ME/6983/22– MICROSOFT/ACTIVISION BLIZZARD
MICROSOFT’S RESPONSE TO THE CMA’S REFERENCE DECISION
This initial submission is made by Microsoft in response to the CMA’s decision to refer
Microsoft’s proposed acquisition of Activision Blizzard (the “Transaction”) for a Phase 2
review, as announced on 1 September 2022 (the “Referral Decision”).
- 1. Executive Summary
- 2.Background and Transaction rationale
- 3.Theory of Harm 1 – no prospect of a substantial lessening of competition through foreclosure of Sony
- A. Sony has had a market leading position for over 20 years
- B. No ability to foreclose Sony
- C.No incentive to withdraw Call of Duty
- (i) Microsoft’s incentives are evidenced by its internal documents and approachfollowing previous acquisitions
- (ii) The CMA cannot base its assessment on an assertion that Xbox would willingly incur substantial losses
- (iii)The Referral Decision fails to take account of cross-platform play
- (iv) Contractual protections provide a powerful disincentive to withdraw content
- (v) Flawed assessment of the Microsoft’s financial modelling
- D. Partial foreclosure is an even weaker theory of harm
- D. No anticompetitive effect
- 4.Theory of Harm 1(b) – no prospect of a substantial lessening of competition through foreclosure of multi-game subscription services
- A. Theory of Harm 2 is novel and without precedent
- B. Cloud gaming is a new and unproven technology
- C. Cloud gaming will lower barriers to entry and network effects
- D. No existing or prospective advantage for Microsoft in cloud gaming owing to its cloud and OS assets
- E. No ability to foreclose rivals through foreclosure of gaming content
- F. No ability or incentive to foreclose access to cloud infrastructure to cloud gaming rivals
- G. No ability or incentive to disadvantage cloud gaming rivals through Windows licensing policies
- H. No anticompetitive effect
1. Executive Summary
1.1 Microsoft’s proposed acquisition of Activision Blizzard will give more people more
choices for how to play games and give developers more choices for how to reach those
people. It reflects Microsoft’s commitment to a future for gaming that expands beyond
consoles in which a diversity of business models, distribution mechanisms and payment
options are available. Microsoft’s gaming business, Xbox, is working to expand choice
in two ways: through Game Pass, a subscription service that lets gamers enjoy the best
of what Xbox has to offer at an affordable price, and through bringing more games to
mobile platforms, including through mobile apps and cloud-based game streaming
technology. The acquisition of Activision Blizzard provides Microsoft with capabilities
and content on mobile, which it currently lacks, while creating new distribution options
for game developers outside of mobile app stores. It also adds new content for Game
Pass subscribers, encouraging them to engage with the service on the device of their
1.2 Notwithstanding the benefits to gamers and developers, the CMA has referred the
Transaction for a Phase 2 review based on potential harms to competitors. First, the
CMA is concerned that the Transaction may harm Sony, the market-leading provider
of PlayStation consoles. The CMA suggests that Microsoft may foreclose competition
in console gaming platforms or multi-game subscription services by withholding
Activision Blizzard content – in particular Call of Duty – from PlayStation. Second,
the CMA has concerns that Microsoft may use Activision Blizzard content either alone
or in combination with its wider “ecosystem” to out-compete rivals such as Google,
Amazon and Nvidia. In a novel theory of harm unsupported by precedent, economic
literature or the evidence, the CMA alleges that a potential increase in network effects
and barriers to entry – without evidence or quantification of any effect – is sufficient to
foreclose major technology companies from cloud gaming services.
1.3 These unsupported theories of harm are not sufficient to justify a reference to Phase 2.
Microsoft is confident that following an in-depth review the CMA will conclude that
there is no prospect that the Transaction will give rise to a substantial lessening of
competition in any market based on Phase 2’s higher “balance of probabilities”
threshold. In summary:
(a)The CMA’s theories of harm relate to one overarching concern: that ActivisionBlizzard’s game catalogue – in particular the Call of Duty franchise – will enableXbox to foreclose its competitors in gaming markets. This concern is misplaced.The Referral Decision fails to recognise the incredible array of popular anddiverse gaming content that is available to market participants and overstatesthe importance of Activision Blizzard’s content to competition in gaming.
(b)Xbox plans to make the Call of Duty franchise available to more gamers in moreways than would have been the case in the counterfactual. While this increasedcompetition will benefit gamers, it has not been welcomed by the market leaderSony. In Theory of Harm 1(a), the CMA adopts Sony’s complaints without theappropriate level of critical review. Indeed, industry commentators have notedthat the Referral Decision is “written like they want to protect Sony’s #1 positionin gaming, while claiming there isn’t much competition to Call of Duty (?!)”.1
(c)The CMA’s concerns are misplaced for the following reasons:
Sony PlayStation has been the largest console platform for over 20 years,with an installed base of over 150 million consoles making it larger thanNintendo and more than double the size of Xbox.2
Sony engages in conduct today which is reflective of its market powerin console gaming, including increasing prices of its consoles withoutfear of losing market share [✂].3
The suggestion that the incumbent market leader, with clear andenduring market power, could be foreclosed by the third largest provideras a result of losing access to one title is not credible. There are morethan 4,000 games available on PlayStation alone.
The evidence shows that less than [✂] of PlayStation’s monthly activeusers (“MAUs”) are playing Call of Duty. Even without all of thosegamers (a highly improbable outcome from a hypothetical foreclosurestrategy), the PlayStation gamer base would remain significantly largerthan Xbox is today.
Since the Transaction was announced, Sony has acquired several gamestudios – including Bungie, developer of the popular online game Destiny2, Haven Studios, Lasengle and Savage Games – and a minority interestin FromSoftware, the developer of the biggest game of 2022, Elden Ringand other hit games. This complements Sony’s existing minorityshareholding in Epic Games, publisher of Fortnite, strong first-partygame catalogue and extensive portfolio of exclusive arrangements with third-party publishers. There were over 280 exclusive first- and third-party titles on PlayStation in 2021, nearly five times as many as on Xbox.
(d)In short, Sony is not vulnerable to a hypothetical foreclosure strategy, and theReferral Decision incorrectly relies on self-serving statements by Sony whichsignificantly exaggerate the importance of Call of Duty to it and neglect toaccount for Sony’s clear ability to competitively respond. The CMA’sassessment of this theory ignores its acknowledgement in the Referral Decisionthat the gaming industry is “dynamic”. While Sony may not welcome increasedcompetition, it has the ability to adapt and compete. Gamers will ultimatelybenefit from this increased competition and choice.
(e)In any event, Microsoft has no intention to take Call of Duty away from gamersand, indeed, it has publicly committed not to do so. The value of Call of Dutydepends on its community of gamers, the majority of whom are on PlayStation.Keeping Call of Duty on PlayStation is, therefore, a commercial imperative forthe Xbox business and the economics of the Transaction. As such, Microsofthas offered Sony a contractual commitment to continue supplying it with Callof Duty, including new releases with feature and content parity. The ReferralDecision fails to explain why, in the CMA’s view, Microsoft would make suchcommitments publicly and privately, if it had no intention of honouring them.Microsoft would not do this. Foreclosure strategies of the type outlined in theReferral Decision would alienate the Call of Duty gamer base and tarnish boththe Call of Duty and Xbox brands, undermining the rationale for the Transaction.Microsoft would place at risk over USD [✂] in annual revenue from sales ofCall of Duty on PlayStation, as well as substantial revenues from other Xboxgames distributed via PlayStation. Microsoft has been clear that it is countingon revenues from the distribution of Activision Blizzard games on SonyPlayStation as part of its business case for the acquisition.
(f)Xbox plans to bring Activision Blizzard content to its multi-game subscription,Game Pass. This will benefit gamers, boosting the value of the subscription andexpanding access to Activision Blizzard content. Activision Blizzard has notallowed its content to be included in third-party subscriptions in any meaningfulway in the past [✂]. Again, this increased competition has not been welcomedby the market leader Sony, which has elected to protect its revenues from salesof newly released games, rather than offer gamers the choice of accessing themvia its subscription, PlayStation Plus.
(g)In Theory of Harm 1(b), the CMA again adopts Sony’s complaints withoutconsidering the potential harm to consumers.
The Referral Decision relies on an assumption that Activision Blizzard’scontent strategy would change in the future as subscription servicesbecome more popular. There are not facts—anywhere—to support thisassertion.
Nor is there any basis for the idea that acquiring Call of Duty could ‘tip’subscription services in Xbox’s favour. Sony has chosen to block GamePass from PlayStation, so it is not available on PlayStation. As all gamesthat are available on Game Pass are also available to purchase, PlayStation gamers will continue to have the ability to buy Call of Duty on PlayStation. And doing so will still cost less than the cost of switching by buying a new Xbox console.
Multi-game subscriptions offer consumers an additional payment option,but there is substantial evidence that customers switch between paymentoptions, with subscription being constrained by buy-to-play and free-to-play games. Accordingly, Game Pass will continue to face significantcompetition across the board, and gamers will continue to havealternative purchasing options post-Transaction.
Should any consumers decide to switch from a gaming platform thatdoes not give them a choice as to how to pay for new games (PlayStation)to one that does (Xbox), then that is the sort of consumer switchingbehavior that the CMA should consider welfare enhancing and indeedencourage. It is not something that the CMA should prevent.
(h)Theory of Harm 2 is novel and without precedent. It suffers from a number offundamental flaws. First, the CMA alleges that Microsoft has an advantage overrivals in cloud gaming by having a broad “multi-product ecosystem”, includinga leading cloud platform and PC operating system (“OS”). However,Microsoft’s cloud gaming service Xbox Cloud Gaming does not use either ofthese offerings. It uses bespoke infrastructure – not provided by Azure – anddoes not stream games from PC hardware. There is no advantage. Indeed, Xboxconsiders that it faces a number of significant disadvantages as compared to rivalproviders of infrastructure for game streaming.
(i)Second, the CMA alleges that Microsoft could leverage this “multi-productecosystem” together with Activision Blizzard’s gaming content to strengthennetwork effects, raise barriers to entry, and hence foreclose rivals in cloudgaming services. This again makes no sense.
The CMA provides no evidence that rivals rely on Microsoft’s “multi-product ecosystem” for cloud gaming. The emergence of cloud gamingproviders that do not rely on Microsoft’s “multi-product ecosystem”shows that Microsoft does not have the ability to foreclose competition.
Consumer adoption of cloud gaming remains low. Harming ordegrading rival services would significantly set-back adoption of thistechnology – protecting market-leading incumbents (i.e., Sony onconsole, Apple and Google on mobile, as well as Steam on PC). Xbox,as a platform which is in last place in console, seventh place in PC andnowhere in mobile game distribution globally, has no incentive to do this–instead its incentive is to encourage the widespread adoption of cloudgaming technologies by as many providers as possible to encourage themajor shift in consumer behaviour required for cloud gaming to succeed.
The evidence shows that streaming services, for example in video, havereduced barriers to entry and encouraged multi-homing, therebydecreasing network effects. As such, the use of Activision Blizzardcontent to promote cloud gaming (which would not occur in the counterfactual) would have the opposite effect to that alleged by the CMA.
(j)As Microsoft does not have a market-leading position in gaming to protect, the“multi-product ecosystem theory of harm” outlined lacks analytical rigour anda limiting principle. A theory of harm that equates size in a particular sector(i.e., digital) with market power discriminates against digital mergers. The“multi-product ecosystem theory of harm” would – if adopted in Phase 2 -represent a significant lowering of the intervention threshold in digital mergers“by the backdoor”. Microsoft strongly urges the CMA to resist this path.
1.4 Microsoft has engaged constructively with the CMA throughout pre-notification and the Phase 1 process. Microsoft looks forward to engaging with the Panel through the Phase 2 process, to explain the benefits that the Transaction will bring to gamers, developers and the gaming industry more broadly.
2.Background and Transaction rationale
2.1 Microsoft has been present in the gaming industry for over twenty years, following the launch of the original Xbox console in 2001. With a primary focus on console gaming, Xbox has been the number three player behind Sony and Nintendo in each successive console generation, in what the gaming industry colloquially refers to as the “console wars”. Far from having any market power, Xbox has been a challenger which has consistently had to offer gamers additional value and champion new propositions, such as multi-game subscriptions and cross-platform play, in order to remain competitive.
2.2 Microsoft’s position on PC is even weaker. While Microsoft developed the Windows operating system (“OS”) for PC, the open nature of the Windows platform has fostered vibrant competition in PC game publishing and distribution. Microsoft has a global share of less than [✂]% in PC game publishing, as well as PC game distribution. The “multi-product ecosystem” alleged by the CMA has not given Microsoft any advantage in PC game distribution, with Microsoft sitting in seventh place globally, well behind leading distributors of PC games.
2.3 The proposed acquisition of Activision Blizzard takes place in the context of an industry which today is experiencing ongoing, significant, expansion beyond the console and PC to mobile devices.4 Mobile has become the largest segment in gaming, with 94% of all gamers globally playing games on mobile devices.5 Consumers now spend more hours gaming on mobile – where Xbox currently has no material presence – than on any other device, and mobile gaming is growing faster than any other part of gaming.6 Innovative new technologies are under development to allow games developed for one type of device to be played across multiple devices (e.g., cloud gaming and compatibility layer software). In addition, powerful new handheld gaming consoles have been launched which allow gamers to play the games they want where they want (e.g., Valve Steam Deck, Logitech G Cloud and Razer Edge 5G).
2.4 These developments do not imply that the industry is at any form of “inflection point”7 or primed to be “reshaped”8, but rather show that it is a growing, innovative, constantly evolving industry. As shown in Figure 1 below, the growth seen in recent years has not come at the expense of the console segment. Rather, it is additive, and Microsoft expects the console business to continue to thrive for the millions of gamers across the globe who choose to play games in this way.9
2.5 Given these developments, the need for a diversity of business models and technology remains core to Xbox’s strategy of reaching the world’s 3 billion gamers, regardless of location, socio-economic status, or device ownership. Xbox’s decision to embrace a gamer-centric, device-agnostic approach requires it to introduce innovative ways of accessing and paying for games that reach the maximum number of gamers possible. Xbox also recognises that growing its presence on mobile is critical to growing consumer engagement beyond console and PC games.
2.6 The Transaction forms part of this broader strategy, with the acquisition of Activision Blizzard bringing several sources of value.
(a)First, the Transaction gives Microsoft a meaningful presence in mobile gaming.Mobile gaming revenues from the King division and titles such as Call of Duty:Mobile, as well as ancillary revenue, represented more than half of ActivisionBlizzard’s revenues and in the first half of 2022.11 Mobile customers accountfor around three-quarters of its MAU. Microsoft currently has no meaningfulpresence in mobile gaming and the Transaction will bring much neededexpertise in mobile game development, marketing and advertising. Activision Blizzard will be able to contribute its learnings from developing and publishing mobile games to Xbox gaming studios.
(b)Second, Activision Blizzard is a well-established business with predictablerevenues and established communities of gamers around its three main gameseries, Call of Duty, World of Warcraft and Candy Crush Saga (which accountfor over three-quarters of its net revenues, and a significantly higher proportionof its operating income).12 Microsoft has consistently voiced its commitment tocontinue to make games available for purchase on any platform on which theyare currently available.13 Microsoft has backed this up, sending Sony a signedcontractual offer to keep all existing Activision Blizzard console titles on SonyPlayStation (including future instalments of Call of Duty) with feature andcontent parity.14 PlayStation is the leading console platform and discontinuingdistribution of Activision Blizzard’s titles on other platforms would costMicrosoft around USD [✂] billion in revenues in 2024 alone, growing to ca.USD [✂] billion by 2032. The deal valuation does not include increasedhardware sales, which do not form part of the deal rationale.
(c)Third, Microsoft’s acquisition of Activision Blizzard will support itsinvestments in Game Pass. Game Pass gives gamers an additional way to payfor games, as the same games that are available on Game Pass are also availableto purchase and gamers frequently switch between the two payment models. Byoffering gamers the ability to try new games for a fixed monthly fee, Game Passencourages gamers to play a broader range of games. The addition of ActivisionBlizzard titles, such as Call of Duty, to Game Pass is projected to increase theGame Pass subscriber base across console and PC by around [✂] millionsubscribers worldwide in FY 2024 and [✂] million by FY 2032, as comparedto the projected base without Activision Blizzard content.15 This represents onlya tiny fraction of the global population of console gamers (850 million) and PCgamers (1.45 billion).16
(d)Fourth, the Transaction will improve Microsoft’s ability to create a nextgeneration game store which operates across a range of devices, includingmobile as a result of the addition of Activision Blizzard’s content.17 Buildingon Activision Blizzard’s existing communities of gamers, Xbox will seek toscale the Xbox Store to mobile, attracting gamers to a new Xbox MobilePlatform. Shifting consumers away from the Google Play Store and App Store on mobile devices will, however, require a major shift in consumer behaviour. Microsoft hopes that by offering well-known and popular content, gamers will be more inclined to try something new. The Open App Store Principles announced by Microsoft will apply to the next generation game store.18
(e)Finally, the Transaction will increase the attractiveness of Microsoft’sadvertising business, as a provider – albeit a small one – of mobile advertisingservices.
3.Theory of Harm 1 – no prospect of a substantial lessening of competition through foreclosure of Sony
3.1 Theory of Harm 1(a) concerns the potential for Microsoft to foreclose Sony, the market leading console platform, by withholding or degrading Activision Blizzard content and in particular Call of Duty.19
3.2 Microsoft does not intend to remove Call of Duty from PlayStation or to degrade access to the franchise. The evidence on this is clear and includes Microsoft’s public statements20 and correspondence with Sony21, public statements by Sony22, the strategic rationale for the Transaction23, the deal valuation24 and Microsoft’s internal documents in relation to the Transaction.25 There is no evidence that Microsoft has considered withdrawing Call of Duty from PlayStation – because it has not. The CMA cannot simply brush this substantial body of evidence aside.26
3.3 However, even if Microsoft were to withdraw Call of Duty from PlayStation (quod non), this would not lead to a substantial lessening of competition. For Sony to be foreclosed, it would have to go from being the clear market leader to being placed at “such a disadvantage that [its] ability to compete is substantially limited”27. The suggestion that the incumbent market leader could be foreclosed by the third largest console provider as a result of losing access to one title is not credible.
3.4 The evidence presented in the Referral Decision does not come close to meeting this standard. The CMA did not even try to quantify the impact on Sony in Phase 1. The Referral Decision goes no further than to assert that a foreclosure strategy could have some unspecified impact on “Sony’s revenues and user base”.28 There is no assessment of the impact on competition or consumers. Indeed, while Sony is mentioned 57 times in the Referral Decision, consumers are only referred to 10 times. In considering this theory at Phase 1 the Referral Decision fails to “consider the effect of a merger on competition in the market rather than on individual competitors” (emphasis added).29
3.5 At Phase 2, the CMA must assess whether it is more likely than not that competition will be substantially lessened – and consumers harmed – as a result of the Transaction. Microsoft is confident that the CMA will find that this is not the case.
A. Sony has had a market leading position for over 20 years
3.6 The Referral Decision’s first theory of harm requires the CMA to find evidence thatSony is a competitor capable of being marginalised as a console platform if it losesaccess to Call of Duty, a single game title. It is implausible that a competitor of Sony’sstrength could be caught, surpassed and marginalised by the loss of Call of Duty (orany other Activision Blizzard content). The suggestion that one title could tip a marketin favour of one provider does not accord with market reality or evidence from the last20 years of console competition.
3.7 Sony is the market leader for consoles and has been for over 20 years. Its PlayStationplatform is ahead of Xbox in all relevant metrics (as is Nintendo).
PlayStation’s installed base of consoles (151.4 million in 2021) is more thandouble the size of Xbox’s (63.7 million in 2021).
PlayStation’s share of console sales both worldwide ([✂] in 2021) and in theUK ([✂]) is significantly higher than Xbox’s worldwide ([✂]) and in the UK([✂]).30
PlayStation has more than double the MAUs (close to 60 million more) of Xbox(and, as explained below, would continue to have significant more even in the hypothetical scenario in which all Call of Duty gamers switched from PlayStation to Xbox31).
3.8 Sony engages in conduct today which is reflective of its market power in console gaming. Such is Sony’s market strength that it has recently imposed a significant price increase for its PlayStation 5 console, of up to 20% in most markets outside of the US (6% in the UK).32 This move has been described by commentators as “displaying an unfortunate arrogance”.33 Xbox has confirmed that it has no plans to raise the price of the Xbox Series X or Series S and that the consoles will remain at their current price in the US, UK and Europe.34
3.9 In considering Sony’s market position, it is important to take account of its large portfolio of high-quality exclusive content. This includes prominent titles such as The Last of Us, Ghosts of Tsushima, God of War, Spider-Man, Demon’s Souls and the Uncharted series.35 These exclusive titles accounted for approximately 17% of consumer spend on PlayStation over the period 2019-2021 and achieved considerably higher average Metacritic scores in 2021 and 2020. In 2021, PlayStation had at least 286 exclusive titles generating revenue, as compared to Xbox’s [✂] exclusive titles.36 PlayStation also has a spate of first- and third-party exclusive titles lined up for launch in 2023, including Spider-Man 2, Wolverine, Horizon, Final Fantasy XVI and Forspoken.37 Sony executives have recently stated that they are growing organically and through acquisitions and this places Sony in “a virtuous cycle where success begets success”.38
3.10 These advantages must be fully considered at Phase 2 in assessing whether Sony would be likely to be foreclosed as a console competitor as a result of the Transaction. Contrary to the concerns outlined in the Referral Decision, the Transaction would help Xbox to continue to offer innovative new content offerings and compete more vigorously with Sony, Nintendo, and other leading gaming companies. This enhanced competition can only benefit gamers.
B. No ability to foreclose Sony
3.11 The Referral Decision concludes that the merged entity may have the ability to foreclose Sony on the basis that: (a) the merged entity would have significant upstream market power; (ii) Activision Blizzard’s content is an “important input” for Sony; and (iii)there are few if any alternative franchises with Call of Duty’s brand awareness andpopularity amongst gamers.39 This is incorrect for the reasons explained below.
(i)The merged entity will not have upstream market power
3.12 First, the merged entity cannot be considered to have upstream market power in game publishing, even on the basis of a narrow segment for console games. Activision Blizzard’s share in this segment is just [✂]% by value globally ([✂]% in the UK) and the combined share would be just [✂]% globally ([✂]% in the UK).40
3.13 The Referral Decision accepts that, based on these shares of supply, no upstream market power can be found. Nevertheless, the CMA goes on to assert that the merged entity would have “significant” upstream market power, since “overall publisher shares do not present a complete picture of the importance of ABK content”. It makes this claim on the basis that “some ABK games, and CoD in particular, are especially important for attracting gamers to a platform” and that “these gamers go on to play other games available on that platform, increasing that platform’s overall revenues”.41
3.14 While these assertions are central to the CMA’s case for referral, they are not supported by any evidence showing that Call of Duty is “especially important for attracting gamers” to Sony’s console platform. Instead, the CMA relies on high level points such as “CoD is currently one of the largest game franchises”, “has a high level of awareness amongst gamers” and “has been consistently successful for nearly a decade”.42 These points may all be true (as they are equally for other popular games such as Grand Theft Auto, FIFA and Fortnite), but do not show that the franchise is critical to attracting gamers to a console platform, nor that the loss of that one title would foreclose the console in question from the market.
3.15 The evidence in fact shows that Call of Duty gamers cannot be considered ‘special’ or ‘unique’ in terms of either their spending or user engagement as compared to gamers that favour other popular franchises.
(a)Xbox data shows that between 2016 and 2022, [✂]% of new Xbox gamersnever played or purchased Call of Duty content and only [✂]% of gamersplayed Call of Duty as their first game on their new Xbox console. Thesenumbers are inconsistent with the idea that Call of Duty drives platformadoption.
(b)The time that Call of Duty gamers spend playing Call of Duty or other games onthe Xbox console is in line with other major franchises. This is shown in Figure2 below. In 2022, Call of Duty gamers spent essentially the same time on Xboxas FIFA, Fortnite, Grand Theft Auto, Minecraft, NBK 2K, and Rocket League gamers. Importantly, they also spent around the same proportion of time playing other individual franchises. This is inconsistent with Call of Duty having a greater ability to drive consumption of non-Call of Duty content on the console platform than other popular franchises, as suggested in the Referral Decision.
Figure 2: [✂]
(c)Call of Duty gamers show preferences in line with gamers of other franchises.As Figure 3 below shows, the ranking of second most played game is essentiallyunchanged across gamers on Xbox regardless of what proportion of their timeis spent playing Call of Duty on Xbox. As such, they are not ‘special’ in termsof the type of content they purchase and consume.
Figure 3: [✂]
(ii)Sony has overstated the importance of Call of Duty to its viability
3.16 Second, and linked to the above points, Sony’s regulatory submissions have vastly exaggerated the importance of Call of Duty to its continued viability. For example, Sony has claimed that “if CoD were exclusively available on Xbox and/or XGP, this could severely adversely impact their ability to compete effectively”.43 This claim is not supported by evidence presented in the Referral Decision and is inconsistent with the evidence base available to the CMA.
3.17 For example, based on data the CMA has on Call of Duty sales and user engagement on PlayStation, as well as publicly available data published by Sony44, Call of Duty represented only [✂]% of Sony’s digital sales worldwide and Call of Duty MAUs represented less than [✂]% of PlayStation’s total MAUs in 2021.
3.18 The Referral Decision states the CMA “has received evidence that CoD has higher levels of user engagement and revenue spend on PlayStation than the Parties estimated”45, without providing further details. Sony’s obvious self-interest in this case means that its submissions must be carefully scrutinised. But even if the revenues and levels of user engagement are somewhat higher than Microsoft’s estimates, they could not be so high that Sony’s viability could depend upon retaining them.46 It is also important to note that Call of Duty’s significance to Sony is fuelled by Sony’s marketing efforts—it would be misleading to compare Call of Duty with games Sony does not market.
3.19 Indeed, as accepted in the Referral Decision, PlayStation currently “has a substantial number of non-CoD MAU”.47 Indeed, even if all of PlayStation’s MAUs that play Call of Duty were, hypothetically, to leave PlayStation it would still have significantly more MAUs than Xbox has today – by Microsoft’s estimate by some [✂] million.48 This holds true even in the purely hypothetical and unrealistic scenario where all these users would switch to Xbox – PlayStation would still have [✂]% ([✂] million) more users than Xbox. The CMA cannot conclude that PlayStation would not be viable with [✂] million MAUs (i.e., if it hypothetically lost all Call of Duty MAUs) when Xbox is clearly viable today with [✂]. See Figure 4 below.
