Sarah Cardell, the recently-appointed CEO and former long-time General Counsel of the UK Competition & Markets Authority (CMA), made the mistake of her (professional) life when ruling against Microsoft's purchase of Activision Blizzard King (ABK) on a nonsensical basis. Whether she trusted the wrong people, overestimated her own understanding of technology markets, or couldn't resist the temptation to use her temporary power over one of the world's largest corporations, her agency's decision--which I diplomatically described as "based on misconceptions and speculative concerns"--cannot be rationalized. A Sky News interview (in which she obviously wasn't going to admit anything) suggests to me that the magnitude of her agency's mistake has dawned on her. I also said so on Twitter.
Contrary to what many analysts say, I still expect the transaction to consummate, albeit with what could (but need not) be a significant delay. In this blog post I'll outline why I think so.
The decision is not merely wrong. It's actually nuts. I'm sure it won't stand.
I called the Phase 1 findings "nonsensical" and thought the CMA was making progress. I did not expect them to go off the deep end. But instead of recognizing that there was only a highly speculative concern left--a concern over competition in cloud gaming--and deciding accordingly, the CMA claimed it had to "protect competition" in cloud gaming.
First, cloud gaming--games being streamed over an internet connection--is just a niche. I play games locally, everyone I know personally who plays games plays them locally. I know people who gave it a try, only to revert to local gameplay. The idea of running the program code of the game in the cloud and reducing the local device to a "dumb terminal" is intriguing in some ways, but the reality is that even a temporary degradation of an internet connection--or a hiccup somewhere along the way that all the data has to travel--will adversely affect gameplay. In a world where everyone has "fiber to the home" and carriers have virtually unlimited capacity that no one else wants to make use of for other purposes, it would work great. But that's not the world we live in, nor are we going to live in such a world in a few years' or even a decade's time. And even in a best-case scenario, there will always be some latency that locally installed games don't experience (no signal can travel faster than light).
It's a nascent market, meaning companies enter and leave all the time. One recent exit was Google Stadia, but Google has shut down hundreds of services over the years, a fact that (rightly) dissuaded game makers from investing in Stadia. The Inquiry Group's inability to understand the market pervades the entire 418-page final report, and one particularly good example is in paragraph 8.299. A submission by an unnamed competitor gets considerably more space than any other input on the question of how important ABK's content is for cloud gaming, though it's a total lunacy that did not even deserve a footnote:
"Another competitor [REDACTED] submitted that the failure of Google Stadia shows the importance of CoD [= Call of Duty] compared to other gaming franchises, and that CoD’s role in attracting consumers to platforms is not directly proportionate to engagement alone. This competitor submitted that [...] further submitted that, despite having these games, Google Stadia did not have CoD, and that this prevented Google Stadia from reaching a meaningful number of [users]."
It's preposterous to attribute the failure of Google Stadia--which was primarily due to Google's lack of commitment to anything other than search and Android--to the fact Stadia didn't have CoD. There was no Stadia competitor that used CoD to make Stadia fail.
The CMA failed to correctly determine market share. While Microsoft told the CMA that its xCloud had a 2021 market share in the single digits, the CMA apparently assumed that every Game Pass Ultimate subscriber was playing games on xCloud. In reality, the vast majority of them download games to their local devices and play there. Many Game Pass Ultimate customers couldn't do that even if they wanted: as a Twitter follower reminded me, xCloud is unavailable in most South American and Asian countries and in all of Africa. On this web page, one just has to click on "Xbox Cloud Gaming (Beta)" to verify that fact.
The most popular cloud gaming service is Nvidia's GeForce Now, and Microsoft and Nvidia entered into a 10-year agreement to bring Call of Duty to that one. Similar agreements were announced with Boosteroid, Ubitus, and British Telecom subsidiary EE. Just today, Microsoft announced an agreement with European cloud gaming service NWARE:
Excited to be working with @Microsoft to bring access to even more titles on the Nware platform to give our gamers the #freedom to play any game, anywhere! 🚀 https://t.co/XL5rdSgo7h
— NWARE (@playnware) April 28, 2023
I could go on and on, but I'll leave that for another day--or for multiple other posts--to show just how bad the CMA decision is. It's an utter absurdity. They conflated stores with streaming as I explained on Twitter (also see this tweet by a gamer with an MBA). They took absurd positions in connection with the Relevant Customer Benefits. But even if one ignored all of that, the remedy--blocking the deal--would be irrationally disproportionate at any rate.
That is the same Inquiry Group that totally embarrassed itself when it compared--in the provisional findings, which had to be revised--the benefits from hypothetical foreclosure over a five-year period to the opportunity cost for only one year. Most eighth-graders would be ashamed of that. One member of that Inquiry Group is actually a professor for financial economics, but I don't want to talk too much about the Inquiry Group now. Suffice it to say for now that they appear clueless with respect to the markets, products, and technologies at issue, which wouldn't have been a problem if they hadn't also decided to buy a number of ridiculous claims made by deal critics while hardly believing anything that Microsoft and ABK told them. Totally biased.
Why is the deal not dead regardless of the UK situation?
Let's talk about the practical and procedural implications.
The merger agreement lists the UK among jurisdictions where approval is required. Therefore, let's first look at how that approval can still be secured, and then talk about a hypothetical scenario in which it could not be, but the deal would be closed anyway.
Competition Appeal Tribunal (CAT)
Microsoft announced its appeal, and ABK said they were already working on it. Parties get about four weeks to do so, and in this case I guess it could be filed even sooner.
While any appeals court in the civilized world can easily see that the CMA's Inquiry Group made an incompetent and irresponsible decision, there are three problems here:
The UK is the only major antitrust jurisdiction in the First World where a competition regulator is afforded enormous deference by the appeals court. If the EU Commission makes findings, the EU General Court can overrule them on questions of fact or law, and purely legal questions can be appealed further to the European Court of Justice. In the U.S., a deal like this can close unless a federal district court orders a preliminary injunction, meaning the regulator has to go to court and ask for it, but a court decides. In the UK, they have that "Judicial Review" standard, which means that unless one shows unlawful conduct or procedural mistakes (a basis on which Apple recently defeated the CMA, which thought it was above the law when it disregarded statutory deadlines), a decision must be shown to be "irrational".
An outright reversal that simply replaces an irrational CMA decision with a rational court ruling is not always possible. Quite often cases have to be remanded to the CMA for further proceedings.
All of this can take a while. Even if the initial appeal is resolved after six or nine months, a remand will take several more months.
While I can understand the UK government's desire not to be restricted in its executive freedoms by courts all the time, the problem in the field of competition enforcement is that when bad things happen, you really need a mechanism to counter regulatory bias and/or excess and/or incompetence. The CMA's CEO and this particular Inquiry Group were likely emboldened by the relatively strong shield that is the Judicial Review standard.
I'm not worried about the standard because the decision is really ridiculous and the mistakes are not going to be hard to show to the judges. I also think that the CMA has produced so much material that the appeals court may even find a way to just resolve the matter directly.
The issue is time. The merger agreement expires on July 18. But there are three possible solutions:
Given how obviously crazy the CMA decision is, I'm wondering whether the appeals court could order an interim measure that would allow Microsoft to close the deal while the appeal would be resolved. I haven't been able to find out whether that is possible, but in such an exceptional case I wouldn't be surprised if an emergency motion succeeded.
The CMA could look at the appeal, recognize its errors like it did in the console foreclosure context, and settle, possibly based on slightly expanded remedies.
Microsoft and ABK could work out an extension of the merger agreement. It looks like both companies are absolutely committed to this deal, and are not going to let a stupid mistake by a small group of people kill a deal that makes so much sense.
ABK's CEO Bobby Kotick said some interesting things in an interview. He noted that the CMA was being constructive and solution-oriented until Sarah Cardell had a meeting with the U.S. FTC's chair Lina Khan, an extremist who would like to be able to prohibit any merger above 5 billion dollars for any reason or no reason. I'm not concerned about that, but about whether the rule of law is sufficiently intact in the UK to correct mistakes or malfeasance. The more interesting part is that there was a case in which a deal that the CMA originally blocked was cleared last year: FNZ / GBST. What I noticed is that three of the four members of the FNZ Inquiry Group were also on the ABK Inquiry Group, among them the chairman as well as the "math whiz" I mentioned before.
In the FNZ case, the CMA opened its investigation in November 2019, and it took more than two years--until January 2022--before the deal finally got cleared. It can't and won't take that long in the ABK case.
Ministerial override
Yesterday I watched a recording of a BBC interview with Microsoft's president, Brad Smith. The two key takeaways were that Microsoft is going to fight--they've been more cooperative with regulators than any Big Tech company in history, but enough is enough in the CMA's case--and that there is a case here for a ministerial override.
That is a possibility I mentioned on Twitter back in February (1, 2). Four years ago, the UK [u]pdated [its] rules to strengthen the government's powers to scrutinise mergers and takeovers that may raise national security concerns." That page also explains the following:
"Ministers can only intervene in mergers and takeovers (foreign or domestic) that give rise to specific public interest concerns of national security, financial stability or media plurality." (emphasis added)
Mr. Smith didn't say "ministerial override", but he did refer to Microsoft's relevance to the UK economy (even including its support of non-profits, which are particularly in the public interest) and its efforts to defend the nation from cybersecurity threats.
The fact that the UK government initially stressed the independence of the CMA means nothing. First, the CMA is not really independent. The government could restructure it or replace people. Second, independence is not the same as immunity from a higher authority overruling an agency that would otherwise be out of control.
In Germany, it's not uncommon at all for the minister of economics to allow a merger that the Bundeskartellamt (Federal Cartel Office) would block.
It would make a lot of sense here, especially in a scenario where the European Commission would have cleared the merger, which would isolate the CMA apart from the ideologically blinded FTC that loses court cases all the time.
British politicians should see that the decision is unbelievably wrong and irresponsible, and that the UK's competition enforcement regime would become much more balanced if the CMA got overruled more often, especially the current CMA that is on the wrong track.
It wouldn't have political implications. We're not talking about compromising the independence of a court. We're just talking about a government agency, and one that is run by someone who has been at the helm for only a short period and already brought about a major crisis. The average British voter would ask: "Sarah Who?"
The question is not one of whether a corporation or a government have more power. It's that a blatant and irrational abuse of governmental power has created a situation that must be addressed.
A ministerial override wouldn't happen overnight, and it could be that the appeal has to be filed first. The CMA would be given a chance to recognize and correct its error. If all else failed, the override could happen, and it would be the right thing to do in that situation. Let's also not forget that someone has political responsibility for certain appointments that have led to this situation.
Closing the deal regardless
The UK-approval requirement in the merger agreement wouldn't even have to be renegotiated; it could simply be waived. It's only there so Microsoft wouldn't have to close the deal against a prohition decision by the CMA if it didn't want to, but if it decided to do that, it would be free to do so.
The consequences of going ahead and closing the deal against the CMA's blocking decision would be massive. It's hard to imagine and there is no indication whatsoever that Microsoft is contemplating that possibility, much less have I asked them whether they would consider it, but in a strictly logical sense it's not 100% inconceivable in a scenario where numerous regulators (seven countries so far, and the EU likely to follow soon) would have cleared the deal.
Let's look at it this way: if the transaction could otherwise go ahead, but the competition authority of a tiny country--say, the Amt für Volkswirtschaft of the Principality of Liechtenstein--suddenly entered a prohibition decision, the deal wouldn't fall through. Even if the government of Liechtenstein--the sixth-smallest nation--banned all of Microsoft's products and services in retaliation, so what?
Now, the UK is the fifth-largest economy, and that's why it's so hard to imagine. The CMA could try to impose fines on Microsoft, but those decisions would be subject to judicial review. And they wouldn't be practically enforceable if Microsoft left the market. Would they do that? Interestingly, Apple's lawyers once told a UK judge they might leave that market. I don't think that Microsoft would do what Apple threatened to do. I'm speaking only in the hypothetical. And I wouldn't even do that if Apple had not contemplated the possibility of simply leaving the UK (in that case the issue was patent royalties).
What would most British companies and citizens prefer? Windows, Office, Teams, Azure, and even the Xbox--or a regulator whose leadership doesn't agree with Spider-Man that great power comes with great responsibility? An Inquiry Group that doesn't even get its basic math right?
Whatever will happen, the transaction is more likely than not to be unblocked. The decision that came down on Wednesday fails consumers and companies alike. Except Sony, Apple, and Google. Sony dominates the worldwide console market, and Apple and Google know that Microsoft is in the best position to challenge their mobile app store monopolies.
