Commission, November 19, 2020, No M.9871

THE EUROPEAN COMMISSION,

Having regard to the Treaty on the Functioning of the European Union (the "TFEU")1,

Having regard to the Agreement on the European Economic Area, and in particular Article 57 thereof,

Having regard to Council Regulation (EC) No. 139/2004 of 20.1.2004 on the control of concentrations between undertakings2 (the "Merger Regulation"), and in particular Article  9(3) thereof, in conjunction with Article 131 of the Agreement on the withdrawal of the United Kingdom of Great Britain and Northern Ireland (“UK”) from the European Union and the European Atomic Energy Community,

Having regard to the notification made by Telefónica, S.A. and Liberty Global PLC on 30 September 2020, pursuant to Article 4 of the Merger Regulation,

Having regard to the request of the Competition & Markets Authority of the UK (the “CMA”) of 8 October 2020 (the “Referral Request”),

Whereas:

(1)     On 30 September 2020, the Commission received notification of a proposed concentration by which Telefónica, S.A. (“Telefónica”) and Liberty Global PLC (“Liberty Global”, together with Telefónica, the “Notifying Parties”) would acquire, within the meaning of Article 3(1)(b) and 3(4) of the Merger Regulation,  joint control of a company constituting a 50/50 joint venture (the “JV”, UK) (the “Transaction”).

(2)     The CMA received a copy of the notification on 30 September 2020.

(3)     By letter dated 8 October 2020, the UK via the CMA requested the referral of the Transaction to its competition authority of the proposed concentration with a view to assessing it under national competition law, pursuant to Article 9(2)(a) of the Merger Regulation.

1. THE PARTIES

(4)     Telefónica is a global telecommunications company that operates fixed and mobile communication networks. It offers mobile, landline, internet and television services under several brands, including Movistar, O2, and Vivo. Telefónica is incorporated under the laws of Spain and is domiciled and registered in Spain.

(5)     In the UK, Telefónica is active through its indirect subsidiary Telefónica UK Limited (“O2”), which is one of the four existing Mobile Network Operators (“MNOs”). It provides voice, SMS and mobile data services to subscribers through its mobile telecommunications network, and provides wholesale network access to Mobile Virtual Network Operators (“MVNOs”). O2 also provides limited fixed services to business customers only.

(6)     The O2 business also includes the following wholly or jointly owned subsidiaries3: i) CTIL: a 50-50 network sharing joint venture between O2 and Vodafone; ii) Tesco Mobile Limited (“Tesco Mobile”): a 50-50 MVNO joint venture with the UK retailer Tesco; iii) Giffgaff Limited (“giffgaff”): a wholly owned MVNO subsidiary  targeting young, tech-savvy, digitally native consumers; and iv) Digital Mobile Spectrum Limited a 4G support joint venture in which each of the four UK MNOs hold 25%.

(7)     Liberty Global is an international video, broadband, and communications company with consolidated operations in the UK and Ireland (under the Virgin Media brand), Belgium (under the Telenet brand), as well as Switzerland, Poland, and Slovakia (under the UPC brand). Liberty Global also owns 50% of the VodafoneZiggo joint venture in the Netherlands.

(8)     In the UK, Liberty Global is active through its Virgin Media Ltd. (“Virgin Media”) and Virgin Mobile Telecoms Ltd. (“Virgin Mobile”) subsidiaries. Virgin Media is a British telecommunications company that provides fixed-line telephony and broadband, as well as pay-TV services, through its own fixed network infrastructure. Virgin Media is also active on the upstream market for the wholesale supply of  leased lines. Virgin Mobile is an MVNO.

2. THE OPERATION AND THE CONCENTRATION

(9)     Both Parties will contribute all of their UK assets to the JV, including O2’s fully or partially owned subsidiaries.4

2.1. Joint control

(10)    Each party to the JV will be allotted 50% of the JV's share capital.5 Decisions on the business conduct of the JV will be vested with the JV’s board of directors, which comprises eight directors.6 Board meetings are quorate where at least three  Telefónica directors and three Liberty Global directors are present or represented by an alternate.7 Simple resolutions of the board of directors will be decided by a simple majority vote of all votes cast by the directors present or represented, provided that at least one Liberty Global director and at least one Telefónica director must vote in favour of the resolution.8 With respect to certain reserved matters (such as the approval  and  amendment of the JV's business plan and any amendment  to the operating budget that forms part of the business plan), the approval of at least three Telefónica directors and three Liberty Global directors will be required. 9

(11)    It follows from these arrangements that each party holds a veto in matters requiring simple majority as well as in strategic matters such as the business plan or the  budget. Therefore, as a result of the Transaction, Telefónica and Liberty Global will jointly control the JV within the meaning of Article 3(1)(b) of the Merger Regulation.

2.2. Full-functionality

(12)    The JV combines already existing client-facing businesses and thus will have its own presence on the market independent from the Notifying Parties. It will have its own management, staff, assets and financial resources. Sale and purchase relationships between the parents and the JV will be restricted to some minimal intra-group services, similar to those already existing within the Liberty Global and Telefónica group entities.10 The JV is established for an indefinite period.11 It  follows that the  JV is full function. Therefore, the Transaction will lead to the creation of a full- function joint venture within the meaning of Article 3(4) of the Merger Regulation.

3. EU DIMENSION

(13)    In the last financial year, Telefónica and Liberty Global achieved a combined aggregate world-wide turnover of more than EUR 5 billion (Telefónica: EUR 48,422 million and Liberty Global: EUR 12,277 million). Each of the undertakings concerned achieved an EU-wide turnover in excess of EUR 250 million (Telefónica: EUR 27,275 million and Liberty Global: EUR […]), but they did not achieve more than two-thirds of their aggregate EU-wide turnover within one and the same Member State. The Transaction therefore has an EU dimension within the meaning  of Article 1(2) of the Merger Regulation.

4. ASSESSMENT UNDER ARTICLE 9(3) OF THE MERGER REGULATION

4.1. Introduction

(14)    By its Referral Request of 8 October 2020, the CMA, on behalf of the UK, requests a referral of the Transaction with a view to assessing the effects of the Transaction in the UK under national competition law, pursuant to Article 9(2)(a) of the Merger Regulation.

(15)    According to Article 9(3) of the Merger Regulation, the Commission may refer the whole or part of the case to the competent authorities of the Member State concerned with a view to applying the Member State’s national competition law if, following a request for referral by that Member State pursuant to Article 9(2) of the Merger Regulation, the Commission considers that the Transaction threatens to significantly affect competition in a market within that Member State, which presents all the characteristics of a distinct market.

(16)    When the criteria laid down in Article 9(2)(a) are met, the Commission will assess whether it is appropriate to refer a given case to a national competition authority. The Commission retains a margin of discretion in deciding whether to refer a case or not.12 In exercising such discretion, the Commission will take into account the need to ensure effective protection of competition in all markets affected by the Transaction.13 The Commission exercises that discretion taking into account the criteria set out in the case law and the Referral Notice.14

(17)    In the following sections, it will be examined whether the criteria of Article 9(2)(a)  of the Merger Regulation are fulfilled (Sections 4.2 - 4.5), and whether it would be appropriate to refer the present case to the UK (Section 4.6).

(18)    In its assessment of the Referral Request, the Commission takes into account the arguments it received from the UK authorities and the Notifying Parties.

4.2. The criteria of Article 9(2)(a) of the Merger Regulation

(19)    In order for a referral request to be issued by a Member State, one procedural and  two substantive conditions must be fulfilled pursuant to Article 9(2)(a) of the Merger Regulation.

(20)    As to the procedural condition, the referral request must be made within 15 working days from the date on which the notification of a concentration to the Commission is received by that Member State. In this regard, the Commission notes that the UK, via the CMA, received a copy of the notification of the Transaction on 30 September 2020. The Referral Request was submitted to the Commission on 8 October 2020,

i.e. within 15 working days following the receipt by the UK authorities of the notification of the Transaction. Consequently, the deadline provided for in Article 9(2) of the Merger Regulation is met.

(21)    As to the substantive conditions, first, in assessing a referral request made pursuant  to Article 9(2)(a) of the Merger Regulation, the Commission is required to determine whether there is a market within the Member State concerned which is affected by the notified concentration and presents all the characteristics of a distinct market. According to Article 9(3) of the Merger Regulation and the case law of the General Court,15 the Commission has to evaluate this on the basis of a definition of the  market for the relevant product or services and a definition of the geographical reference market. Second, the Commission is required to verify whether the Transaction threatens to significantly affect competition in that market. Article 9(7)  of the Merger Regulation gives further indications as to which area the geographically relevant market shall consist of, and which elements the Commission must take particular account of when assessing that issue. These conditions are assessed in turn in the following sections.

4.3. Markets within the UK which present all the characteristics of distinct markets

4.3.1.The CMA’s submission

(22)    The CMA submits that the requirements for referral under Article 9(2)(a) EUMR are met.

(23)    The CMA considers that, without prejudice to a full investigation, the Transaction threatens to significantly affect competition in the following markets within the UK presenting all the characteristics of distinct markets:

–                   Retail mobile telecommunications services;

–                   Retail fixed internet, voice and TV services;

–                   Retail business connectivity and other retail business services;

–                   Wholesale access and call origination services on public mobile networks; and

–                   Wholesale leased lines.

(24)    With regard to these markets, the CMA considers that the geographic scope can be defined at the national level, or narrower, in line with the Commission’s practice.

4.3.2. The Notifying Parties’ view

(25)    In the Form CO, the Notifying Parties submit that the relevant markets (see  paragraph (27) below) for the purpose of assessing the Transaction are national in scope, i.e. the UK.16

(26)    In their response17 to the CMA’s Referral Request, the Notifying Parties do not touch again on the various geographic market definitions of the relevant markets. Instead, they only discuss the fact that the CMA believes that the affected markets potentially have regional characteristics.18 The Notifying Parties focus on the leased lines  market and claim that the CMA’s argument is not a compelling reason to refer. In  any event, the Commission has consistently defined the market for leased lines as national.19

4.3.3.Commission’s assessment

(27)    The Transaction concerns the following markets:

–                   The retail mobile telecommunications market, where O2 is currently active as an MNO and Virgin Media is active as an MVNO;

–                   The retail fixed internet market,20,21 where O2 currently provides limited services to business customers only, and Virgin Media is active;

–                   The market for wholesale access and call origination on public mobile networks, where O2 is active on the supply side and Virgin Media is active on the demand side;

–                   The retail market for multiple play services, where Virgin Media is active on the supply side with a fixed-mobile bundle;

–                   The wholesale leased lines market, where Virgin Media is active on the supply side and O2 is active on the demand side.

4.3.3.1. Retail mobile telecommunications services

(a) Product market definition

(28)    Mobile communications services to end customers (also referred to as "retail mobile telecommunication services") include services for national and international voice calls, SMS (including MMS and other messages), mobile internet with data services, and access to content via the mobile network and retail international roaming services.22

Previous Commission decisions

(29)    The Commission has previously considered that there is an overall retail market for mobile telecommunications services constituting a separate market from retail fixed telecommunication services.23 The Commission did not further segment the overall retail mobile telecommunications market based on the type of service (voice calls, SMS, MMS, mobile internet data services), or the type of network technology (for example, 2G/3G/4G). The Commission has considered possible distinctions within the overall retail market for mobile telecommunication services between pre-paid or post-paid services and private customers or business customers, concluding that these did not constitute separate product markets but represent rather market segments within an overall retail market and leaving the question of these possible further segmentations open.24

The Notifying Parties’ views

(30)    The Notifying Parties consider that the Transaction should be assessed on the basis of an overall retail mobile telecommunications market, without further distinction by reference to type of service (voice calls, SMS, MMS, mobile data services), network technology (2G, 3G, 4G or 5G), or distribution channel (online and different type of offline retailers).25

(31)    The Notifying Parties consider that it is appropriate to identify separate segments, rather than markets, for the provision of retail mobile telecommunications services to consumer and business customers, within an overall market for retail mobile telecommunications services.26 Similarly, the Notifying Parties identify separate sub- segments for the provision of pre-paid and post-paid services, within an overall market for retail mobile telecommunications services.27

(32)    Finally, the Notifying Parties agree with the Commission’s decisional practice that the overall retail mobile market excludes machine-to-machine subscriptions (M2M services), and WiFi services.28 However, while the Commission has consistently held that Over-The-Top (“OTT”) services29 do not fall within the retail mobile telecommunications market since OTT services rely on mobile data, or fixed broadband services or Wifi to function, the Notifying Parties submit that OTT services should, at the very least, be taken into account as an out of market constraint.30

Commission’s assessment

(33)    The market investigation in the present case did not provide any indication that the Commission should depart from its findings in previous cases. The vast majority of respondents indicated that there is indeed an overall retail market for mobile telecommunications services, with a distinction between private and business customers and pre-paid and post-paid services representing at most, market segments within the overall retail market rather than separate product markets.31

(34)    On the basis of the information in its possession and in light of the foregoing, the Commission considers that there are no grounds to depart from its previous  decisional practice and concludes, for the purposes of this Decision and without prejudice to further investigation by the CMA, that the relevant product market definition is the market for overall retail mobile telecommunications services.

(b) Geographic market definition

Previous Commission decisions

(35)    In previous decisions, the Commission has consistently concluded that the market for the provision of retail mobile telecommunications services is national in scope.32

The Notifying Parties’ views

(36)    In line with the Commission’s previous practice, the Notifying Parties submit that the market for mobile telecommunications services to end customers is national in scope.33

Commission’s assessment

(37)    The market investigation in the present case did not provide any indication that the Commission should depart from its findings in previous cases. A large majority of respondents agreed that competitors are active in the whole of UK and their strength does not vary across regions.34

(38)    On the basis of the information in its possession and in light of the foregoing, the Commission considers that there are no grounds to depart from its previous  decisional practice and concludes, for the purpose of this Decision and without prejudice to further investigation by the CMA, that the relevant market for the retail provision of mobile telecommunications services is national in scope.

4.3.3.2. Wholesale access and call origination on public mobile networks

(a) Product market definition

(39)    On the market for wholesale access and call origination on public mobile networks, MNOs sell access to their mobile network and the ability to make calls and exchange data traffic. MNOs that own mobile networks constitute the supply side, whereas MVNOs (which do not own a mobile network and thus seek access to one or more of the MNO networks in order to provide their mobile retail services) constitute the demand side of this market. The type of service sold can be different depending on the type of customers. For example, compared to a full MVNO, a light MVNO  would require more services from the host MNO as it does not own any network infrastructure. MNOs typically provide network access and call origination jointly to MVNOs.35

Previous Commission decisions

(40)    In its previous decisions, the Commission defined a single wholesale market including both access and call origination services on public mobile networks.36  The Commission noted that MNOs generally supply these services jointly to MVNOs and both services are essential for MVNOs to be able to provide retail mobile telecommunication services to end users.37

The Notifying Parties’ views

(41)    The Notifying Parties concur with the Commission’s view in its previous decisions that there is a single wholesale market including both access and call origination services on public mobile networks. They submit that the transaction should be assessed on the basis of a market for network access and call origination on public mobile networks, without any sub-division as regards access provided to different types of MVNOs or branded resellers.38

Commission’s assessment

(42)    The market investigation in the present case did not provide any indication that the Commission should depart from its findings in previous cases. The vast majority of respondents indicated that there is indeed an overall market for wholesale access and call origination on public mobile networks without separating out call origination and without any distinction of wholesale services by customer type.39

(43)    On the basis of the information in its possession and in light of the foregoing, the Commission considers that there are no grounds to depart from its previous  decisional practice and concludes, for the purposes of this Decision and without prejudice to further investigation by the CMA, that the relevant product market definition is the market for wholesale access and call origination on public mobile networks.

(b) Geographic market definition

Previous Commission decisions

(44)    In its past decisions, the Commission has consistently held that the geographic scope of the market for wholesale network access and call origination on public mobile networks is national, namely in this case, the UK.40 This is due to regulatory barriers stemming from the fact that licenses granted to MNOs are generally national in scope.41

The Notifying Parties’ views

(45)    In line with previous Commission decisions, the Notifying Parties submit that the relevant geographic scope of the market for wholesale access and call origination on mobile networks is national, limited to the territory of the UK.42

Commission’s assessment

(46)    Nothing on the Commission’s case file indicates that a departure from the conclusions reached by the Commission in previous cases would be justified. The Commission’s past approach was also supported by the market investigation.43

(47)    In light of the above, for the purpose of this Decision and without prejudice to further investigation by the CMA, the Commission concludes that the relevant geographic market for the wholesale access and call origination services on public mobile networks, and all its possible sub-segments, is national in scope.

4.3.3.3. Retail supply of fixed internet access services44

(a) Product market definition

(48)    Fixed internet access services at the retail level consist of the provision of subscriptions enabling customers to access the internet through a fixed telecommunications connection.45

Previous Commission decisions

(49)    In recent cases, within the supply of retail fixed internet access services, the Commission considered but ultimately left open possible segmentations according to

(i) product type (distinguishing between narrowband, broadband, and dedicated access), and (ii) distribution technology (distinguishing between xDSL, fibre, cable, and mobile broadband).

(50)    Furthermore, the Commission also considered, but ultimately left open, possible segmentations as to customer type, distinguishing between residential and small business customers, on the one hand, and larger business and public authorities, on the other hand.46

(51)    Conversely, the Commission acknowledged that the retail market for fixed internet access services should not be divided according to download speed.47 Further, the Commission acknowledged that fixed internet services to large business and government customers form part of a separate market for the retail supply of business connectivity.48

(52)    Finally, the Commission previously considered that fixed internet access services provided through mobile network infrastructure (so-called “fixed wireless access” or FWA services) do not form part of the retail fixed internet access market.49

The Notifying Parties’ views

(53)    The Notifying Parties submit that there is an overall market for retail fixed internet access services, including all broadband speeds (standard, superfast and ultrafast broadband) and both residential and SoHo50/SME51 business customers.52 The Notifying Parties further submit that FWA services do not form part of the retail  fixed internet access market in the UK.53

Commission’s assessment

(54)    With regards to the provision of fixed internet services to residential and SoHo customers on the one hand, and large business and government customers on the other hand, the majority of respondents to the market investigation considered that these are not interchangeable both from a demand54 and supply55 perspective. This is because the requirements of residential and small business customers are rather standardised and similar, while large business customers need special infrastructure, such as VPNs and dedicated lines, and require customised solutions in terms of transmission quality (availability, latency, bandwidth), service level agreements (conditions for interference elimination), connection concepts (redundant connection), etc. Therefore, in line with the Commission’s previous practice and the results of the market investigation,56 the Commission considers for the purposes of  the present decision and without prejudice to further investigation by the CMA, that the provision of fixed internet services to large business and government customers  is part of a separate market. This is also referred to as the market for the provision of retail business connectivity services. As this market is not affected by the Transaction, it will not be addressed separately.57

(55)    In relation to the inclusion of FWA services in the relevant market of fixed internet services), the majority of respondents to the market investigation consider that these services as not interchangeable, either from a demand or a supply-side perspective.58 For example, certain respondents indicated that they would not accept mobile broadband delivery for security and resilience reasons, as well as due to lack of speed.59  Other respondents indicated that for the  vast majority of customers, FWA would not be a viable alternative in view of the key differences between the services in terms of speed, performance (capacity constraints related to spectrum) and geographic coverage.60

(56)    A number of respondents noted that latest evidence suggests that 5G FWA services have the potential, depending on location, to place a degree of competitive constraint on the pricing of fixed broadband.61 However, some of these respondents also acknowledged that this potential is currently mostly limited.62 On this basis, it  appears that FWA services are likely to be compliments rather than substitutes to internet services provided through fixed infrastructure as they are typically used when a fixed connection is not available. When both services are available, the difference in speed, reliability and price generally render them non-substitutable. Therefore, in line with the Commission’s previous practice and the results of the market investigation,63 for the purposes of the present decision and without prejudice to further investigation by the CMA, the Commission considers that fixed internet access services provided through mobile network infrastructure (FWA services) do not form part of the retail fixed internet access market.64

(57)    With regard to the possible segmentation of the market according to product type (narrowband, broadband and dedicated line), the distinction becomes irrelevant as  the difference between standard broadband and dedicated lines is captured in the distinction by customers (only large business and government customers use dedicated lines), while narrowband is virtually non-existent in the UK.

(58)    Finally, the distinctions by speed (30 Mbit/s – 300 Mbit/s, 300 Mbit/s and above) or distribution mode (xDSL, fibre, cable) are unlikely to be justified as above a certain speed consumers are likely to find different speeds as substitutable and they are  likely to be agnostic as to the delivery method. However, the market investigation  did not provide clear evidence on these points.

(59)    In light of the above, for the purposes of this Decision and without prejudice to further investigation by the CMA, the Commission concludes that the market for the retail supply of fixed internet access services excludes FWA services and retail business connectivity services and that there is no need to distinguish different product types. The Commission leaves open the distinction by distribution mode and speed, since doing so would not change the outcome of the Commission’s evaluation of the Referral Request.

(b) Geographic market definition

Previous Commission decisions

(60)    In its previous decisions, the Commission concluded that the retail market for the provision of fixed internet services was national in scope, and in some cases, the Commission left open whether markets should potentially be defined on a regional basis or by reference to the footprint of the operators’ networks.65

The Notifying Parties’ views

(61)    The Notifying Parties submit that the relevant geographic market for the assessment of the Transaction is national in scope, in line with the Commission’s previous decisional practice.66

Commission’s assessment

(62)    In relation to the overall retail fixed internet access market defined above in recital (59), a large number of respondents submitted that, taking into account retail offers via regulated wholesale access, the number and the identity of competitors vary strongly across different regions in the UK.67

(63)    Specifically, respondents highlighted that number and strength of competitors are different in and outside of Virgin Media areas.68 Inside Virgin Media’s footprint, there are four significant retail operators (i.e. BT, Sky, Virgin Media and TalkTalk), whereas outside that area there are only three operators. Further, operators other than Virgin Media such as BT, Sky and TalkTalk use exclusively or mostly the BT Openreach network, while Virgin Media uses mostly its own network.

(64)    By way of exception, Hull is in a separate geographic market, where the incumbent network provider is Kcom rather than BT/Openreach.69 Likewise, there are a large number of providers in the Central London area. Finally, in some remote areas BT Openreach is the only provider.

(65)    These facts indicate appreciably different competitive conditions within and outside Virgin Media’s footprint. Outside this footprint BT is only constrained by two players who use BT’s own network, i.e. the competitors’ offers are not very differentiated. Inside Virgin Media’s footprint BT is constrained by three players, including one strong player (Virgin Media) with its own network and thus a much more differentiated offer.

(66)    The Commission therefore considers that, for the purpose of this Decision and without prejudice to further investigation by the CMA, the market is sub-national and Virgin Media’s footprint is a separate market. Hull, central London and the BT-only areas also have very different competitive dynamics but Virgin Media is not present there, so they are not of interest for the purposes of this assessment.

4.3.3.4. Retail supply of multiple play services

(a) Product market definition

(67)    The term "multiple play" relates to products comprising two or more of the following services provided to retail consumers on the basis of a single or multiple contracts by the same provider: mobile telecommunications services, fixed telephony services, fixed internet access and TV services. Multiple play offers comprising two, three or four of these services are referred to as dual play ("2P"), triple play ("3P") and quadruple play ("4P") respectively. FMC bundles refer to fixed-mobile converged bundles, i.e. the combination of a mobile service with one or more fixed services.

Previous Commission decisions

(68)    In previous decisions, the Commission has considered but ultimately left open the question as to whether there exist one or more multiple play markets, which are distinct from each of the underlying individual telecommunication services.70

(69)    The Commission has previously also noted that in some of these bundles the  different services are delivered over different infrastructures (fixed for dual play and triple play or fixed and mobile for quadruple play). Thus, instead of one possible market for multiple play, there could be several possible multiple play markets: a market for fixed bundles (dual play, and triple play) and another separate market for fixed-mobile convergence bundles. The Commission has also acknowledged that the possibility for several mobile subscriptions to be included in a quadruple play bundle further complicates the picture.71

(70)    Finally, the Commission noted that questions around FMC bundles can be addressed as possible conglomerate effects of a transaction, rather than through defining separate markets for bundles.

The Notifying Parties’ views

(71)    The Notifying Parties submit that it would not be appropriate to define separate markets for fixed-mobile bundles of any configuration.

(72)    According to the Notifying Parties, customers remain willing to ‘unbundle’ their purchases, and this would constrain any attempt to impose a price increase in respect of FMC bundles.72 In addition, the Notifying Parties note that while a large  proportion of UK consumers take some form of bundle, the take-up of FMC bundles is limited in the UK, in particular compared with certain other European markets.73

(73)    Furthermore, the Notifying Parties consider that there is no indication of any lack of supply-side substitutability, since suppliers offering fixed and mobile services separately (either using their own networks or via wholesale access to other operators’ networks) can, without incurring extra costs, start offering these services as part of a bundle.74

(74)    Finally, the Notifying Parties submit that from a practical perspective, given the large number of possible permutations of double-play, triple-play, or quadruple-play offers that are increasingly being sold in the market (including various permutations of fixed-mobile bundles), it is unclear which package of FMC services should be taken as a possible candidate market and which different types of bundles are seen by customers as potential substitutes in the event of a price rise.75

Commission’s assessment

(75)    Given that O2’s presence in fixed services is de minimis, from the point of view of the Transaction, the only question is whether FMC bundles form a separate market. The results of the market investigation in this case are not sufficiently conclusive to establish with the required degree of certainty the existence of a separate market for FMC bundles and which combinations of services would be included in such a market, if it were to exist.76

(76)    Most notably, there was no clear feedback with respect to the question whether 5- 10% price increase of an FMC bundle would be profitable or whether it would be defeated by a demand-side reaction, i.e. whether enough customers would unpick the bundle following a price increase.77 The rest of the responses overall suggested that a separate market should not be distinguished. First, respondents noted that FMC bundles, and in particular dual and triple play FMC offers, remain niche in the UK. It was pointed out that the UK has had a slower uptake of FMC bundles than continental Europe, although this is evolving,78 partially driven by the Transaction, which creates a strong converged player.79 Ofcom reported that only 6% of telecoms subscribers took fixed and mobile services together as triple or quad play offers in 2018.80 This suggests that a large part of consumers in the UK still have a preference for separate purchasing of mobile services. Second, switching barriers do not seem to be very high and mainly consist of contractual obligations, which also apply to standalone products. A majority of respondents considered that there are no significant switching barriers between FMC packages and standalone mobile and fixed services. 81 Although these responses suggest that there is no separate market for FMC bundles, they do not provide decisive and clear evidence to dismiss this possibility.

(77)    Consequently, as the existence of a separate market, or the lack thereof, would not change the outcome of the Commission’s evaluation of the Referral Request, in light of the foregoing the Commission considers that, for the purpose of this Decision and without prejudice to further investigation by the CMA, the question as to whether there exist one or more multiple play markets which are distinct from each of the underlying individual telecommunications services can be left open.

(b) Geographic market definition

Previous Commission decisions

(78)    In previous decisions, the Commission considered that the geographic scope of any possible retail market for multiple play services would be national since the components of the multiple play offers are offered individually at a national level, and the bundling of the services would not change the geographic scope of the components.82

The Notifying Parties’ views

(79)    The Notifying Parties submit that it would not be appropriate to define separate markets for fixed-mobile bundles of any configuration in the UK,83 and thus did not address the question of the geographic scope of such a market.

Commission’s assessment

(80)    A majority of the respondents to the market investigation of this case considered that a possible market for FMC bundles (irrespective of what type of multiple play bundles are included in such possible market) would be national in scope.84 This view, however, does not necessarily align with the result from the market investigation obtained in relation to retail fixed internet services, which would be one of the components of such a hypothetical market. For example, if Virgin Media does not compete across the whole of UK in retail fixed internet, a hypothetical market of FMC bundles is also unlikely to be national. As the precise geographic market would not change the outcome of the Commission’s evaluation of the Referral Request, the Commission leaves the question open.

4.3.3.5. Wholesale supply of leased lines

(a) Product market definition

(81)    A leased line is a dedicated connection between two points in a communications network. Leased lines are inputs for the provision of fixed telecommunications services (retail fixed internet access), retail business connectivity (connectivity services to large business customers) and mobile communications services. Telecommunications service providers (MNOs, fixed internet providers, business connectivity providers) purchase leased lines to complement their network either because they lack physical infrastructure at a certain location or they need to increase the capacity of their infrastructure. MNOs use leased lines as mobile backhaul, i.e. to connect their base stations and their core network. Suppliers of leased lines are infrastructure owners or suppliers who themselves purchase leased lines from infrastructure owners on a wholesale basis.

(82)    A leased line service can be provided using three main hardware  technologies: copper  cables,  fibre  cables  and  wireless  microwave  links.85 Especially with the advent of 5G, the role of copper and microwave is expected to decrease rapidly, while the opposite is true in the case of fibre.

(83)    Leased lines can be supplied at three levels of the network, namely the i) access layer (i.e. when the leased line connects the base station/business customer to aggregation nodes); ii) the aggregation layer (when the leased line connects aggregation nodes to backhaul aggregation nodes) and iii) core network layer (i.e. when the leased line connects core nodes to other nodes in the core network). Since all MNOs and fixed network operators operate their own core network, only the access and aggregation layers are of interest.

(84)    Fibre leased lines can be supplied as a raw piece of infrastructure or as a more or less ready-to-use product. Namely, it can be supplied as i) a physical line itself without any electronic equipment (i.e. unlit optical cables connecting two nodes, otherwise called “dark fibre”); ii) the line can be leased together with electronic equipment that gives access to part of the optical spectrum of the fibre (“optical access”); or iii) the line can be leased with electronic equipment that supports a transmission protocol (predominantly “Ethernet access”).86 Dark fibre is also referred to as a “passive” leased line whereas optical and Ethernet access leased lines are sometimes collectively referred to as “active” leased lines.

Previous Commission decisions

(85)    In a recent decision, the Commission distinguished between fibre and non-fibre leased lines.87  The Commission considered further segmentations, notably by

(a)             network layers: i) access layer, ii) aggregation layer;

(b)             by bandwidth: (i) below 2 Mbit/s, ii) between 2 Mbit/s and 10Mbit/s, iii) above 10Mbit/s), and

(c)             by the active or passive nature of the leased line: (i) passive leased line i.e. dark fibre, and ii) active lease line, i.e. leased lines that provide optical or Ethernet access.

(86)    However, in each case the Commission left the market definition open.88 The Commission's decisional practice does not distinguish between the provision of wholesale leased lines for the purposes of mobile backhaul and provision of wholesale leased lines for other purposes.

Notifying Parties’ views

(87)    The Notifying Parties submit that fibre and non-fibre leased lines constitute separate markets in line with the Commission’s precedents.89

(88)    Other than the hardware technology, the Notifying Parties consider that none of the other distinctions are justified and the supply of all fibre leased lines constitute one overall market.

(89)    The Notifying Parties consider, first, that the distinction by bandwidth/speed (with speeds below 2 Mbit/s, between 2 Mbit/s and 10 Mbit/s) is not warranted. They submit that all new wholesale connections are fibre based and operate at higher speeds than these cut-off points would imply. In other words, technological progress has made the distinction of lines above and below 2 Mbps and 10 Mbps,  respectively, obsolete.90

(90)    Second, as regards the distinction by network layer, the Notifying Parties submit that bandwidth requirements are not necessarily different in the access and aggregation layers in that a rural aggregation layer can have the same capacity as an urban access layer.91 If the bandwidth requirements are similar, the same products are likely to be deployed in the access and aggregation layer. In addition, all products are easily substitutable by suppliers and, due to regulation, the suppliers are able to offer leased lines for both layers at similar conditions.92

(91)    Third, as to the distinction between passive leased lines (dark fibre) and active leased lines (optical or Ethernet access), the Notifying Parties submit that all products have the technical suitability and bandwidth to provide a similar service.93 Although certain differences do arise from the structure of the pricing, the impact on purchasers’ balance sheets and the nature of the capacity risk, these elements are not technical requirements of the underlying products and should be seen as product differentiation, which occurs in any market with customers that have heterogeneous requirements. From a supply perspective, the Notifying Parties submit that all products in the market are based on fibre and are therefore identical before adding electronic equipment to them. Thus, suppliers can easily substitute between offering dark fibre, optical access and Ethernet.94

(92)    Finally, in relation to the distinction by customers, the Notifying Parties submit that in 2016 MNOs preferred leased lines with specific features, namely an aggregation function that made the management of a larger number of leased lines easier (the CMA referred to this as “managed” leased lines in the BT/EE decision95). However, the Notifying Parties note that today all suppliers are able to supply this function. Further, while in 2016 dark fibre was found not to be substitutable with this managed product by the CMA, in the Notifying Parties’ view, the role of dark fibre is clearly greater today as […]. Dark fibre, which is unmanaged as it lacks any equipment, is thus substitutable with managed products.96

Commission’s assessment

Fibre vs. microwave vs. copper

(93)    In line with precedents and the Notifying Parties’ views, the market investigation indicated that there are substantial differences between fibre, microwave and copper leased lines. A majority of respondents considered that having regard to the intended use case, price and performance, these forms of leased lines are not interchangeable.97 Namely, respondents considered that fibre connections are significantly more expensive, performant and reliable than either copper or microwave. By contrast, copper and wireless connections cannot handle the same bandwidths as fibre and they have much higher fault rates.98 It was submitted that only fibre can meet most customers’ needs and that copper is being phased out as it  is no longer relevant in business applications (i.e. also in leased lines).99 In addition, copper and wireless lines are not widely available in wholesale markets.100 These responses suggest that, from a demand perspective, lower performance fibre and wireless connections do not constrain fibre, which is in line with the fact that customers are willing to pay a much higher price for the latter.

(94)    The investigation also indicated that there is no supply-side substitution either as building out a fibre connection or network is a lengthy and costly process.101 Indeed, the very nature of the product – physical infrastructure – suggests that switching between two types cannot be done at a short notice and with minimal costs. Therefore, on the basis of the information in its possession and in light of the foregoing, the Commission considers, for the purposes of this Decision and without prejudice to further investigation by the CMA, that fibre, microwave and copper leased lines form separate markets.

Distinction by bandwidth

(95)    An overwhelming large majority of respondents also indicated that the distinctions based on the bandwidth (speed) points of 2 Mbit/s and 10 Mbit/s are no longer relevant.102 It was submitted that customers are interested in bandwidths of at least 100 Mbit/s and that leased lines below these levels are rarely supplied. Consequently, the Commission considers, for the purposes of this Decision and without prejudice to further investigation by the CMA, that it is no longer justified to distinguish the market on the basis of the bandwidth points of 2 Mbit/s and 10 Mbit/s. Some respondents suggested that 1 Gbit/s could be a new cut off-point103 on the basis that higher bandwidths are only used by MNOs and fixed networks and not by end users, i.e. large enterprises. Further, higher bandwidths command a significant price premium, suggesting the lack of demand-side substitutability between different bandwidths.

(96)    However, a large majority of respondents indicated that fibre leased line suppliers are able to supply all speeds as requested by the customer.104 Thus, the Commission considers, for the purposes of this Decision and without prejudice to further investigation by the CMA, that, due to supply-side substitution, different bandwidths do not constitute separate markets within fibre based leased lines.

Active vs. passive lines

(97)    As regards the distinction between active and passive leased lines (i.e. fibre line with optical  equipment  and  dark  fibre  respectively),  the  market  investigation  did not produce clear results. Although on a simple count the majority of respondents consider them interchangeable,105 the qualitative responses do not  necessarily  support the existence of demand side substitutability. For example, certain respondents submitted that substitutability depends on price, speed or market circumstances, which are ambiguous responses.106 Another, more specific view was that below 1 Gbit/s dark fibre is more expensive but the situation is the reverse when dark fibre is compared to higher bandwidth active lines.107 A respondent submitted that in general dark fibre uses a different architecture than active lines (active lines are point-to-point, whereas dark fibre lines are often not), which results in increased resilience and makes the densification of a network easier and cheaper.108

(98)    From a supply side perspective, the market investigation indicated that suppliers of active lines can easily switch to supplying dark fibre as switching simply implies removing the electronic equipment necessary for an active line.109 Switching in the opposite direction, i.e. from dark fibre to active lines would, however, involve sunk costs and could not be done swiftly as the supplier would need to place equipment on the line.110 That being said, all dark fibre suppliers also offer active lines, so a separate active line market would not result in fewer competitive constraints and thus may not be warranted.

(99)    Nonetheless, the Commission notes that these results are incomplete. First, while demand substitution was unclear, it appears that MNOs, which are a subset of customers, prefer dark fibre because it is cheaper, its architecture is better suited to MNOs and its pricing structure is more favourable in light of their financial profile.111 Second, the capacity of a dark fibre line can be upgraded with a modest one-off cost, whereas in the case of active lines MNOs may be forced to acquire higher bandwidth active line products that are significantly more expensive.112 In other words, dark fibre offers bandwidth flexibility, which active products do not.

(100)    Second, although suppliers of active lines can constrain suppliers of dark fibre, they may lack the incentives to do so.113 Indeed, BT, the largest supplier of leased lines only supplies active lines, although it has the ability to supply dark  fibre.114  It resisted regulatory efforts to introduce mandatory access to its dark fibre, which shows that it would supply dark fibre only if it is forced to by regulation.115 One plausible explanation is that, as a former monopoly, BT is under non-discrimination obligations116 and thus cannot provide dark fibre selectively, i.e. only to certain customers or only in certain areas of the country. Thus it has to decide on supplying dark fibre on a national basis and taking all customers into account and it seems that in that context supplying active products is more profitable. The fact that BT does not supply dark fibre shows that competitive constraints cannot be established solely on the basis of ability.

(101)    It appears therefore that for a proper market definition it is necessary to establish MNOs’ preferences and suppliers’ incentives to supply dark fibre. For example, if there were no demand-substitution, the market definition would turn on supply-side substitution. Based on the market investigation it is easy to switch from supplying active lines to dark fibre, which normally would imply that there is supply-side substitution at least in one direction. However, concluding that this is the case would imply that BT, one of the most important leased line suppliers, is a competitive constraint on dark fibre suppliers whereas this is not the case because it refuses to  sell dark fibre in general.

(102)    In any event, as the existence of a separate market, or the lack thereof, would not change the outcome of the Commission’s evaluation of the Referral Request for the purposes of this Decision, the Commission leaves this market definition question open.

Segmentation by customers

(103)    As indicated above, in the BT/EE decision the CMA found that Ethernet based managed mobile backhaul is not part of a wider market of all end-to-end Ethernet leased line services.117 For example, the CMA noted that TalkTalk does not supply (managed) mobile backhaul whereas it does supply (unmanaged) Ethernet leased lines.

(104)    However, the market investigation indicated that dark fibre, which is not managed, currently plays a significant role in mobile backhaul.118 Consequently, separating out mobile backhaul on the basis of the managed/unmanaged distinction does not appear justified.

(105)    The market feedback also indicated that, to the extent that they use active lines, MNOs generally need higher bandwidth products.119 However, as discussed in relation to the distinction by bandwidth (see paragraphs (95)-(96) above), suppliers can supply any bandwidth products, so an MNO’s bandwidth requirements also do not serve as a basis to define leased lines sold to MNOs as a separate market.

(106)    Respondents to the market investigation also submitted that MNOs have stringent requirements in terms of latency, jitter, wander and packet loss and that not all suppliers can meet these requirements.120 This could, in principle, result in a  narrower market as it would imply that a supplier to MNOs is faced with fewer effective competitors than a supplier to less demanding customers. The Commission notes, however, that, as a practical matter, in the relevant geographic market where there could be a threat to competition the only leased line supplier other than Virgin Media is BT Openreach, which would not be disregarded as a constraint based on quality concerns. Thus, for the purposes of this decision, distinguishing separate markets on the basis of quality differences does not appear warranted.

(107)    Apart from high bandwidth and strong quality requirements, the market feedback did not bring up any other differences between leased lines used by MNOs and leased lies used by other users. Indeed, the majority of respondents indicated that leased lines supplied to MNOs are interchangeable with leased lines supplied to other customers.121

(108)    It appears therefore that based on product characteristics a separate market for MNO backhaul need not be distinguished.

(109)    However, this conclusion still leaves the possibility that MNOs as customers form a separate market if prices can be raised selectively vis-à-vis MNOs, i.e. if they can be price discriminated. In line with the Market Definition Notice,122 the two conditions that need to be fulfilled in this regard is that it is possible to identify MNOs clearly and that trade among customers or arbitrage by third parties is not feasible.123 The first condition is clearly met as MNOs are easily identifiable. There is insufficient information in the market feedback to firmly verify the second condition. MNOs do not trade lease lines among themselves but the practice to buy leased lines on a wholesale basis from infrastructure owners and then resell it to leased line customers exists.124 However, the leased lines used by MNOs are often location specific in that they often need lines to connect their base stations with their core network. As the base stations have quite distinct locations, which are often away from the locations where other users are located (e.g. business parks in the case of end users),125 it appears plausible that infrastructure owners such as Virgin Media can control the conditions at which MNOs receive leased lines even from third parties. As no firm conclusion can be made in this regard, the Commission leaves the question open. Leaving the question open does not change the assessment of the Referral Request.

(110)    In any event, as the existence of a separate market, or the lack thereof, would not change the outcome of the Commission’s evaluation of the Referral Request, for the purposes of this Decision, the Commission leaves the question open whether supply of leased lines to MNOs can be regarded as a separate market.

Access vs aggregation layer

(111)    As regards this distinction, demand substitution is inapplicable by nature as the two layers or segments are different parts of the network and a circuit in one segment will not substitute for a circuit in the other segment. From a supply-side perspective, the market investigation indicated that switching from supplying leased lines in one segment to do the same in the other segment without prior presence in that segment  is prohibitively expensive.126 This would lead to separate markets unless separating these markets would be unnecessary from a practical point of view because the same suppliers are present in both segments with roughly equal strength.

(112)    In this regard, the market feedback indicates that most suppliers are present in both segments and that regulated wholesale access to BT Openreach’s network gives other leased line suppliers the possibility to complement their own network.127 Nonetheless,  competitive  constraints  do  seem  to  differ  substantially  in  the two segments. First, it appears unlikely that wholesale access can completely make up for the lack of a physical network, which is clear from the fact that the two largest networks (BT Opeanreach and Virgin Media) are also considered by market participants as the two strongest players.128 Second, the investigation also indicated that certain suppliers focus on either one or the other segment, creating different constraints and competitive dynamics in the two segments.129 Thus, the two segments appear to be separate markets.

Summary

(113)    Based on the market investigation, fibre, microwave and copper leased lines very likely fall into separate markets. Of these markets only fibre is of interest for the purpose of the assessment in Section (26) as, of the Notifying Parties, only Virgin Media is active in supplying leased lines, and it only supplies fibre leased lines. Within fibre leased lines, there is no need to distinguish separate markets based on bandwidth or speed. Fibre leased lines should be further segmented between the access and aggregation layers. Within each of these segments, the questions whether active and passive lines fall into one market and whether leased lines supplied to MNOs form a separate market are left open, since this would not change the outcome of the Commission’s evaluation of the Referral Request.

(b) Geographic market definition

(114)    The Notifying Parties submit that the market for leased lines is national in scope. The Notifying Parties acknowledge that absent regulation, competitive conditions would differ based on the footprint of the different networks. However, due to Ofcom regulation BT Openreach leased lines are available to all competitors subject to a price cap, which allows all competitors to be present nationally beyond their networks.130

Access layer

(115)    A large majority of respondents indicated that, even taking into account regulated access, the number and the identity of competitors vary strongly across different regions in the UK.131 Concretely a respondent submitted that different areas in the  UK have markedly different characteristics, including as a result of the number of present or nearby networks.132 Even more specifically, several respondents explained that, in line with the geographic market definition of Ofcom, there are four distinct areas:133

·       The Central London Area (CLA) where several suppliers are present and where the market is competitive.

·       The High Network Reach (HNR) areas, i.e. centres of other cities and towns, which are also characterised by more than one network in addition to BT Openreach.

·       Area 2, where BT Openreach and Virgin Media are the two suppliers, i.e. the market is characterized by a duopoly.134

·       Area 3, where BT Openreach has a monopoly.

(116)    As indicated in the responses, these areas match the geographic markets defined by Ofcom in the latest Wholesale Fixed Telecoms Market Review (“WFTMR”).135

(117)    The respective sizes of the areas are 275 postcodes for Central London, 304 postcodes for HNR, 5538 postcodes for Area 2 and 3515 postcodes for Area 3.136 In terms of population coverage Area 2 and Area 3 represent 55.1% and 44.8% of premises in the UK respectively, the balance of 0.2% of premises being in the CLA and HMR areas.137

(118)    The Commission notes that the responses were given on the understanding that regulated access exists, which suggests that regulated access does not overcome the regional variations resulting from the network footprints. In addition, as mentioned before the two largest networks (BT Openreach and Virgin Media) are also considered by market participants as the two strongest players,138 which also  suggests that wholesale access does not create homogenous competitive conditions across the UK.

(119)    Given that a majority of respondents pointed to strong regional variations in competitive conditions, that this view is aligned with Ofcom’s geographic market definition and that several respondents explicitly referred to the Ofcom classification, the Commission considers, for the purpose of this Decision and without prejudice to further investigation by the CMA, that the geographic market is sub-national and the markets in question correspond to the four areas identified by Ofcom, as discussed above.

Aggregation layer

(120)    The market investigation did not produce clear results in relation to the geographic scope of the aggregation layer.

(121)    Under its regulatory review, referred to as Business Connectivity Market Review (“BCMR”),139 Ofcom defined each BT exchange (the network  facility  that aggregates lines) as a separate geographic market.140  There are a total of 5 573 BT exchanges in the UK with the following distribution in terms of competitive dynamics:141

·       BT is the only provider at 4269 exchanges (BT only exchanges)

·       BT and an additional provider are present at 733 exchanges (BT+1 exchanges)

·       BT and two or more providers are present at 571 exchanges (BT+2 or more exchanges)

(122)    There is nothing in the Commission’s file that would call into question this market definition and thus the Commission adopts it for the purposes of this Decision and without prejudice to further investigation by the CMA.

4.3.4. Conclusion on the first substantive condition

(123)    In addition to the Commission’s findings set out above (Section (26)), the Commission notes that the JV’s services will only be offered in the UK. In the Form CO, the Notifying Parties have not indicated any relevant or affected markets outside the UK. The JV’s activities will, at least in the foreseeable future, be restricted to the UK territory.

(124)    In light of the above, the Commission considers that the markets identified in the Referral Request (Section 4.3.1) present the characteristics of distinct markets in the UK, as required under Article 9(2)(a) of the Merger Regulation, also in light of Article 9(7) thereof.

4.4. Markets within the UK in which the Transaction threatens to significantly affect competition

4.4.1. The CMA’s submission

(125)    In its Referral Request, the CMA submits that the Transaction threatens to significantly affect competition in the UK as a result of the following:

–                   Vertical concerns:142 first, the CMA claims that there is a risk of input foreclosure in relation to the wholesale market for access and call origination  on public mobile networks (upstream) and MVNOs on the retail mobile telecommunications market (downstream) in the UK. In that respect, the CMA points to the JV’s ability and incentive to restrict the supply of these services in order to weaken competition in retail telecommunications mobile services in general. Second, the CMA submits that there are input foreclosure concerns in relation to the wholesale leased lines and mobile backhaul (upstream) and retail mobile telecommunications services (downstream) in the UK. In that respect, the CMA stresses the fact that Virgin Media is the second largest provider of wholesale leased lines and the only player of significant scale aside from BT.

–                   Horizontal concerns:143 the CMA notes that, any review of the Transaction will need to consider potential horizontal unilateral effects in relation to retail mobile telecommunications services in the UK. Although Virgin Mobile is a relatively small retail mobile supplier, the Transaction would strengthen O2 as the UK player with the largest (or second largest) market share.

(126)    In light of the above, the CMA prima facie considers that the Transaction is likely to have significant effects in the UK market.

4.4.2. The Notifying Parties’ view

(127)    The Notifying Parties disputes the CMA’s findings in its Referral Request.

(128)    In relation to the CMA’s horizontal concerns, the Notifying Parties’ views are set out in detail in Section 4.4.3.2.

(129)    Second, in relation to the CMA’s vertical concerns, in particular the input foreclosure in relation to the wholesale market for access and call origination on public mobile networks (upstream) and MVNOs on the retail mobile telecommunications market (downstream) in the UK, the Notifying Parties views’ are set out in detail in Section 4.4.3.4.

(130)    Third, in relation to the CMA’s vertical concerns, in particular the input foreclosure in relation to the wholesale leased lines and mobile backhaul (upstream) and retail mobile telecommunications services (downstream) in the UK, the Notifying Parties’ views are set out in detail in Sections 4.4.3.6 and 4.4.3.7.

4.4.3. Commission’s assessment

4.4.3.1. Overview

(131)    In the following sections, the Commission discusses all of the horizontal and non- horizontal relationships between the Notifying Parties, which give rise to affected markets within the meaning of Section 6.3 of Annex I to Commission Regulation No 802/2004,144 and carries out a preliminary assessment as to whether the Transaction threatens to significantly affect competition in these markets.

(132)    The Transaction gives rise to the following horizontal and vertical relationships that involve affected markets.

Horizontal relationships:

·       In the retail mobile telecommunications market O2 is one of the four MNOs, while Virgin Mobile is the largest MVNO. O2’s market share is [30-40]% by subscribers and [30-40]% by revenues. Virgin Media’s market share is [0-5]% and [0-5]% by subscribers and revenues respectively. The retail mobile telecommunications market is thus horizontally affected.

·       In the retail fixed internet market Virgin Media’s share is around [20-30]% on a national basis and well in excess of this figure in the relevant geographic  market,

i.e. Virgin Media’s footprint. O2 has some minimal fixed internet activities, which technically makes this market a horizontally affected market.

Vertical relationships:

·       O2’s share on the upstream wholesale market for call origination and access on public mobile networks is [30-40]% based on subscribers, while Virgin Mobile’s share on the downstream retail mobile telecommunications market is [0-5]% by the same metric. As a result the downstream retail mobile telecommunications market is vertically affected on account of potential input foreclosure.

·       O2’s share in the retail mobile telecommunications market is [30-40]% by subscribers and [30-40]% by revenues and this is increased by Virgin Media’s share of [0-5]% by subscribers and [0-5]% by revenues. Thus the upstream wholesale market for call origination and access on public mobile networks is vertically affected on account of customer foreclosure.

·       Virgin Media’s share on the upstream access layer fibre leased line market in Area 2 is 40-50% if active and passive lines are in one market and a monopoly if passive lines (dark fibre) is a separate market.145 Its share would be even higher than 40-50% if passive and active lines are one market but supply of leased lines to MNOs was a separate market. Given that MNOs use leased lines as mobile backhaul, the downstream retail mobile telecommunications market is vertically affected on account of input foreclosure.

·       In the upstream aggregation layer fibre leased line market each BT exchange is a separate geographic market. Out of the 5 573 BT exchanges 733 are BT+1 exchanges. Virgin Media is the only other supplier next to BT in a subset of these exchanges. Market shares are not available but Virgin Media clearly has market power as a duopolist if passive and active lines belong to one market. Its market power would be even greater if the market only included MNOs. Finally, if dark fibre was a separate market it would again be a monopolist at all these exchanges and a duopolist at a subset of the 571 BT+2 or more exchanges, notably at the exchanges where there are only two dark fibre suppliers. Given that MNOs use aggregation layer leased lines as mobile backhaul, the downstream retail mobile telecommunications market is vertically affected on account of input foreclosure.

·       O2’s share on the downstream retail mobile telecommunications market is [30- 40]% by subscribers and [30-40]% by revenues. Given that MNOs use access layer fibre leased lines as mobile backhaul, this makes the upstream access layer fibre leased line market in Area 2 vertically affected on account of customer foreclosure.

·       O2’s share on the downstream retail mobile telecommunications market is [30- 40]% by subscribers and [30-40]% by revenues. Given that MNOs use aggregation layer fibre leased lines as mobile backhaul, this makes all those upstream aggregation layer fibre leased line markets (exchanges) affected where Virgin Media is present vertically affected on account of customer foreclosure.

(133)    Based on the above, the markets affected by the Transaction are the following:

·       Retail mobile telecommunications market (horizontal relationship and three vertical relationships)

·       The fixed internet market in Virgin Media’s footprint (horizontal relationship)

·       The wholesale market for call origination and access on public mobile networks (vertical relationship)

·       The access layer fibre leased line market in Area 2 (vertical relationship)

·       A number of aggregation layer fibre leased line markets, namely the BT exchanges where Virgin Media is present (vertical relationship)

(134)    As it is clear on the basis of this list, the retail mobile telecommunications market  can be affected by an elimination of a competitor (horizontal effect) and three different types of input foreclosure strategies (namely those related to the wholesale market for market for call origination and access on public mobile networks in the UK; the access layer fibre leased line market in Area 2; and a number of aggregation layer fibre leased line markets). The Commission notes that these effects can accumulate and thus it may be the case that the Transaction threatens to significantly affect competition in this market even if individually none of the four effects would reach that threshold. As a result a small effect in one relationship cannot simply be disregarded on its own but has to be looked at together with other effects.

(135)    In the subsequent sections the Commission will present the results of the market investigation and will use those results to draw more precise conclusions as regards Article 9(2)(a) of the Merger Regulation. The Commission also notes that, in accordance with footnote 21 of the Referral Notice, the existence of “affected markets” within its accepted meaning (i.e. within the meaning of Annex III to Commission Regulation No 802/2004) would generally be considered sufficient to demonstrate that the transaction in question may significantly affect competition in that market within the meaning of Article 4(4) of the Merger Regulation. By analogy, the same applies to Article 9(2)(a) of the Merger Regulation, which applies an equivalent test. Consequently, it is against the background that each of the markets listed in recital (133) above already constitute “affected markets” that the results of the investigation need to be assessed.

4.4.3.2. Horizontal non-coordinated effects in the retail mobile telecommunications market in the UK

(a) Market shares

(136)    As stated above in recital (132), in the retail mobile telecommunications market O2  is one of the four MNOs, while Virgin Mobile is the largest UK MVNO. The Notifying Parties’ combined market share is [30-40]% by subscribers and [30-40]% in terms of revenues in 2019. Individually, O2’s and Virgin Mobile’s market share is [30-40]% and [0-5]% by subscribers and [30-40]% and [0-5]% by revenues respectively. The retail mobile telecommunications market is thus horizontally affected.

(b) The Notifying Parties’ views

(137)    The Notifying Parties submit that the Transaction cannot credibly give rise to unilateral effects concerns.146

(138)    First, while the JV’s combined retail mobile telecommunications market share will  be around [30-40]%, the Transaction involves an MVNO, Virgin Mobile, with a market share lower than 5% against a market background where the JV will continue to compete with three similarly sized MNOs and over 150 MVNOs.147

(139)    Second, according to the Notifying Parties, all the available diversion ratios show  that O2 and Virgin Mobile are distant competitors, reflecting their distinct commercial strategies (O2 with a ‘mobile first’ strategy that targets other MNOs; Virgin Media with a converged strategy, cross-selling mobile services to its fixed customer base and targeting other converged providers such as BT/EE and Sky).148

(140)    Finally, the Notifying Parties further submit that based on their Indicative Price Rise (IPR) analysis,149 in which two different counterfactuals were considered ((i) Virgin Mobile continues to be hosted on BT/EE’s network; and (ii) Virgin Mobile migrates to Vodafone’s network (as was intended prior to the creation of the JV), weighted across both Parties’ brands, the results of the IPRs were generally below [0-5]%. The Notifying Parties note that this % is well below the IPRs identified in any Commission investigation where remedies were required and thus confirms that the horizontal relationship between O2 and Virgin Media does not raise any unilateral effects concerns.150

(c) Commission’s assessment

(141)    For the reasons set out below, the Commission preliminarily considers  that horizontal effects in the retail mobile telecommunications market in the UK are unlikely to threaten competition to a significant degree.

(142)    First, the Transaction does not involve an appreciable market share increment. Virgin Media’s increment of [0-5]% by subscribers ([0-5]% by revenues) is modest. Virgin Mobile has been an MVNO on the UK market for over 20 years and its share has decreased over time from around 6% to around [0-5]% in 2009, where it has remained largely stable since this time.151 Furthermore, its gross adds share ([0-5]% in 2019)152 is lower than its overall market share, which also indicates that it is not a strong constraint.153

(143)    Second, O2 and Virgin Media do not appear to compete closely. O2’s approach is a customer-centric, mobile-first strategy,154 while in Virgin Mobile’s strategy mobile is only an add-on to make its triple-play bundles more attractive.155 Diversion ratios based on mobile number portability (“MNP”) data (even if such data is imperfect)  are consistent with this finding.156

(144)    Third, the retail mobile telecommunications market in the UK is well functioning and very competitive.157 O2 is one of four MNOs in the UK, competing with BT/EE, Vodafone, and Three. The UK also features over 150 MVNOs. The post-merger UK market will continue to have four MNOs and a number of large MVNOs. This is in line with the Commission’s findings in previous cases.158

(145)    Finally, the vast majority of respondents to the market investigation did not raise any concerns.159 The majority of respondents also noted that it is easy for customers on the UK retail mobile telecommunications market to port their numbers and switch.160

(146)    Nonetheless, as in all cases where a horizontal competitor is eliminated, there is an internalisation effect as the merged entity will no longer lose customers to Virgin Media and as a result will be able to afford slightly higher prices. While the market investigation indicates that this effect would not be significant in itself, it can add to other effects on the retail market that are brought about by the Transaction. In other words this effect can still be taken into account when the different effects on the  retail market are cumulated.

(d) Conclusion

(147)    Therefore, based on its preliminary analysis and without prejudice to further investigation by the CMA, the Commission considers that the Transaction would not threaten to significantly affect competition in the retail mobile telecommunications market in the UK a result of horizontal non-coordinated effects caused by the Transaction. The non-significant effect can still, however, be taken into account when effects on the retail mobile telecommunications market are cumulated.

4.4.3.3. Horizontal non-coordinated effects in the fixed internet market

(148)    On a national basis, Virgin Media is the second largest player after BT (around 40% market share) in the UK with a market share of approx. [20-30]% by revenues ([10- 20]% by subscribers).161 In the likely relevant geographic market (Virgin Media’s footprint), Virgin Media’s share is not available but it is certainly significantly above 20%.

(149)    O2’s activities are limited and relate to business customers only (which in this case means small businesses, i.e. home office and SME customers, as services to large businesses and government customers fall into the retail business connectivity market). O2 has a market share of approx. [0-5]% by revenues and subscribers on a national basis and equally insignificant under any geographic market, including virgin Media’s footprint.162

(150)    Thus, the retail fixed internet market is horizontally affected.

(151)    However, as the increment is insignificant (around [0-5]%) and without prejudice to further investigation by the CMA, the Commission considers that the Transaction would threaten to significantly affect competition in the retail fixed internet market  as a result of horizontal non-coordinated effects caused by the Transaction.

(152)    For the reasons set out in detail in Section 4.6.3, while partial referrals are possible under Article 9, the Commission nonetheless considers it appropriate that the retail fixed internet market in Virgin’s footprint is also dealt with by the CMA. A partial referral would inherently risk fragmentation of the case, which is undesirable in view of the one-stop-shop principle (see recital (245)).163

4.4.3.4. Foreclosure of rival MNOs by foreclosing access to the merged entity’s wholesale call origination and access services on public mobile networks (input foreclosure)

(153)    For the reasons set out below, the Commission considers that the Transaction threatens to affect significantly competition on the downstream retail mobile telecommunications market.

(a) Market shares

(154)    As discussed above in paragraph (132), O2’s share on the upstream wholesale market for call origination and access on public mobile networks is [30-40]% based on subscribers, while Virgin Mobile’s share on the downstream retail mobile telecommunications market is [0-5]% by the same metric.

(b) The Notifying Parties’ views

(155)    The Notifying Parties submit that the JV will not have the ability or incentive to i) restrict or degrade the supply of wholesale call origination and access services to MVNOs, including suppliers of fixed-mobile services (fixed-MVNOs), by not bidding or bidding less competitively for future contracts; or ii) degrade, within the existing contract, the quality of the wholesale call origination and access services that O2 currently provides to Sky.164

(156)    As regards future contracts, the Notifying Parties submit that the JV will lack the ability to foreclose MVNOs from the retail mobile telecommunications market because there are a number of credible alternative providers of wholesale call origination and access services which any MVNO could choose from in the event the JV refused to provide access to its network.165 Moreover, even in a scenario where  the JV did not bid or bid more weakly, absent market transparency other MNOs would not recognise that competition was weaker and would not degrade their wholesale offers materially in response.166 As regards incentives, the Notifying Parties submit that the JV would not have the incentive to foreclose MVNOs as it could not recoup the lost wholesale revenues on the downstream retail market. The lack of incentives to foreclose is clearly observable today as O2 competes strongly  on the wholesale market.167 This remains true even as regards foreclosing fixed MVNOs since any refusal to supply or degradation in the terms on which wholesale call origination and access services are supplied would not lead to material switching by those MVNOs' fixed customers to the JV’s fixed products. This was confirmed by BT / EE’s behaviour on the wholesale market following the BT/EE merger in that EE continued to offer competitive terms to fixed MVNOs despite the fact that, theoretically, it could gain fixed customers downstream by degrading the mobile element in the bundle.168 Even if the JV had the ability and incentive to foreclose, there  would  be  no  impact  on  downstream  competition  as  the  retail  mobile telecommunications  market  is  competitive  and  a  number  of  players  (BT, Sky, Vodafone) could offer FMC bundles.169

(157)    As regards the foreclosure of Sky’s MVNO by degrading its mobile service, the Notifying Parties submit that the same arguments apply.170 In addition, […],171 […], and Sky can also retaliate by refusing to supply to the JV its “must-have” premium TV content.172 Further, Sky’s loyal TV and broadband customers will not switch to the JV’s fixed products just because their mobile service is degraded but will rather cancel their mobile subscription and purchase a different one.173 Consequently, the  JV will not be able to recoup its lost revenues upstream, which means it will not have the incentive to follow this strategy.

(c) Commission’s assessment

(158)    The market investigation raised concerns with regard to this vertical link. Overall, market participants are concerned that the merged entity will have the incentive to foreclose MVNOs from access to its mobile wholesale services.174 In particular, market respondents indicated that considering that the merged entity will be an FMC player and that it will also be present in the wholesale call origination and access market, it will have the incentive to foreclose MVNOs offering FMC bundles.175

(159)    In the hypothetical scenario where the quality of the mobile component in an FMC bundle is degraded, half of the market respondents who expressed a view considered that FMC customers would most likely switch providers for the entire bundle, while the other half of the market respondents consider that most FMC customers would cancel the mobile subscription and switch mobile provider.176

(160)    Specifically, a respondent submitted that FMC uptake in the UK has lagged behind that in many EU Member States, and only a small number of UK customers currently purchase services through FMC bundles. However, this is changing and the UK is witnessing more fixed-mobile cross selling and discounting across services. There is also evidence to demonstrate an emerging trend within the UK market towards single-price FMC bundles. The Transaction would accelerate this trend as it would create a strong converged player (in particular, by adding O2’s successful mobile service/brand and customer base to Virgin Media’s current offering). The  Transaction is therefore likely to increase competition in FMC bundles, making them more attractive and popular. As it is easy to switch between the various components in an FMC bundle, this respondent submits that it is possible that the majority of switching FMC customers would switch the provider for the entire bundle.177  These responses suggest that a strategy focusing on FMC customers by degrading the quality of the mobile component might be effective.

