Wednesday, January 16, 2019

The FTC has rested its likely-winning case with a final hand grenade destroying a Qualcomm mantra

Yesterday (Tuesday, January 15) was Day 6 of the FTC v. Qualcomm antitrust trial in San Jose (Northern District of California), which will continue in Judge Lucy H. Koh's court on Friday, January 18.

When FTC lead counsel Jennifer Milici said "Your Honor, the FTC rests its case," you could have heard a pin drop even if there had been thousands of people in the audience, provided that all of them would have been reasonably knowledgeable. That's because seconds before that procedural notice, former Qualcomm licensing president Marvin Blecker had been confronted, in a videotaped deposition, with an internal email in which a colleague confirmed to him that Qualcomm's chip division had actually held product shipments to a customer who had not yet accepted Qualcomm's license terms.

It had been Qualcomm's mantra all the time that they had never actually carried out the "No License-No Chips" threat against existing customers. They couldn't deny that they wouldn't accept a new customer prior to taking a license from them. And an Apple witness said that after Apple sued Qualcomm (in January 2017), Qualcomm refused to even discuss a potential 5G product partnership with Apple. But again, Qualcomm's lawyers had over and over again--you could set your watch by them--elicited testimony from current and former Qualcommm executives that the company had never carried out the threat of holding chipset shipments to an existing customer due to disagreements on licensing.

In the seconds before resting its case, the FTC made them all look like...well, I don't want to use the harsh words that I actually think would be warranted here, so let me just call them "unreliable witnesses and lawyers you better don't trust in this context."

CDRinfo, a hardware-focused website, has also reported on that testimony.

The actual legal significance of whether or not the threat ever was carried out shouldn't be overrated. It's not like robbing someone at gunpoint is allowed; illegality starts long before pulling the trigger. Also, "No License-No Chips" is just one of the FTC's four theories in this case, though it is, as the FTC's economic expert Professor Carl Shapiro (UC Berkeley) explained yesterday, one of three types of conduct that are closely interrelated.

I've previously used chess terminology in the context of this trial. With respect to No License-No Chips, the FTC has taken out what Qualcomm portrayed as its queen, even if in my opinion it was a smokescreen, at best a pawn. With respect to Qualcomm's credibility in this case, it was at least a rook.

Qualcomm's case-in-chief began yesterday with testimony that's largely irrelevant to the outcome-determinative issues. Judge Koh wasn't nearly as busy as she normally is during testimony. More about that later.

During the FTC's case-in-chief, Qualcomm's cross-examinations were largely ineffective, with only two exceptions. The first one was MediaTek's testimony, which was well-meant but which Qualcomm was able to use in part to make a small dent in the FTC's argument with respect to the licensing of rival chipset makers. The second and far more significant one reminded me of Davis Cup tennis matches where the strongest player of one team played against the weakest one of the other. Here, the unbalanced matchup pitted Cravath lawyer Gary Bornstein (for Qualcomm) against one of three FTC expert witnesses, Michael Lasinski, whose methodologies are indeed questionable in at least some respects. Maybe his report should be called the "Lazynski report" because he didn't make the necessary effort to underpin his top-down analysis more solidly, even if it might have been painstaking. The way Mr. Bornstein handled the cross-examination was a demonstration of clever legal work, measured but forceful courtroom demeanor, and raw analytical brilliance. After that experience, I wonder what I may hear from Mr. Lasinski later this year when attending Huawei v. Samsung (which I probably will), where he's actually supporting Qualcomm-like behavior on Huawei's part and, apparently, taking unreasonable positions on royalty levels.

This leads us to the FTC's third (and by far the best) expert witness, Professor Shapiro, who testified yesterday. He was relaxed, poised, and you could see his experience in teaching because he explained everything in the simplest, most understandable terms possible. The only audience that mattered here was one person: Judge Koh. (And later on, the appellate judges, who will see the transcript.) If her FTC v. Qualcomm judgment becomes a landmark ruling (like her summary judgment decision on rival chipset licensing was one of the most important pretrial decisions in the history of this industry), she'll probably be widely regarded as the world's number one technology industry judge, a position for which one obviously needs to be in Silicon Valley, but just being there doesn't suffice.

