Sunday, January 27, 2019

Changes in Qualcomm's market power over the years and their relevance to antitrust remedies

Tomorrow (Monday) the FTC will get a couple of hours for its rebuttal, and its sole live witness will be Professor Carl Shapiro. Qualcomm's experts tried to discredit his analysis. Reasonable people can always disagree on methodology: economics is not an exact science. But those Qualcomm experts taking aim at Professor Shapiro's approach by claiming it wasn't sufficiently numbers-driven were, to put it mildly, a case of the pot calling the kettle black. One of them, Dr. Chipty, couldn't deny having failed to consider more than 200 million baseband chips in Samsung phones; two others, Professor Snyder and Professor Nevo, had little more to offer than a convoluted version of "you can't argue with success." The latter spouted nonsense about Microsoft's past per-hardware vs. per-software operating system licensing strategy and the practical implications of chipset licensing, and in a language spoken widely in San Jose and even more widely in Qualcomm's town, he was simply "hundido" after he had to concede his FRAND "analysis" ignored key contract terms and multiple major licensees to the point he might have excluded most of the market.

As I said before, I thought Qualcomm could win the battle of the experts, and the FTC would overcompensate for a defeat there (provided it would have been narrow) with its huge victory related to industry testimony. But where things stand now, even prior to Professor Shapiro's rebuttal, the FTC has the upper hand in both respects. It does have the burden of proof, and I don't mean to say the FTC would necessarily win on all counts, but I really can't imagine Qualcomm could successfully defend itself all the way. My prior predictions in this case played out as follows:

The FTC and Professor Shapiro have to take Qualcomm seriously, just like the way to win a sprint is to keep going at full speed for a few more yards. But I'm just an observer, and I'm already looking past the question of merits and on to the subject of remedies.

In order to preserve the record for an appeal, Qualcomm filed an 11th-hour offer to submit evidence related to the post-cutoff period, such as dozens of additional (5G-related) license agreements. Judge Koh denied the motion (as Qualcomm knew she would), just like she had denied a closely-related one in December. And just like before, Qualcomm stressed that it primarily wanted the additional material to be considered with a view to the prospective remedy that is injunctive relief.

Qualcomm argues that injunctive relief is no longer warranted because, according to testimony it elicited, the big three device makers that use premium LTE chips--Samsung, Apple, Huawei--either buy them from Intel (Apple) or build a high percentage of them internally (Huawei has HiSilicon, Samsung has Exynos). Therefore, Qualcomm says it no longer has market power in premium LTE chips, and injunctive relief wouldn't be warranted anymore even in the event of the FTC convincing the court of anticompetitive conduct and harm in the past. During the trial, Qualcomm also elicited such testimony on every occasion, though it had to ask witnesses to limit their answer to the state of affairs as per March 2018, just to comply with the cutoff date.

The FTC already addressed this, without a lot of detail, in the final section of its pretrial brief.

I don't want to go into full detail either, but I want to explain a few general principles here, especially since some Qualcomm-aligned Internet trolls and fake analysts try to make this question of current (or March 2018) market power look like a "get out of jail free" card, which it definitely isn't:

  1. The relevance of market power is a statute-by-statute question. As the DOJ explains, "a finding of market power is a prerequisite to a section 2 [of the Sherman Act] violation." But the FTC also has Section 1 theories in this case.

  2. My favorite cause here is the duty to extent SEP licenses on FRAND terms to rival chipset makers. The market power from Qualcomm's standard-essential patents, however, is monopoly power (because if a patent is truly standard-essential, you could exclude someone from an entire market with it), and it is powerful enough even without a single chipset being sold. All FRAND-related rulings in the case law were based on the power of SEPs, not that of chips.

  3. As the FTC explained toward the end of its pretrial brief, antitrust injunctions have been granted in other cases based on past market power.

  4. In practical terms, courts must have some discretion with respect to cutoff dates. Otherwise an antitrust defendant could constantly produce new "evidence," and justice delayed would be justice denied.

  5. Even if one or more of the FTC's injunctions were denied for a combination of diminished market power and Section 2, Qualcomm would still be held responsible for its past conduct. The FTC is seeking not only injunctive but, prior to that, also declaratory relief. The FTC itself isn't seeking damages. But:

    • A consumer class action seeking $5 billion to be distributed to up to 250 million people is part of this case. Judge Koh has stayed it because Qualcomm successfully petitioned the Ninth Circuit to hear an interlocutory appeal against class certification. This means Qualcomm has a chance to get rid of the consumer class, but it's far from a foregone conclusion that the class certification will be overruled. Ironically, if Apple loses Apple v. Pepper in the Supreme Court (my impression of the November hearing was that only the Chief Justice is firmly on Apple's side, and no one appeared to agree with Apple on a statutory basis--at best there's just some case law that could be overruled), it will adversely affect the prospects of Qualcomm's class certification appeal. In the most extreme case, Qualcomm might then be left with just a feasibility argument based on the size of the class, pointing to pre-Internet-era case law on how to manage large classes that just doesn't apply to our cloud age.

    • Apple, regardless of that App Store case, has its claims against Qualcomm in San Diego (the trial will start in mid-April). MLex attended a motion hearing there on Friday while also following the FTC case.

    • Anyone else such as Intel or Huawei could bring damages claims against Qualcomm following a declaratory judgment in the FTC's favor regardless of whether or not the FTC is granted injunctive relief.

Some of the stuff in this case may be water under the bridge. For an example, Qualcomm's exclusive deal with Apple is history because Apple is now buying baseband chips from Intel--but even the question of whether there would (if not for Qualcomm's conduct) have been true competition in the merchant (= chipmakers willing to sell to third parties) market for premium LTE chips at an earlier stage could still give rise to significant damages claims. In that particular context, Qualcomm's best shot is its claim that the deal allegedly didn't actually cause anticompetitive harm because, in Qualcomm's opinion that was supported by some of its witnesses, Intel couldn't have supplied chips meeting Apple's requirements. But the question of licenses to rival chipset makers is going to remain important for many years--even decades--to come. The procompetitive effects of the FTC prevailing on that claim would be immeasurable.

Share with other professionals via LinkedIn: