and frequency of the checks involved (A, B, C and D-checks). (160)

(166)   The Parties submit that the precise scope of the product market definition for MRO can be left open as no serious doubts would arise under any plausible market definition. However, in line with the Commission’s decisional practice, they provided data for each MRO segment and notably for line maintenance services where the Parties’ activities overlap.

(167)   In light of the above, the Commission concludes that the precise scope of the product market definition for MRO services can be left open since the  Transaction does not raise serious doubts as to its compatibility with the internal market under any plausible product market definition.

4.2.4.2. Relevant geographic market

(168)   In prior decisions, the Commission considered that the geographic scope of the market for heavy maintenance services might be at least EEA-wide, whereas line maintenance services could be local in scope and even limited to the airport  where services are provided. (161) Indeed, line maintenance services are usually carried out at the airport of origin or destination, or at the aircraft’s operational base. (162) As regards to engine maintenance services and components maintenance services, the Commission has considered these services to be worldwide in scope.  (163)

(169)   The Parties submit that the precise scope of the geographic market definition for all MRO segments can be left open.

(170)   For the assessment of the Transaction, the Commission concludes that the precise geographic market definition for MRO services can be left open since the Transaction does not raise serious doubts as to its compatibility with the internal market under any plausible geographic market definition.

4.2.4.3.  Conclusion

(171)   Given that the Transaction does not raise serious doubts as to its compatibility with the internal market in respect of MRO services under any plausible market definition, the Commission concludes that the precise product and geographic market definitions can be left open.

5. COMPETITIVE ASSESSMENT

5.1. Passenger air transport services under the O&D approach

5.1.1. Analytical framework

5.1.1.1. Alliances and profit-sharing joint ventures

(172)   Air France and Delta were founding members of the SkyTeam alliance in 2001. Since 2004, shortly after the merger of Air France and KLM, KLM has also been a full member of the SkyTeam alliance. (164) Consistent with the Commission’s practice, alliance partners are generally not considered as forming a single economic entity for the purpose of determining affected markets. (165)

(173)   Delta, Air France-KLM and Alitalia are parties to a revenue-sharing joint venture (or “metal neutral joint venture”), covering routes between Europe (166) and North America (the “TATL” Joint Venture”). (167)

(174)   Delta and Virgin Atlantic are parties to a metal neutral joint venture covering direct routes between the United Kingdom and North America (the “DL-VS Joint Venture”). (168)

(175)   Delta, AFKL, Alitalia and Virgin Atlantic are also parties to a Coordination Agreement, covering routes between the United Kingdom and North America. (169) By final order issued on 23 September 2013, the American Department of Transportation granted antitrust immunity with respect to the activities contemplated under the DL-VS Joint Venture and  the  Coordination Agreement. (170)

(176)   Metal neutral joint ventures involve extensive cooperation such as revenue- sharing and joint management of schedules, pricing and capacity. (171)

(177)   There are two other metal neutral joint ventures on transatlantic routes, namely (i) the oneworld Transatlantic Joint Business Agreement entered into  by American  Airlines,  British  Airways  (BA)  and  Iberia  (IB)  and  Finnair(AY) (“oneworld”) and (ii) the Star Alliance A++ Joint Venture of United (UA), Lufthansa (LH) and Air Canada (AC) (“Star Alliance”).

(178)   Consistent with the Commission’s approach, each of the metal neutral joint ventures are treated as a single entity for the purposes of assessing market shares on all relevant routes. (172)

5.1.1.2. Methodology used to calculate market shares

(179)   The Commission has previously used Marketing Information Data Tapes (“MIDT”) data (173) and PaxIS PLUS data (174) as appropriate proxies to estimate market shares for air transport of passengers.

(180)   The Parties have first submitted data on market size and market shares for each relevant O&D on the basis of MIDT while recognising that MIDT has certain limitations: MIDT data is based on bookings made through the  Global Distribution Systems (GDS) and does not capture the direct sales made direct by airline to their customers. (175)

(181)   To take account of the increasing number of direct sales, the Commission has asked the Parties in the pre-notification phase to use a data source that captures direct sales by the Parties and their competitors.

(182)   The Parties submitted that none of the Parties subscribes to PaxIS PLUS and proposed not to use PaxIS PLUS since PaxIS PLUS [details about usability of PaxIS PLUS in the context of the case at hand]. (176)

(183)   Instead of PaxIS Plus, the Parties proposed to use Direct Data Solutions (“DDS”) data, another database developed by IATA. (177) The Parties explained that DDS includes actual bookings data from GDS and ACR data (US country of sale agency settlement data) as well as contributed data from the carriers subscribing  to it. IATA would apply an algorithm to estimate volumes for carriers that do not contribute data or do not allow for their contributed data to be published. According to the Parties, DDS is intended to capture all passenger bookings. (178) The Parties submitted data on market size and market shares for each relevant O&D on the basis of DDS for the last two IATA Seasons. (179)

(184)   The Parties have also submitted their actual bookings data for the last two IATA Seasons and submit that the DDS data is often similar to the actual sales data in terms of number of bookings. (180)

(185)   The Commission is of the view that DDS data are the best proxy to estimate the market shares and are appropriate for the assessment of the affected routes in this case.

5.1.1.3. “Filters”

(186)   Consistent with previous Commission practice, (181) the Parties have applied the following filters to exclude likely unproblematic routes from the scope of its investigation (all criteria must have been met in the four last completed IATA Seasons and for all passenger segments for a route to be excluded under the filters): (182)

(a)  For direct/indirect overlaps:

(i)      the Parties’ combined market share was below 25%; or

(ii)   one of the Parties had a market share below 2%; (183) or

(iii) short-haul routes where the total share of indirect operations in the relevant market was below 10%; or

(iv) at least one end of the city pair is outside the EU and the total annual traffic was below 30 000 passengers; or

(v)    the route was below the HHI thresholds of paragraph 20 of the Horizontal Merger Guidelines. (184)

(b)  For indirect/indirect overlaps:

(i)      the Parties’ combined market share was below 25%; or

(ii)   one of the Parties had a market share below 2%; (185) or

(iii) as regards short-haul routes where the total annual traffic was below 15 000 passengers or as regards long-haul routes where the total annual traffic was below 30 000 passengers; or

(iv) the route was below the HHI thresholds of paragraph 20 of the Horizontal Merger Guidelines.

(187)   As a result of the above criteria, the Parties submitted that 32 routes (direct/indirect overlaps and indirect/indirect overlaps) would be affected by the Transaction.

5.1.1.4. Closeness of competition

(188)   The Transaction gives rise to 19 affected horizontal direct/indirect overlaps. On eight affected routes from the United Kingdom to North America, Virgin Atlantic/Delta operates a direct service, while AFKL operates an indirect service. These routes are: London-Seattle, Manchester-New York, London-Atlanta, London-Minneapolis, London-Detroit, Edinburgh-New York, London-Portland, Manchester-San Francisco. On seven affected routes from Continental Europe/Ireland to North America, AFKL/Delta operates a direct service, while Virgin Atlantic operates an indirect service. These routes are: Amsterdam-Los Angeles, Amsterdam-Miami, Amsterdam-San Francisco, Paris-San Francisco, Paris-Boston, Nice-New York and Dublin-New York. On three affected routes (London-Lagos, London-Shanghai, London-Havana), Virgin Atlantic operates a direct service, while AFKL operates an indirect service and on one affected route, Virgin Atlantic operates a direct service, while AFKL and Delta operate an indirect service (London-Montego Bay).

(189)   The Commission will first assess the closeness of competition in general between the Parties on the 19 horizontal direct/indirect overlap routes, before assessing the effects of the Transaction on these routes.

(190)   For the reasons explained below, the affected direct/indirect overlap routes will not give rise to competition concerns. If direct and indirect flights were to be considered as two separate markets, there would be no affected overlap routes as  a result of the Transaction. Therefore, it is not necessary to reach a conclusion on whether direct and indirect passenger air transport services constitute part of the same market, as it would not change the conclusion of the competitive assessment as already stated in section 4.2.1.2 above.

(191)   The Commission has in previous airline cases analysed the closeness of competition between the parties to the concentration. (186) 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. (187) It is therefore relevant to assess whether a) Virgin Atlantic/Delta’s direct service and AFKL’s indirect service on the routes from the United Kingdom to North America, b) AFKL/Delta’s direct service and Virgin Atlantic’s indirect service on the routes from Continental Europe/Ireland to North America and c) Virgin Atlantic’s direct services and AFKL’s and Delta’s indirect services on the routes from London to certain rest of the world-destinations would be considered as close substitutes.

(192)   As explained above, the market investigation gave mixed results concerning the question if of whether passengers consider direct flights as an alternative to indirect flights: while the majority of competitors (airlines) considered this to be the case, the views of the responding customers (travel agencies and corporate customers) were mixed. (188)

(193)   The majority of customers (travel agencies and corporate customers) stated that they sell tickets to both direct and indirect flights to their customers and that they buy tickets for both direct and indirect flights on the affected direct/indirect overlap routes. (189) However, around 1/3 of the corporate customers responding to the market investigation explained that they generally only buy tickets to direct flights. (190) When asked which criteria would make customers choose an indirect flight over a direct flight, respondents to the Commission’s market investigation identified most frequently the price difference, followed by the schedule of the respective flight and the total travel duration. (191)

(194)   The majority of respondents to the market investigation (competitors, travel agents, corporate customers) indicated that none of the Parties are each other’s closest competitor on any of the 19 affected direct/indirect overlap routes. (192) In  the majority of cases, British Airways, American Airlines, United Airlines or Lufthansa was identified as either Virgin Atlantic’s, AFKL’s or Delta’s closest competitor.

(195)   In the light of the above, the Commission considers that none of the Parties are each other’s close competitor on the 19 direct/indirect overlap routes.

5.1.1.5. Significant competitive pressure from other airlines

(196)   The market investigation in previous cases has shown that indirect routes are  often modified from one IATA Season to the next and that price increases or reductions of capacity could be countered by competitors who could start operating on these routes more easily than on direct/direct overlap routes which require the deployment of aircraft dedicated to the O&D route. (193) Therefore, in general, any attempt by the Parties to raise prices on routes, where the overlap is between direct/indirect services, would likely be short-lived and ineffective. This was also confirmed by the market investigation. The majority of respondents having expressed a view consider that there will be sufficient competition to prevent the Parties from raising prices on all affected direct/indirect overlap routes. The Commission thus considers that other carriers like the members of the Star Alliance or oneworld would exert significant competitive pressure on the Parties post-Transaction on the affected direct/indirect overlap routes.

5.1.1.6. De minimis increment

(197)   In previous cases, the Commission has considered increments up to 5% as low increments on direct/indirect overlap routes. (194) In line with its prior decisional practice, the Commission considers that, also taking into account that other airlines would exert competitive pressure on the Parties post-Transaction, an increment of up to 5% on an affected direct/indirect overlap is de minimis.

5.1.2.   Direct/Indirect affected markets between the United Kingdom and North America

(198)   As explained in paragraph (178) above, the Commission treats each of the metal neutral joint ventures as a single entity for the purpose of assessing the market shares on all relevant routes. Consistent with the approach, Delta and Virgin Atlantic are treated as a single entity on every route between the United Kingdom and North America.

5.1.2.1. London – Seattle

(199)   In Summer 2017 IATA Season, [100.000-200.000] passengers travelled between London Heathrow and Seattle airports and [100.000-200.000] passengers flew between London Heathrow, Gatwick and City and Seattle airports. (195) In Winter 2017/2018 IATA Season, [60.000-80.000] passengers travelled between London Heathrow and Seattle and [80.000-100.000] passengers flew between London Heathrow, Gatwick and City and Seattle airports. (196)

(200)   In Summer 2017 and Winter 2017/2018 IATA Seasons, Virgin Atlantic/Delta operated direct services on the LHR-SEA airport pair, while AFKL offered only indirect services on this airport pair via Paris-Charles de Gaulle or Amsterdam airports.

(201)   As explained in section 4.2.1.4 above, the Commission will assess the effect of  the Transaction on the following airport pairs: LHR-SEA and LHR/LGW/LCY- SEA.

(202)   The table below provides the market shares of the Parties on the LHR-SEA airport pair.

(203)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. (197)

(204)   The table below provides the market shares of the Parties on the LHR/LGW/LCY-SEA airport pair.

(205)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(206)   In addition, the Parties will remain constrained by the competition of the year- round direct service of oneworld as well as the daily all-year round indirect  service operated by oneworld, Star Alliance and Icelandair on LHR-SEA. (198) In addition, Norwegian operated a direct service on the LGW-SEA airport pair in Summer 2017 IATA Season. (199)

(207)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (200)

(208)   In light of the above and of all available evidence, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the London – Seattle route under any plausible market definition.

5.1.2.2. Manchester – New York

(209)   In Summer 2017 IATA Season, [200.000-300.000] passengers travelled between Manchester and John F. Kennedy airports and [20.000-40.000] passengers flew between Manchester and Newark airports. (201) In Winter 2017/2018 IATA Season, [100.000-200.000]    passengers    travelled     between     Manchester     and   John F. Kennedy airports while [40.000-60.000] passengers travelled between Manchester and Newark airports. (202)

(210)   In Summer 2017 and Winter 2017/2018 IATA Seasons, Virgin Atlantic/Delta operated direct services on the MAN-JFK airport pair, while AFKL offered only indirect services on this airport pair via Paris-Charles de Gaulle or Amsterdam airports.

(211)   As explained in section 4.2.1.4 above, the Commission will assess the effect of  the Transaction on the following airport pairs: MAN-JFK/EWR and MAN-JFK. (203)

(212)   The table below provides the market shares of the Parties on the MAN-JFK/EWR airport pair.

(213)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment attributable to the Transaction would be [0-5]% in Summer 2017 IATA Season and [0-5]% in Winter 2017/2018 IATA Season.

(214)   The table below provides the market shares of the Parties on the MAN-JFK airport pair.

(215)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment attributable to the Transaction would be [0-5]% in Summer 2017 IATA Season and [0-5]% in Winter 2017/2018 IATA Season.

(216)   In addition, the Parties will remain constrained by the competition of the daily all- year round direct service operated by Star Alliance on the MAN-EWR airport pair. (206) With respect to indirect flights, oneworld and Star Alliance, Aer Lingus and Icelandair operate on the MAN-JFK and MAN-EWR airport pairs all-year round. (207)

(217)   Thomas Cook/Condor also provides three to four weekly direct services all-year round on the MAN-JFK airport pair. (208) The Parties consider that Thomas Cook/Condor  exert  a  meaningful  competitive  constraint  on  the  Parties   post-Transaction, particularly in the non-time-sensitive segment of the market. (209)

(218)   Irrespective of whether seats sold by Thomas Cook/Condor are included or not, in both IATA Seasons, under any plausible market definition, the increment brought about by the Transaction is de minimis. Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (210)

(219)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the Manchester – New York route under any plausible market definition.

5.1.2.3. London – Atlanta

(220)   In Summer 2017 IATA Season, [100.000-200.000] passengers travelled between London Heathrow and Atlanta airports (211) and [100.000-200.000] passengers flew between London Heathrow, Gatwick and City and Atlanta airports. In Winter 2017/2018 IATA Season, [80.000-100.000] passengers travelled between London Heathrow and Atlanta airports and [80.000-100.000] passengers flew between London Heathrow, Gatwick and City and Atlanta airports. (212)

(221)   In Summer 2017 and Winter 2017/2018 IATA Seasons, Virgin Atlantic/Delta operated direct services as well as indirect services on the LHR-ATL airport pair, while AFKL offered only indirect services on this airport pair as well as the LGW-ATL airport pair via Paris-Charles de Gaulle or Amsterdam airports.

(222)   As explained in section 4.2.1.4 above, the Commission will assess the effect of  the Transaction on the following airport pairs: LHR-ATL and LHR/LGW/LCY- ATL.

(223)   The table below provides the market shares of the Parties on the LHR-ATL airport pair.

(224)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer 2017 IATA Season and [0-5]% in Winter 2017/2018 IATA Season. With regard to  NTS passengers, the increment attributable to the Transaction would be [0-5]% in Summer 2017 IATA Season and [0-5]% in Winter 2017/2018 IATA Season.

(225)   The table below provides the market shares of the Parties on the LHR/LGW/LCY-ATL airport pair.

(226)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction was [0-5]% in Summer 2017 IATA Season and [0-5]% in Winter 2017/2018 IATA Season. With regard to NTS passengers, the increment attributable to the Transaction was [0-5]% in Summer 2017 IATA Season and [0-5]% in Winter 2017/2018 IATA Season.

(227)   In addition, the Parties will remain constrained by the competition of the daily all- year round direct service operated by oneworld on the LHR-ATL airport pair. (213) In addition, oneworld operates an indirect service on the LHR-ATL airport pair. (214)

(228)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction .(215)

(229)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the London – Atlanta route under any plausible market definition.

5.1.2.4. London – Minneapolis

(230)   In Summer 2017 IATA Season, [60.000-80.000] passengers travelled between London Heathrow and Minneapolis airports and [60.000-80.000] passengers flew between London Heathrow, Gatwick and City and Minneapolis airports. (216) In Winter 2017/2018 IATA Season, [20.000-40.000] passengers travelled between London Heathrow and Minneapolis airports and [20.000-40.000] passengers flew between London Heathrow, Gatwick and City and Minneapolis airports. (217)

(231)   In Summer 2017 and Winter 2017/2018 IATA Seasons, Virgin Atlantic/Delta operated direct services as well as indirect services on the LHR-MSP airport pair, while AFKL offered only indirect services on this airport pair via Paris-Charles  de Gaulle or Amsterdam airports.

(232)   As explained in section 4.2.1.4 above, the Commission will assess the effect of  the Transaction on the following airport pairs: LHR-MSP and LHR/LGW/LCY- MSP.

(233)   The table below provides the market shares of the Parties on the LHR-MSP airport pair.

(235)   In both IATA Seasons and on both airport pairs, the increment brought about by the Transaction is rather limited and does not exceed [5-10]%. (220) In the TS segment the increment is de minimis (below [0-5]%).

(236)   On both airport pairs, the Parties will remain constrained by the competition of other carriers. In particular, Star Alliance, oneworld and Icelandair provide a year-round indirect service on the LHR-MSP airport pair; in addition, Icelandair provides a year-round indirect service on the LGW-MSP airport pair.

(237)   The market shares of the Parties’ competitors are as follows:

(238)   The increment brought about by the Transaction is lower than the market shares  of the Parties’ biggest competitor post-Transaction (with the exception of the  NTS market segmentation in Winter 2017/2018 IATA Season on both airport pairs, on which the increment is around 1 percentage point higher than the market share of the biggest competitor).

(239)   As already stated above (Section 5.1.1.5), the Commission considers that other carriers like the members of the Star Alliance or oneworld exert significant competitive pressure on the Parties post-Transaction on the affected direct/indirect overlap routes.  Besides, the Parties submit that the market shares  of Virgin might be overstated as, over the last two seasons, between 40% and  50% of passengers were connecting at Minneapolis to or from other destinations. (221)

(240)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction.

(241)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the London - Minneapolis route under any plausible market definition.

5.1.2.5. London – Detroit

(242)   In Summer 2017 IATA Season, [40.000-60.000] passengers travelled between London Heathrow and Detroit airports (222); [40.000-60.000] passengers travelled between London Heathrow, Gatwick and City and Detroit airports. In Winter 2017/2018 IATA Season, [20.000-40.000] passengers travelled between London Heathrow and Detroit airports; [20.000-40.000] travelled between London Heathrow, Gatwick and City and Detroit airports. (223)

(243)   In both Summer 2017 and Winter 2017/2018 IATA Seasons, Virgin Atlantic/Delta operated direct services on the LHR-DTW airport pair, as well as indirect services on this airport pair. AFKL serves LHR-DTW on an indirect basis, via Amsterdam or Paris Charles-de-Gaulle airports. (224)

(244)   Given that the Parties’ market shares are similar on both LHR-DTW and LHR/LGW/LCY airports pairs, (225) the Commission will assess the effect of the Transaction on the LHR/LGW/LCY-DTW airport pair.

(245)   As explained in section 4.2.1.4 above, The Commission will assess the effects of the Transaction on the following airport pairs: LHR-DTW and LHR/LGW/LCY- DTW.

(246)   The table below provides the market shares of the Parties on the LHR/ LGW/LCY-DTW airport pair.

(250)   It is noteworthy that the increment attributable to the Transaction (overall and in respect of the TS and NTS segments) is lower than the market shares of the Parties’ biggest competitor post-Transaction in both Summer 2017 and Winter 2017/2018 IATA Seasons.

(251)   As already stated above (Section 5.1.1.5), the Commission considers that other carriers like the members of the Star Alliance or oneworld exert significant competitive pressure on the Parties post-Transaction on the affected direct/indirect overlap routes.

(252)   Besides, the Parties submit that the market share attributable to Virgin Atlantic/Delta is likely to overstate its position on that route as more than 60% of passengers on the London – Detroit route are connecting to another service. (227)

(253)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (228)

(254)   In light of the above and of all available evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the London - Detroit route under any plausible market definition.

5.1.2.6. Edinburgh – New York

(255)   In Summer 2017 IATA Season, [60.000-80.000] passengers travelled between Edinburgh and John F. Kennedy airports (229) and [40.000-60.000] passengers flew between Edinburgh and Newark airports. In Winter 2017/2018 IATA Season, [20.000-40.000] passengers travelled between Edinburgh and John F. Kennedy airports while [20.000-40.000] passengers travelled between Edinburgh and Newark airports. (230)

(256)   In Summer 2017 and Winter 2017/2018 IATA Seasons, Virgin Atlantic/Delta operated both direct and indirect services on the EDI-JFK airport pair, while AFKL  offered  only  indirect  services  on  the  EDI-JFK  airport  pair  via  Paris-

Charles de Gaulle or Amsterdam airports. (231)

(257)   As explained in section 4.2.1.4 above, the Commission will assess the effect of  the Transaction on the following airport pairs: EDI-JFK/EWR and EDI-JFK. (232)

(258)   The table below provides the market shares of the Parties on the EDI-JFK/EWR airport pair.

(261)   The increment brought about by the Transaction is de minimis. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter 2017/2018 IATA Season. With regard to NTS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(262)   While oneworld operated a direct service on the EDI-JFK airport pair in the Summer 2018 IATA Season, it ceased operating this route and will operate the EDI-PHL airport pair instead. However, the Parties will remain constrained by  the direct service of Star Alliance on EDI-EWR as well as the indirect services  ofoneworld, Star Alliance and Aer Lingus on the EDI-JFK and EDI-EWR airport pairs. (235)

(263)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (236)

(264)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the Edinburgh – New York route under any plausible market definition.

5.1.2.7. London – Portland

(265)   In Summer 2017 IATA Season, [40.000-60.000] passengers travelled between London Heathrow and Portland airports (237) and [40.000-60.000] passengers flew between London Heathrow, Gatwick and City and Portland airports. In Winter 2017/2018 IATA Season, [0-20.000] passengers travelled between London Heathrow and Portland and [0-20.000] passengers flew between London Heathrow, Gatwick and City and Portland. (238)

(266)   In Summer 2017 IATA Season, Virgin Atlantic/Delta operated direct services as well as indirect services on the LHR-PDX airport pair, while Virgin Atlantic/Delta operated only indirect services on the LHR-PDX airport pair in Winter 2017/2018 IATA Season. AFKL offered only indirect services on this airport pair via Paris-Charles de Gaulle or Amsterdam airports.

(267)   As explained in section 4.2.1.4 above, the Commission will assess the effect of  the Transaction on the following airport pairs: LHR-PDX and LHR/LGW/LCY- PDX.

(268)   The table below provides the market shares of the Parties on the LHR-PDX airport pair.

(271)   The increment brought about by the Transaction is de minimis. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(272)   In addition, the Parties will remain constrained by the competition of other  airlines operating an indirect service on the LHR-PDX airport pair, for example oneworld and Star Alliance. (241)

(273)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (242)

(274)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the London – Portland route under any plausible market definition.

5.1.2.8. Manchester – San Francisco

(275)   In Summer 2017 IATA Season, [60.0000-80.000] passengers travelled between Manchester and San Francisco airports. (243) In Winter 2017/2018 IATA Season, [0-20.000] passengers travelled between Manchester and San Francisco airports. (244)  The only plausible airport pair is MAN-SFO. (245)

(276)   In both Summer 2017 and Winter 2017/2018 IATA Seasons, Virgin Atlantic/Delta operated direct services on the MAN-SFO airport pair, while  AFKL offered indirect services on this route via Paris Charles-de-Gaulle or Amsterdam airports. (246)

(277)   The table below provides the market shares of the Parties on the MAN-SFO airport pair.

(278)   The increment brought about by the Transaction is de minimis. With regard to TS passengers,  the  increment  attributable  to  the  Transaction  would  be  [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(279)   In addition, the Parties will remain constraint by the competition of other airlines operating this route. Oneworld and Star Alliance offer indirect services on this route year-round. (248)

(280)   Non-stop services were provided by Thomas Cook/Condor in Summer 2017 IATA Season. (249) Irrespective of whether seats sold by Thomas Cook/Condor are included or not, in both IATA Seasons, under any plausible market definition, the increment brought about by the Transaction is de minimis.

(281)   The majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (250)

(282)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the Manchester – San Francisco route under any plausible market definition.

5.1.3. Direct/Indirect affected markets between Continental Europe/Ireland and North America

(283)   As explained in paragraph (178) above, the Commission treats each of the metal neutral joint ventures as a single entity for the purpose of assessing the market shares on all relevant routes. Consistent with the approach, Delta and AFKL are treated as a single entity on every route between the Continental Europe/ Ireland and North America.

5.1.3.1. Amsterdam – Los Angeles

(284)   In Summer 2017 IATA Season, [100.000-200.000] passengers travelled between Amsterdam and Los Angeles airports. (251) In Winter 2017/2018 IATA Season, [40.000-60.000] passengers travelled between Amsterdam and Los Angeles airports. (252)  The only plausible airport pair is AMS-LAX. (253)

(285)   In both Summer 2017 and Winter 2017/2018 IATA Seasons, AFKL/Delta operated direct services on the Amsterdam – Los Angeles route, while Virgin Atlantic offered indirect services on this route via London. (254)

(286)   The table below provides the market shares of the Parties on the AMS-LAX airport pair.

(287)   The increment brought about by the Transaction is de minimis. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(288)   In addition, both oneworld and Star Alliance offer indirect services on this route all year long. (255)

(289)   The majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (256)

(290)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the Amsterdam – Los Angeles route under any plausible market definition.

5.1.3.2. Amsterdam – Miami

(291)   In Summer 2017 IATA Season, [60.000-80.000] passengers travelled between Amsterdam   and   Miami   airports. (257)     In   Winter   2017/2018   IATA    Season, [40.000-60.000] passengers travelled on this route. (258) The only plausible airport pair is AMS-MIA. (259)

(292)   In both Summer 2017 and Winter 2017/2018 IATA Seasons, AFKL/Delta operated direct services on the AMS-MIA airport pair, while Virgin Atlantic offered indirect services on this airport pair via London. (260)

(293)   The table below provides the market shares of the Parties on the AMS-MIA airport pair.

(294)   The increment brought about by the Transaction is de minimis. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(295)   In addition, the Parties will remain constrained by other airlines operating this route. Oneworld, Star Alliance, TAP Portugal and Air Europa offer indirect services on this airport pair all year long. (262)

(296)   Non-stop services were provided by TUI in both Summer 2017 and Winter 2017/2018 IATA Season. Irrespective of whether seats sold by TUI are included or not, in both IATA Seasons, under any plausible market definition, the increment brought about by the Transaction is de minimis.

(297)   The majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (263)

(298)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the Amsterdam – Miami route under any plausible market definition.

5.1.3.3. Amsterdam – San Francisco

(299)   In Summer 2017 IATA Season, [100.000-200.000] passengers travelled between Amsterdam and San Francisco airports. In Winter 2017/2018 IATA Season, [40.000-60.000] passengers travelled between Amsterdam and San Francisco airports. (264)  The only plausible airport pair is AMS-SFO. (265)

(300)   In both Summer 2017 and Winter 2017/2018 IATA Seasons, AFKL/Delta operated direct services on the AMS-SFO airport pair, while Virgin Atlantic offered indirect services on this airport pair via London. (266)

(301)   The table below provides the market shares of the Parties on the AMS-SFO airport pair.

(302)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(303)   In addition, the Parties will remain constrained by other airlines operating this route. oneworld and Star Alliance offer indirect services on this route all year long. (267)

(304)   The majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (268)

(305)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the Amsterdam – San Francisco route under any plausible market definition.

5.1.3.4. Paris – San Francisco

(306)   In Summer 2017 IATA Season, [200.000-300.000] passengers travelled between Paris-Charles de Gaulle and San Francisco airports and [200.000-300.000] passengers flew between Paris-Charles de Gaulle, Orly and San Francisco  airports. In Winter 2017/2018 IATA Season, [80.000-100.000] passengers travelled between Paris-Charles de Gaulle and San Francisco airports and [80.000-100.000] passengers flew between Paris-Charles de Gaulle, Orly and San Francisco airports. (269)

(307)   In Summer 2017 and Winter 2017/2018 IATA Seasons, AFKL/Delta operated direct services on the CDG-SFO airport pair, while Virgin Atlantic offered only indirect services on this airport pair.

(308)   As explained in section 4.2.1.4 above, the Commission will assess the effect of  the Transaction on the following airport pairs: CDG-SFO and CDG/ORY-SFO. (270)

(309)   The table below provides the market shares of the Parties on the CDG-SFO airport pair.

(310)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(311)   he table below provides the market shares of the Parties on the CDG/ORY-SFO airport pair.

(312)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(313)   In addition, the Parties will remain constrained by the competition of the daily all- year round direct service operated by Star Alliance on the CDG-SFO airport pair. (271) XL Airways also operated a weekly non-stop service in Summer 2017 IATA Season. (272) In addition, oneworld and Star Alliance operate an indirect service on the CDG-SFO airport pair. (273) On the ORY-SFO airport  pair, Frenchbee commenced to operate a weekly non-stop service and Star Alliance operate an indirect service. (274)

(314)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (275)

(315)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the Paris – San Francisco route under  any  plausible market definition.

5.1.3.5. Paris – Boston

(316)   In Summer 2017 IATA Season, [100.000-200.000] passengers travelled between Paris Charles de Gaulle and Boston airports and [100.000-200.000] passengers flew between Paris Charles de Gaulle, Orly and Boston airports. (276) In Winter 2017/2018 IATA Season, [40.000-60.000] passengers travelled between Paris Charles de Gaulle and Boston airports and [60.000-80.000] passengers flew between Paris Charles de Gaulle, Orly and Boston airports. (277)

(317)   In Summer 2017 and Winter 2017/2018 IATA Seasons, AFKL/Delta operated direct services on the CDG-BOS airport pair, while Virgin Atlantic offered only indirect services on this airport pair.

(318)   As explained in section 4.2.1.4 above, the Commission will assess the effect of  the Transaction on the following airport pairs: CDG-BOS and CDG/ORY-BOS.

(319)   The table below provides the market shares of the Parties on the CDG-BOS airport pair.

(320)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(321)   The table below provides the market shares of the Parties on the CDG/ORY-BOS airport pair.

(322)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(323)   In addition, the Parties will remain constrained by the other airlines operating on this route. On the CDG-BOS airport pair, oneworld operated a weekly direct service in Summer 2017 IATA Season and Norwegian started to operate a non- stop service in 2018. (278) In addition, Star Alliance, oneworld, Aer Lingus and Icelandair operate indirect services all year round. (279) On the ORY-BOS airport pair, oneworld, Icelandair and TAP Portugal operate indirect services. (280) Air Europa operated a one-stop service in Summer 2017 IATA Season. (281)

(324)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (282) A competitor  also indicated it intends to provide several weekly frequency on the route on a seasonal basis. (283)

(325)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the Paris – Boston route under any plausible market definition.

5.1.3.6. Nice – New York

(326)   In Summer 2017 IATA Season, [60.000-80.000] passengers travelled between Nice and John F. Kennedy airports and [5.000-10.000] passengers flew between Nice and Newark airports. (284) In Winter 2017/2018 IATA Season, [0-20.000] passengers travelled between Nice and John F. Kennedy airports while [0-5.000] passengers travelled between Nice and Newark airports. (285)

(327)   In Summer 2017 Season, AFKL/Delta operated direct services on the NCE-JFK airport pair and an indirect service all-year round; Virgin Atlantic operates an indirect service on this airport pair. (286) AFKL/Delta and Virgin Atlantic also operate indirect services on the NCE-EWR airport pair on a year-round basis. (287)

(328)   As explained in section 4.2.1.4 above, the Commission will assess the effect of  the Transaction on the following airport pairs: NCE-JFK, NCE-EWR and NCE-JFK/EWR.

(329)   The table below provides the market shares of the Parties on the NCE-JFK airport pair.

(330)   In Summer IATA Season, the increment brought about by the Transaction is de minimis under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% with respect to the NTS segment. In Winter IATA Season, the increment brought about by the Transaction is moderate on the overall and NTS segments (below [10-20]%). With regard to TS passengers in Winter IATA Season, the increment attributable to the Transaction would be de minimis ([0-5]%).

(331)   The market shares of the Parties’ competitors are as follows:

(335)   In Summer IATA Season, the increment brought about by the Transaction is de minimis under any plausible market definition. With respect to the TS passengers, the increment attributable to the Transaction would be [0-5]%. With regard to NTS passengers, the increment would be [0-5]%.

(336)   In Winter IATA Season, the Parties’ combined market shares are below [40-50]% under any plausible market definition and the increment brought about by the Transaction is moderate. The increment is [10-20]% in respect of NTS passengers and [10-20]%

(337)   On all plausible airport pairs, the Parties will remain constrained by the competition of other carriers’ services. In particular, oneworld operates a year round indirect service on the NCE-JFK airport pair (288) and Star Alliance is operating a year-round indirect service on  NCE-EWR. (289)  In  addition,  as  of  May 2019, La Compagnie will offer a five-time weekly direct service on NCE-EWR on a seasonal basis. (290)

(338)   As already stated above (Section 5.1.1.5), the Commission considers that other carriers like the members of the Star Alliance or oneworld exert significant competitive pressure on the Parties post-Transaction on the affected direct/indirect overlap routes.

(339)   It is noteworthy that the increment attributable to the Transaction is lower than  the market share of the Parties’ biggest competitor post-Transaction in both Summer 2017 and Winter 2017/2018 IATA Seasons under any plausible market segmentation.

(340)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (291)

(341)   In light of the above and of all available evidence, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the Nice – New York route under any plausible market definition.

5.1.3.7. Dublin – New York

(342)   In Summer 2017 IATA Season, [100.000-200.000] passengers travelled between Dublin and John F. Kennedy airports and [100.000-200.000] passengers flew between Dublin and Newark airports. (292) In Winter 2017/2018 IATA Season, [100.000-200.000] passengers travelled between Dublin and John F. Kennedy airports while [60.000-80.000] passengers travelled between Dublin and Newark airports. (293)

(343)   In Summer 2017 and Winter 2017/2018 IATA Seasons, AFKL/Delta operated direct services on the DUB-JFK airport pair, while Virgin Atlantic offered only indirect services on this airport pair via London. (294) Furthermore, Virgin Atlantic also offer a one-stop service on the DUB-EWR airport pair. (295)

(344)   As explained in section 4.2.1.4 above, the Commission will assess the effect of  the Transaction on the following airport pairs: DUB-JFK/EWR and DUB-JFK. (296)

(345)   The table below provides the market shares of the Parties on the DUB-JFK/EWR airport pair.

(346)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers,   the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(347)   The table below provides the market shares of the Parties on the DUB-JFK airport pair.

(348)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(349)   In addition, Aer Lingus offers a daily all-year round non-stop service on the DUB-JFK airport pair. (297) With respect to indirect flights, oneworld and Star Alliance operate all-year round. (298) On the DUB-EWR airport pair, oneworld and Aer Lingus offer a daily all-year round non-stop service; Star Alliance and oneworld also offer a one-stop service on this airport pair. (299)

(350)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (300)

(351)   In light of the above and of all available evidence, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the Dublin – New York route under any plausible market definition.

5.1.4.  Direct/Indirect affected markets between the United Kingdom and destinations excluding North America and Continental Europe/Ireland

5.1.4.1.London – Lagos

(352)   In Summer 2017 IATA Season, [100.000-200.000] passengers travelled between London Heathrow and Lagos airports and [200.000-300.000] passengers flew between London Heathrow, Gatwick and City and Lagos airports. (301) In Winter 2017/2018 IATA Season, [100.000-200.000] passengers travelled between London Heathrow and Lagos airports and [100.000-200.000] passengers flew between London Heathrow, Gatwick and City and Lagos airports. (302)

(353)   In both Seasons, Summer 2017 IATA Season and Winter 2017/2018 IATA Season, Virgin Atlantic operated direct services as well as indirect services on the LHR-LOS airport pair. AFKL offers indirect services on this airport pair as well as on the LCY-LOS airport pair. Delta does not operate this route.

(354)   As explained in section 4.2.1.4 above, the Commission will assess the effect of  the Transaction on the following airport pairs: LHR-LOS and LHR/LGW/LCY- LOS.

(355)   The table below provides the market shares of the Parties on the LHR-LOS airport pair.

(356)   The    table   below     provides     the    market   shares   of    the    Parties   on    the LHR/LGW/LCY-LOS airport pair.

(357)   On both airport pairs, the Parties’ combined market share would be below 60% in both IATA Seasons.

(358)   Post-Transaction, the Parties will remain constrained by the competition from other airlines operating the London – Lagos route. In particular, oneworld is a significant competitor operating a direct daily all-year round service on the LHR- LOS airport pair, whose market share (on all plausible market) is higher than the increment brought about by the Transaction. (303)

(359)   The Parties’ competitors’ market share is as follows:

(360)   The increment brought about by the Transaction is lower than the market shares  of the Parties’ biggest competitor post-Transaction.

(361)   As already stated above (Section 5.1.1.5), the Commission considers that other carriers like the members of the Star Alliance or oneworld exert significant competitive pressure on the Parties post-Transaction on the affected direct/indirect overlap routes.

(362)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (304)

(363)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the London – Lagos route under any plausible market definition.

5.1.4.2. London – Shanghai

(364)   In Summer 2017 IATA Season, [300.000-400.000] passengers travelled between London Heathrow and Shanghai airports and [300.000-400.000]passengers flew between London Heathrow, Gatwick and City and Shanghai airports. (305) In Winter 2017/2018 IATA Season, [100.000-200.000] passengers travelled between London Heathrow and Shanghai and [100.000-200.000] passengers flew between London Heathrow, Gatwick and City and Shanghai. (306)

(365)   In both Seasons, Summer 2017 IATA Season and Winter 2017/2018 IATA Season, Virgin Atlantic operated direct services on the LHR-PVG airport pair. AFKL offers indirect services on this airport pair as well as on the LCY-PVG airport pair. Delta does not operate on this route.

(366)   As explained in section 4.2.1.4 above, the Commission will assess the effect of  the Transaction on the following airport pairs: LHR-PVG and LHR/LGW/LCY- PVG.

(367)   The table below provides the market shares of the Parties on the LHR-PVG airport pair.

(370)   The increment brought about by the Transaction is de minimis in both IATA Seasons and under any plausible market definition. With regard to TS passengers, the increment attributable to the Transaction would be [0-5]% in Summer IATA Season and [0-5]% in Winter IATA Season. With regard to NTS passengers, the increment attributable to the Transaction would be [5-10]% in Summer IATA Season and [0-5]% in Winter IATA Season.

(371)   In addition, post-Transaction, the Parties will remain constrained by the all year round direct service of China Eastern and oneworld on the LHR-PVG airport pair. In addition, several other airlines provide indirect services on the LHR-PVG and LGW-PVG airport pairs. (307)

(372)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction .(308)

(373)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the London – Shanghai route under any plausible market definition.

5.1.4.3. London – Havana

(374)   In Summer 2017 IATA Season, [40-000-60.000] passengers travelled between London Heathrow, Gatwick and City and Havana airports. (309) In Winter  2017/2018 IATA Season, [40-000-60.000] passengers flew between London Heathrow, Gatwick and City and Havana airports.  (310)

(375)   Virgin Atlantic operates direct flights on the LGW-HAV airport pair, as well as  an indirect service all-year round. AFKL offers indirect flights on the LHR-HAV airport pair in both IATA Seasons. Delta does not operate on this route. (311) Therefore, the Parties’ activities do not overlap on the LHR-HAV airport pair.

(376)   In view of the above and as explained in section 4.2.1.4 above, the Commission will assess the effect of the Transaction on the LHR/LGW/LCY-HAV airport  pair.

(377)   The table below provides the market shares of the Parties on the LHR/LGW/LCY-HAV airport pair. (312)

(378)   Virgin Atlantic has high market shares on the London – Havana route, being the only carrier offering direct flights on this route. (314) In Winter IATA Season, on the LHR/LGW/LCY-HAV airport pair, the increments are moderate. In Summer IATA Season, the increments are between [5-10] and [10-20]%. In addition, the Parties submit that Virgin Atlantic has decreased its direct services from three to two weekly services in Winter 2018/2019 IATA Season. (315)

(379)   It is noteworthy that Virgin Atlantic operated a twice weekly direct flight in Summer 2017 IATA Season; it increased its offering to three direct weekly frequencies in Winter 2017/2018 but it has decreased its direct services to two weekly services in Winter 2018/2019 IATA Season. (316) Despite high combined market shares, the Parties will remain subject to an effective competitive constraint post-Transaction on the above airport pairs. Close to 25% of passengers on the London – Havana route flew on indirect services in IATA year 2017. (317) Indirect flights therefore exercise an effective competition constraint on Virgin Atlantic’s direct flights, notably because they are more frequent than its twice- weekly direct offering. (318) In particular, Spanish carriers Iberia and Air Europa offer indirect flights on the LGW-HAV airport pair and IAG (Iberia and British Airways), Aeroméxico and Alitalia offer indirect flights on the LHR-HAV airport pair. (319)

(380)   The market shares of the Parties’ main competitors are as follows:

(381)   Besides, the Parties submit that charter carriers offering dry seats not captured by the DDS data are active on the route and exert a meaningful competitive constraint on carriers operating scheduled services, in particular, given the leisure nature of the route. (320) Therefore, it is very likely that the Parties’ market shares  are overstated.

(382)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (321)

(383)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the London - Havana route under any plausible market definition.

5.1.4.4.  London – Montego Bay

(384)   In Summer 2017 IATA Season, [40.000-60.000] passengers travelled between London Heathrow and Gatwick and Montego Bay airports. (322) In Winter 2017/2018 IATA Season, [40.000-60.000] passengers flew between London Heathrow and Gatwick and Montego Bay airports. (323)

(385)   Virgin Atlantic operates direct flights on the LGW-MBJ airport pair, as well as an indirect service all-year round on the LHR-MBJ airport pair. AFKL offers an indirect service on the LHR-MBJ airport pair, but has a minimal number of bookings. (324) Therefore, the Parties’ activities do not overlap on the LHR-MBJ airport pair. Based on the information provided by the Parties, no airline  operates the LCY-MBJ airport pair. (325) Therefore, the Commission will assess the effects  of the Transaction on the LHR/LGW-MBJ airport pair combination.

(386)   The table below provides the market shares of the Parties on the LHR/LGW-MBJ airport pair.

(387)   The Parties submit Virgin Altantic has high market shares in the TS segment in Summer and Winter IATA Seasons, because it is the only scheduled carrier to operate a direct service on this route. (326)

(388)   The Parties consider that both TUI and Thomas Cook/Condor operate (charter) direct services on this route and exert a meaningful competitive constraint on the Parties post-Transaction, particularly in the non-time-sensitive segment of the market and in the overall market through the sale of dry seats (the latter being particularly relevant in respect of the leisure nature of the route). (327)

(389)   Irrespective of whether seats sold by TUI and Thomas Cook/Condor are included or not, in both IATA Seasons, under any plausible market definition, the increment brought about by the Transaction is de minimis.

(390)   Finally, the majority of respondents to the market investigation having expressed a view considers that there will be sufficient competition on the route to prevent the Parties’ from raising prices on the route post-Transaction. (328)

(391)   In light of the above and of all evidence available to it, the Commission considers that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the London – Montego Bay route under any plausible market definition.

5.1.5.  Indirect/Indirect overlaps

(392)   As explained above, the Commission treats each of the metal neutral joint ventures as a single entity for the purpose of assessing the market shares on all relevant routes. Consistent with the approach, Delta and AFKL are treated as a single entity on every route between the Continental Europe/ Ireland and North America and Delta and Virgin Atlantic are treated as a single entity on every  route between the United Kingdom and North America.

(393)   The activities of the Parties also give rise to affected markets on 13 routes on a city-pair level (329), where both (only) operate indirect flights. It is noteworthy that some of these routes are affected only on some segments or during only one of  the two relevant IATA Seasons.

(394)   These routes are analysed with respect to combined market shares of the Parties and the market share increment brought about by the Transaction.

(395)   Given that the Transaction does not give rise to serious doubts as to its compatibility with the internal market under any plausible market definition, the Commission will, for the purpose of this Decision, take a conservative approach and assess the effects of the Transaction on the airport pair where the Parties’ market shares are the highest.

(396)   An overview of the Parties’ market shares on the relevant routes is provided in the table below.

(397)   Indirect routes are often established in an opportunistic way by carriers and are modified from one IATA season to the next. Furthermore, price increases or reductions of capacity could be countered by competitors who could start operating on these routes more easily than on direct/direct routes, which require the deployment of aircraft dedicated to the O&D route.

(398)   On all of these routes no competition concerns arise because either the Parties combined market shares are below 60% or the increment brought about by the Transaction is below 5% (so that no material merger-specific effect would likely exist). Given the low competitive constraint between indirect services, market shares below 60% on routes indicate that there is already prima facie sufficient competition from other carriers.

(399)   The majority of the respondents having expressed a view considers that there will be sufficient competition on the indirect/indirect overlap routes to prevent the Parties from raising prices in the markets for passenger air transport services after the Transaction. (333)

(400)   In light of the above, the Commission considers that the Transaction does not raise any serious doubts as to its compatibility with the internal market under any possible market definition on any of the 13 indirect/indirect overlap routes.

5.1.6.  Conclusion

(401)   In light of the above, the Commission considers that, under the O&D approach, the Transaction does not raise any serious doubts as to its compatibility with the internal market with regard to passenger air transport services, under any plausible market definition.

5.2. Passenger air transport services under the airport-by-airport approach

5.2.1. Introduction

(402)   According to paragraph 36 of the Horizontal Merger Guidelines, (334) "some proposed mergers would, if allowed to proceed, significantly impede effective competition by leaving the merged firm in a position where it would have the ability and incentive to make the expansion of smaller firms and potential competitors more difficult or otherwise restrict the ability of rival firms to compete. In such a case, competitors may not, either individually or in the aggregate, be in a position to constrain the merged entity to such a degree that it would not increase prices or take other actions detrimental to competition. For instance, the merged entity may have such a degree of control, or influence over, the supply of inputs or distribution possibilities that expansion or entry by rival firms may be more costly."

(403)   As explained in section 4.1.2 above, the Transaction entails the increase of the Parties’ combined slot holding at airports where their slot portfolios overlap, namely London Heathrow and Manchester airports.

(404)   Accordingly, the Commission will first assess whether the Transaction, by reinforcing the Parties' slot holding at a number of airports and granting it broader access to their infrastructure, gives the Parties the ability and incentive to prevent other air carriers from getting access to airport infrastructure and therefore to the markets for the provision of passenger air transport services from those airports, preventing or reducing competition on those markets (sections 5.2.2 and 5.2.3 below). The Commission will then analyse the overall effects of the Parties’ slot holding position on the ability of the Parties’ rivals to compete at the relevant airports (section 5.2.4).

5.2.2. Ability of the Parties to foreclose access to the market for the provision of passenger air transport services

5.2.2.1. Conditions for the ability to foreclose access to the market for the provision of passenger air transport services

(405)   In line with its prior decisional practice, the Commission considers that, for the Parties to be able to foreclose their competitors post-Transaction, the following conditions must be fulfilled: (i) the slots that the Parties would hold post- Transaction represent a significant share of the airport capacity, in particular at peak times; (ii) the Transaction has a material impact on the Parties' slot holding at the airport, in particular at peak times; and (iii) the Parties’ slot holding could negatively affect the overall availability of slots as an input for the passenger air transport markets to or from the relevant airport. (335)

(406)   In the present case, the combined slot holdings post-Transaction do not represent  a significant share of the capacity of any of the airports where the Parties’ respective portfolios overlap. In addition, the Transaction has no material impact on any of the Parties’ slot holding at any relevant  airport,  as  described  in  section 5.2.2.3 below. Therefore, it is not necessary to assess whether the third condition (i.e. the negative impact on the overall availability of slots) is fulfilled.

(407)   In assessing whether the merged entity would have the ability to foreclose, the Commission considers whether rival firms would be likely to deploy effective and timely counter-strategies in case of foreclosure. In this case, the Commission  notes that there are limited effective and timely counter-strategies that the Parties’ competitors would be likely to deploy in the case of a foreclosure strategy by any of them. There is no possibility for an air carrier to be less reliant on access to airport infrastructure and very limited possibility to sponsor the expansion of airport capacity or the opening of new airports. (336)

(408)   In light of the above, the Commission will assess the ability of the Parties post- Transaction to foreclose access to the markets for the provision of passenger air transport at the relevant airports by taking account of the following two factors together: (337) (i) the share of slots held by the Parties post-Transaction at the airport or at substitutable airports being high, in particular at peak times and (ii) the increment in the Parties' slot holding brought about by the Transaction at the airport or at the substitutable airports being material, in particular at peak times. Considering that the Parties' slot holdings at the relevant airports vary between the Summer and Winter IATA Seasons, the Commission will carry out separate assessments for each IATA Season.

(409)   Before conducting an airport-by-airport assessment of the Parties’ ability to foreclose access to the markets for the provision of passenger air transport (section 5.2.2.3), the Commission will detail the methodologies for determining the slot holding post-Transaction and the increment brought about by the Transaction.

5.2.2.2. Methodologies

(410)   A slot holding is defined as the ratio between the number of slots held by an air carrier (or the air carriers that are part of the same group) at an airport and the  total available slots at that airport (i.e. the airport runway capacity).

(411)   Delta and AFKL have estimated their respective slot holding (338) on the basis of AFKL’s internal slot data, while VAL estimated its slot holding through data from the slot coordinator and cross-checked against its internal data. (339) The runway capacity is based on the information provided by Airport Coordination Limited, which is the slot coordinator at the relevant overlap airports. (340)

(412)   The Commission has calculated two values for the Parties’ slot holding post- Transaction: (i) their average slot holding during the opening hours of the airport, and (ii) their average slot holding during peak times. (341)

(413)   In order to adopt a conservative approach, the Commission has calculated the Parties’ combined highest slot holding at any given hour band throughout the whole week (including any peak hour), which exceeds their combined average slot holding during peak times.

(414)   The increment brought about by the Transaction corresponds to the difference between the combined slot holding post-Transaction and VAL/Delta’s slot  holding pre-Transaction.

(415)   To assess the impact of the Transaction, the Commission therefore considers the Parties’ combined slot holding post-Transaction, as well as the increment in AFKL’s slot holding as a result of the Transaction.

5.2.2.3. Airport-by-airport assessment

(416)   As explained in section 5.2.1 above, the Commission will assess the effects of the Transaction on the Parties’ slot holding in both IATA Seasons at London Heathrow and Manchester airports.

(419)   In Summer 2019 IATA Season, the Parties estimate that their combined slot holding post-Transaction at London Heathrow and Manchester airports would be no more than [5-10]% on average during the relevant airport’s opening hours,  with a gross increment of [0-5]% at most. Such levels of slot holding indicate that the Transaction is unlikely to be problematic.

(420)   In Summer 2019 IATA Season, the Parties’ highest slot holding during any specific hour band at London Heathrow would not exceed [10-20]%. At Manchester, the Parties’ combined highest slot holding during any relevant hour band would be [5-10]%.

(421)   Therefore, in both IATA Seasons, the level of the Parties’ slot holdings post- Transaction is unlikely to give rise to serious doubts as to its compatibility with the internal market. Furthermore, the limited increment in slot holdings brought about by the Transaction at London Heathrow and Manchester airports is unlikely to have any material impact on the Parties’ slot holdings at these airports and thus their ability to foreclose access to other air carriers.

(422)   Finally, the majority of competitors having expressed a view consider that the Parties would not have the ability to prevent other carriers from providing passenger air transport services to/from London Heathrow and Manchester airports. (345)

5.2.2.4. Conclusion

(423)   In light of the above, given the Parties’ combined slot holding at London Heathrow and Manchester airports post-Transaction during the relevant IATA Seasons, the Commission considers that the Parties would likely not have the ability to foreclose competitors’ access to the markets for the provision of passenger air transport services.

5.2.3.   Incentive of the Parties to foreclose access to the markets for the provision of passenger air transport services

(424)   Although the Commission considers that the Parties will not have the ability to foreclose their competitors in either IATA Season, the Commission will briefly assess whether the Parties would have the incentive to foreclose access to market for the provision of passenger air transport at London Heathrow and Manchester.

(425)   A dominant carrier at a relevant airport would in principle have a strong incentive to pursue a foreclosure strategy, as any new service or expansion by another carrier would be likely to introduce or increase competition on one of the dominant carrier's routes. Such dominant carrier would have a greater incentive than any other carrier at the airport to keep slots out of reach of other carriers. The incentive to foreclose would also grow with the increased size of the slot portfolio it would control at the airport.

(426)   The Commission notes that the presence of the Parties at London Heathrow and Manchester airports is limited and they are evidently not dominant. At both airports, the Parties face the competition of the oneworld alliance, which holds a larger slot portfolio than the Parties. (346) At London Heathrow, British Airways operates a base of 233 aircraft and more than 50% of the slots are held by oneworld alliance’s members British Airways and American Airlines. (347)

(427)   Finally, respondents to the market investigation have not raised substantial concerns according to which the Parties had previously engaged or would engage in exclusionary practices such as "slot hoarding", (348)  "slot babysitting" (349)  or use of its "slot shuffling power". (350) A minority of respondents to the  market investigation expressed concerns about the ability of the Parties, post-Transaction, to engage in exclusionary practice at London Heathrow through their ability to share slots through their joint venture. In particular, a competitor submits that the merged entity would have the ability to “saturate capacity at times [the competitor] would prefer to operate”. (351) Nevertheless, considering the Parties’ limited slot holding, such a practice appears unlikely (as it would likely require the Parties’ to cancel or retime their own operations) and could be countered by British Airways, given its strong position at London Heathrow.

(428)   In light of the above, the Commission considers that the Parties’ would not have the incentive to foreclose competitors’ access to the markets for the provision of passenger air transport services at London Heathrow and Manchester airports post-Transaction during either IATA Winter and Summer Season.

5.2.4. Overall effect on competition for passenger air transport services

(429)   Effective competition would be significantly harmed if the foreclosed air carriers played a sufficiently important role in the competitive process on the passenger  air transport markets from and to the overlap airports. The higher the proportion  of carriers which would be foreclosed on these markets, the more likely it would be that the merger would result in a significant price increase in the passenger air transport markets and, therefore, to significantly impede effective competition therein.

(430)   On that basis, the Commission will assess whether the strengthening of the Parties’ slot holdings post-Transaction would result in less effective competition for passenger air transport services to or from London Heathrow and Manchester airports.

(431)   The Commission notes that, when an air carrier holds a significant slot portfolio  at a given airport while the remaining slot holding is very fragmented and slots  are held by a large number of small air carriers, the latter are unable to translate these slots into a viable alternative to dominant air carriers. However, as noted in sections 5.2.2 and 5.2.3 above, the Parties do not hold a slot portfolio at London Heathrow and Manchester airports post-Transaction that would enable them, or give them the incentive, to foreclose other air carriers.

(432)   In this respect, the Commission notes that the Parties will continue to face strong competition at London Heathrow and Manchester airports. At London Heathrow airport, more than 70 airlines compete with the Parties. (352)  At Manchester  airport, the Parties face competition from more than 60 airlines. (353) These competitors will likely have the ability to react to any competitive behaviour by the merged entity.

(433)   The Commission therefore considers that it is unlikely that the Transaction will harm effective competitive on the provision of passenger air transport services at London Heathrow and Manchester airports, where the Parties’ slot portfolios overlap.

5.2.5.   Conclusion

(434)   In light of the above, and considering in particular the Parties’ lack of ability to foreclose access of competitors to the markets for the provision of passenger air transport services at any relevant airports, the Commission concludes that, under the airport-by-airport approach, the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the markets for  passenger air transport to or from London Heathrow and Manchester airports, under any plausible market definition.

5.3. Cargo air transport

5.3.1. Analytical framework for the assessment of air transport of cargo – Alliances and joint ventures

5.3.1.1. Alliances and joint ventures

(435)   Delta, AF, KL and Alitalia are all parties of the TATL Joint Venture, which covers combined passenger/cargo flights, but excludes cargo-only flights. This Joint Venture involves cargo operations at seven main hubs: Amsterdam- Schiphol, Atlanta, Detroit, Minneapolis-St. Paul, New York-JFK, Paris-CDG and Rome-Fiumicino.

(436)   Delta and Virgin Atlantic are also the two members of the DL-VS Joint Venture, which includes coordination on pricing, sales and capacity in respect of UK – North America cargo flows.

(437)   In addition, AFKL and Delta are members of the SkyTeam Cargo alliance, which provides for a cooperation between participating SkyTeam members in transporting cargo across the SkyTeam global network. The SkyTeam alliance offers a less integrated cooperation than the TATL and DL-VS Joint Ventures and aims at developing the traffic between members by means of reciprocal  interlining agreements. (354)

(438)   For the purpose of the Decision, the market position of Delta and Virgin Atlantic will be analysed as being a single one on all routes covered by the DL-VS Joint Venture and the market position of AFKL and Delta will be analysed as being a single one on all routes covered by the TATL Joint Venture. (355)

5.3.1.2. Methodology used to calculate market shares

(439)   The Commission has previously accepted to take into account in its assessment estimated market shares based on World Air Cargo Data (“WACD”) and the Cargo Accounts Settlement System (“CASS”). (356) However, the Commission has also noted that CASS and/or WorldACD data do not reflect the entire air cargo markets, (357) and therefore underestimate the total market sizes and overestimate the Parties’ market shares. (358) The Parties also acknowledge the inherent limitations in the different datasets available to them with respect to the air transport of cargo. (359)

(440)   In fact, CASS does not include sales to agents that are not registered with IATA, nor direct sales to end-customers, whereas WorldACD data reflect an incomplete representation of the market as only sixty-six carriers provide their data to WorldACD.

(441)   The Parties have submitted data based on their own estimates on the basis of World Air Cargo Data (“WACD”) database. (360) However, the Parties have also compared the WACD data with alternative datasets, (361) and have indicated in their submission, where those other data sources reinforce the position as shown by the WACD data or where they suggest a different competitive position. (362)

(442)   For the purpose of the Decision, the Commission is of the view that the Parties’ best estimates on the basis of WACD data is appropriate for the assessment of the affected routes in this case.

5.3.2.   Horizontal overlaps

(443)   Air France-KLM Martinair Cargo is AFKL’s air cargo division. For the air transport of cargo, AFKL primarily uses belly-hold space in its passenger aircraft or combi-aircraft (passenger planes with main deck cargo capacity) as well as  six full-freighter aircraft. (363) Both Delta and Virgin Atlantic do not operate full- freighter aircraft and use only belly-hold space in its passenger aircraft. (364)

(444)   The Transaction gives rise to fifteen affected markets with respect to air transport of cargo. (365) The market shares of the Parties on the affected markets in 2017 were as follows:

5.3.2.1. The Parties’ views

(445)   The Parties submit that their combined market shares in any of the relevant affected routes do not give rise to competition concerns in that they are generally below [30-40]%. (366)

(446)   The Parties note that, given the limited market coverage of cargo market data sources, the Parties’ market shares on all the affected routes are likely to be overestimated. (367)

5.3.2.2. The Commission’s assessment

(447)   As can be seen in Table 45 above, the Parties’ combined market share post- Transaction will remain below [30-40]% for most of the routes and reach at most around [40-50]% with small increments of below [10-20]% and not higher than 12% in all affected routes. The only exception is UK-Mexico on which the combined market shares will be slightly above [40-50]% with an increment of [10-20]%.

(448)   In addition, on all routes, there are several credible and well-established competitors capable of imposing a significant competitive constraint on the  Parties post-Transaction, such as combination airlines (i.e. Lufthansa, British Airways, American Airways), belly-hold cargo carriers (i.e. TUI, Thomas Cook), cargo carriers (i.e. LH Cargo, Cargolux) (368) A majority of respondents to the market investigation confirmed the    existence of alternative credible suppliers in the affected routes. (369) As regards to the UK-Mexico flow, the Parties will remain under a significant competitive constraint, notably from competitors such as CargoLogicAir’s, Panalpina and Turkish Airlines. (370) This was also confirmed by  a majority of respondents to the market investigation having expressed a view. (371)

(449)   The Parties are not close competitors, since only AFKL operates both cargo-only freighters and belly-hold cargo services while Delta and Virgin only provide belly-hold cargo services. A majority of respondents to the market investigation confirmed that neither AFKL, Delta nor Virgin Atlantic are each other’s closest competitor on any affected routes. (372)

(450)   The majority of the respondents to the market investigation also indicated that switching between air cargo service suppliers is easy, given the limited costs and time frame required for a customer to switch to another cargo supplier (373) and the fact that cargo service contracts are normally concluded for six months to one  year and do not contain exclusivity clauses. (374)

(451)   Moreover, respondents to the market investigation having expressed a view did not raise any material concern as regards the effects of the Transaction in terms of prices or levels of services on any of the affected routes. (375)

5.3.2.3. Conclusion

(452)   In light of the above considerations and of all evidence available to it, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the market for air transport of cargo services under any plausible market definition.

5.4. MRO services

(453)    AFKL, Delta and, to a limited extent only, VAL are all active in the provision of maintenance, repair and overhaul (“MRO”) services. AFKL and Delta provide all the four types of MRO services (i.e. line maintenance, heavy maintenance, engine maintenance and components maintenance) to third-party carriers at various locations. The TATL Joint Venture between AFKL and Delta does not cover MRO services.

(454)   VAL is not active in heavy, engine or components maintenance and only offers line maintenance services to third-parties at three airports (JFK, JNB and LHR). (376)

5.4.1. Horizontal overlaps

(455)   In the EEA, the Transaction does not give rise to any horizontal overlaps with respect to the provision of heavy maintenance services, engine maintenance services and components maintenance services. (377) The Parties’ activities overlap at LHR airport for the provision of line maintenance services to third parties, however, the market for the supply of line maintenance services at LHR is not horizontally affected.

(456)   In light of the above considerations and all evidence available to it, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market due to horizontal effects in the markets for MRO services, under any plausible market definition.

5.4.2.Vertical relationships

(457)   The Parties are active on some vertically related markets since the provision of MRO services is an input for the provision of air transport services. However,  The Transaction does not give rise to any vertically affected markets with respect to the provision of MRO services to third parties.

(458)   In light of the above considerations and all evidence available to it, the Commission concludes that the Transaction does not raise does not raise serious doubts as to its compatibility with the internal market due to vertical effects in the market for the provision of MRO services, under any plausible market definition.

5.4.3. Conclusion

(459)   In light of the above considerations and all evidence available to it, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the markets for MRO services, under any plausible market definition.

5.5. Conclusion

(460)   In light of the above considerations and all evidence available to it, the Commission concludes that the Transaction does not raise serious doubts as to its compatibility with the internal market with respect to the markets for passengers air transport services, under both the O&D the airport-by-airport approaches, cargo air transport services and MRO services, under any plausible market definition.

6. CONCLUSION

(461)   For the above reasons, the European Commission has decided not to oppose the notified operation and to declare it compatible with the internal market and with the EEA Agreement. This decision is adopted in application of Article 6(1)(b) of the Merger Regulation and Article 57 of the EEA Agreement.

1        OJ L 24, 29.1.2004, p. 1 (the 'Merger Regulation'). With effect from 1 December 2009, the Treaty on the Functioning of the European Union ('TFEU') has introduced certain changes, such as the replacement of 'Community' by 'Union' and 'common market' by 'internal market'. The terminology of the TFEU will be used throughout this decision.

2        OJ L 1, 3.1.1994, p. 3 (the 'EEA Agreement').

3        Publication in the Official Journal of the European Union No C 020, 16.01.2019, p. 6.

4        Delta holds an 8.8% (non-controlling) stake in AFKL.

5        See paragraph (11).

6        See paragraph (11).

7        By final order issued on 23 September 2013, the DOT granted antitrust immunity with respect to the activities contemplated under the DL-VS Joint Venture and to the 5-Way Coordination Agreement.

8        Amended and Restated Transatlantic Joint Venture Agreement, dated 15 May 2018. Form CO, Annex A.09e.

9        Memorandum of Understanding in respect of the Enhanced Strategic Cooperation Agreement (“ESCA”), dated 5 October 2018. Form CO, Annex A.09f.

10     See M.6828 – Delta Airlines/Virgin Group/Virgin Atlantic Limited.

11     Sale and Purchase Agreement in respect of 31 per cent equity interest in Virgin Atlantic Limited, dated 15 May 2018. Form CO, Annex A.09b.

12     [Details on potential parties to new transatlantic joint venture agreement].

13     A metal neutral joint venture arrangement covering non-stop routes between the UK and North America.

14     Alitalia is not a party to the JVA. [Details about implementing the JVA] (Form CO, footnote 6).

15     [Details about implementation of cooperation].

16     Commission’s Decision of 12 May 2015 under Article 9(1) of Council Regulation 1/2003 in case AT.39964 – Air France/KLM/Alitalia/Delta.

17     See [reference to confidential document], Form CO, Annex A.09g.

18     See [reference to clause] of the Shareholders’ Agreement, Form CO, Annex A.09c.

19     See [reference to clause] of the Shareholders’ Agreement, Form CO, Annex A.09c.

20     See [reference to clause] of the Shareholders’ Agreement, Form CO, Annex A.09c.

21     See [reference to confidential document], Form CO, Annex A.09g.

22     See [reference to clause] of the Shareholders’ Agreement, Form CO, Annex A.09c, and [reference to confidential document], Form CO, Annex A.09g.

23     See [reference to clause], Form CO, Annex A.09c.

24     See [reference to clause] of the JVA, Form CO, Annex A.09c.

25     See [reference to clause] of the Shareholders’ Agreement, Form CO, Annex A.09c.

26     The Commission recalls that, regarding the EU air transport licensing provisions, pursuant to paragraph 23 of the Consolidated Jurisdictional Notice, "the concept of control under the Merger Regulation may be different from that applied in specific areas of Community and national legislation concerning, for example, prudential rules, taxation, air transport or the media. The interpretation of ‘control’ in other areas is therefore not necessarily decisive for the concept of control under the Merger Regulation."

27     Turnover calculated in accordance with Article 5 of the Merger Regulation.

28     AFKL, Delta and VAL calculated their turnover based on the “point of sales” methodology, but the Parties confirmed that the turnover thresholds would also be met under the “point of origin” and the “50/50 split” methods (with the exception of AFKL’s MRO turnover, which was calculated based on the  “point-of-origin”  methodology,   but   AFKL   confirmed   that   its   turnover   exceeds   the   EUR 250 million threshold under any calculation method).

29     The Parties’ activities give rise to overlap in the provision of passenger air transport services under both the route-by-route (or “O&D”) approach and the airport-by-airport approach.

30     See e.g. Cases M.7541 - IAG/Aer Lingus, paragraph 14; M.7333 - Alitalia/Etihad, paragraph 63; M.6447 - IAG/bmi, paragraph 31.

31     See e.g. Cases M.8869 – Ryanair/LaudaMotion, paragraphs 96 and 116, M.8672 – easyJet/Certain Air Berlin Assets, paragraph 41; M.8633 – Lufthansa/Certain Air Berlin Assets, paragraph 58; M.6447 – IAG/bmi, paragraph 483. For Cases M.8672 – easyJet/Certain Air Berlin Assets and M.8633 – Lufthansa/Certain Air Berlin Assets, the Commission only carried out an airport-by-airport  assessment, since the target assets did not constitute an active air carrier (Air Berlin had definitively ceased its flight operations on all routes due to its insolvency).

32     See e.g. Case M.8633 – Lufthansa/Certain Air Berlin Assets, paragraphs 164 et seq.

33     Council Regulation (EEC) No 95/93 of 18 January 1993 on common rules for the allocation of slots at Community airports (the “Slot Regulation”).

34     See e.g. Cases M.7541 – IAG/Aer Lingus, paragraphs 188 et seq.; M.6447 – IAG/bmi, paragraphs 174 and 663.

35     Recital (4) of the Commission Proposal for a Regulation of the European Parliament and of the  Council on common rules for the allocation of slots at European Union airports (COM/2011/827 final of 01 December 2011).

36     Slot Regulation, first recital: "Whereas there is a growing imbalance between the expansion of the air transport system in Europe and the availability of adequate airport infrastructure to meet that demand; whereas there is, as a result, an increasing number of congested airports in the Community."

37     Slot Regulation, Article 8(2).

38     Slot Regulation, Article 10(6).

39     The coordinator is the person responsible for the allocation of slots (Slot Regulation, Article 4(5)). According to the first sentence of Article 8a(2) of the Slot Regulation, "[t]he transfers or exchanges referred to in paragraph 1 shall be notified to the coordinator and shall not take effect prior to the express confirmation by the coordinator."

40     Form CO, paragraph 293.

41     COM/2011/827 final of 01 December 2011.

42     See e.g. Cases M.8869 – Ryanair/LaudaMotion, paragraph 116; M.8672 – easyJet/Certain Air Berlin Assets, paragraph 41; M.8633 – Lufthansa/Certain Air Berlin Assets, paragraph 58; M.6447 – IAG/bmi, paragraph 483.

43     See e.g. Cases M.8869 – Ryanair/LaudaMotion, paragraph 506 et seq. ;M.8672 – easyJet/Certain Air Berlin Assets, paragraphs 91 et seq.; M.8633 – Lufthansa/Certain Air Berlin Assets, paragraphs 160 et seq.; M.6447 - IAG/bmi, paragraph 483.

44     Form CO, paragraph 204.

45     Form CO, paragraphs 306-307. Delta holds a negligible amount of slots at AMS and CDG, namely [20-30] daily slot pairs at AMS and [20-30] daily slot pairs at CDG.

46     Form CO, paragraph 305.

47     Without prejudice to any assessment conducted under Article 101 TFUE, the Commission considers that the Transaction will not bring any change with regard to slot holding at non-overlap airports.

48     For the sake of completeness, it can be noted that, in the United Kingdom, AFKL has [5-10] daily slot pairs at Birmingham, [5-10] daily slot pairs at London City, [5-10] daily slot pairs at Newcastle, [5-10] daily slot pairs at Aberdeen, [0-5] daily slot pairs at Bristol, [0-5] daily slot pairs at Leeds Bradford, [0-5] daily slot pairs at Southampton and [0-5] daily slot pair at Belfast City. Virgin Atlantic/Delta has no slot pairs at any of these airports (Form CO, footnote 89).

49     See e.g. Cases M.8046 – TUI/Transat France, paragraphs 66-88; M.5867 – Thomas Cook/Öger Tours, paragraph 16; M.4601 – KarstadtQuelle/MyTravel, paragraph 43; M.4600 – TUI/First Choice, paragraph 57; M.3770 – Lufthansa/Swiss, paragraph 20.

50     Decision of 27 February 2013, case M.6663 – Ryanair/Aer Lingus III, paragraph 409.

51     See Reply of the Parties to RFI 3 of 23 January 2019, question 5.

52     On routes from Europe (excluding the UK) to North America, because of the TATL Joint Venture, Virgin Atlantic’s market share equate to the increment brought about by the Transaction. On all other routes, the increment brought about by the Transaction corresponds to AFKL’s market share given that pre-Transaction, VAL is co-controlled by Virgin Group and Delta.

53     In the absence of data on the total size of the market for supply of airline seats to tour operators, the Commission considers that AFKL (or Virgin when relevant) can be considered as only marginally active on that market when the number of seats sold to tour operators is on average below 5 per day or when the number of seats sold to tour operators represent less than 5% of the number of seats sold by Virgin Atlantic/Delta (or AFKL/Delta when relevant).

54     Should these markets considered as relevant, the Commission notes that the Transaction would not raise serious doubts as to its compatibility with the internal market with respect to the supply of airline seats to tour operators, under any plausible market definition.

55     See Form CO, annex B.01, Wholesale supply of seats data of each of AFKL, Virgin Atlantic, and Delta; Reply of the Parties to RFI 3 of 23 January 2019.

56     See Reply of the Parties to RFI 3 of 23 January 2019, questions 1 and 4.

57     See Reply of the Parties to RFI 3 of 23 January 2019, question 1. AFKL has a joint venture agreement on that route with China Eastern which covers Amsterdam – Shanghai and Paris – Shanghai but not London-Shanghai. China Eastern can therefore also be considered as a competitor of the Parties operating a direct flight London – Shanghai.

58     The Parties submit that a large part of their wholesale bookings are attributable to Chinese tourists travelling in organised groups (for instance, [90-100]% of AFKL’s wholesale activities on that route were point-of-sale China). In 2016, over 260 000 Chinese tourists visited the UK, of which [40-50]% were part of a tour group. The number of seats sold by the Parties to tour operators represents [40-50]% of the number of Chinese tourists who visited the UK as part of a tour group (Form CO, paragraph 212 and Reply of the Parties to RFI 3 of 23 January 2019, question 2).

59     See Section 5.1.4.2 below.

60     See e.g. Cases M.7541 – IAG/Aer Lingus, paragraphs 188 et seq.; M.6447 – IAG/bmi, paragraphs 174 and 663.

61     Form CO, paragraphs 237-239.

62     See e.g. Cases M.7333 – Alitalia/Etihad, paragraphs 70 et seq.; M.7270 – Cesky Aeroholding/Travel Service/Ceske Aerolinie, paragraphs 20 et seq.; M.6663 – Ryanair/Aer Lingus III, recital 382;  M.6607 – US Airways/American Airlines, paragraph 8; M.6447 – IAG/bmi, paragraph 36.

63     See e.g. Cases M.8361 – Qatar Airways/Alisarda/Meridiana, paragraphs 20-23; M.7541 – IAG/Aer Lingus, paragraphs 28-29; M.7333 – Alitalia/Etihad, paragraphs 70-74; M.6828 – Delta Air Lines/Virgin Group/Virgin Atlantic Limited, paragraphs 23-28.

64     Replies to eQ1 – Questionnaire 1 for Competitors, questions 8 and 8.1; eQ2 – Questionnaire 2 to corporate customers and travel agencies, questions 3 and 3.1.

65     Replies to eQ1 – Questionnaire 1 for Competitors, questions 8 and 8.1; eQ2 – Questionnaire 2 to corporate customers and travel agencies, questions 3 and 3.1.

66     "Non-stop" flights are flights that take off at airport A and land at airport B where they load off passengers without any stops in between. By contrast, "direct" flights may entail a refuelling stop and/or a disembarking/re-embarking stop, but are marketed under a single flight code and are flown with a single aircraft. "One-stop" flights include direct flights that do not qualify as "non-stop", as well as indirect flights which are journeys that require a change of aircraft or a change of flight code.

67     Form CO, paragraph 245.

68     Cases M. 8361 – Qatar Airways/Alisarda/Meridiana, paragraph 24; M.7541 – IAG/Aer Lingus, paragraph 30; M.7333 –Alitalia/Etihad, paragraph  75;  M.6663  –  Ryanair/Aer  Lingus  III,  paragraph 373.

69     Cases M.7541 – IAG/Aer Lingus, paragraph 31; M.7333 –Alitalia/Etihad, paragraph 76; M.6447 – IAG/bmi, paragraph 68; M.5440 – Lufthansa/Austrian Airlines, paragraph 24; M.5335 – Lufthansa/SN Airholding, paragraph 36.

70     Form CO, paragraph 246.

71     Cases M.8361 – Qatar Airways/Alisarda/Meridiana, paragraph 26; M.7541 – IAG/Aer Lingus, paragraph 33; M.6828 – Delta Air Lines/Virgin Group/Virgin Atlantic Limited, paragraph 29;  M.6447 -   IAG/bmi, paragraph 69.

72     Replies to eQ1 – Questionnaire 1 for Competitors, questions 9 and 9.1.

73     Replies to eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 5 and 6.

74     Replies to eQ2 – Questionnaire 2 to corporate customers and travel agencies, questions 7 and 7.1.

75     Form CO, paragraphs 17 and 18.

76     Form CO, paragraph 251.

77     Form CO, paragraph 251.

78     See case M.6663 – Ryanair/Aer Lingus III, paragraph 388.

79     See case M.6663 – Ryanair/Aer Lingus III, paragraph 389.

80     Cases M.6663 – Ryanair/Aer Lingus III, paragraph 418; M.4439 – Ryanair/Aer Lingus, paragraph 311.

81     Cases M.6663 – Ryanair/Aer Lingus III, paragraph 419; M.4439 – Ryanair/Aer Lingus, paragraph 311; M.6828 – Delta Air Lines/Virgin Group/Virgin Atlantic Limited, paragraph 71.

82     See Form CO, paragraph 214.

83     See e.g. Cases M.8361 – Qatar Airways/Alisarda/Meridiana, paragraph 29; M.7333 –Alitalia/Etihad, paragraph 82; M.6663 – Ryanair/Aer Lingus III, paragraph 65.

84     The 100 km/1 hour driving time is nevertheless used as a first proxy only. It was defined by the Commission in the specific case of routes served out of Dublin by two airlines with typical attributes  of low-frills point-to-point carriers. This "rule" is thus not necessarily strictly applicable for other cases (see Case M.6663 – Ryanair/Aer Lingus III, paragraph 82). In this regard, In "Airport Competition: Myth or Reality? IATA Economics Briefing", November 2017, it is noted that "[w]hile isochrones are a simple and powerful visual tool, they are of limited use in understanding the choices that passengers actually make. (…) Moreover, the proximity of an alternative airport can only represent a relevant choice if the airlines which compete with each other offer a substitutable service, for instance a comparable itinerary. Isochrone maps do not reflect the availability of services at comparator airports and are therefore likely to overstate the extent of effective airport competition."

85     Shanghai, Lyon, Düsseldorf, Manchester and San Francisco are cities served by several airports. Nevertheless, the question of airport substitutability does not arise given that the Parties operate the affected routes to/from the same airport (namely Shanghai Pudong International, Lyon, Düsseldorf, Manchester and San Francisco airports) and no competitor operates a direct service to/from the potential substitutable airport(s). Should the Commission take account of potential substitutable airports, the outcome of the competitive assessment would remain unchanged.

86     Case M.5830 – Olympic/Aegean Airlines, paragraph 1676: the Commission concluded that CDG and ORY are substitutable for TS and NTS passengers on the Athens – Paris route.

87     Form CO, paragraphs 702-704.

88     Form CO, paragraphs 731-732.

89     Replies to eQ1 – Questionnaire 1 for Competitors, question 15; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 9.

90     Replies to eQ1 – Questionnaire 1 for Competitors, question 16.

91     Cases    M.7333    –     Alitalia/Etihad,    paragraphs    102-105;    M.4439    – Ryanair/Aer          Lingus, paragraphs 263-267; Case M.6663 – Ryanair/Aer Lingus III, paragraphs 245-252.

92     Form CO, paragraphs 978-979 and 1034-1035.

93     Replies to eQ1 – Questionnaire 1 for Competitors, question 15; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 9.

94     Replies to eQ1 – Questionnaire 1 for Competitors, question 16.

95     Given that no direct services on any of the relevant routes are operated to/from STN, LTN and SEN and that a negligible number of passengers booked indirect flights provided by the Parties’ competitors to/from these airports, these airports are not considered as plausible substitutable airports for the purpose of the Decision (Form CO, paragraphs 258, 261 and 1140).

96     M.6828 – Delta Air Lines/Virgin Group/Virgin Atlantic Limited, paragraph 43.

97     Form CO, paragraphs 256 and 265.

98     Form CO, paragraphs 441-442.

99     Form CO, paragraphs 496-498.

100  Form CO, paragraphs 521-522.

101   Form CO, paragraphs 545-546.

102   Form CO, paragraphs 596-597.

103   Form CO, paragraphs 1064-1066.

104   Form CO, paragraphs 1096-1098.

105   Form CO, paragraphs 1127-1129.

106   Form CO, paragraphs 1151-1152.

107   Replies to eQ1 – Questionnaire 1 for Competitors, question 15; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 9.

108   Replies to eQ1 – Questionnaire 1 for Competitors, question 16.

109   See e.g. Form CO, paragraph 488.

110   Cases M.7333 – Alitalia/Etihad, paragraph 110; M.6828 – Delta Air Lines/Virgin Group/Virgin Atlantic Limited, paragraph 47. In case M.7541 – IAG/Aer Lingus, paragraph 104, the Commission considered that JFK and EWR were substitutable for TS and NTS passengers with respect to the Dublin – New York route.

111   Form CO, paragraphs 471-473.

112   Form CO, paragraphs 568-570.

113   Form CO, paragraphs 756-758.

114   Form CO, paragraphs 789-791.

115   Form CO, paragraphs 899-900.

116   Form CO, paragraphs 922-923.

117   Form CO, paragraphs 959-960.

118   Form CO, paragraph 270.

119  Replies to eQ1 – Questionnaire 1 for Competitors, questions 15 and 16.

120   Replies to eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 9.

121   M.6828 – Delta Air Lines/Virgin Group/Virgin Atlantic Limited, paragraph 52.

122   Form CO, paragraphs 278 and 845.

123   Form CO, paragraphs 862-863.

124   Replies to eQ1 – Questionnaire 1 for Competitors, questions 15 and 16. The question of whether air carriers monitors competitors flying between Dublin and Sanford airports is irrelevant as no air carrier provides direct services on that airport pair.

125   Replies to eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 9.

126   See Cases M.8869 – Ryanair/LaudaMotion, paragraph 222; M.8672 – easyJet/Certain Air Berlin assets, paragraph 52; M.8633 – Lufthansa/Certain Air Berlin assets, paragraph 58; M.6447 - IAG/bmi, paragraphs 492-506.

127   See Cases M.8869 – Ryanair/LaudaMotion, paragraphs 223-226 ; M.8672 – easyJet/Certain Air Berlin assets, paragraphs 53 et seq.; M.8633 – Lufthansa/Certain Air Berlin assets, paragraphs 59 et seq.

128   See in particular paragraphs (72) and (89) above.

129   Press release of 1 December 2011 accompanying the Airport Package (http://europa.eu/rapid/press- release  MEMO-11-857 en.htm).

130   See e.g. Cases M.7270 - Český Aeroholding/Travel Service/České aerolinie, paragraph 50; M.7008   –Aena International/Axa PE/LLAGL, paragraph 12.

131   See e.g. Case M.7398 - MIRAEL/ Ferrovial/NDH1, paragraph 19.

132   See e.g. Cases M.5652 – GIP/Gatwick Airport, paragraph 14; M.4164 –   Ferrovial/Quebec/GIC/BAA,

paragraphs 15-17; M.3823 – MAG/Ferrovial Aeropuertos/Exeter Airport, paragraphs 16-19.

133   See Cases M.8869 – Ryanair/LaudaMotion, paragraphs 238 et seq., M.8672 – easyJet/Certain Air Berlin assets, paragraphs 73 et seq.; M.8633 – Lufthansa/Certain Air Berlin assets, paragraphs 117     et seq.

134   Form CO, paragraph 300. In addition, none of the Parties holds slots at the other plausible airports. Taking account of these airports would therefore dilute the Parties’ slot holdings.

135   Heathrow airport’s “Facts and figures”: https://www.heathrow.com/company/company-news-and- information/company-information/facts-and-figures .

136   Gatwick Key Facts https://www.gatwickairport.com/business-community/about-gatwick/company- information/gatwick-by-numbers/ .

137   London City Airport’s website https://www.londoncityairport.com/media-centre/passengers-up-by-6- 4-in-2018-to-4-8m .

138   Heathrow airport’s “Facts and figures”: https://www.heathrow.com/company/company-news-and- information/company-information/facts-and-figures                       and “History” https://www heathrow.com/company/company-news-and-information/company-information/our- history .

139   Gatwick Key Facts https://www.gatwickairport.com/business-community/about-gatwick/company- information/gatwick-key-facts/ .

140   2018           LCY       Customer             Profile: https://assets.ctfassets net/ggj4kbqgcch2/7KiRJ9oud2wEsKyAcSie60/66c269783be6dcbc5d604a512af 57364/2018  LCY  Customer  Profile.pdf .

141   For the sake of completeness, the Parties have provided their slot holding for all plausible airport combination and submit that the competitive assessment concerning their slot holding at LHR is valid under any plausible market definition (Form CO, paragraph 300).

142    […].

143   See e.g. Case M.6663 – Ryanair/Aer Lingus III, paragraph 238.

144   Under the Slot Regulation, a schedules facilitated airport is "an airport where there is potential for congestion at some periods of the day, week or year which is amenable to resolution by voluntary cooperation between air carriers and where a schedules facilitator has been appointed to facilitate the operations of air carriers operating services or intending to operate services at that airport."

145   Under the 8th Edition of Worldwide Slot Guidelines issued by IATA (effective 1 January 2017), a Level 2 airport "is one where there is potential for congestion during some periods of the day, week or season, which can be resolved by schedule adjustments mutually agreed between the airlines and facilitator." A Level 2 airport under the IATA Worldwide Slot Guidelines corresponds to a schedules facilitated airport under the Slot Regulation.

146   Leeds airport website: https://www.leedsbradfordairport.co.uk/our-destinations/airlines.

147   Leeds airport’s Strategic Development Plan, available https:// www.leedsbradfordairport.co.uk/media/2522/masterplan-2017-update.pdf .

148   https://www.liverpoolairport.com/news/2019/01/double-celebration-for-ljla-with-growing-passenger- numbers-and-the-uk-s-best-on-time-performance .

149   Cases M.8361 – Qatar Airways/Alisarda/Meridiana, paragraph 37; M. 6828 – Delta Airlines/ Virgin Group / Virgin Atlantic Limited, paragraph 76; M.6447 – IAG/bmi, paragraph 92; M.5747 - Iberia/British Airways, paragraph 40.

150   M.  6828  –  Delta  Airlines/  Virgin  Group  /  Virgin  Atlantic  Limited,  paragraphs  73-74;  M.6447 –

IAG/bmi, paragraphs 91-92; M.5747 – Iberia/British Airways, paragraph 40.

151   M. 6828 – Delta Airlines/ Virgin Group / Virgin Atlantic Limited, paragraph 75; M.6447 – IAG/bmi, paragraph 89; M.5747 – Iberia/British Airways, paragraph 38.

152   Cases M.8361 – Qatar Airways/Alisarda/Meridiana, paragraph 38; M.6828 – Delta Airlines/ Virgin Group / Virgin Atlantic Limited, paragraph 72; M.6447 – IAG/bmi, paragraph 87; M.5747 – Iberia/British Airways, paragraph 36.

153   Cases M.6828 – Delta Airlines/ Virgin Group / Virgin Atlantic Limited, paragraph 73; M.6447 – IAG/bmi, paragraph 88; M.5747 – Iberia/British Airways, paragraph 37.

154   Cases M.8361 – Qatar Airways/Alisarda/Meridiana, paragraph 39; M.6828 – Delta Airlines/ Virgin Group / Virgin Atlantic Limited, paragraph 77; M.6447 – IAG/bmi, paragraph 90; M.5747 – Iberia/British Airways, paragraph 39.

155   Form CO, paragraphs 1418 and 1420.

156   Cases M.8361 – Qatar Airways/Alisarda/Meridiana, paragraph 39; M.6447 – IAG/bmi, paragraph 93; M.5747 – Iberia/British Airways, paragraph 41.

157   Cases M.8361 – Qatar Airways/Alisarda/Meridiana, paragraph 39; M.6828 – Delta Airlines/ Virgin Group / Virgin Atlantic Limited, paragraph 79; M.6447 – IAG/bmi, paragraph 94 ; M.5747 – Iberia/British Airways, paragraph 42.

158   Cases M.8425 – Safran/Zodiac Aerospace, paragraph 289; M.6447 – IAG/bmi, paragraph 105; M.6554 –  EADS/STA/Elbe Flugzeugwerke JV, paragraph 25; M.6410 – UTC/Goodrich, paragraph 174; M.5747 –  Iberia/British Airways, paragraph 48; M.3374 – SR Technics/FLS Aerospace, paragraph 9; M.3280 – Air France/KLM, paragraph 39.

159   Cases M.8425 – Safran/Zodiac Aerospace, paragraph 289.

160   Cases M.8425 – Safran/Zodiac Aerospace, paragraph 289; M.6554 – EADS/STA/Elbe Flugzeugwerke JV, paragraph 25; M.6410 – UTC/Goodrich, paragraph 174; M.3280 – Air France/KLM, paragraph 39. An A-check is performed approximately every 800 flight hours and requires around 200-300 man- hours to complete. Bchecks are performed approximately every 4-6 months and are usually performed within 3 days at an airport hangar. C-checks are performed approximately every 18 to 24 months or after a specific amount of actual Flight Hours as defined by the manufacturer. D-checks/structural checks are the most comprehensive and demanding checks, since the entire aircraft structure is taken apart for inspection and overhaul. Intermediate structural checks occur after 5-6 years and heavy structural checks occur after 10- 12 years. Such checks will usually demand around 15,000 to 20,000 man-hours and around 1 month to complete at suitably equipped maintenance bases.

161   Cases M.7545 – IAG/Aer Lingus, paragraph 121; M.6554 – EADS/STA/Elbe Flugzeugwerke JV, paragraph 26; M.6447 – IAG/bmi, paragraph 108;  M.3374  –  SR  Technics/FLS  Aerospace,  paragraph 12.

162   Cases M.7545 – IAG/Aer Lingus, paragraph 121; M.6554 – EADS/STA/Elbe Flugzeugwerke JV, paragraph 26; M.3374 – SR Technics/FLS Aerospace, paragraph 12; M.3280 – Air France/KLM, paragraph 40.

163   Cases M.7545 – IAG/Aer Lingus, paragraph 121; M.6554 – EADS/STA/Elbe Flugzeugwerke JV, paragraph 26; M.6447 – IAG/bmi, paragraph 108.

164   Form CO, paragraph 11.

165   See e.g. Case M.7541 – IAG/Aer Lingus, paragraph 145.

166   For the purpose of this Decision, it is understood that routes to/from the United Kingdom are not covered by the TATL Joint Venture.

167   Form CO, paragraph 402.

168   Form CO, paragraph 403.

169   Form CO, paragraphs 13 and 402.

170   Form CO, paragraph 13.

171   See e.g. Case M. 7541 – IAG/Aer Lingus, paragraph 146.

172   See e.g. Cases M. 7541 – IAG/Aer Lingus, paragraph 147, M.6828 – Delta Air Lines/Virgin Group/Virgin Atlantic Limited, paragraph 87.

173   See  e.g.  Cases  M.6447  –  IAG/bmi;  M.5889  –  United  Airlines/Continental  Airlines;  M.5747     –Iberia/British Airways.

174   See e.g. Cases M.7333 – Alitalia/Etihad, paragraphs 135 et seq.; M.7541 – IAG/Aer Lingus,  paragraphs 135 et seq.

175   Form CO, paragraph 410 et seq.

176   Form CO, paragraph 417.

177   Form CO, paragraph 418.

178   See email of Parties to EC case team of 29 November 2018.

179   Form CO, paragraph 420.

180   Form CO, paragraph 423.

181   See e.g. Cases M.7541 – IAG/Aer Lingus, paragraphs 151 et seq.; M.7333 – Alitalia/Etihad,  paragraphs 171 et seq.; M.6828 – Delta Air Lines/Virgin Group/Virgin Atlantic Limited, footnote 76; M.6607 – US Airways/American Airlines, paragraph 32.

182   The Parties have applied the filters to the MIDT data and filtered out around 700 routes. Considering that the Commission used MIDT data as an appropriate data source in several previous cases and considering the Parties submission that a comparison of the MIDT and the DDS shows that the DDS data is often similar to the MIDT data in terms of the Parties’ combined market shares and the market increment, the Commission has accepted the routes filtered out based on MIDT data.

183   The Parties have treated the parties to each of the TATL Joint venture and the DL-VS Joint Venture as a single entity: for the routes between the UK and North America, Delta and Virgin Atlantic were treated as a single entity and for routes between Europe (excluding the UK) and North America, Delta and AFKL were treated as a single entity. For the other O&Ds, each of AFKL, DL and VAL’s market share was analysed separately (see Form CO, paragraph 16 and footnote 18). Therefore, for the  purpose of applying this filter (“one of the Parties had a market share below 2%”), the market share of either each of the Parties, or of the TATL Joint Venture, or the DL-VS Joint Venture has been taken into account, depending on the route (see Annex B.06).

184   Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, OJ C31, 5.2.2004, p.5.

185   The Parties have treated the parties to each of the TATL Joint venture and the DL-VS Joint Venture as a single entity: for the routes between the UK and North America, Delta and Virgin Atlantic were treated as a single entity and for routes between Europe (excluding the UK) and North America, Delta and AFKL were treated as a single entity. For the other O&Ds, the Parties’ market share was analysed separately, see Form CO, paragraph 16 and footnote 18.

186   See e.g. Cases M.7541 - IAG/Aer Lingus, paragraphs155 et seq.; M.6828 - Delta Air Lines/Virgin Group/Virgin Atlantic Limited, paragraphs 168 et seq.

187   See paragraphs 28-30 of the Horizontal Merger Guidelines.

188   Replies to eQ1 – Questionnaire 1 to competitors, question 9; eQ2 – Questionnaire to corporate customers and travel agencies, question 4.

189   Replies to eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 5 and 6.

190   Replies to eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 6.

191   Replies to eQ1 – Questionnaire 1 to competitors, question 3; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 7.

192   Replies to eQ1 – Questionnaire 1 for Competitors, question 17; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 11.

193   See case M.7333 – Alitalia/Etihad, paragraph 279.

194   See e.g. case M.7541 – IAG/Aer Lingus, paragraphs 412 et seq.

195   Seattle-Tacoma Airport (SEA).

196   Form CO, paragraphs 449 and 460.

197   The Commission has applied the filters described in section 5.1.1.3 above on the basis of the MIDT data that were provided for the four last IATA Seasons rather than DDS data available only for the  past two IATA Seasons. Therefore, the London – Seattle route is considered as “affected” although it might have been considered as “unproblematic”, should AFKL’s market shares have been below 2% in the past four IATA Seasons.

198   Form CO, paragraph 441.

199   Form CO, paragraph 442.

200   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

201   Manchester Airport (MAN).

202   Form CO, paragraphs 480 and 486.

203   The Parties note that they have a small number of indirect bookings on the MAN-EWR airport pair in both Summer 2017 and Winter 2017/2018 IATA Seasons (Form CO, footnote 26). The effect of the Transaction on the MAN-EWR airport (indirect/indirect overlap) pair will not be further assessed as  the Transaction is unlikely to give rise to competition concerns given that the Parties’ combined market shares will be below [5-10]% under any plausible market definition (Form CO, paragraph 488).

204   Market shares excluding seats sold by Thomas Cook/Condor do not significantly differ (Form CO, paragraph 480).

205   Market shares excluding seats sold by Thomas Cook/Condor do not significantly differ (Form CO, paragraph 486).

206   Form CO, paragraph 479.

207   Form CO, paragraphs 472-473.

208   Form CO, paragraph 471.

209   Form CO, paragraph 477.

210   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

211   Hartsfield-Jackson Atlanta Airport (ATL).

212   Form CO, paragraphs 503 and 509.

213   Form CO, paragraph 496.

214   Form CO, paragraph 497.

215   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

216   Minneapolis-Saint Paul Airport (MSP).

217   Form CO, paragraphs 528 and 534.

218   Market shares excluding seats sold by Thomas Cook/Condor do not significantly differ (Form CO, paragraph 534), in particular, the analysis is valid irrespective of whether these seats are included or not.

219   Market shares excluding seats sold by Thomas Cook/Condor do not significantly differ (Form CO, paragraph 528), in particular, the analysis is valid irrespective of whether these seats are included or not.

220   When excluding Thomas Cook/Condor’s seats, AFKL’s market shares are slightly higher but below [5-10]% in any market definition.

221   Form CO, paragraph 521.

222   Detroit Metropolitan Wayne County Airport (DTW).

223   Form CO, paragraphs 552 and 557 and Form CO, Annex B.04, DDS Data.

224   For the sake of completeness, the Parties note that they do not overlap on LGW-DTW and that they have a small number of indirect bookings on the LCY-DTW, the latter being a thin route with less than 30 000 passengers per year (Form CO, paragraph 559).

225   Form CO, paragraphs 552 and 557. The difference is negligible as the market shares on LHR-DTW   are 0.1% higher than on LHR/LGW/LCY-DTW.

226   Form CO, paragraph 546.

227   Form CO, paragraph 545.

228   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

229   Edinburgh airport (EDI).

230   Form CO, paragraphs 577 and 583.

231   Form CO, paragraphs 568 et seq.

232   The effects of the Transaction on the EDI-EWR airport pair (indirect/indirect overlap) will not be further assessed as the Transaction is unlikely to give rise to competition concerns given that the Parties’ combined market shares will be below [0-5]% under any plausible market definition.

233   Market shares excluding seats sold by Thomas Cook/Condor do not significantly differ (Form CO, paragraph 577), in particular, the analysis is valid irrespective of whether these seats are included or not.

234   Market shares excluding seats sold by Thomas Cook/Condor do not significantly differ (Form CO, paragraph 583), in particular, the analysis is valid irrespective of whether these seats are included or not.

235   Form CO, paragraphs 568 et seq.

236   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

237   Portland International Airport (PDX).

238   Form CO, paragraphs 503 and 509.

239   Market shares excluding seats sold by Thomas Cook/Condor do not significantly differ (see Form CO, paragraph 614), in particular, the analysis is valid irrespective of whether these seats are included or not.

240   Market shares excluding seats sold by Thomas Cook/Condor do not significantly differ (see Form CO, paragraph 608), in particular, the analysis is valid irrespective of whether these seats are included or not.

241   Form CO, paragraph 597.

242   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

243   San Francisco International Airport (SFO).

244   Form CO, paragraph 636.

245   See section 4.2.1.4 above and Form CO, paragraph 626.

246   Form CO, paragraphs 626 et seq.

247   Market shares excluding seats sold by Thomas Cook/Condor do not significantly differ (see Form CO, paragraph 614), in particular, the analysis is valid irrespective of whether these seats are included or not.

248   Form CO, paragraph 633.

249   Form CO, paragraph 626. There were no regular direct services on this route in Winter 2017/2018.

250   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

251   Amsterdam Airport Schipol (AMS) and Los Angeles International Airport (LAX).

252   Form CO, paragraph 668.

253   See section 4.2.1.4 above and Form CO, paragraph 662.

254   Form CO, paragraphs 662 et seq.

255   Form CO, paragraphs 663 and 668.

256   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

257   Miami International Airport (MIA).

258   Form CO, paragraph 682.

259   See section 4.2.1.4 above and Form CO, paragraph 674.

260   Form CO, paragraphs 674 et seq.

261   Market shares excluding seats sold by TUI do not significantly differ (see Form CO, paragraph 614), in particular, the analysis is valid irrespective of whether these seats are included or not.

262   Form CO, paragraphs 674 et seq. and 682.

263   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

264   Form CO, paragraph 694.

265   See section 4.2.1.4 above and Form CO, paragraph 688.

266   Form CO, paragraphs 688 et seq.

267   Form CO, paragraphs 689 and 694.

268   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

269   Form CO, paragraphs 712 and 720.

270   The Parties’ activities do not overlap on ORY-SFO. Indeed, Virgin had a negligible number of booking in Summer 2017 IATA Season but discontinued this indirect service from Winter 2017/2018 IATA Season. The impact of the Transaction on this airport pair will therefore not be further assessed.

271   Form CO, paragraph 702.

272   Form CO, paragraph 702.

273   Form CO, paragraph 703.

274  Form CO, paragraph 704.

275   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

276   Boston Logan International Airport (BOS).

277   Form CO, paragraphs 741 and 747.

278   Form Co, paragraph 731.

279   Form Co, paragraph 732.

280   Form Co, paragraph 733.

281   Form Co, paragraph 733.

282   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

283   Reply to eQ1 - Questionnaire 1 for Competitors, question 19.

284   Nice Côte d’Azur airport (NCE).

285   Form CO, paragraphs 774 and 780.

286   Form Co, paragraph 756 et seq.

287   Form Co, paragraph 758.

288   Form CO, paragraph 774.

289   Form CO, paragraph 780.

290   La Compagnie’s press release, available https://www.lacompagnie.com/media/1627/la-co-nice- announcement-  -final.pdf .

291   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

292   Dublin airport (DUB).

293   Form CO, paragraphs 798 and 804.

294   Form Co, paragraph 789.

295   Form Co, paragraph 791.

296   The effects of the Transaction on the DUB-EWR airport pair (indirect/indirect overlap) will not be further assessed as the Transaction is unlikely to give rise to competition concerns given that the Parties’ combined market shares will be below  [0-5]%  under  any  plausible  market  definition  (Form CO, paragraph 807).

297   Form CO, paragraph 789.

298   Form CO, paragraph 790.

299   Form CO, paragraph 791.

300   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

301   Murtala Muhammed International Airport (LOS).

302   Form CO, paragraphs 1074 and 1080.

303   On LHR/LCY/LGW-LOS, in the NTS segment in Summer IATA Season, the share of oneworld  would be equivalent to the increment brought about by the Transaction (Form CO, paragraphs 1074 and 1080.

304   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

305   Shanghai Pudong Airport (PVG).

306   Form CO, paragraphs 1106 and 1113.

307   See Form CO, paragraphs 1096 et seq.

308   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

309   José Martí International Airport (HAV).

310   Form CO, paragraph 1135.

311  Form Co, paragraph 1140.

312   For the sake of completeness, it can be noted that the Parties’ market shares on a narrower market LHR/LGW-HAV are similar and therefore, the competition assessment will remain unchanged.

313   Market shares excluding seats sold by e.g. Thomas Cook/Condor or AeroMexico do not significantly differ (Form CO, paragraph 1131).

314   Form CO, paragraph 1132.

315   Form CO, paragraph 1127.

316   Form CO, paragraph 1127.

317   Form CO, paragraph 1132.

318   Form CO, paragraph 1132.

319   Form CO, paragraph 1132.

320   Form CO, paragraph 1134.

321   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

322   Sangster International Airport (MBJ).

323   Form CO, paragraph 1135.

324   Form CO, paragraph 1152.

325   Form CO, paragraph 1162.

326   Form CO, paragraph 1151.

327   Form CO, paragraph 1158.

328   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

329   Los Angeles – Manchester, Dublin – Las Vegas, Dublin – Orlando, Amsterdam – Orlando,  Barcelona –  Los Angeles, Lyon – New York, Marseille – New York, Madrid – San Francisco, Hamburg – New York, Los Angeles – Milan, Düsseldorf – Los Angeles, Hamburg – Miami and Milan – San Francisco.

330   For the Amsterdam-Orlando route and the Manchester-Los Angeles route, the data submitted by the Parties includes seats sold by Thomas Cook/Condor. Market shares excluding these seats do not significantly differ (see Form CO, paragraphs 651 and 877), in particular, the analysis is valid irrespective of whether these seats are included or not.

331   For the Winter 2017/2018 IATA Season the NTS combined market share is [40-50]% on the plausible airport-pair LIN – SFO and [30-40]% on the plausible airport-pair LIN+MXP – SFO and therefore higher than on the plausible airport pair MXP – SFO. However, the market shares are still below 60%, which indicates that there is already prima facie sufficient competition from other carriers.

332   For the Winter 2017/2018 IATA Season the overall combined market share is [30-40]% on the plausible airport-pair LIN – SFO and [30-40]% on the plausible airport-pair LIN+MXP – SFO and therefore higher than on the plausible airport pair MXP – SFO. However, the market shares are still below 60%, which indicates that there is already prima facie sufficient competition from other carriers.

333   Replies to eQ1 – Questionnaire 1 for Competitors, question 26; eQ2 – Questionnaire 2 to corporate customers and travel agencies, question 13.

334   Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, OJ C 31, 5 February 2017, p. 5.

335   See e.g. Cases M.8869 – Ryanair/LaudaMotion, paragraph 508; M.8672 – easyJet/Certain Air Berlin Assets, paragraph 105; M.8633 – Lufthansa/Certain Air Berlin Assets, paragraph 182.

336   Without prejudice to the exceptional cases of joint ventures between an airport manager and an airline (see e.g. the joint venture between Lufthansa and Flughafen München GmbH, the company managing Munich airport).

337   As explained in paragraph (406) above, given that the three conditions are cumulative, there is no need to assess the third condition given that the first two conditions are not fulfilled.

338   The Parties have submitted slot holding data that captures all the slots which the parties used for their own operations in Winter 2017/2018 and Summer 2018. This includes remedy slots any party benefits from, as well as slots leased to or transferred to any of the Parties, as applicable. The data does not include slots leased to, transferred to, swapped out to or exchanged with another carrier (Reply of the Parties  to  RFI  4  of  28  January  2019).  Nevertheless,  as  described  in  footnote  339  below, the Commission has crosschecked the data of the Parties against the data of the slot coordinator and did not note any material difference in the slot holding of the Parties.

339   Form CO, paragraph 296. The Commission has cross-checked the data on slot holding provided by the Parties against Airport Coordination Limited Seasonal Report information, which contains a breakdown of the number of slots allocated to an airline at a given airport. These reports are available online https://www.acl-uk.org/airport-info/. See for instance Heathrow airport, Winter 2018/19 Start  of Season Report available on ACL’s website https://www.acl-uk.org/wp- content/uploads/2018/09/LHR-W18-Start-Of-Season-Report.pdf .

340   The runway capacity of LHR and MAN is available online https://www.acl-uk.org/airport-info- details/?aid=1, section “Capacity Declaration”.

341   London Heathrow airport is open 24 hours with some restrictions on night flights between 23:30 and 6:00. Similarly, Manchester airports operates on a 24-hour basis with restrictions at night, in particular between 23:00 and 6:59. Given that the Parties hold some slots at both airports during the restricted hour bands, in order to adopt a conservative approach, the Commission will take those slots into account for the sake of calculating the highest slot holding.

342   Rounding effects.

343   Rounding effects.

344   It can be noted that runway movements are limited during the hour band 4:00-4:59. In particular, landing is prohibited until 4:30 (Form CO, footnote 90). The Parties therefore submit that their highest slot holding is actually [10-20]% during the hour band 9:00-9:59.

345   Replies to eQ1 – Questionnaire 1 for Competitors, question 21.

346  Form CO, paragraph 294.

347   Form CO, footnote 90.

348   "Slot hoarding" involves the operation of small aircraft and/or low load factors in order to keep slots rather than lose them under the "use-it-or-lose-it" principle described in section 4.1.2.1 above.

349   "Slot babysitting" involves the transfer of slots to non-competing airlines, such as partner airlines within an alliance before shifting them to uses that are more profitable.

350   "Slot shuffling power" refers to the ability to move a flight to a timing close to the timing envisaged by a new entrant.

351   Replies to eQ1 – Questionnaire 1 for Competitors, question 22.

352   Heathrow airport “Facts and figures » https://www.heathrow.com/company/company-news-and- information/company-information/facts-and-figures.

353   Route             Development             at             Manchester             airport,             https://live-webadmin- media.s3.amazonaws.com/media/3734/maav-0007-mag-routes-brochure-updated-web.pdf

354   For the avoidance of doubt, the Parties submit that [details about strategic cooperation] (Form CO, Chapter C, footnote 6).

355   The Parties submitted data for AFKL that included AFKL’s cargo-only freighters. Since cargo-only is excluded from the scope of the cooperation, the Parties combined market shares under the TATL Joint Venture, which covers only belly-hold cargo, are inflated.

356   M.5440 – Lufthansa/Austrian Airlines, paragraph 283.

357   M.5440 – Lufthansa/Austrian Airlines, paragraph 284.

358   M.6447 – IAG/bmi, paragraph 555.

359   Form CO, paragraph 1427.

360   Form CO, paragraph 1427.

361   The other datasets taken into account by the Parties are the following: the IATA Cargo Accounts Settlement System (“CASS”), T100 (the U.S. Department of Transport dataset) and Seabury.

362   Form CO, paragraph 1428.

363   Form CO, paragraphs 1402 and 1404.

364   Form CO, paragraphs 1405-1406.

365   In line with the Commission’s established geographic market definition, the Parties have provided data on an EEA-basis. However, in light of the respective geographic focus of the parties’ activities, the Parties separately provided data on a UK-only basis in order to allow the Commission to assess the increment brought about by Virgin Atlantic in a more meaningful way. The Commission accepts to take account of this additional dataset in its assessment since even on this narrower basis, the Transaction does not give rise to any competition concerns.

366   Form CO, paragraph 1401.

367   Form CO, paragraph 1432.

368   Form CO, paragraph 1401, Replies to eQ3 – Questionnaire 3 to customer in  cargo  transport,  questions 6.1, 6.2 and 6.3. The Parties submit that they are also constrained post-Transaction by integrators such as e.g. FedEX, UPS and DHL. However, the Parties have also submitted that they cannot be certain that the WACD data covers integrators such as FedEX and UPS, see Form CO, paragraph 1430. The Commission notes, that neither FedEX nor UPS or DHL are mentioned as competitors for the affected cargo flows in the data submitted by the Parties. However, to the extent that these integrators are operating on the affected cargo flows, the Parties’ combined market shares post-Transaction are overstated. The competitive analysis does not change, irrespective of whether integrators are considered or not.

369   Replies to eQ3 – Questionnaire 3 to customer in cargo transport, questions 6.1, 6.2 and 6.3.

370   Form CO, paragraphs 1449 and 1461.

371   Replies to eQ3 – Questionnaire 3 to customer in cargo transport, question 13.

372   Replies to eQ3 – Questionnaire 3 to customer in cargo transport, questions 6.1, 6.2 and 6.3.

373   Replies to eQ3 – Questionnaire 3 to customer in cargo transport, question 8.

374   Replies to eQ3 – Questionnaire 3 to customer in cargo transport, question 7.

375   Replies to eQ1 – Questionnaire 1 for competitors, questions 25, 28 and 29; eQ3 – Questionnaire 3 to customer in cargo transport, questions 12 and 13.

376   Form CO, paragraph 1610.

377   Form CO, paragraph 1607.