There's so much to say about Professor Shapiro's testimony that I'll do a follow-up post at some point, either based on my notes or on the transcript whenever I obtain it. In this post here, I'll just summarize what's most important with a view to how strong the FTC's case is:

  • Obviously, Professor Shapiro supported the FTC's theories.

  • He's an expert in antitrust economics, also at the intersection of antitrust and IP. According to his co-author, Stanford Professor Mark Lemley, Qualcomm has already spent millions of dollars attacking them over their papers on standard-essential patents (SEP) holdup:

  • On the basis of Qualcomm having imposed supra-FRAND royalties (he didn't say supra-FRAND yesterday, but meant it), the economic effect of the allegedly anticompetitive conduct had three components:

    1. reduction of rival chipset makers' volume/market share;

    2. reduction of device makers' margins (that's because of them absorbing part of the increased cost); and

    3. the most critical part in U.S. antitrust law: higher prices paid by consumers than otherwise (that's the part device makers passed on).

    Professor Shapiro didn't call it a vicious circle, but he did point out that the effects of these dynamics get worse because of a reduced competitiveness by other chipset makers.

  • He testified to Qualcomm having had a premium LTE baseband chip monopoly until including 2016. He acknowledged a slowly-decreasing market share, but he did not take a position on Qualcomm's position in LTE in 2017 or 2018, much less on Qualcomm's future 5G market share in 2019. To be clear, he did not say Qualcomm didn't have or wasn't going to have market power in those other years; those years were simply outside the scope of his task.

  • Professor Shapiro pointed out that the fact Qualcomm was and is innovative and made and makes great products is unrelated to the behavioral issues in this case. When Qualcomm argued that other companies just didn't have equally good chips, he said that healthy competition in a market also means that "companies who are not quite as good" have a chance to challenge the incumbent.

  • He described No License-No Chips as "a very heavy hammer that Qualcomm is bringing down on OEMs at least as a threat."

  • He basically pre-empted the testimony we were then going to hear from a true technology giant, Qualcomm's co-founder and first (and long-time) CEO, Dr. Irwin Jacobs. Qualcomm called Dr. Jacobs, and he told great stories, but mostly they had nothing to do with the issues before Judge Koh. Large parts of his testimony would have been perfect just one block away from the courthouse: for a presentation at the Tech Museum. To be clear, I really do have the greatest respect for Dr. Jacobs and I love people who successfully swim against the tide, as he and his company did with CDMA. I also appreciate the Twitter like I got from him, provided that @jacobs_irwin is not a fake account (never know). But the case is not about how Qualcomm started; it's about what it did after it succeeded. Professor Shapiro said the following:

  • Qualcomm's lead counsel Bob van Nest tried something that was just a non-starter: when cross-examining Professor Shapiro, he argued that Intel simply had enough money to compete with Qualcomm in the baseband chipset business. This was a typical point to make in front of a jury, just like he (or someone else from his firm) asked Oracle President Safra Catz (in Oracle v. Google) about the tens of millions she makes per year. With such "deep pocket" stories you can appeal to jurors' envy and/or make jurors subscribe to the fallacy that money buys everything, but you won't impress a pro like Judge Koh. Professor Shapiro calmly and with a smile pointed out that as an economist, he does not look at it that way: what matters is not whether someone has money (let me explain: in economic theory, the assumption is that money will ultimately find its way to where more money can be made, through whatever funding mechanism, be it lending or venture capital or public markets), but whether (not exactly his words) a business would be viable on its own.

  • Professor Shapiro had to suffer some more economic nonsense gladly--because of a Qualcomm expert report we'll hear about more during Qualcomm's case-in-chief. He was asked about Qualcomm's expert's claim that rival chipset makers were "freeriders" because Qualcomm doesn't proactively sue them. The FTC wanted to know how he'd respond to that, and with a smile he said he was "responding badly" because it's economically wrong. Two weeks ago I addressed the same idiocy (but without reference to any expert report, just to another pretrial filing) on this blog.

  • In one (and only one) significant respect, Mr. van Nest's cross-examination effort was successful: Professor Shapiro ultimately had no choice but to acknowledge that his economic theory of harm (ultimately also consumer harm) is based, as an essential "building block" as he conceded, on the starting point of a supra-FRAND patent royalty being charged by Qualcomm. Mr. van Nest wanted to get him to confirm that any conclusions would be "irrelevant" without that part, and I agree with Mr. van Nest in that regard (for a change).

One can sum it up like this: there was nothing Qualcomm was able to do to reduce Professor Shapiro's credibility. Nothing that Qualcomm could really do to seriously question his methodologies (unlike in the "Lazynski" case). The only thing they can try is to leverage the shortcomings of the Lasinski report and testimony against the Shapiro report--but that's not nearly as easy to achieve as it sounds:

  1. Judge Koh would have thrown out the Lasinski testimony in its entirety if it had seemed wholly unreliable to her. In other words, whatever he got to present and testify here is stuff that she may give more or less weight to, and Qualcomm made strong points against relying on that report in every respect. But realistically, some parts will still get at least a bit of weight. Qualcomm's problem is that its royalties are so out of line that even if one discounted the numbers in a report like Mr. Lasinski's, one would still find them to be supra-FRAND.

  2. Professor Shapiro made two clever moves to maximize the independence of his report from the Lasinski report:

    • He explained that using general economic bargaining theory, his report arrives at the conclusion of a supra-FRAND royalties on its own.

    • He furthermore pointed again and again to the trial testimony of device makers and rival chipset players. I believe that testimony is so incredibly strong that I'm not sure the court even needs any expert report at all to conclude that supra-FRAND royalties were imposed. It's not just that virtually the whole industry says so, but also that witnesses pointed to the disproportionate percentage of overall cellular SEP royalties paid in this industry that goes to Qualcomm (80%-90% in Huawei's case, for instance). Even Ericsson, which has interests partly aligned with Qualcomm's, undermined Qualcomm's FRAND-compliance claims.

  3. There's also the Donaldson report on "atypical" licensing terms. Qualcomm wasn't nearly as successful attacking this licensing expert with 31 years of experience at Texas Instruments as they were when attacking Mr. Lasinski.

The Lasinski report is, and will be (presumably also on appeal, where Qualcomm can reargue any Daubert points), Qualcomm's primary attack vector, but it's not like a chain that's as strong as weakest link: it's more like there are various other ways to build a chain, and Professor Shapiro has shown the way already.

By insisting on a FRAND determination, Qualcomm sought to reargue its 2017 motion to dismiss, which Judge Koh denied. The related order already took the perspective that many roads lead to Rome, to put it that way. The FTC doesn't need to have three perfect expert reports. "Two out of three ain't bad."

The FTC's litigation staff can be proud of the tremendous work they've done (and imagine the distractions resulting from the government shutdown). The case the government has rested is extremely strong, and I already have this feeling that this "Your Honor, the FTC rests its case" moment may have been one of the most important moments in worldwide antitrust history.

Nobody is perfect, and no case is perfect. Qualcomm's case-in-chief has begun, and we'll now hear testimony from partly the same witnesses (but parts of their videotaped depositions that Qualcomm views as helpful) and Qualcomm's preferred witnesses. It's just that almost all the major players have spoken already, including Qualcomm's CEO and Qualcomm's president. And yesterday we got history lessons and general talk about engineering culture (from Qualcomm's senior VP of engineering, Dr. Durga Malladi) that suggested Qualcomm doesn't even know how to kill all the trial time it has, while the FTC (which had even more testimony on the list, but then ran out of time and stopped at the best possible moment) could presumably have gone on an on for weeks with evidence buttressing its theories.

In order to turn this around, or at least to have a reasonable record for an appeal, Qualcomm will now need several hand grenades of the kind the FTC detonated seconds before resting its case. I will stay tuned. I know many of you will, too. (By the way, I wish to thank the readers who said hi at the courthouse.)

Share with other professionals via LinkedIn: