Tuesday, January 15, 2019

After mixed Day 5 of FTC v. Qualcomm antitrust trial, irrelevant German ruling to be handed down later

[Update] The Mannheim Regional Court indeed threw out Qualcomm's complaint for non-infringement based on claim construction (the first of three possibilities I discussed further below). [/Update]

Day 5 of the FTC v. Qualcomm antitrust trial just concluded. It was the most eventful and interesting day of the first half of this bench trial. The morning and the early afternoon were a total disaster for Qualcomm, but toward the end Qualcomm had its strongest hour to date.

I usually write about these long and tiring trial days the next morning. I'll do so again, given that we learned some interesting numbers and other relevant facts, but here's a very short summary--and then I'll provide a preview of the Qualcomm v. Apple patent infringement ruling coming up in Mannheim at 9 AM local time there (midnight Pacific Time):

  • The first witness this morning was Apple COO Jeff Williams. Some of what he said effectively nuanced what Qualcomm CEO Steve Mollenkopf said on Friday about Apple's alleged proposal to agree to exclusivity. Mr. Williams provided the court and trial watchers with interesting and credible information on various other issues. He was the first witness to be rudely interrupted by Keker van Nest & Peters' Eugene Paige, but in his low-key, friendly, unpretentious way drove all the important points home.

  • The FTC staff then read into the record some testimony by an LG Electronics IP executive that showed, among other things, how Qualcomm bullied LG when it challenged in arbitration (which wasn't even remotely as threatening to Qualcomm's business model as a challenge in a court of law would have been) Qualcomm's licensing terms.

  • Then came an FTC licensing expert, Richard Donaldson. Against his background of 31 years (!) of patent licensing work at Texas Instruments, and many years of consulting and expert testimony since he retired from TI, Mr. Donaldson explained how atypical the licensing terms Qualcomm imposes on licensees are, and he also explained that chipset-level licensing is as feasible as it is commercially viable.

    The aforementioned lawyer, Mr. Paige, cut off Mr. Donaldson's answers, and even though I think Mr. Donaldson could have given a tactically better answer to the question of whether TI sought to avoid the exhaustion of system-level (= device-level) patents through chipset licensing (he could have pointed out that this is a non-issue if a chip-level license deal involves only specified or otherwise clearly-defined chip-level patents), I think he dealt with Mr. Paige's onslaught very well. At some point Judge Lucy H. Koh intervened against those interuptions of the witness.

  • After the lunch break (which was in the middle of Mr. Donaldson's testimony) Judge Koh came back to the issue of Mr. Paige's constant interruptions and told him not to cut off a witness "after two words." She called this "improper" (I couldn't agree more--it was extremely annoying) and warned that the next party to do so would lose two minutes of its trial time.

  • The FTC played some testimony by Ericsson's licensing president Christina Petersson. While Ericsson is also very much into patent monetization, especially standard-essential patent (SEP) monetization, she said various things that seriously undermine some of Qualcomm's positions and defenses, such as with respect to Qualcomm's licensing terms being unusual (the only chipset-level outbound license deal Ericsson still has in place is with Qualcomm). Ericsson sought to position itself as a reasonable SEP licensor (I've repeatedly criticized Ericsson on this blog, but they're not the worst for sure) who understands every SEP holder has to consider he doesn't own the standard alone "in order for the system to work." In other words, Ericsson disagrees with Qualcomm's extreme demands because if everyone did that, the collective impact of royalty-stacking would be devastating. Ericsson even undermined Qualcomm's position with respect to chipset-level licensing, saying that despite the agreement it still has in place with Qualcomm, Ericsson continued to invest in research and development as well as standard-setting.

    With respect to LTE, Ericsson believes it has the strongest portfolio in the industry, with Qualcomm "definitively" weaker in Ericsson's opinion, and with some allegedly believing Nokia is also ahead of Qualcomm.

    We've now heard pretty much every significant industry player. No one is on Qualcomm's side in every respect. There are some who are 100% against Qualcomm, and some who at least disagree with Qualcomm on some key issues.

  • Then came Michael Lasinski's testimony. He's another licensing expert testifying on the FTC's behalf. His background is impressive: a former president of the U.S. & Canada chapter of the Licensing Executives Society, and a former American Bar Association IP division chair. A couple of my Twitter followers immediately vouched for him when I mentioned him.

    After an otherwise awful day for Qualcomm, we saw a world-class performance by Cravath Swaine & Moore lawyer Gary A. Bornstein. He struck just the right balance between being assertive and respectful, and he managed to highlight a number of issues relating to Mr. Lasinski's methodologies. Question by question, Mr. Bornstein cornered Mr. Lasinski and forced him to concede limitations and shortcomings, and in one context even an outright contradiction. Mr. Bornstein won this fight by a wide margin, though Mr. Lasinski made a stronger showing initially than the other expert, Mr. Donaldson, on cross-examination.

    While the positions Qualcomm takes on economic expert testimony are way too demanding (no government agency or company in the world could possibly satisfy them), there are indications that Mr. Lasinski's analysis could be the, relatively speaking, weakest link in the FTC's chain. It remains to be seen tomorrow to what extent Professor Carl Shapiro's economic analysis renders those deficiencies less relevant.