(161)     […].

(162) […]. According to the Notifying Parties’ own public statements, fixed-mobile convergence is a key rationale for the Transaction at a time when interest in such services in the UK is increasing.

(163)     […].

(164)    In summary, respondents in the market investigation […] were concerned that the Transaction may lead to foreclosure of MVNOs.

(165)    The Commission considers that such concerns are not well founded with regard to mobile-only MVNOs. O2 already has its “own” MVNO, Tesco Mobile, and its own retail business and thus could have already followed the strategy of foreclosing mobile-only MVNOs if it had been profitable and possible. The fact that it did not do so (it is a big MVNO host) indicates that it lacks the ability and/or the incentive to do so. The issue is less straightforward with regard to MVNOs offering FMC bundles (fixed MVNOs), in particular Sky. This is because the […]178 and thus it may be profitable for the merged entity to specifically target fixed MVNOs and in particular Sky.

(166)    As regards its ability to foreclose fixed MVNOs and Sky in particular, the Commission considers that nothing prevents the merged entity from withdrawing its wholesale services or to increase its price (reduce its quality) one way or another. It can chose to break its contractual obligations such that Sky is forced to terminate its contract and then not bid for Sky or any other fixed MVNO (scenario 1). The elimination of one out of four competitors on the upstream market will likely result  in a price increase. The JV can also degrade service quality to the level that does not result in a breakdown of the contract, the effect of which equals to a price increase (scenario 2). Finally, the JV can withdraw from bidding for Sky once the contract expires or when the relationship ends for any reason other than the JV breaching its contract (scenario 3). The withdrawal of a major competitor would result in a price increase in this case too. For completeness, the JV can also withdraw from bidding for fixed MVNOs other than Sky, although other than Sky and Virgin Mobile (which will be one entity with the JV post-merger) there are no fixed MVNOs of significant size (scenario 4). In summary the JV is likely to be able to force an upstream price increase. If the JV completely disappears from the upstream market (with respect to Sky or other fixed MVNOs), then the price increase is likely to be significant as the number of players is reduced from four to three. Even a quality decrease that does  not result in a breakdown of contract relations can prompt a significant number of FMC customers to switch downstream providers. Whether this strategy is likely to be implemented depends on the expected losses upstream and the gains downstream, in other words the incentives.

(167)    As regard the incentives, the expected losses upstream and the gains downstream differ in the different scenarios indicated above. In scenario 1 the merged entity loses all the upstream profits, while in scenario 2 it loses only the profits that it earns on  the departing customers. In scenarios 3 and 4 it would lose only the expected profits from not bidding. The downstream profits may also be different in each scenario and depend on a number of things such as the magnitude of upstream price increase, the pass through rate of that increase, the downstream price increase of the FMC bundle and the standalone product, the number of FMC and other customers that would leave as a result of the downstream price increase and the ratio of leavers that would choose the JV (diversion ratio downstream). It is also an important factor how many FMC customers who leave would unpick the bundle as only the capture of customers with FMC bundles has the hope of making the strategy profitable. Further, in scenario 1, many customers would leave Sky as soon the service is seriously degraded or disrupted, i.e. even before the switching to another MNO host takes place given that such a switch involves considerable time and organization. Subsequently, when the new tender is run and the price increases (because there would be three bidders instead of four) there would be additional leavers as the price increase trickles down.

(168)    Standard modelling of some of these scenarios by the Parties indicates that foreclosure does not seem to be an economically rational strategy as the loss of (expected) upstream profits could not be recouped downstream.179 However, the modelling assumed a competitive wholesale market, whereas […] the market investigation […] points out that Sky would not achieve very favourable terms.180 […] has quality issues, while BT and the merged entity would be close converged competitors of Sky and unwilling to offer good terms. […] appears to be a willing competitor but one single willing competitor may not suffice for a good competitive outcome. In other words, the price increase upstream may be substantial. Further, the modelling also did not take into account that in certain scenarios (e.g. scenario 1) a large number of customers would leave even before the switch to another MNO  could take place. Another unclear point that can have an influence on the incentives is the nature of the MVNO agreement between Sky and O2 […]. The legal assessment of this document can influence the assessment as it has an effect on the upstream losses (penalties that O2 would have to pay if the contract is breached) in scenario 1 and the scope of quality decrease (i.e. price increase) without triggering a contract cancellation (scenario 2).

(169)    The results of the market investigation specific to incentives do not lead to firm conclusions. First, the results indicate that Sky’s customers are loyal to Sky and  value highly its TV content,181 which would limit the number of customers  that would switch the whole bundle as opposed to just switching the mobile content in the case of Sky. However, Sky content is also on Virgin Media, i.e. the same content would also be available to switching customers unless Sky withdraws that content as retaliation. Second, respondents described Sky’s content as “must have”,182 which implies that Sky can at least threaten to retaliate by withdrawing that content. However, carrying out that threat may be too costly for Sky and thus such a threat may not be credible.

(170)    Overall, the Commission considers that in view of the fact that open questions  remain as to whether such strategy would be profitable and thus whether the merged entity would be likely to pursue it, the Transaction threatens to have a significant impact on competition cannot be excluded. The outcome would not be different if FMC bundles were a separate market. The assessment on the ability and incentive to foreclose would be similar but the impact would be relatively larger as the same impact would occur on a smaller market.

(d) Conclusion

(171)    In light of the above, based on its preliminary analysis and without prejudice to further investigation by the CMA, the Commission considers that the Transaction threatens to significantly affect competition in the downstream market for retail mobile telecommunications services as a result of input foreclosure strategies at the upstream market for wholesale call origination and access services. This effect can  be cumulated with other effects such as the horizontal non-coordinated effects discussed in Section 4.4.3.2 and any other effect on the retail mobile telecommunications market.

4.4.3.5. Foreclosure of rival wholesale call origination and access providers on the upstream wholesale market for call origination and access services on public mobile networks by foreclosing access to a wholesale customer on the retail mobile telecommunications market (customer foreclosure)

(a) Market shares

(172)    As stated above in paragraph (132), O2’s share on the upstream wholesale call origination and access market is [30-40]% based on subscribers, while Virgin Mobile’s share on the downstream retail mobile telecommunications market is [0- 5]% by the same metric.

(173)    Virgin Media is currently hosted on EE's network, but has agreed to migrate its mobile customers to the Vodafone network, starting at the end of 2021. As a result of the Transaction, Virgin Media will transition its subscriber base to O2.

(b) The Notifying Parties’ views

(174)    The Notifying Parties submit that customer foreclosure could only arise if other MNOs depend on wholesale revenue from Virgin Media in order to compete on the wholesale access market.183 Virgin Media is currently hosted on EE's network, but entered into an agreement in October 2019 to migrate its mobile customers to the Vodafone network, starting at the end of 2021.

(175)    Neither EE nor Vodafone depends on revenues from Virgin Media in order to compete on the wholesale access or retail markets. First, EE would have lost these revenues in any event absent the Transaction. Second, it is implausible that the loss of these revenues could materially affect Vodafone’s business, which indeed was able to compete for the Virgin Media contract without the benefit of being Virgin Media's host network.184

(c) Commission’s assessment

(176)    The Commission’s preliminary view is that the Transaction does not lead to credible customer foreclosure concerns.

(177)    The only change brought about by the Transaction is that Virgin Mobile will no longer be available as a customer. This is unlikely to affect wholesale providers to such a degree that they would have to raise their prices. In concrete terms it is unlikely that Vodafone, the MNO that stands to lose Virgin Mobile as a customer as the latter would have been hosted by Vodafone absent the Transaction, will no longer be able to compete effectively on the wholesale market. Indeed, the vast majority of the respondents to the market investigation indicated that Vodafone is actively competing on the wholesale call origination and access market.185

(178)     […].

(179)    However, the Commission considers that the loss of Virgin Mobile as a customer is unlikely to affect Vodafone to such degree that it would have to cancel or delay investments. This is because, even without Virgin Mobile, Vodafone has larger scale than Three, which is still a formidable retail competitor and is rolling out 5G.

(d) Conclusion

(180)    Therefore, based on its preliminary analysis and without prejudice to further investigation by the CMA, the Commission considers that the Transaction would not threaten to significantly affect competition in the upstream wholesale market for call origination and access on public mobile networks as a result of customer foreclosure strategies.

(181)    For the reasons set out in detail in Section 4.6.3, while partial referrals are possible under Article 9, the Commission nonetheless considers it appropriate that the UK wholesale market for call origination and access on public mobile networks is also dealt with by the CMA. A partial referral would inherently risk fragmentation of the case, which is undesirable in view of the one-stop-shop principle (see recital (245)).186

4.4.3.6. Foreclosure of rival MNOs by foreclosing leased lines in the access layer (input foreclosure)

(182)    This section discusses the vertical relationship between the access layer fibre leased line market in Area 2 upstream and the national retail mobile telecommunications market downstream.

(a) Market shares

(183)    As discussed in paragraph (132) above, Virgin Media’s share is 40-50% under the widest possible market definition, i.e. if dark fibre and active lines fall into one market and if the market is not restricted to MNO customers. If the market is restricted to MNO customers, Virgin Media’s share would be higher as it is a particularly important leased line supplier of MNOs. In both cases its only competitor would be BT Openreach. Finally, regardless of the distinction by customers, Virgin Media would be a monopoly if dark fibre was a separate market.

(b) The Notifying Parties’ views

(184)    In the Form CO, the Notifying Parties submit, first, that the JV does not enjoy any market power for its backhaul services as customers can obtain these services from BT at regulated prices at all areas where there is a limited number of suppliers.187

(185)    Second, the Notifying Parties submit that most of the MNOs already have access to cost-effective infrastructure such that they could not be vertically foreclosed. Specifically EE has access to the BT infrastructure, Vodafone can self-supply a proportion of its backhaul and Three partnered with SSE.188

(186)    Third, from an MNO perspective, costs for backhaul services account only for a small share of the total costs to provide mobile services and only a small share of these are supplied by Virgin Media.189 Concretely, the Notifying Parties submit that Ofcom estimated backhaul costs to be 8% of total operating costs. However, the Notifying Parties note that a more appropriate metric would be to compare backhaul costs to incremental costs as only these influence pricing or revenues. Using these metrics and based on O2 data, the share of backhaul costs would be [0-5]% and [0- 5]%, respectively.190 Even if Virgin Media fully withheld access to its backhaul services, quod non, the costs for backhaul services previously provided by Virgin Media would not increase significantly, due to the ability of MNOs to switch to BT  or to alternative suppliers, and moreover any resulting cost increase would only have a negligible effect on the MNO’s overall cost structure.191

(187)    Fourth, customers are protected against full input foreclosure through long-term contracts or investments in self-provision. Notably, Three and Vodafone have contracts in place with Virgin Media.192

(188)    In subsequent submissions the Notifying Parties further elaborated on these arguments and pointed out that Vrigin Media has little scope to increase prices.193 Virgin Media’s dark fibre is priced […] and is more expensive than […] product. As MNOs will not need capacities above 10 GB, Virgin Media is […]. The Notifying Parties further underline that the upstream market is national based on the  availability and uniform pricing of BT Openreach’s active products.194 In any event, only […] of O2’s sites fall in the BT/Virgin Media-only areas. The Notifying Parties also argue that Virgin Media can already align its prices with BT Openreach’s transparent prices and thus if was profitable to withdraw dark fibre it would already do so.195 As this is not the case, any foreclosure would involve sacrificing upstream profits for downstream gains. Given the minimal share of backhaul costs in MNOs’ costs, in the Notifying Parties’ view the downstream cost increases and thus the downstream benefits are minimal.196  The leverage to foreclose is further reduced by the JV’s limited footprint, the existing contractual protections, and the presence of competing suppliers in the most densely populated areas of the country. Given that O2 shares a network with Vodafone, it is also unlikely that the JV would foreclose Vodafone, which reduces the leverage even more. Finally the Notifying Parties submit that withdrawal of dark fibre cannot disrupt the rollout of 5G networks because of the ubiquity of BT Openreach’s network.197

(c) Commission’s assessment

(189)    At the outset, the Commission notes that it sees no reason for revising the geographic market definition. The Notifying Parties’ arguments do not introduce anything new compared to their views presented in Section 4.3.3.5 (iii), which have already been assessed in detail in Section 4.3.3.5 (iv).

MNOs prefer dark fibre

(190)    The Commission notes, first, that, as indicated in the section on market definition, MNOs appear to have a strong preference for dark fibre as opposed to active lines.

(191)    An MNO respondent submitted that dark fibre connections tend to be cheaper and have more attractive pricing structures. In addition, dark fibre circuits allow for scaling up capacity with demand, while only bearing the incremental cost (often a one-off software update). With active leased lines, operative expenses will increase significantly as the bandwidth bracket is increased. Furthermore, dark fibre networks often use a ring architecture, rather than point-to-point. This adds greater resilience (critical for mobile networks) and allows mobile operators to connect directly to the loop without the need for aggregation points, which in turn means it is easier and cheaper to densify the network.198 The comment on the bandwidth bracket refers to the fact that active lines are sold in discreet bandwidths of 1 Gb/s, 10 Gb/s and 100 Gb/s, which are progressively more expensive. Another MNO echoed these comments and submitted that dark fibre would enable it to upgrade capacity as required at a much lower cost.199

(192)    The fact that dark fibre is preferred by MNOs because of its flexibility to upgrade capacity is also supported by […]200 Likewise, in relation to 5G rollout the […]201 Although the Notifying Parties also mention that optical access solutions202 also give the required flexibility, there is no evidence on the file that would support this claim. The Commission notes that […]

(193)    The MNOs and the Notifying Parties’ views are in line with the fact that […] already procured dark fibre from Virgin Media.

(194)    The evidence shows, therefore, a consistent picture: 5G requires a significant  increase in backhaul capacity and MNOs prefer to use dark fibre for this purpose because it has a network architecture that is more suitable to MNOs and because MNOs can upgrade their backhaul capacity at a lower cost relative to upgrading it with active lines. Based on the market feedback the cost advantages relative to active lines come from the fact that dark fibre is cheaper, that it can be upgraded at minimal cost (as opposed to being forced to change bandwidth bracket) and that it is less costly to densify the network with dark fibre than with active lines.

(195)    These results contradict the Notifying Parties’ assertion that dark fibre is already priced […] and that, as a result, there is no scope for cost increases. Based on the market feedback, if Virgin Media were to fully withdraw dark fibre, MNOs would be forced to use active lines with the resulting cost increases. In light of these results,  the Commission cannot exclude that the withdrawal of dark fibre would make the upgrading of rival MNOs’ backhaul networks significantly more expensive and thereby threaten to affect significantly competition on the retail mobile telecommunications market.

(196)    Furthermore, even independently of costs, MNO respondents also submitted that a withdrawal of dark fibre by the merged entity would force them to revise their network planning and thus slow down and compromise their ability to deliver 5G.203

Ability to foreclose

(197)    As Virgin Media is the only supplier of dark fibre in Area 2, the Commission considers that it has the ability to withdraw access to it.

(198)    The fact that […] have both already contracts in place does not necessarily imply a lack of ability. […] and the rollout of 5G can lead to cell densification, which would lead to further purchases. Moreover, frustrating supply under an existing contract also cannot be ruled out.

Incentives to foreclose

(199)    As regards incentives, the JV could start to supply active lines instead of dark fibre. Doing so would not necessarily imply that all the profits it made on dark fibre would be lost. By the Notifying Parties’ own admission, the market is transparent and BT’s prices are well-known nationally.204 Thus, nothing seems to prevent Virgin Media from keeping its active line prices slightly below BT Openreach’s price and maintain an unchanged volume of supplies, which could minimize its losses on the upstream market or keep upstream profits unchanged. BT’s ubiquity would not prevent this as MNOs can and do divide their purchases between suppliers based on areas.

(200)    The move may even be profitable. MNOs’ feedback suggests that active lines are more expensive than passive mines, which may indicate that it is more profitable to supply them. Further, BT’s behaviour of not supplying dark fibre also suggests that, taking all bandwidth brackets together, active lines are more profitable than passive lines. In addition, the market feedback also indicates that if capacity is ramped up with active lines, MNOs would be forced to the more profitable higher bandwidth brackets. Moreover, if it is indeed the case that it is easier to densify the network  with dark fibre, then it is likely that more active lines are needed than passive lines for the same level of densification.

(201)    Naturally, if the move is profitable, the question arises why Virgin Media has not switched to active lines already or why it would not do so in the future even absent the merger. One plausible explanation would be that MNOs have bargaining power, which pre-merger prevented Virgin Media from extracting all the rents it could. Post-merger, however, the merged entity would be much less concerned by an MNO’s threat to walk away.

(202)    In any event, it is at least very plausible that the upstream losses could be minimized by switching to supplying active lines.

(203)    The Commission also notes that under the assumption that Virgin Media already extracts all the rents it could, as the Notifying Parties argue, it would not be true that switching to active lines would make the ramping up of MNOs’ backhaul capacity significantly more expensive. This is because under this assumption Virgin Media would be expected to charge MNOs a dark fibre price that would make the use of dark fibre to upgrade the network only marginally less costly for MNOs than using active lines for the same purpose. This, however, completely contradicts the consistent picture of MNOs’ preference for dark fibre outlined above, which is partly supported by the Notifying Parties themselves. Thus the Notifying Parties’  arguments (there is no scope for cost increases by switching to active  lines  and Virgin Media is already extracting all the rents it could), while consistent, would contradict the equally consistent picture that MNOs prefer dark fibre and that they could upgrade their network with dark fibre at a lower cost. A more detailed investigation appears necessary to clarify the facts and the incentives in this regard. However, the Commission cannot exclude that MNOs’ costs would significantly increase by using active lines.

(204)    As regards the downstream effects, the Commission notes that the timing of introducing 5G and its quality are major differentiators among MNOs. Thus,  delaying or compromising the rollout of 5G by foreclosing access to dark fibre leased lines can have significant impact on the competitive position of a competing MNO.  A significant impact would translate to a significant number of customers being lost and thus potentially sufficient downstream gains to compensate the relatively modest upstream losses, if any losses are incurred anyway.

(205)    If a switch to active lines does not delay or compromise 5G rollout but simply makes it more expensive, the downstream profits depend on the magnitude of cost  increases, the resulting price increases and customer losses by rival MNOs as well as the customer gains by the merged entity. The Notifying Parties submitted that backhaul costs amount to no more than [0-5]% of incremental costs, that part of backhaul costs relate to microwave backhaul (i.e. not all backhaul costs are fibre) and that these costs cannot increase much. By contrast, the market investigation indicates that backhaul costs are a significantly larger share of incremental costs and that a large part of these costs is fibre backhaul.205

(206)    In addition, it is difficult to judge the extent of cost increases in case an MNO is forced to use active lines to upgrade its backhaul. MNO respondents submitted that the source of this increase could be the need to move to a higher bandwidth bracket  or making it more expensive to densify the network. The extent of these cost increases would depend on data demand, the number of lines that would need to be upgraded and the required densification. This is a complicated assessment, which would require a more detailed investigation206 but it cannot be excluded that the cost increases would be  significant.  The  average  UK mobile customer used  2.9 GB of data monthly in 2019 and this data usage could increase dramatically with 5G as a result of more advanced and connected applications, high quality video, internet of things and potentially an FWA home broadband service. Only by 2022, average mobile data consumption could be around 20 GB per month.207 Growth by 2024  could even dwarf these figures. Large data growth in turn implies upgrading many backhaul circuits and building out new backhaul circuits because of cell densification. The limit on MNOs spectrum holdings would force cell densification and if this is achievable at a lower cost with dark fibre, the lack of dark fibre could increase costs substantially. To a certain extent the acknowledgment by the Notifying Parties that MNOs prefer flexible capacity backhaul solutions and that active lines represent a “price risk” is also an acknowledgment that, depending on the evolution of data usage, costs can increase significantly.

(207)    If the share of backhaul costs are not insignificant and the cost increase is substantial, it is plausible that the price increases and the resulting customer churn could lead to enough profits to compensate for the modest losses upstream, provided there are any. Thus it cannot be excluded that the JV would also have the incentive to withdraw dark fibre.

(208)    The Commission also adds that, to the extent some of these additional costs are fixed costs as opposed to incremental costs, then a significant increase in fixed costs would have a chilling effect on the investments necessary for 5G, especially at a time when there is some uncertainty over future demand for these services.

(209)    Finally, the Commission notes that the merger will align BT Openreach’s and Virgin Media’s incentives in relation to the supply of dark fibre to MNOs in that neither will have the incentives to support MNOs with dark fibre. Further, several respondents described the competition in active leased lines as weak and pointed to the existence of coordination.208 The Notifying Parties also pointed out that the market is transparent and that pricing […].209 A withdrawal of dark fibre could reinforce a coordination in Area 2 in active leased lines as both suppliers would supply the same active product and neither would have the incentive to supply dark fibre. This enhanced coordination may be the mere consequence of the withdrawal based on input foreclosure. Alternatively, it may be an additional incentive to withdraw dark fibre. In the latter case it remains important to clarify why this increased coordination would not happen independently of the merger.

Overall effect

(210)    Even if the JV has the ability and the incentive to withdraw dark fibre, the effect on overall competition in the retail mobile market could potentially be modest. The most important factors in this regard are that Area 2 is just part of the national market and that Vodafone shares a network with O2.

(211)    As regards the first factor, as discussed in paragraph (206) above, the effects on prices and investments depend on the extent of cost increases, which in turn depend on the projected data usage and network developments. These developments are difficult to ascertain but Area 2 is large enough so that it cannot be excluded that a significant cost increase in this area would have a negative impact on national pricing or nationwide investments. Moreover, the Commission also noted that the rollout of 5G could be delayed or compromised in Area 2 independently of costs, just because MNOs would be forced to reassess their network plans. In that regard it would already be damaging to the competitive position of an MNO if this were to happen in a large enough part of the national market.

(212)    As regards the second factor, the Notifying Parties claim that the JV would not foreclose Vodafone as they share a network and by compromising its 5G development it would compromise its own. The evidence on the file is not clear in this respect but in any event the foreclosure […] would threaten to affect  significantly competition on the retail mobile telecommunications market.

(d) Conclusion

(213)    Therefore, based on its preliminary analysis and without prejudice to further investigation by the CMA, the Commission considers that the Transaction threatens to significantly affect competition in the retail mobile market in the United Kingdom a result of input foreclosure.210 This potential effect can be cumulated with the horizontal effect described in Section 4.4.3.2 and the effect of potential vertical foreclosure of MVNOs described in 4.4.3.4.

4.4.3.7. Foreclosure of rival MNOs by foreclosing leased lines in the aggregation layer  (input foreclosure)

(a) Market shares

(214)    Market shares are not available for each BT exchange where only Virgin Media and BT are present. However, such market shares are not necessary as in a duopoly Virgin Media would necessarily have market power and it is the only supplier of  dark fibre.

(b) The Notifying Parties’ view

The Notifying Parties did not analyze separately the access and aggregation layers as they consider these two segments to be one market.

(c) Commission’s assessment

(215)    In relation to the BT exchanges where BT and Virgin Media are the only two suppliers, the same assessment applies as in the case of the access layer. First, as the sole supplier of dark fibre, the JV will have the ability to withdraw it and existing contractual protections do not provide full protection in this regard. Second, incentives need to be further investigated but it cannot be excluded that based on minimal upstream losses and potential downstream gains the JV would have incentives to foreclose downstream competitors. Third, while the size of the overall effect is unclear, a significant impact cannot be excluded.

(216)    The Commission also notes that in BT-only exchanges Ofcom already imposed a dark fibre remedy,211  i.e. an obligation to supply dark fibre at a regulated price. As a result of the merger the BT+1 exchanges where the other operator present is Virgin Media would functionally become like BT-only exchanges in relation to MNOs. In addition foreclosure concerns could also arise in relation to exchanges where Virgin Media is one of the two dark fibre suppliers.

(d) Conclusion

(217)    Therefore, for the purpose of this Decision and without prejudice to further investigation by the CMA, the Commission considers that the Transaction threatens to significantly affect competition in the retail mobile market in the United Kingdom as a result of withdrawal of dark fibre in the aggregation layer.212 Although it is unclear whether the effect of foreclosing access to dark fibre in the aggregation layer would be significant on its own, even an insignificant effect would add and reinforce the effect of dark fibre withdrawal in the access layer (Section (180)), the horizontal effect described in Section 4.4.3.2 and the effect of potential vertical foreclosure of MVNOs described in Section 4.4.3.4.

4.4.3.8. Foreclosure of competing leased line providers by foreclosing access to a leased line customer (customer foreclosure)

(a) Market shares

(218)    As discussed in Section 4.4.3.1 O2’s share on the downstream retail mobile market is [30-40]% by subscribers and [30-40]% by revenues. Given that MNOs use access layer fibre leased lines as mobile backhaul, this makes the upstream access layer  fibre leased line market in Area 2 and the upstream aggregation layer markets vertically affected on account of customer foreclosure.

(b) The Notifying Parties’ views

(219)    The Notifying Parties submit that the proportion of O2’s backhaul demand that, absent the Transaction, might have been won by players other than BT and Virgin Media is simply too small in the context of the overall market for such concerns to be credible. Concretely, the Notifying Parties submit that based on available data estimates the size of the retail leased line market to be GBP 1 015 million. This contrasts with a total cost of operating O2's mobile backhaul aggregation and access layers of […] in 2019.213

(c) Commission’s assessment

(220)    The figures presented by the Notifying Parties were not contradicted by the market investigation. Although in line with 5G rollout MNOs’ spend is expected to increase significantly, O2 still represents a small share of overall leased line demand as leased lines are also supplied to fixed telecom providers and end users such as large enterprises. Thus the loss of O2 as a customer is unlikely to have a significant impact on other leased line suppliers. Moreover, outside Virgin Media’s footprint O2 will still have to purchase leased lines.

(d) Conclusion

(221)    Therefore, based on its preliminary analysis and without prejudice to further investigation by the CMA, the Commission considers that the Transaction does not threaten to significantly affect competition in any of the access layer or aggregation layer fibre leased line markets within the UK as a result of customer foreclosure strategies.

4.4.4. Conclusion on the second substantive condition

(222)    In light of the above considerations, following the Commission’s preliminary assessment,214 the Commission concludes that the Transaction threatens to significantly affect competition in the retail mobile market in the UK. The Commission identified the following standalone threats:

·             Foreclosure of MVNOs offering FMC bundles by withdrawing from hosting such MVNOs (input foreclosure described in Section 4.4.3.4.)

·             Foreclosure of rival MNOs by withdrawing dark fibre leased lines in the access layer (input foreclosure described in (Section (180))

(223)    These threats are standalone in a sense each of them alone could threaten to significantly affect competition. However, their potential effect is cumulative.

(224)    In addition, the foreclosure of rival MNOs by withdrawing dark fibre leased lines in the aggregation layer (Section 4.4.3.7) may or may not be a standalone threat but the potential effect adds to the combined effect of the two standalone effects.

(225)    Finally, while the horizontal non-coordinated effects on the retail market (Section 4.4.3.2) would not be significant on their own, they nevertheless contribute to the cumulative effect of all three potential foreclosures above.

(226)    As regards the prima facie competition concerns presented by the CMA in relation to possible foreclosure of (1) rival wholesale call origination and access providers on  the upstream wholesale market for access and call origination on public mobile networks by denying access to a wholesale customer on the retail mobile market (Section 4.4.3.5) and (2) rival leased line access providers upstream by denying access to a leased line customer on the retail mobile market (Section 4.4.3.8) as well as horizontal concerns on the retail fixed internet market (Section 4.4.3.3) the Commission’s finding that the Transaction does not pose a threat is without prejudice to the CMA’s assessment.

4.5. Conclusion on the criteria of Article 9(2)(a) and (3) of the Merger Regulation

(227)    In light of the above, the Commission considers that the legal requirements for a referral laid down in Article 9(2)(a) and (3) of the Merger Regulation are fulfilled with regard to the Transaction.

4.6. The Commission's discretion whether to refer

(228)    Pursuant to Article 9(3) of the Merger Regulation, in the event that the criteria provided for in Article 9(2)(a) are fulfilled with regard to a proposed transaction, the Commission has discretion whether to refer a given case to a national competition authority.

4.6.1. CMA’s submission

(229)    The CMA considers that the factors set out in the Notice that are intended to inform the re-attribution of cases between the Commission and EU Member States are satisfied in the present case.

(230)    First, the CMA is the best placed authority to assess the Transaction for various reasons such as its recent direct experience in the markets covered by the  Transaction, such as the 2016 acquisition by BT of EE (BT/EE), and its knowledge  of the characteristics of the UK and/or local markets. A review by the CMA would also ensure legal certainty with regard to the application of the UK legal test (“substantial lessening of competition”), but it would also ensure continuity in light  of Brexit.

(231)    Second, the effects of the Transaction are wholly UK-centric. The Transaction solely concerns undertakings operating fixed and mobile telecommunications services in  the UK. Any adverse effects resulting from a loss of competition will be felt exclusively by consumers in the UK (and/or regions within) and would not have any impact on competition in markets in any other EU Member States, including on any cross-EU telecommunications initiatives.

(232)    Third, the “one-stop-shop” principle of the EUMR is not applicable for this Transaction as no more than one National Competition Authority is competent to review this case.

(233)    Fourth, legal certainty is not compromised given the fact that from the beginning of this Transaction there were substantial, compelling and clear reasons to depart for the Commission’s original jurisdiction.

4.6.2. The Notifying Parties’ view

(234)    The Notifying Parties consider that the Commission is better placed to review the Transaction, for the following reasons:

(235)    First, the Notifying Parties explain that Brexit is not a factor that should be taken in  to account when assessing the referral request. The Withdrawal Agreement addresses this specific situation by codifying and endorsing the existing system of case referral throughout the transitional period (currently scheduled to expire on 31 December 2020).

(236)    Second, the Notifying Parties point out that the Commission can and routinely does review national markets, and explain that the existence of national markets does not as such justify the referral of transactions to Member States.

(237)    Fourth, according to the Notifying Parties, the Commission has ample telecommunications expertise and therefore the CMA is not “uniquely well placed”  to assess the Transaction. In addition, the Commission has consistently rejected referrals of telecommunications transactions, including in the following cases: Telefónica  Deutschland/E-Plus  (Germany),215   Orange/Jazztel  (Spain),216 Altice/PT Portugal (Portugal),217 Hutchison 3G/Telefónica UK (United Kingdom)218 and Vodafone/Liberty Global/Dutch JV (Netherlands).219

(238)    Fifth, the Notifying Parties explain that a review by the Commission would be in the best interest of UK consumers. The Transaction will create substantial benefits for consumers. It will, in particular, reduce O2’s backhaul and Virgin Mobile’s  wholesale access costs through the elimination of double marginalisation and allow the JV to cross-sell its combined services into the Notifying Parties’ current bases as FMC bundles, thus increasing competition with the incumbent provider BT/EE. A referral to the CMA would inevitably delay these benefits and the achievement of the UK Government’s ultrafast broadband and 5G coverage aims.

4.6.3. Commission’s assessment

(239)    In the following recitals, the Commission assesses the appropriateness of a referral220 in the present case in light of the principles set out in the Referral Notice.

(240)    According to paragraph 9 of the Referral Notice, “[i]n principle, jurisdiction should only be reattributed to another competition authority in circumstances where the latter is more appropriate for dealing with the merger, having regard to the specific characteristics of the case as well as the tools and expertise available to the authority”. The Referral Notice also states that “particular regard should be had to the likely locus of any impact on competition resulting from the merger” and that “[r]egard may also be had to the implications, in terms of administrative effort, of any contemplated referral”.

(241)    Moreover, paragraph 13 of the Referral Notice states that “referral should normally only be made when there is a compelling reason for departing from 'original jurisdiction' over the case in question, particularly at the post-notification stage”.

(242)    Contrary to the Notifying Parties’ view, the Commission considers that there are compelling reasons for departing from the original jurisdiction over the present case, by referring the Transaction to the UK.

(243)    First, Brexit is a unique and exceptional circumstance that inevitably affects the Transaction. The Commission considers that any effects on competition arising as a result of the Transaction would appear exclusively in the UK and would do so only after the expiry of the Transition Period. In particular, the JV would be active in the UK, and several respondents to the market investigation have indicated that, in their view, the Transaction would have a negative impact on these markets in the UK. The Commission retaining jurisdiction could potentially raise a risk of conflicting assessment between this case and subsequent cases in the UK telecoms sector, which will from 2021 fall under the exclusive jurisdiction of the CMA with respect to their effects in UK markets. The CMA applies the legal test of a “substantial lessening of competition” whereas the Commission applies “significant impediment of effective competition” (“SIEC”) test. As the CMA submits, given that the “substantial lessening of competition” test was applied in the last major transaction in this sector in  the  UK  (BT/EE),  and  will  be  applied  in  all  future  transactions,  the  risk of conflicting approaches to assessment would be minimised by the transfer of the case to the CMA. Therefore, the Commission considers that, in light of the fact that the Transaction would only affect the UK and after the expiry of the Transition period, it would be more appropriate that the CMA examine the Transaction in accordance  with its own legal test.

(244)    Second, as to the point raised by the Notifying Parties in relation to the  Commission’s experience in the relevant markets affected by the case, it should be noted that the CMA also has significant knowledge and expertise in relation to the UK telecom markets, in particular through the 2016 BT/EE case. In addition, the CMA is well placed to cooperate with OFCOM that can assist with the review of the Transaction at any point.

(245)    Third, the requested referral will preserve the one-stop-shop principle, as the whole case, including the retail fixed internet access market and the call origination and access on public mobile networks market where the Commission preliminarily did not find a significant threat, will be referred to a single competition authority, which contributes to administrative efficiency. The Commission considers that a fragmentation of the case by which only the retail fixed internet access and the call origination and access on public mobile networks market, the geographic scope of which is also limited to the UK, is examined by the Commission, while the  remainder is examined by the CMA, would be at odds with the one-stop-shop principle.221

(246)    Fourth, it appears likely that any additional administrative effort for the Notifying Parties due to a referral will not be disproportionate. The CMA has followed closely the pre- and post-notification process before the Commission and already had a thorough understanding of the main characteristics of the case and potential competition concerns prior to the filing of its Referral Request.

(247)    Fifth, while the Notifying Parties note that the Commission has in the past  maintained jurisdiction for transactions in the telecom sector, the unique circumstances of Brexit support the Commission’s decision exceptionally to depart from its previous practice and refer the case to the CMA. The present Decision is therefore without prejudice to future referral requests with regard to similar transactions. .

(248)    In light of the above, the Commission considers that the CMA is in the best position to investigate the effects of the Transaction in the UK.

5. CONCLUSION

(249)    From the above it follows that the conditions to request a referral under Article 9(2)(a) Merger Regulation are met. The Commission also considers that, given the local scope of the markets affected by the Transaction, the competent authorities of the UK are better placed to carry out a thorough investigation of the whole case, and that it is therefore appropriate for the Commission to exercise its discretion under Article 9(3)(b) Merger Regulation so as to grant the referral.

HAS ADOPTED THIS DECISION:

Article 1

The notified concentration is referred in its entirety to the competition authority of the UK, pursuant to Article 9(3)(b) of Council Regulation (EC) No 139/2004.

Article 2

This Decision is addressed to the UK.

Done at Brussels,

1                  OJ C115, 9.8.2008, P.47.

2                  OJ L 24, 29.1.2004, p.1. 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.

3                  Form CO, Sections 1-5, paragraph 48.

4                  Form CO, Sections 1-5, paragraph 73.

5                  Form CO, Sections 1-5, paragraph 73.

6                  Form CO, Sections 1-5, paragraph 76.

7                  Form CO, Sections 1-5, paragraph 76.

8                  Form CO, Sections 1-5, paragraph 77.

9                  Form CO, Sections 1-5, paragraph 78.

10                 Form CO, Sections 1-5, paragraph 80 c).

11                 Form CO, Sections 1-5, paragraph 80 d).

12                 See  also  Commission Notice  on Case  Referral  in respect  of concentrations  (hereafter, the "Referral Notice"), OJ C 56, 05.03.2005, p. 2, paragraph 7.

13                 Referral Notice, paragraph 8.

14                 Referral Notice, paragraph 7.

15                 Joined Cases T-346/02 and T-347/02 Cableuropa SA and Others v Commission [2003] EU:T:2003:256, paragraph 105.

16                 With regards to the retail mobile market, see Form CO, Sections 6-8, Chapter 1, paragraph 160. With regards to the retail fixed internet access market, see Form CO, Section 6-8, Chapter 3, paragraph 110. With regards to the wholesale call origination and access market on public mobile networks, see Form CO, Sections 6-8, Chapter 2, paragraph 114. With regards to the wholesale leased lines market, see Form CO, Sections 6-8, Chapter 5, paragraph 213.

17                 “Response to the CMA’s referral request” of 12 October 2020.

18                 CMA Referral Request, recital 25.

19                 “Response to the CMA’s referral request” of 12 October 2020, page 13.

20                 The CMA also referred to the retail voice and TV markets (see recital    (23) above). In relation to fixed voice services, Virgin Media has a share of approximately [10-20]%, to which O2 adds an increment of around [0-5]%. The JV’s share will therefore remain well below 20% and, accordingly, no horizontally affected market arises. In relation to TV services, O2 does not supply retail TV services. Virgin   Media has a market share of approx. [20-30]% in terms of revenues and [20-30]% in terms of subscribers. In view of the lack of horizontal overlap between O2 and Virgin Media, there is no horizontally affected market.

21                 Both Virgin Media and O2 provide retail business connectivity services falling within a  national “retail business connectivity” market identified in Commission precedents (e.g. Commission decision of 18 July 2019 in case M.8864 - Vodafone / Certain Liberty Global Assets). However, due to the low market shares of the Parties on this market, this market is not affected and the Commission will therefore not discuss it separately.

22                 Commission decision of 10 October 2014 in Case M.7000 – Liberty Global/Ziggo, recital 137.

23                 Commission  decisions  of  8  October  2018  in  case  M.8842  –  Tele2/Com  Hem,  recital  10;  of 11 May 2016 in case M.7612 – Hutchison 3G UK/Telefónica UK, recital 252; of 10 October 2014 in case M.7000 – Liberty Global/Ziggo, recital 141; of 2 July 2014 in case M.7018 – Telefónica Deutschland/E-Plus, recital 64.

24                 Commission  decisions  of  8  October  2018  in  case  M.8842   –  Tele2/Com  Hem,  recital        47;  of 1 September 2016 in case M.7758 – Hutchison 3G Italy/Wind/JV, recitals 149 and 161; of 3 August 2016 in case M.7978 – Vodafone/Liberty Global/Dutch JV, recital 74; of 11 May 2016 in case  M.7612 - Hutchison 3G UK/Telefónica UK, recitals 255, 261, 270, 279, 287; of 2 July 2014 in case M.7018 –Telefónica Deutschland/E-Plus, recitals 31 to 55; of 10 October 2014 in case M.7000 – Liberty Global/Ziggo, recital 141; of 28 May 2014 in case M.6992 – Hutchison 3G UK/Telefónica Ireland, recital 141; of 12 December 2012 in case M.6497 – Hutchison 3G Austria/Orange Austria, recital 58.

25                 Form CO, Sections 6-8, Chapter 1, paragraphs 150, 151, and 154.

26                 Form CO, Sections 6-8, Chapter 1, paragraph 143.

27                 Form CO, Sections 6-8, Chapter 1, paragraph 146.

28                 Form CO, Sections 6-8, Chapter 1, paragraphs 155-157.

29                 Electronic communication services provided over the internet such as messaging apps and    email (also known as 'over-the-top' or 'OTT' services).

30                 Form CO, Sections 6-8, Chapter 1, paragraph 158.

31                 See  Replies  to  Q1  –  mobile  network  operators,  question  12  and  Replies  to  Q2  –  fixed  network operators, question 12.

32                 Commission decision of 15 July 2019 in case M.9370, Telenor/DNA; Commission decision of   27 July 2018 in case M.8883, PPF/Telenor Target Companies; Commission decision of 27 November 2018 in case M.8792, T-Mobile NL/Tele2 NL.

33                 Form CO, Sections 6-8, Chapter 1, paragraph 160.

34                 See  Replies  to  Q1  –  mobile  network  operators,  question  39  and  Replies  to  Q2  –  fixed  network operators, question 39.

35                 Commission decision of 27 November 2018 in Case M.8792 – T-Mobile NL/Tele2 NL, recital 235.

36                 Commission decision of 12 December 2012 in Case M.6497 – Hutchison 3G Austria/Orange Austria, recitals 34-35; of 2 July 2014 in Case M.7018 – Telefonica Deutschland/E-Plus, recitals 34-35; of 11 May 2016 in Case M.7612 - Hutchison/O2, recitals 278-279; of 1 September 2016 in Case M.7758 – Hutchison 3G Italy/WIND/JV, recitals 155 – 157; of 28 May 2014 in Case M.6992 – Hutchison 3G UK/Telefonica Ireland, recital 149; and of 27 November 2018 in Case M.8792 - T-Mobile NL/Tele2 NL, recitals 181 – 190.

37                 Commission decision of 27 November 2018 in Case M.8792 – T-Mobile NL/Tele2 NL, recital 236.

38                 Form CO, Sections 6-8, Chapter 2, paragraph 113.

39                 See  Replies  to  Q1  –  mobile  network  operators,  question  27  and  Replies  to  Q2  –  fixed  network operators, question 27.

40                 Commission decision of 12 December 2012 in Case    M.6497 – Hutchison 3G Austria/Orange Austria, recitals 34-35; of 2 July 2014 in Case M.7018 – Telefonica Deutschland/E-Plus, recitals 34-35; of 11 May 2016 in Case M.7612 - Hutchison/O2, recitals 278-279; of 1 September 2016 in Case M.7758 – Hutchison 3G Italy/WIND/JV, recitals 155 – 157; of 28 May 2014 in Case M.6992 – Hutchison 3G UK/Telefonica Ireland, recital 149; and of 27 November 2018 in Case M.8792 - T-Mobile NL/Tele2 NL, recitals 181 – 190.

41                 Commission decision of 27 November 2018 in Case M.8792 – T-Mobile NL/Tele2 NL, recital 240.

42                 Form CO, Sections 6-8, Chapter 2, paragraph 114.

43                 See  Replies  to  Q1  –  mobile  network  operators,  question  40  and  Replies  to  Q2  –  fixed  network operators, question 40.

44                 Both Parties are also active on the fixed voice services market and Virgin Media is also active on the retail TV services market. Neither market is an affected market, therefore, the Commission will not discuss these markets in this decision. See Form CO, Chapter 3, paragraphs 6-7.

45                 Commission decision of 18 July 2019 in case M.8864 – Vodafone/Certain Liberty Global Assets, recital 47.

46                 Commission  decisions  of  8  October  2018  in  case  M.8842  -  Tele2/Com  Hem,  paragraph    26;  of 7 October 2016 in case M.8131 - Tele2 Sverige/TDC Sverige, recital 32; of 19 May 2015 in case M.7421 - Orange/Jazztel, recital 42; of 10 October 2014 in case M.7000 - Liberty Global/Ziggo, recital 132.

47                 Commission decisions of 12 November 2019 in case M.9064    – Telia Company/Bonnier Broadcasting Holding,  recital  218;  of  8  October  2018  in  case  M.8842  –  Tele2/Com  Hem,  paragraph  26;  of   3 August 2016 in case M.7978 – Vodafone/Liberty Global/Dutch JV, recital 38; of 20 September 2013 in case M.6990 - Vodafone/Kabel Deutschland, recital 194.

48                 Commission decision of 18 July 2019 in case M.8864 – Vodafone/Certain Liberty Global Assets, recital 51.

49                 Commission decision of 18 July 2019 in case M.8864 – Vodafone/Certain Liberty Global Assets, recital 56.

50                 Small-office/home office.

51                 Small- and medium-sized enterprises.

52                 Form CO, Sections 6-8, Chapter 3, paragraph 106.

53                 Form CO, Sections 6-8, Chapter 3, paragraph 107.

54                 See  Replies  to  Q1  –  mobile  network  operators,  question  15  and  Replies  to  Q2  –  fixed  network operators, question 15.

55                 See Replies to Q1 – mobile network operators, questions 16-17 and Replies to Q2 – fixed network operators, questions 16-17.

56                 See Replies to Q1 – mobile network operators, questions 13-17 and Replies to Q2 – fixed network operators, questions 13-17.

57                 See footnote 20.

58                 See Replies to Q1 – mobile network operators, questions 13-14 and Replies to Q2 – fixed network operators, questions 13-14.

59                 See Replies to Q1 – mobile network operators, questions 13-14 and Replies to Q2 – fixed network operators, questions 13-14.

60                 See Replies to Q1 – mobile network operators, questions 13-14 and Replies to Q2 – fixed network operators, questions 13-14.

61                 See Replies to Q1 – mobile network operators, questions 13-14 and Replies to Q2 – fixed network operators, questions 13-14.

62                 See Replies to Q1 – mobile network operators, questions 13-14 and Replies to Q2 – fixed network operators, questions 13-14.

63                 See Replies to Q1 – mobile network operators, questions 13-14 and Replies to Q2 – fixed network operators, questions 13-14.

64                 Consequently, as the market for FWA services would not be affected, the Commission will  not discuss it further.

65                 Commission decisions of 12 November 2019 in case M.9064    – Telia Company/Bonnier Broadcasting Holding, recital 239; of 8 October 2018 in case M.8842 – Tele2/Com Hem; of 3 August 2016 in case M.7978 – Vodafone/Liberty Global/Dutch JV, recital 40; of 20 September 2013 in case M.6990 - Vodafone/Kabel Deutschland, recital 197.

66                 Form CO, Section 6-8, Chapter 3, paragraph 110.

67                 See Replies to Q1 – mobile network operators, questions 41-42 and Replies to Q2 – fixed network operators, questions 41-42.

68                 See Replies to Questionnaire Q2 – fixed network operators, question 41.

69                 See  Replies  to  Q1  –  mobile  network  operators,  question  41  and  Replies  to  Q2  –  fixed  network operators, question 41.

70                 Commission decisions of 8 October 2018 in case M.8842 – Tele2/Com Hem, recital 60; of 3 August 2016 in case M.7978 – Vodafone/Liberty Global/Dutch JV, recital 108; of 4 February 2016 in case M.7637 – Liberty Global/BASE Belgium, recital 96; of 19 May 2015 in case M.7421 – Orange/Jazztel, recitals 86 and 91.

71                 Commission decisions of 8 October 2018 in case M.8842 – Tele2/Com Hem, recital; of 3  August 2016 in case M.7978 – Vodafone/Liberty Global/Dutch JV, recital 107.

72                 Form CO, Sections 6-8, Chapter 3, paragraphs 135-138.

73                 Form CO, Sections 6-8, Chapter 3, paragraph 136.

74                 Form CO, Sections 6-8, Chapter 3, paragraph 137.

75                 Form CO, Sections 6-8, Chapter 3, paragraph 137.

76                 See Replies to Q1 – mobile network operators, questions 18-26 and Replies to Q2 – fixed network operators, questions 18-26.

77                 See  Replies  to  Q1  –  mobile  network  operators,  question  24  and  Replies  to  Q2  –  fixed  network operators, question 24.

78                 See  Replies  to  Q1  –  mobile  network  operators,  question  19  and  Replies  to  Q2  –  fixed  network operators, question 19.

79                 See  Replies  to  Q1  –  mobile  network  operators,  question  20  and  Replies  to  Q2  –  fixed  network operators, question 20.

80                 See  Replies  to  Q1  –  mobile  network  operators,  question  18  and  Replies  to  Q2  –  fixed  network operators, question 18. See also Ofcom Communications Market Report 2018, Figure 1.5, at https://www.ofcom.org.uk/    data/assets/pdf file/0022/117256/CMR-2018-narrative-report.pdf.

81                 See Replies to Q1 – mobile network operators, questions 25-25.2 and Replies to Q2 – fixed network operators, questions 25-25.2.

82                 Commission  decisions  of  3  August  2016  in  case  M.7978  –  Vodafone/Liberty    Global/Dutch  JV, paragraphs 112; of 19 May 2015 in case M.7421 - Orange/Jazztel, recitals 89-90; of 10 October 2014 in case M.7000 - Liberty Global/Ziggo, recitals 152-153; of 20 September 2013 in case M.6990 - Vodafone/Kabel Deutschland, paragraphs 263-265; of 16 June 2011 in case M.5900 - LGI/KBW, paragraphs 183-186.

83                 Form CO, Sections 6-8, Chapter 3, paragraph 138.

84                 See  Replies  to  Q1  –  mobile  network  operators,  question  43  and  Replies  to  Q2  –  fixed  network operators, question 43.

85                 Form CO, Section 6-8, Chapter 5, paragraph 11.

86                 Form CO, Section 6-8, Chapter 5, paragraph 12.

87                 M.9674 Vodafone Italia/Tim/Inwit JV, paragraphs 124-129.

88                 M.8864 Vodafone/Certain Liberty Global Assets paragraphs 198 - 201; M.8808 T-Mobile Austria/UPC Austria (T-Mobile/UPC), paragraph 90. M.6584 - Vodafone/Cable & Wireless, paragraph 30; M.7637 - Liberty Global/BASE Belgium, recital 146; M.7109 - Deutsche Telecom/GTS, paragraph 70;

89                 Form CO, Section 6-8, Chapter 5, paragraph 181

90                 Form CO, Section 6-8, Chapter 5, paragraph 176-177.

91                 Form CO, Section 6-8, Chapter 5, paragraph 186.

92                 Form CO, Section 6-8, Chapter 5, paragraphs 186-187.

93                 Form CO, Section 6-8, Chapter 5, paragraphs 192.

94                 Form CO, Section 6-8, Chapter 5, paragraphs 193.

95                 CMA, BT/EE Final Report, paragraphs 15.41 and 15.43.

96                 Form CO, Section 6-8, Chapter 5, paragraphs 204.

97                 Questionnaire Q1 to mobile operators, question 28; Questionnaire Q2 to fixed operators, question 28.

98                 Questionnaire Q1 to mobile operators, question 28; Questionnaire Q2 to fixed operators, question 28.

99                 Questionnaire Q1 to mobile operators, question 28; Questionnaire Q2 to fixed operators, question 28.

100               Questionnaire Q1 to mobile operators, question 28; Questionnaire Q2 to fixed operators, question 28.

101               Questionnaire Q1 to mobile operators, question 29; Questionnaire Q2 to fixed operators, question 29.

102               Questionnaire Q1 to mobile operators, question 30; Questionnaire Q2 to fixed operators, question 30.

103               Questionnaire Q1 to mobile operators, question 31; Questionnaire Q2 to fixed operators, question 31.

104               Questionnaire Q1 to mobile operators, question 31; Questionnaire Q2 to fixed operators, question 31.

105               Questionnaire Q1 to mobile operators, question 32; Questionnaire Q2 to fixed operators, question 32.

106               See the responses […] Questionnaire Q2 to fixed operators, question 32.

107               […] Questionnaire Q2 to fixed operators, question 32.

108               […] Questionnaire Q1 to mobile operators, question 32.

109               Questionnaire Q1 to mobile operators, question 33; Questionnaire Q2 to fixed operators, question 33.

110               Questionnaire Q1 to mobile operators, question 34; Questionnaire Q2 to fixed operators, question 34.

111               Questionnaire Q1 to mobile operators, question 32; Questionnaire Q2 to fixed operators, question 32.

112               Questionnaire Q1 to mobile operators, question 32; Questionnaire Q2 to fixed operators, question 32.

113               Questionnaire Q1 to mobile operators, question 33; Questionnaire Q2 to fixed operators, question 33.

114               Parties response to the State of Play meeting, submitted on 22 October 2020.

115               Parties response to the State of Play meeting, submitted on 22 October 2020.

116               Form CO Sections 6-8, Chapter 5, paragraph 23.

117               CMA, BT/EE Final Report, paragraphs 15.41 and 15.43.

118               See above the potential distinction between active and passive lines.

119               Questionnaire Q1 to mobile operators, question 37; Questionnaire Q2 to fixed operators, question 37.

120               Questionnaire Q1 to mobile operators, question 37; Questionnaire Q2 to fixed operators, question 37.

121               Questionnaire Q1 to mobile operators, question 37; Questionnaire Q2 to fixed operators, question 37.

122               Commission Notice on the definition of relevant market for the purposes of Community competition law (97/C 372 /03) (Market Definition Notice). See https://eur-lex.europa.eu/legal- content/EN/TXT/PDF/?uri=CELEX:31997Y1209(01)&from=EN.

123               Market Definition Notice, paragraph 43.

124               Form CO, Section 6-8, Chapter 5, paragraph 57.

125               […] Questionnaire Q1 to mobile operators, question 37.1

126               Questionnaire Q1 to mobile operators, question 35; Questionnaire Q2 to fixed operators, question 35.

127               Questionnaire Q1 to mobile operators, question 36; Questionnaire Q2 to fixed operators, question 36.

128               Questionnaire Q1 to mobile operators, question 36; Questionnaire Q2 to fixed operators, question 36.

129               Questionnaire Q1 to mobile operators, question 36; Questionnaire Q2 to fixed operators, question 36.

130               Form CO, Section 6-8, Chapter 5, paragraphs 211-213.

131               Questionnaire Q1 to mobile operators, question 44; Questionnaire Q2 to fixed operators, question 44.

132               […] Questionnaire Q2 to fixed operators, question 44.

133               Questionnaire Q1 to mobile operators, question 44; Questionnaire Q2 to fixed operators, question 44.

134               Duopoly on the condition that active lines and dark fibre are in one market as BT Openreach does not supply dark fibre – see the section on product market definition.

135               Promoting competition and investment in fibre networks: Wholesale Fixed Telecoms Market Review 2021-26, Volume 2, paragraphs 7.91-93. https://www.ofcom.org.uk/    data/assets/pdf file/0029/188822/wftmr-volume-2-market-assessment.pdf

136               WFTMR, Volume 2, Table 7.1.

137               WFTMR, Volume 2,  Table 7.2. The figures only represent the Multi-Service  Networks    (MSNs) and disregard leased line-only networks. However, leased lines only networks only affect the competitive situation marginally such that Ofcom did not modify the areas drawn based on MSNs (see paragraph 7.75)

138               Questionnaire Q1 to mobile operators, question 36; Questionnaire Q2 to fixed operators, question 36.

139               https://www.ofcom.org.uk/consultations-and-statements/category-1/review-physical-infrastructure-and- business-connectivity-markets

140               BCMR, Volume 2, paragraphs 7.74 - 7.77 https://www.ofcom.org.uk/    data/assets/pdf file/0025/154591/volume-2-bcmr-final-statement.pdf

141               BCMR, Volume 2, Table 8.1

142               CMA Referral Request, paragraph 11.

143               CMA Referral Request, paragraph 12.

145               WFTMR, Volume 2, Table 8.3 (Service shares).

146               Response to CMA Referral Request, page 9.

147               Response to CMA Referral Request, page 9 and Form CO, Sections 6-8, Chapter 1, paragraphs 3 and 7.

148               Response to CMA Referral Request, page 9 and Form CO, Sections 6-8, Chapter 1, paragraphs 3 and 8.

149               Form CO, Sections 6-8, Chapter 1, paragraphs 216-218 and Annex Ch1.4.3.

150               Response to CMA Referral Request, page 9.

151               Form CO, Sections 6-8, Chapter 1, Figure 1.14 and paragraph 201.

152               Form CO, Sections 6-8, Chapter 1, Table 1.10.

153               Form CO, Sections 6-8, Chapter 1, recital 201.

154               See https://www.o2.co.uk/abouto2/company-history.

155               Form CO, Sections 6-8, Chapter 1, recital 176.

156               Form CO, Sections 6-8, Chapter 1, recitals 191-196 and Annex Ch1.4.2.

157               See also Replies to Q1 – mobile network operators, question 46 and Replies to Q2 – fixed network operators, question 46.

158               Commission decision of 11 May 2016 in case M.7612 – Hutchison 3G UK/Telefónica UK, recital 1175.

159               See also Replies to Q1 – mobile network operators, questions 46-51.1 and 76.1.1 and Replies to Q2 – fixed network operators, question 46-51.1 and 78.1.1.

160               See also Replies to Q1 – mobile network operators, questions 49.1 and Replies to Q2 – fixed network operators, question 49.1.

161               Form CO, Sections 6-8, Chapter 3, Table 3.2.

162               Form CO, Sections 6-8, Chapter 3, Table 3.2.

163               Referral Notice, recital 50.

164               Form CO, Sections 6-8, Chapter 3, paragraph 4.

165               Form CO, Sections 6-8, Chapter 3, paragraph 5.

166               Form CO, Sections 6-8, Chapter 3, paragraph 6.

167               Form CO, Sections 6-8, Chapter 3, paragraph 7.

168               Form CO, Sections 6-8, Chapter 3, paragraph 7.

169               Form CO, Sections 6-8, Chapter 3, paragraph 9.

170               Form CO, Sections 6-8, Chapter 3, paragraphs 10-21.

171               Form CO, Sections 6-8, Chapter 3, paragraph 11.

172               Form CO, Sections 6-8, Chapter 3, paragraph 12.

173               Form CO, Sections 6-8, Chapter 3, paragraphs 14-17.

174               Questionnaire Q2 to fixed operators, question 54.

175               Questionnaire Q2 to fixed operators, question 54. One respondent explained that the history of   the UK mobile market demonstrates that MNOs have incentives to foreclose all MVNOs, not just those offering FMC bundles. Each additional MNO increases the competition in the retail mobile market, and as all MNOs are vertically integrated into retail operations, there are incentives to foreclose. Given the lack of importance of FMCs in the UK communications sector, there is unlikely to be any additional incentive to foreclose due to a particular MVNO offering FMC bundles.

176               Questionnaire Q2 to fixed operators, question 56, Questionnaire Q1 to mobile operators, question 56.

177               Questionnaire Q2 to fixed operators, question 56.1.

178               Notifying  Parties’  presentation  of  4  August  2020  “Incentives  to  foreclose  access  to O2’s Mobile Network”, page 9.

179               Notifying  Parties’  presentation  of  4  August  2020  “Incentives  to  foreclose  access  to O2’s Mobile Network”

180               Questionnaire Q2 to fixed operators, questions 61-63, Questionnaire Q1 to mobile  operators, questions 61-63.

181               Questionnaire Q2 to fixed operators, questions 60, Questionnaire Q1 to mobile operators, questions 60.

182               Questionnaire Q2 to fixed operators, questions 59, Questionnaire Q1 to mobile operators, questions 59.

183               Form CO, Section 6-8, Chapter 2, paragraphs 20 and 232-233.

184               Form CO, Section 6-8, Chapter 2, paragraph 20.

185               Questionnaire Q1 to mobile operators, question 63; Questionnaire Q2 to fixed operators, question 63.

186               Referral Notice, recital 50.

187               Form CO, Section 6-8, Chapter 5, paragraph 257.

188               Form CO, Section 6-8, Chapter 5, paragraph 258. SSE was chosen for the aggregation layer. The reason for its inclusion here is that the Notifying Parties did not distinguish the access and aggregation layers  as separate markets.

189               Form CO, Section 6-8, Chapter 5, paragraph 258.

190               Parties response to RFI 5, question 4.

191               Form CO, Section 6-8, Chapter 5, paragraph 258.

192               Form CO, Section 6-8, Chapter 5, paragraph 258.

193               Parties response to the State of Play meeting, submitted on 22 October 2020, section 2.

194               Parties response to the State of Play meeting, submitted on 22 October 2020, section 3.

195               Parties response to the State of Play meeting, submitted on 22 October 2020, section 5.

196               Parties response to the State of Play meeting, submitted on 22 October 2020, section 6

197               Parties response to the State of Play meeting, submitted on 22 October 2020, section 7.

198               Questionnaire Q1 to mobile operators, question 32; Questionnaire Q2 to fixed operators, question 32.

199               Response of an MNO to Commission inquiry submitted on 22 October by email.

200               Section 6-8 Chapter 5, paragraph 21.

201               Form CO, Section 6-8, Chapter 5, paragraph 67.

202               See paragraph (90)

203               Questionnaire Q1 to mobile operators, question 74; Questionnaire Q2 to fixed operators, question 74.

204               From CO, Section 6-8 Chapter 5, paragraph 94; Parties’ Response to State of Play Meeting,  paragraphs 3.2. and 5.3.

205               Questionnaire Q1 to mobile operators, question 68; Questionnaire Q2 to fixed operators, question 68.

206               Unless contrary to the investigation results it was not true that with active lines MNOs are forced to higher brackets and that it is cheaper to densify a network with dark fibre.

207               https://5g.co.uk/guides/how-much-data-does-5g-use/

208               Questionnaire Q2 to fixed operators, question 73.

209               From CO, Section 6-8 Chapter 5, paragraph 94; Parties’ Response to State of Play Meeting,  paragraphs

3.2. and 5.3.

210               The same assessment would apply regardless of whether dark fibre and active lines are in one market or they fall into separate markets. In both cases MNOs prefer dark fibre and Virgin Media would be the only supplier of dark fibre in Area 2. Thus Virgin Media would have the ability to withdraw dark fibre and this would involve the same losses and downstream gains i.e. the same assessment of the incentives with the same caveats. The effect on the retail mobile market would be the same. The assessment is also similar whether or not the market is restricted to MNOs.

211               Form CO, Sections 6-8, chapter 5, paragraph 211.

212               The same assessment would apply regardless of whether dark fibre and active lines are in one market or they fall into separate markets. In both cases MNOs prefer dark fibre and Virgin Media would be the only supplier of dark fibre in Area 2. Thus Virgin Media would have the ability to withdraw dark fibre and this would involve the same losses and downstream gains i.e. the same assessment of the incentives with the same caveats. The effect on the retail mobile market would be the same.

213               Form CO, Sections 6-8, chapter 5, paragraph 314.

214               Commission Notice on Case Referral in respect of concentrations, recital 35.

215               Case COMP/M.7018, Commission decision of 30 January 2014.

216               Case COMP/M.7421, Commission decision of 26 January 2015.

217               Case COMP/M.7499, Commission decision of 4 April 2015.

218               Case COMP/M.7612, Commission decision of 4 December 2015.

219               Case COMP/M.7978, Commission decision of 3 August 2016.

220               The arguments set out in this section apply to all markets examined in this Decision (Section 4.3.3), therefore the Commission considers that a full referral of the case to the CMA is appropriate in this case.

221               Referral Notice, recital 12 and footnote 16.