3.20 The Referral Decision fails to draw the obvious conclusion from the evidence, instead arguing that “the CMA believes that some of these non-CoD gamers may also switch away from PlayStation following any total or partial foreclosure strategies” because “some of these non-CoD gamers would want to continue to play other games with their friends who are CoD gamers”.49 This is, however, pure assertion. No evidence is put forward to support this proposition, nor the allegation that material numbers of Call of Duty gamers themselves would switch console platform if the game were to become exclusive to Xbox.50
3.21 In any event, even if this were true to some extent, Sony would need to lose a [✂] higher number of non-Call of Duty gamers than actual Call of Duty gamers for its total MAU to fall to Xbox’s current level (i.e., [✂] million MAUs). This is patently not credible, and yet even in such an unrealistic scenario the CMA could not conclude that Sony would be likely to be foreclosed, given that Xbox is clearly a viable competitor today at this level (and the CMA does not suggest otherwise).
Figure 4 – [✂]
(iii)The success of platforms without Call of Duty lays bare the fallacy of the Sony theory of harm
3.22 As well as lacking evidence that a foreclosure strategy could credibly threaten Sony’s viability (as opposed to simply affecting its revenues), the CMA has disregarded clear evidence from the market today demonstrating that a console platform’s competitiveness cannot be affected by an inability to offer Call of Duty to gamers.
3.23 Nintendo’s console business is highly successful, without a single version of Call of Duty being available to play on its latest console, the Nintendo Switch (launched in 2017).51 This shows that the overall quality and appeal of a console platform are ultimately far more important to its success than any particular game being available on it. A further example of a platform that has succeeded without Call of Duty is Steam, which is the largest digital storefront, with a [✂] share of PC game digital sales in the UK. Steam has not carried any newly released Activision Blizzard games for the last three years following Activision Blizzard’s commercial decision to only sell its PC games on Battle.net – but this has not affected its leading position. Indeed, Steam’s success is so great that Activision Blizzard has recently decided to reverse its previous decision and distribute the next release of Call of Duty on the platform in October 2022.52
3.24 The Referral Decision does not address Steam, but the CMA seeks to dismiss the relevance of Nintendo’s console success absent Call of Duty by claiming that “this is because Nintendo generally offers a differentiated gaming experience to Xbox and PlayStation” and that “Xbox and PlayStation are closer substitutes to each other than to Nintendo Switch”.53 The suggestion that Xbox, PlayStation and Nintendo are not close competitors is incorrect54, but in any event these claims miss the relevant point. It is precisely the fact that Nintendo successfully offers a “differentiated” console platform, without Call of Duty, which demonstrates that console platforms have many options available to compete effectively which are not dependent on any particular gaming content.
3.25 Indeed, the Referral Decision in general overlooks the vast array of gaming content which is available to console providers (and multi-game subscription service providers). Just by way of example, there are over 4,000 games available on PlayStation alone, with new games continually being released. The suggestion that just one of these games can make or break a platform’s success is not supported by the evidence.
3.26 The Referral Decision also ignores the wide range of options which Sony would have to respond to a hypothetical foreclosure strategy, as well as the steps it has taken to bolster its gaming business even in the short period since the Transaction was announced. Sony has acquired several game studios – including Bungie, developer of the popular online game Destiny 255, Haven Studios, Lasengle and Savage Games56 – and a minority interest in FromSoftware, the developer of the biggest game of 2022, Elden Ring and other hit games.57 This complements Sony’s existing minority shareholding in Epic Games, publisher of Fortnite, strong first-party game catalogue and extensive portfolio of exclusive arrangements with third-party publishers. There were over 280 exclusive first- and third-party titles on PlayStation in 2021, nearly five times as many as on Xbox. Sony executives have made it clear that there are more to come.58 And it launched its enhanced multi-game subscription service, PlayStation Plus, in June 2022 – see further section 4 below.
3.27 With access to Call of Duty guaranteed through to at least [✂] under the existing contract with Activision Blizzard (and through at least the end of 2027 if it were to accept Microsoft’s current contractual offer), Sony has more than sufficient time to ensure that its console platform and content portfolio is competitively positioned to withstand any impact from a hypothetical foreclosure strategy. The fact that the CMA declines to even consider Sony’s competitive response is especially concerning given that it is prepared to speculate regarding Activision Blizzard’s likely approach to subscription services absent the Transaction (see section 4 below) and is at odds with the CMA’s acknowledgment that the gaming industry is dynamic in nature.59
(iv)Call of Duty’s continued success is not guaranteed
3.28 While Call of Duty is one of a number of popular franchises, its success over time is not guaranteed. Relevance with gamers is earned or lost with every release. This dynamic is shown by the performance of last year’s Call of Duty: Vanguard release, which was heavily criticized by the trade press and gamers alike, resulting in significantly lower sales than reflected in the internal documents cited by the CMA at paragraphs 142 – 144 of the Referral Decision.
3.29 The CMA responds to this in the Referral Decision by asserting that the “persistent high revenues and player engagement across all CoD titles, even after the release of Vanguard, indicates that gamers who did not like Vanguard most likely continued to play older CoD titles rather than switch away to a different game”. Again, however, no evidence is presented to support this assertion, which is contradicted by Activision Blizzard’s update to investors which shows that MAUs fell across the entire Call of Duty franchise following the launch of Vanguard:
“Average MAUs decreased by 47 million or 12% for the three months ended June 30, 2022, as compared to the three months ended June 30, 2021. The decrease was primarily due to lower average MAUs for Activision, driven by the Call of Duty franchise”.60
C.No incentive to withdraw Call of Duty
3.30 As noted above, Microsoft’s incentives to continue distributing Call of Duty to Sony are clear from its public statements and correspondence with Sony, public statements by Sony, the strategic rationale for the Transaction, the deal valuation and Microsoft’s internal documents in relation to the Transaction. The Referral Decision fails properly to consider the evidence that Microsoft’s commercial incentives are to continue broadly distributing Call of Duty.
3.31 First, it is clear from Microsoft’s valuation model that Microsoft’s intent is to “[✂]”.61 This evidence is overlooked entirely in the Referral Decision. However, as the Merger Assessment Guidelines make clear, in assessing incentives to foreclose the CMA must take an undertaking’s “business strategy” and “deal rationale” into account. Just as “if the merger firms’ internal documents show that it would be strategically beneficial to stop supplying rivals, it may not be necessary to try to infer their behaviour from their financial incentives”, it is equally the case that where internal documents evidence a strategic rationale and intent to continue supplying rivals this must be given significant weight in the CMA’s assessment.62
3.32 Microsoft’s incentives to continue to make Call of Duty available cross-platform are also clear from its business strategy following prior acquisitions. In particular, the most directly comparable game franchise previously acquired by Microsoft is Mojang’s Minecraft. Minecraft, like Call of Duty, is a globally popular multi-player franchise with a strong player community and social element that was available on multiple platforms when Microsoft acquired it. Microsoft not only kept Minecraft available on PlayStation after its 2014 acquisition (including new versions), but it has also expanded platform access to the franchise.63 As well as benefiting gamers, Minecraft’s multi-platform strategy has been a significant financial success, with over 200 million copies sold by 2020, approximately a 300% increase in sales since its acquisition in 2014.64
3.33 The games that the CMA cites as evidence of prior conduct at paragraph 192 of the Referral Decision are fundamentally different to Minecraft or Call of Duty and cannot be used as evidence of Microsoft’s “broad incentives” as they relate to Activision Blizzard content.65 Microsoft has never taken an existing game away from a rival platform post-acquisition.
3.34 Second, as noted in Microsoft’s deal model, “continued sales of Activision Blizzard’s portfolio on all platforms (console, PC, mobile)” accounted for [✂] of the estimated “Value to Microsoft” from the Transaction.66 Given that ~[✂]% of Call of Duty’s total MAUs (and [✂]% of console MAUs) are on PlayStation, a hypothetical foreclosure strategy would involve putting at risk a significant portion of Activision Blizzard’s revenues. This would be a commercially irrational strategy, in particular in circumstances where Microsoft could not conceivably expect to foreclose Sony from the market.67
3.35 The Referral Decision does not respond to these basic, intuitive points, instead focusing its attention on a critique of the financial modelling put forward by Microsoft (addressed further below) and basing its overall assessment of Microsoft’s incentives on the high-level assertion that “Microsoft has shown itself to be willing to make losses in the short term in order to build scale and increase its user base”.68
3.36 It makes this claim solely on the basis that Xbox has confirmed publicly that it makes a loss on the sale of console hardware (with this loss made up for through game sales and other transactional revenues). However, as the same article cited by the CMA notes, this “follows the traditional console business model” which has also been adopted by Sony (including before the launch of the Xbox)69, whereby profitability depends on earning revenues from the sale of first- and third-party games for the console. Indeed, this is a model that has been well known in competition law and economics for the past three decades.70
3.37 The CMA cannot read across from Xbox’s adoption of this “traditional” business model to assert that it would be willing to forgo substantial revenues from the distribution of Call of Duty in pursuit of gains at some unspecified point in the future. Rather, at Phase 2 the CMA must assess whether there is evidence showing that it is more likely than not that it would be in Xbox’s interests to put at risk over half of Call of Duty’s gamer base – and to degrade the user experience for all gamers (see below) – in the hope that a sufficient number would switch to Xbox to make such a strategy profitable. The Referral Decision does not present any such evidence.
(iii)The Referral Decision fails to take account of cross-platform play
3.38 Third, the Referral Decision fails to take account of the importance to gamers of cross-platform play (i.e., the ability for gamers to play together online irrespective of which device each of the gamers playing is using to access the game).
3.39 While the Referral Decision specifically acknowledges that there are direct network effects between gamers playing the multi-player mode of a game – as gamers want to play games with large communities, and moreover to play with their friends and family71 – it completely disregards the fact that Call of Duty’s multi-player feature is cross-platform. This implies that direct network effects transcend platform boundaries and operate at the market level, not the individual console or platform level.
3.40 Importantly, cross-platform play leads to material and measurable improvements in gamers’ user experience. This is because the more gamers that want to play a game online, the more efficient is the “matchmaking” algorithm that places gamers in the same multi-player instance of the game, and which can be based on the skill and the experience of the gamers.72 Cross-platform play increases the total gamer base available for matchmaking (i.e., because gamers are not limited to playing with others on the same platform that they are using) thereby amplifying the impact of direct network effects.
3.42 In fact, these network effects provide a powerful disincentive to withhold content from other platforms. Call of Duty game titles have allowed gamers to cross-play since 2019. This means that not only can PlayStation gamers play with their friends that use Xbox or PC, they can also be automatically matched by the matchmaking system against Xbox gamers and PC gamers. Microsoft has provided data showing that cross-play on Call of Duty is widespread on PlayStation. For example, over the last twelve months, [✂]% of PlayStation gamers of Call of Duty: Modern Warfare and Call of Duty: Warzone had cross-play enabled while playing online, and were therefore able to play with gamers on Xbox and PC.74
3.43 As all gamers significantly benefit from cross-play, a hypothetical foreclosure strategy involving withdrawing Call of Duty from PlayStation would be self-defeating because it would hurt the whole Call of Duty gamer base, including Xbox and PC gamers. This, among other factors, means that any strategy to make Call of Duty exclusive to Xbox would inevitably result in significant gamer backlash, damaging both the Xbox and Call of Duty brands and risking the loss of gamers to other competing franchises, such as Fortnite, Apex Legends, Grand Theft Auto, FIFA and others.75
3.44 Accordingly, the existence of cross-play strengthens Xbox’s incentives to broaden the scope of the Call of Duty gamer base as much as possible. The Referral Decision fails to consider this point, addressing cross-play only in the context of partial foreclosure (see below).76
(iv) Contractual protections provide a powerful disincentive to withdraw content
3.45 Fourth, the Referral Decision does not properly take account of Sony’s contractual protections under its existing contracts. Paragraph 7.15 of the Merger Assessment Guidelines states that, even if the CMA is unlikely to place material weight on contractual protections in assessing the ability to engage in foreclosure strategies, it “may consider any financial or reputational costs of terminating contracts in its assessment of foreclosure incentives”. However, the CMA has given no consideration to how the financial and reputational costs of breaching or terminating the existing contractual arrangements with Sony would impact Microsoft’s incentives to engage in a hypothetical foreclosure strategy. Nor does the CMA give appropriate weight to Sony’s public statements that “[w]e expect that Microsoft will abide by contractual agreements and continue to ensure Activision games are multiplatform”.77
(v) Flawed assessment of the Microsoft’s financial modelling
3.46 Finally, Microsoft strongly disagrees with the CMA’s critique of Microsoft’s financial modelling, which is highly selective and does not properly reflect the number of decidedly conservative estimates used and sensitivities presented in Microsoft’s analysis. At root, the CMA’s issues with Microsoft’s analysis appear to be based on its position that “the number of gamers that would switch to Xbox if Microsoft made ABK’s content exclusive could be significantly higher than the Parties predict, given the importance of ABK’s content to gamers, which again would make the model inaccurate”.78 However, like much of the Referral Decision, no further explanation or evidence is provided to substantiate the CMA’s view on the likely extent of switching, with the Referral Decision limited to a one-sided, high-level application of dynamic considerations to a static analysis.
3.47 The CMA’s unwarranted criticisms do not alter the conclusion that it is highly implausible that a hypothetical foreclosure strategy could bring about sufficient diversion of PlayStation users to be profitable. Microsoft is also seriously concerned that the CMA appears to have placed considerable weight on alternative modelling presented by Sony without providing transparency to Microsoft of the content of that modelling or the assumptions used.79 The CMA has not given Microsoft access to Sony’s alternative modelling or an opportunity to comment on it.
D. Partial foreclosure is an even weaker theory of harm
3.48 The Referral Decision contemplates four partial foreclosure strategies that Microsoft could hypothetically engage in post-Transaction, in particular: (i) making Activision Blizzard content available for release on rival console gaming platforms at a later date compared to Xbox (i.e., ‘timed exclusivity’), (ii) degrading the quality of such content available on PlayStation, (iii) making features or upgrades of Activision Blizzard games unavailable to Sony (i.e., ‘content exclusivity’), and/or (v) raising the wholesale price of Activision Blizzard content on PlayStation.80
3.49 Any such strategies would inherently have less impact on the competitiveness of a rival than total foreclosure. Indeed, as the Referral Decision accepts, past timed exclusivity and content exclusivity arrangements in relation to Call of Duty – which have historically benefitted both Xbox and, more recently, PlayStation – have never resulted in foreclosure of any rival console platform.81 In particular, Call of Duty has provided exclusive or timed-exclusive downloadable content for either PlayStation or Xbox since Call of Duty 2 in 2005.82 Yet, even in the period where these arrangements benefitted PlayStation, Xbox – which already trailed PlayStation [✂] in consoles sold – was not foreclosed.
3.50 Given this evidence, the CMA appears to concede that its theory of harm in fact only relates to “partial foreclosure strategies that remain untested” – in particular degrading the quality of Activision Blizzard content on PlayStation or raising wholesale prices.83 However, even in relation to these “untested strategies”, the Referral Decision fails to comply with paragraph 7.12 of the Merger Assessment Guidelines, which states that “[i]n assessing the ability of the merged entity to foreclose its rivals, the CMA will go beyond examining simply whether it could supply its inputs to them on worse terms and will consider whether it would be able to harm their competitiveness by doing so” (emphasis added).
3.51 The Referral Decision fails to assess the potential impact on Sony of any partial foreclosure strategy in any meaningful way, simply asserting without further explanation that “these strategies could significantly impact the ability of Sony to compete”.84 There is no evidence to support such an assertion. On the contrary:
- (a)It is unclear what the Referral Decision means by “degrading the quality” ofCall of Duty on PlayStation. For example, if the CMA means that the versionof Call of Duty made available on Xbox would have additional features orfunctionality then this is no different than content exclusivity, which the CMAaccepts has never resulted in foreclosure.85
- (b)Furthermore, Microsoft would have no incentive to degrade the quality of thegame made available on PlayStation.
- In particular, the Referral Decision fails to properly acknowledge thefact that any partial foreclosure strategy of this nature would beextremely costly for Xbox given the increasing importance of cross-platform play (as explained above) which relies on a consistentexperience across the gamer base and therefore across platforms.86
- The CMA’s suggestion that “given the high number of CoD players,high-quality matchmaking would continue to be possible even if cross-play were compromised”87 is, again, a simple assertion and reflects a failure to understand the relevance and importance of cross-play, which goes well beyond improving matchmaking.88
- The fact that Activision Blizzard does not allow Xbox and PC players toentirely switch off cross-platform play for Call of Duty matches on theseplatforms shows that it is not the case that the high number of Call ofDuty players is sufficient to allow smooth cross-platform play.
- (c)As regards potential raising of wholesale prices, any such theory of harm wouldappear (contrary to the Merger Assessment Guidelines) solely concerned withprotecting the profitability of Sony, a company whose gaming business earnedrevenues of USD 24.4 billion in 2021. It is implausible that paying a higherprice for Call of Duty could threaten the viability of the market leader, nor hasthe CMA sought to explain how it could.
D. No anticompetitive effect
3.52 For the reasons set out above, the Referral Decision is wrong to conclude that Microsoft would be incentivised to withdraw (in whole or in part) Activision Blizzard content from Sony. However, even if it were hypothetically to do so, it is implausible that this could “substantially limit” the market leader’s ability to compete. It is therefore not plausible that a substantial lessening of competition could arise as a result of total or partial foreclosure of Sony.
4.Theory of Harm 1(b) – no prospect of a substantial lessening of competition through foreclosure of multi-game subscription services
4.1 Theory of Harm 1(b) relates to Microsoft’s ability and incentive to harm rival multi-game subscription services through complete or partial withdrawal of ActivisionBlizzard content.
4.2 A key benefit of the Transaction for gamers lies in Xbox’s plan to make ActivisionBlizzard content available on Game Pass on the day of the game’s release (known as“day and date” access).89
- This will offer gamers on Xbox, PC, Steam Deck, Logitech G Cloud and otherdevices that support Game Pass a choice as to how they pay for this content (asall games that are available on Game Pass are also available to purchase).
- For gamers who choose to subscribe to Game Pass, they will have access to abroader menu of games to choose from and enjoy.
- Xbox has been clear that this is subject to existing contractual obligations thatActivision Blizzard may have with other platforms, which will be honoured.90
- Xbox may differentiate Game Pass by including Activision Blizzard games inGame Pass, whilst not making the titles available in the same manner or at thesame time on other subscription services.91
4.3 It is not plausible that such a strategy could result in a substantial lessening of competition.
While Xbox expects the addition of Activision Blizzard content to increase theappeal of Game Pass, it is not expected to [✂] increase subscriber numberscompared to the counterfactual and will not enable Xbox to build an“unmatchable” advantage in multi-game subscription services, which will inany event continue to be constrained by traditional buy-to-play and otherpayment models.
Activision Blizzard does not make its games (including Call of Duty) availablein any meaningful sense to any multi-game subscription services, [✂].
It is therefore not possible that its games are an “important input” forsubscription services or that they could be “foreclosed” by not having access tothem post-Transaction, which would be the case in the counterfactual (i.e.,absent the acquisition).
In any event, Sony and other multi-game subscription providers have a broadand deep library of available content and a range of options available to bolstertheir own services in response to Xbox improving its own offering and couldnot be foreclosed by not having access to Activision Blizzard content.
4.4 As explained below, the Referral Decision’s logic on this theory of harm is undermined by the CMA’s conclusion that multi-game subscriptions constitute a separate frame of reference, distinct from the many other payment models adopted by platforms and publishers to monetize the sale of gaming content.
A. Multi-game subscriptions are a means of payment – not a market
4.5 The Referral Decision fails to properly take account of the broader industry contextwhen considering how the Transaction might impact competition in multi-gamesubscriptions, which represent just one of many different monetization models forgaming content.
4.6 Subscription and other payment models will coexist and compete alongside each otherfor gamers’ spend and game-time for the foreseeable future. This is clear from industryreports, Microsoft’s internal documents, public statements by competitors and user data.92 Moreover, all market forecasts show that multi-game subscriptions will continue to account for only a modest portion of overall consumer spend on gaming for the foreseeable future, and the games made available through subscriptions will continue to also be available on a buy-to-play basis.93 The Referral Decision’s suggestion that Microsoft could “become a de facto gatekeeper between game publishers and gamers” as a result of the Transaction is, therefore, incorrect. Game publishers have, and will continue to have, multiple avenues by which to deliver their content to gamers.94 And gamers have, and will continue to have, numerous channels through which to access their favourite games.
4.7 The Referral Decision finds – “on a cautious basis” – that it is appropriate to consider a separate frame of reference for multi-game subscription services.95 Microsoft strongly disagrees with this approach. Subscription services cannot be considered as a separate market, since they are merely an alternative pricing model for the exact same content. Multi-game subscription services do not offer gamers any content that cannot be found on a buy-to-play basis. Every game in Game Pass is also separately available for purchase. The only difference is that subscribers have access to a wider catalogue of games at a lower price than if they were to purchase each game individually, albeit only so long as they maintain their subscription. The Referral Decision acknowledges evidence of direct competition between buy-to-play games and multi-game subscription services, but fails to draw the correct conclusion from it.96
4.8 In any event, the Referral Decision’s characterisation of Game Pass as a “strong incumbent” in this area is not consistent with market reality.97 Subscription-pricing models have seen limited adoption by consumers, and Xbox, like all other providers, faces significant headwinds. A range of industry players have introduced some version of a multi-game subscription service – including Sony, Nintendo, Amazon, Meta, Apple, Electronic Arts and Ubisoft – with significant differentiation between each business model in terms of the type of content, pricing, and technical features.98
4.9 It is difficult to comprehend how the CMA has reached a conclusion that Game Pass has ‘incumbency’ in such a fast-evolving space.
Game Pass was launched in 2017 and has grown gradually over the last fiveyears, reaching [✂] 25 million subscribers by January 2022.
As a proportion of total revenue from the digital distribution of gaming content,Game Pass accounted for less than [✂]% by value in 2021 globally ([✂]% inthe UK).
4.10 On a global basis, Xbox has a smaller share of supply in the segment by value than Sony ([✂]% vs 38.2%).99 As Game Pass offers the same service and catalogue of content globally, a UK frame of reference is not relevant.100 It makes no sense to characterise Game Pass as a “strong incumbent” in these circumstances. And as set out below, Sony has by far the largest subscriber base of any multi-game subscription service. Sony engages in conduct today which is reflective of its market power in multi-game subscriptions, including charging a significant pricing premium for its subscription services without fear of losing share in either subscriptions or consoles, [✂].101
4.11 It should also be recognised that Game Pass’ position today reflects the fact that it was Xbox that sought to pioneer this innovative payment model. As the smallest console provider, consistently lagging behind its competitors by some margin, it has the least to lose from offering alternative pricing models to the traditional “buy-to-play” model. However, this does not mean that competitors are or will be unable to rival Game Pass following the Transaction, as explained below.
B. Activision Blizzard content does not feature in multi-game subscriptionservices
4.12 With regard to the importance of Activision Blizzard content for multi-game subscription services, Activision Blizzard currently only makes limited back catalogue titles available on subscription, and only on a limited time-period basis (e.g., for 30 days); Activision Blizzard has never made a new release available on a subscription service; and Call of Duty has never been extensively available on any subscription service. A limited number of past releases have been made available for very short periods on Sony subscription services. The revenue accrued by Activision Blizzard for making its games available on Sony’s subscription services is [✂].
4.13 Such content cannot, therefore, be considered an ‘important input’ for existing or potential multi-game subscription providers and the Referral Decision’s assertion, on which Theory of Harm 1(b) depends, that “ABK’s content would likely be available on [multi-game subscription services] in future, in which case it would be expected to be as important as it is to consoles today on a buy-to-play basis” is unsupported by the facts.102 [✂].
4.14 Accordingly, the only counterfactual available to the CMA is the prevailing conditions in which Activision Blizzard content (specifically Call of Duty) is not available via subscription services. Against this counterfactual, it is not credible to maintain any theory of harm based on the unavailability of Activision Blizzard content to rival platforms to Game Pass, since that is already the case today.103
C. No prospect of foreclosing rival providers of multi-game subscriptions
4.15 Given that Activision Blizzard does not make its most valuable content available via multi-game subscriptions, even if such content were made available on Game Pass post-Transaction, but not on other multi-game subscriptions in the same manner or time, this would not foreclose rivals. As set out above, the merged entity will not have market power in the upstream game publishing market and there will be a range of alternative content available to build a viable multi-game subscription service.
4.16 Activision Blizzard’s reluctance to make its content available through multi-game subscription services has not deterred several companies from launching subscription services in recent years. This includes Sony, Nintendo, Amazon, Meta, Netflix104, EA, Ubisoft, Apple, Nvidia and many others.105 Nor has the announcement of the Transaction deterred these rivals from investing in their services.
4.17 In particular, Sony has already responded to competition from Game Pass by launching a new PlayStation Plus (“PS+”) offering in June 2022. The Premium tier of this service includes the ability to stream certain games from the cloud to PlayStation consoles and PCs. The first-party and third-party Sony titles included in the PS+ Premium tier will include successful hits (exclusive to PlayStation) like Ghost of Tsushima Director’s Cut, Horizon Zero Dawn, Marvel’s Spider-Man, Marvel’s Spider-Man: Miles Morales, Demon’s Souls, Infamous: Second Son, Bloodborne and Days Gone. Even before the launch of the new tiers, PS+ had reached 50 million members (almost double the number of subscribers to Game Pass)106, with Sony marketing itself as having “built the world’s largest game subscription service”.
Figure 5 – [✂]
4.18 This competitive dynamic benefits gamers, and is only likely to accelerate post-Transaction. For example, Sony has the ability to respond competitively to Xbox adding new content to Game Pass by including more of its exclusive content into PS+ and/or making its newer content available in PS+, which to date it has declined to do despite charging a higher monthly subscription fee than Game Pass.
D. No anticompetitive effect
4.19 For the reasons set out above, it is not plausible that the Transaction could give rise to a substantial lessening of competition in relation to multi-game subscription services. The Referral Decision appears to be premised on an underlying concern that the growth of multi-game subscription services is somehow bad for gamers. It fails to recognise that these services offer gamers greater choice and introduce greater competition against the traditional buy-to-play model.
4.20 The Transaction will actually bring specific benefits to gamers by allowing Microsoft to place Activision Blizzard’s games available on Game Pass on “day and date” of release, which would not occur otherwise. The Referral Decision ignores this actual, immediate, merger-specific benefit and instead focuses on speculative harm caused by a hypothetical withholding of an asset that is not even available as an input in the market today.
5. Theory of Harm 2 – No prospect of a substantial lessening of competition through foreclosure of cloud gaming services
5.1 In a novel theory of harm, unsupported by citations to legal precedent or economicliterature, the CMA alleges that:
(a)Cloud gaming is a nascent and rapidly developing market that is expected togrow significantly in the future;
(b)Microsoft already has an advantage over rivals by having a broad multi-productecosystem, including a leading cloud platform and PC OS; and
(c)Microsoft could leverage this ecosystem together with Activision Blizzard’sgaming content to strengthen network effects, raise barriers to entry, and henceforeclose rivals in cloud gaming services.107
5.2 Like Theories of Harm 1(a) and 1(b), this theory focusses on the concerns of competitors, rather than consumers. In particular, the Referral Decision references high-level concerns raised by Sony and other competitors regarding Microsoft’s alleged advantages due its cloud and OS assets. The alleged benefits are based on a misunderstanding of the technology underpinning Microsoft’s own cloud gaming service and do not exist.
5.3 The Referral Decision fails to recognise that the Transaction aims to facilitate the adoption of innovative services (including cloud gaming) for which there is currently no proven consumer demand. As the CMA recognised in its Mobile Ecosystem Market Study, cloud gaming would – if adopted by consumers – have the potential to deliver benefits to gamers and developers. The benefits which the CMA recognised in the Mobile Ecosystem Market Study are notable for their omission in the Referral Decision.108
A. Theory of Harm 2 is novel and without precedent
5.4 Theory of Harm 2 is a novel theory of harm without any precedent. Neither the Merger Assessment Guidelines nor previous CMA decisions refer to a “multi-product ecosystem theory of harm”.109 The CMA cites no economic literature in support of its theory, referring only in passing to the Furman Report (“Furman”)110, which notes that:
“large digital companies have also used acquisitions to develop strong ecosystems across multiple layers of value chains in order to cement their position in their main market, though this is not to say that every acquisition should be viewed from this perspective” (emphasis added).111
5.5 Furman goes on to state that:
“most acquisitions made by digital companies are likely to be benign or beneficial to consumers due to efficiencies, and the potential for innovative products and services to be brought more quickly to market. However, a minority of acquisitions are likely to have been anticompetitive” (emphasis added).112
5.6 In articulating its novel “multi-product ecosystem theory of harm”, the CMA fails to reflect Furman’s conclusion that most acquisitions by digital companies are not anticompetitive. Instead, the CMA’s articulation is so broad that most acquisitions by digital companies (whatever their position in the relevant market) would be caught. A theory of harm that equates size in a particular sector (i.e., digital) with market power discriminates against digital mergers. If adopted in Phase 2, the “multi-product ecosystem theory of harm” would represent a significant lowering of the intervention threshold in digital mergers by the backdoor, contrary to government policy.113
B. Cloud gaming is a new and unproven technology
5.7 Cloud gaming allows gamers to stream games from a remote server to any device (e.g., mobile, PC, console, smart TV) over the internet. This is a new and immature technology which the CMA has recognized faces significant challenges, particularly on mobile devices, as “users may be unaware of the choices available to them or find it difficult to access a provider’s services since web apps are not currently discoverable on, or distributed by, the App Store, which is how users are accustomed to discovering apps”.114 In considering Theory of Harm 2, the CMA must recognise that cloud gaming currently accounts for a de minimis proportion of consumer spend on gaming: just [✂]% in 2021.
5.8 While this may grow, particularly on mobile devices, adoption is not expected to be rapid as it requires a significant change in consumer behaviour. Research published by the CMA show that, both worldwide and in the UK, where cloud gaming app users had a choice between a provider’s native or web app on Android, around 99% of users used the native app, with 1% using either the web app or a combination of the web and native app.115 Microsoft and many industry experts expect that gamers on PC and console will continue to download the vast majority of the games they play.116
5.9 The Referral Decision finds a separate product market for cloud gaming services. However, such services do not provide different gaming content to what is available to download. Rather than a distinct product or market, cloud gaming represents a feature which provides an alternative way for gamers to access content that is not tied to a specific device. It should not be considered separately from the other ways of accessing and playing games (at least on console and PC).
5.10 Gamers choose a gaming experience based on whether it provides enjoyment at an attractive price point. Gamers are not motivated by the location of the content or means of delivery (e.g., whether they are using computing power locally or in the cloud). Rather, gamers care about subject matter, storylines, graphical performance, speed (e.g., loading times and latencies), mechanics, game selection, and game cost – and that is true across all gaming scenarios, streaming being no exception. Streaming services therefore need to compete effectively with downloadable gaming options across these metrics if they are to grow.
5.11 The Referral Decision rejects Microsoft’s views on the basis that they “fail to recognise the impact of cloud gaming services on demand for consoles, PCs, and games”, arguing that “cloud gaming services can be seen as an alternative for gamers to owning a console or PC”. This vastly overstates the relevance and importance of cloud gaming services in the gaming space at present and over the medium term. Microsoft agree that in future cloud gaming services may mean that hardware distinctions will become less important. However, the reality is that today cloud gaming remains in its infancy and unproven as a consumer proposition.117 Evidence from Microsoft’s internal documents, data and third-party reports shows that cloud gaming services are not relevant in any meaningful way to gamers’ “demand for consoles, PCs, and games”, nor is this expected to change in next few years.118 No evidence is presented in the Decision to suggest otherwise.
C. Cloud gaming will lower barriers to entry and network effects
5.12 In applying its novel “multi-product ecosystem theory of harm”, the CMA fails to consider that the Transaction is aimed at bringing innovative services (including cloud gaming) more quickly to market. The CMA does not cite, or appear to have considered, the work of its own Mobile Ecosystem Market Study which recognises that cloud gaming “could provide benefits to consumers and developers, as well as increasing the level of competition between operating systems”.119
5.13 The CMA also fails to recognise that by promoting cloud gaming, the Transaction would lower barriers to entry and network effects. The Referral Decision states that cloud gaming “means that gamers can access games that were previously available only on console through a wider range of less powerful devices (e.g. smart TVs, mobiles)” and that “cloud gaming services could be attractive to a different pool of customers who do not have access to the current hardware”.120 However, as cloud gaming is effectively device-agnostic and does not require investment in hardware, gamers can easily switch and multi-home across services regardless of the device they choose to play on.
5.14 There is substantial evidence from video streaming that consumers multi-home and subscribe to many services at the same time, thereby reducing the importance of network effects (as smaller platforms can still attract content creators to distribute on their services) and barriers to entry (as users are not “captive” and are willing to try new services). Data published by Ofcom confirms that multi-homing is widespread among subscription video-on-demand (SVOD) users with 13.2 million UK households (46%) subscribing to two or more video on demand services and 5.2 million households subscribing to Netflix, Amazon Prime Video and Disney+.121
5.15 As noted in the economic literature on multi-product ecosystems “[i]n general, the more multi-homing users there are, the easier it will be for multiple platforms to co-exist in a market, and the less ‘tippy’ the market will be”.122 This is particularly the case for gaming as evidenced by the economics and business literature.123 By promoting the take-up of cloud gaming, the Transaction would make it easier for gamers to multi-home and lower barriers to entry and network effects, to the benefit of competition and consumers.
D. No existing or prospective advantage for Microsoft in cloud gaming owing to its cloud and OS assets
5.16 Theory of Harm 2 postulates that Microsoft has an existing advantage in providing cloud gaming services owing to its cloud as a result of its “broad multi-product ecosystem, including a leading cloud platform and PC OS”, which it could leverage alongside Activision Blizzard content to strengthen network effects, raise barriers to entry and foreclose cloud gaming rivals post-Transaction.124 The CMA is, however, incorrect to assert as its starting position that Microsoft’s “multi-product ecosystem already places it in a much stronger position than rivals”.125
5.17 As acknowledged in the Referral Decision, Microsoft’s xCloud service is provided on dedicated Xbox consoles located in Microsoft data centres, streaming console games on the Xbox operating system, rather than servers running Azure services and streaming Windows PC games.126 [✂].
5.18 The CMA nevertheless considers that, because Microsoft has considered migrating its cloud gaming service to Azure, it may have a significant advantage over rivals without a cloud platform. [✂]127 [✂].128 [✂]129 [✂].
5.19 Moreover, while the CMA asserts that Microsoft may have a “significant advantage” over rivals without a leading OS owing to its ability to promote its gaming services through Windows OS130, this is not supported by the evidence. The evidence cited in the Referral Decision clearly demonstrates that promotional strategies are not effective customer acquisition tools. These must, moreover, be considered in context and the equivalent promotional strategies available to cloud gaming rivals such as Amazon (which owns Twitch, the leading game streaming site, as well as Amazon Prime), Google (which owns YouTube and the Google Play Store, as well as a leading digital advertising business), Sony (which has a multi-product entertainment ecosystem, including Sony Pictures and Sony Music) and Meta (which owns Facebook, Instagram, WhatsApp, as well as a leading digital advertising business).131 These social media and entertainment related surfaces are more suited to promoting a mature-rated game, such as Call of Duty, than Windows which is used across a broad range of enterprise, school government and domestic settings.
5.20 The emergence of several cloud gaming providers that do not rely on Microsoft’s “multi-product ecosystem” demonstrates that Microsoft does not have the ability to foreclose competition, as explained in more detail below.
E. No ability to foreclose rivals through foreclosure of gaming content
5.21 As set out in sections 3 and 4 above, it is not credible that the merged entity could foreclose rivals in console and multi-game subscription services through total or partial foreclosure of Activision Blizzard content. The same is true of rival cloud gaming services (many, but not all, of which are also providers of multi-game subscriptions).
5.22 The Referral Decision does not specify how the merged entity could target cloud gaming rivals with foreclosure strategies, instead referring to its analysis under Theory of Harm 1.132 Activision Blizzard has chosen not to make its content available via cloud-based game streaming to date. Industry analysts have noted that Activision Blizzard has “sat out the cloud developments”.133
5.23 As such, Activision Blizzard content cannot be considered an important input which would drive cloud-based game streaming developments absent the Transaction. Nor could any decision by Microsoft not to make Activision Blizzard content available to rival cloud gaming services have a foreclosure effect. Indeed, Activision Blizzard’s reluctance to make its content available through cloud gaming services has not deterred several companies from launching cloud gaming services in recent years.
5.24 Furthermore, as noted above, there is a vast array of gaming content potentially available to cloud gaming providers to build viable catalogues. As the CMA notes, cloud gaming technology will “become feasible for most game titles in the foreseeable future”.134 The Referral Decision presents no evidence to suggest that such providers could be foreclosed due to lack of available streaming content.
F. No ability or incentive to foreclose access to cloud infrastructure to cloud gaming rivals
5.25 The Referral Decision alleges that Microsoft could raise rivals’ costs by denying access to Microsoft’s Azure cloud platform, or offering Azure services on worse terms.135 This assertion does not reflect the market reality that if Microsoft sought to raise prices or technically degrade cloud infrastructure to cloud gaming companies, it would just push these customers to procure elsewhere. Amazon and Google (along with Alibaba, Huawei, IBM, Oracle, Tencent, Nvidia, and many others) have made large investments in cloud infrastructure and compete head-to-head as they seek to reduce costs through increasing utilization.
5.26 Workloads like cloud gaming rely on Infrastructure-as-a-Service (“IaaS”). To provide cloud game streaming services to end-users, providers deploy for virtual machines (“VMs”) running on powerful graphical processing units (“GPUs”) with fast data transmission capability to enable the complicated graphics, low latency, and high bandwidth that games require. All major IaaS providers have the capability to design and offer VMs to meet these needs. All providers use the same commodity hardware from third-party suppliers (e.g., AMD and Nvidia), and the prices for VMs running Windows in the cloud are similar on a cost per hour basis across major cloud providers. Moreover, customers regularly “multi-home” by procuring services from two or more service providers.136
5.27 Microsoft’s commercial incentives – like any cloud provider – are to drive workload volumes on Azure to achieve economies of scale and scope and reduce costs. Azure’s pricing structures (in line with industry practice) are not end-use specific. Microsoft’s entire corporate focus is on the opportunity to drive down the cost of cloud services overall to attract more customers to those services, which would be undermined by pursuing any discriminatory strategy aimed at indirectly favouring Microsoft’s own cloud gaming service. Any such strategy would be highly unlikely to be successful, particularly in the presence of credible cloud infrastructure alternatives which Microsoft’s cloud gaming rivals could turn to, while putting at risk billions of dollars of revenue from Microsoft’s wider cloud computing business.
G. No ability or incentive to disadvantage cloud gaming rivals through Windows licensing policies
5.28 The Referral Decision further alleges that Microsoft could foreclose rivals through denying access to a Windows OS licence, or offering it on worse terms (including price).137 This is incorrect. Microsoft’s extensive and well-established licensing program for Windows OS does not alter Windows license terms based on end use-case and Microsoft could not effectively target cloud gaming rivals.
5.29 Windows Server is primarily licensed to cloud computing providers under the Service Provider License Agreement (“SPLA”) program. These licenses are available to a wide variety of providers across many industries and sectors, including data centre providers, telecom operators, media streaming services, and many others. Microsoft does not alter its pricing based on the type of use-case a service provider serves to its end customers. Indeed, companies which license Windows Server from Microsoft are not required to inform Microsoft of the services which they offer using the licensed software, meaning Microsoft has no way of selectively charging more for cloud gaming use.
5.30 Furthermore, even if Microsoft were hypothetically to attempt to target rival cloud gaming companies (quod non), they could in any event get access to Windows Server from other cloud providers. In particular, cloud service providers such as AWS and GCP use SPLA licensing for a wide array of workloads to serve end customers across many industries. Microsoft has no way of selectively charging these customers more for cloud gaming end-uses. Service providers focused exclusively on cloud gaming can therefore build those services on the infrastructure of other SPLA licensees, like AWS, GCP and others, to obtain access to Windows Server.
5.31 The Referral Decision also notes concerns raised by one cloud gaming provider that “the Windows Client version of OS is superior to the Windows Server version for gaming”.138 This is incorrect. Windows Client OS [✂]. In particular, Windows Client contains hundreds of services for client computing such as the desktop graphical user interface, browser, virtual assistant, and timeline features that consume significant resources and are unnecessary for cloud gaming. By contrast, Windows Server is purpose-built to serve multi-user cloud workloads in virtualized environments. Claims that Windows Client is required are, therefore, misleading. In any event, as noted above, [✂].139
5.32 Finally, there are a number of established and emerging technologies for games to be built on Windows OS alternatives, such as Linux OS. Compatibility tools including Proton, WINE, and Vulcan have proven commercially viable for eliminating porting costs to Linux OS for top PC games, eliminating the need for cloud gaming providers to license Windows OS at all. The top Windows-based PC games are already widely available through Proton. Valve (which owns Steam, the leading distributor of PC games) has been a long-standing supporter of Linux adoption having integrated Proton into the Steam client to allow gamers to play Windows PC games on Linux devices. Valve’s successful new Steam Deck console, launched in 2022, has a Linux operating system and uses Proton to make Windows PC games available on the device (in addition to Linux games).
5.33 As such, Microsoft has no ability to foreclose cloud gaming rivals by denying access to Windows OS licenses, given the alternatives available to them to run Windows OS or alternative operating systems to support their streaming services. It equally has no incentive to compromise its sophisticated Windows OS licensing system in a vain attempt to foreclose a nascent cloud gaming segment.
H. No anticompetitive effect
5.34 For the reasons set out above, it is not plausible that the Transaction could give rise to a substantial lessening of competition in relation to cloud gaming services. The Referral Decision fails to articulate how Microsoft could meaningfully raise its rivals’ costs or how its incentives to do so could change as a result of the Transaction. Moreover, the CMA wrongly ignores the benefits that cloud gaming has the potential to deliver to gamers, by lowering barriers to entry and network effects and expanding access to content previously available only on consoles and high-end specification PCs.
5.35 The Referral Decision also raises a concern that if “Microsoft were to acquire significant market power in cloud gaming services, it could become a de facto gatekeeper between game publishers and gamers”.140 As explained above (see paragraph 4.7), this is patently incorrect. Cloud gaming services will co-exist alongside (and compete with) other forms of access to the same gaming content for the foreseeable future and game publishers will therefore have multiple avenues by which to deliver their content to gamers.
6.1 Microsoft has engaged constructively with the CMA throughout pre-notification and the Phase 1 process. Microsoft was disappointed that it was not offered the opportunity to explain these points to the CMA in detail prior to the Phase 1 issues meeting and supplementary issues meeting, despite several requests to provide teach-in sessions for the benefit of the case team. Microsoft looks forward to engaging with the Panel through the Phase 2 process, to explain the benefits that the Transaction will bring to gamers, developers and the gaming industry more broadly.
6.2 For the reasons set out in this response, the theories of harm and evidence presented in the Referral Decision do not provide any plausible basis on which it could be found that the Transaction will give rise to a substantial lessening of competition. Certainly, the CMA cannot find that this would be the case to the requisite standard at Phase 2.141