  • Apple contract manufacturer Wistron (which was spun off from Acer a long time ago) confirmed what others said about Qualcomm simply not making concessions on key licensing terms. He said they ended up agreeing on a huge upfront payment to Qualcomm, and in order to recoup that one as soon as possible, they couldn't work with other baseband chipset makers though there would been economically attractive options.

So there was a lot of shadow for Qualcomm, but also a silver lining for them with respect to Mr. Lasinski's methodologies.

After the first half of the trial (not counting the closing argument scheduled for February 1, 2019), I believe Qualcomm made the mistake of having too many cooks in the kitchen and partly the wrong cooks in certain places. The name of the game is not how many law firms you get involved. It's how effective they are, how "suitable to task" in tech lingo.

Cravath is Qualcomm's lead counsel against Apple, but the lead here in San Jose was given to Bob van Nest. Mr. van Nest himself has class and style, though he's primarily good at jury trials, and Judge Koh is the very opposite in terms of competence and professional coolness from a layperson jury. He's got a great reputation in this district, and Cravath is HQ'd in New York State.

But in retrospect I believe Qualcomm should have given Mr. Bornstein the lead here. He's been lead counsel in other high-profile antitrust cases. It seems to me that he's the smartest member of Qualcomm's trial team here, and generally all the Cravath lawyers here appear classier and more effective than the Keker van Nest lawyers apart from Mr. van Nest himself. There really is a very noticeable difference between one of the most reputable law firms in the country and a regional player good at misleading juries such as in the Oracle v. Google case, where they had a great jury strategy and benefited from a judge who made some key decisions against Oracle that the appeals court unanimously overruled in two different years. By contrast, even if Qualcomm loses (and I still think that's more likely than not to happen), I don't think anyone can blame Cravath.

The next news cycle related to Qualcomm's patents is only about six hours away. At midnight Pacific Time, or 9 AM Central European Time, the Mannheim Regional Court will announce decision, which can be a final judgment or a procedural order, on one of Qualcomm's German infringement lawsuits against Apple targeting Intel-powered iPhones. The patent-in-suit is EP2460270 on a "switch with improved biasing" ("biasing" in this context basically meaning that one voltage gets to control another).

I'm not going to stay up, or get up at midnight, for a nuisance lawsuit (which is all that this one is in practical terms), but I'm sure many will hear about the decision, so I'll quickly explain what may happen and why it's a pointless lawsuit in any event.

  • If the court clears those iPhones of infringement, it will most likely be because the court, more or less sua sponte if I understood it correctly (I watched the trial a few months ago), developed a claim construction approach based on an unasserted parallel claim.

  • The court might also stay the proceedings pending a parallel nullity action. The Swedish patent office provided an opinion according to which there's nothing inventive about that patent. And it did so on an independent basis, without any specific theory being presented to the examiner (just the prior art references).

  • Even if the court refrains from deciding the case based on its great claim construction idea and also declines to attach weight to the Swedish patent examiner's analysis, and formally enters an injunction, so what? At trial it was undisputed between the parties, and mentioned in open court, that Pegatron, one of Apple's contract manufacturers, is licensed to the patent. So there wouldn't be any bottom-line impact on the availability of any iPhone model in Germany.

Among various strong points, the weakest point Mr. Bornstein made in his cross-examination of Mr. Lasinski related to a hypothetical question they had asked him in early 2018 about whether he'd consider Qualcomm's non-SEPs more valuable if Qualcomm obtained injunctions against Apple in foreign courts. We know that some injunctions have come down: two in China (over two patents), and two in Munich, Germany (over the same patent in both Munich cases, just targeting different Apple entities). The problem with Mr. Bornstein's attempt to leverage those decisions now is that those injunctions still don't give Qualcomm any serious leverage.

The Chinese patents-in-suit have, according to Aple, been worked around by iOS 12. We'll have to wait until it becomes clearer whether the Chinese court agrees, but given that those are non-standard-essential software patents, it's perfectly plausible.

The Munich decision affects only 3% of Apple's German sales of the iPhone 7 and of the iPhone 8: direct sales to end users through its 15 retail stores and its German online store. Even those 3% of the sales of the two oldest iPhone generations on sale in a market that generally isn't huge for Apple (significant, but far from substantial) aren't really lost because some will buy other iPhone models instead and others will simply buy those iPhone models from resellers.

Anyone can see on the Internet that the "enjoined" iPhone models are still widely available in Germany.

Even the best-case outcome of the Mannheim case up for a decision in about six hours would fall short of the minimal impact of the Munich injunction: the worst-case scenario for Apple, based on what was said at trial (and in the court's preliminary opinion militated against an antitrust issue int his case), would simply be that all German iPhones (of the affected generations) would have to be manufactured by Pegatron.

Another Mannheim decision is scheduled for February 19, 2019 in a case in which the court strongly suggested to Qualcomm to stipulate to a stay given serious doubts about the validity of the asserted claim.

In between those two Mannheim decisions, the Munich I Regional Court will rule (on January 31) on a bunch of lawsuits related to a patent family that Qualcomm is asserting against Apple's Spotlight search. iOS 12 contains a workaround based on what was discussed at trial.

Share with other professionals via LinkedIn: