Thursday, January 24, 2019

Comments on Kerrisdale Capital's analysis of Qualcomm's antitrust-related risks

Yesterday I became aware of a 19-page document (PDF) explaining why research-centric investment group Kerrisdale Capital has decided to shortsell Qualcomm's (QCOM) shares, with a particular emphasis on the fallout from the ongoing FTC v. Qualcomm antitrust trial in the Northern District of California. Unlike a lot of fake analysis on the Internet that looks like having been funded and more or less dictated by Qualcomm, Kerrisdale has done some very thorough work (they appear to have obtained all trial transcripts; I must admit that, since I attended eveything, I haven't even done so yet), and most importantly, they put their money where their mouth is. Even more than that, the risk of shortselling is substantial (theoretically infinite, though one can limit the risk with call options, which they may or may not have purchased). All I can say is that I'm extremely risk-averse and don't short as a matter of principle.

Kerrisdale's PDF document links to this blog on several occasions, and this is what they have to say about yours truly:

"Florian Mueller, an anti-software-patent advocate who has followed the various legal cases against Qualcomm very closely for years and is attending the current trial in person, [...]"

"To be sure, Mueller can be regarded as a biased observer in light of his personal opposition to Qualcomm's business model, but his analytical track record is nothing to scoff at; for instance, he correctly predicted that the FTC would win its important motion for summary judgment."

All of the above is fair--or "FRAND" :-)

It's true that I got into the patent business almost 15 years ago as an anti-software-patent campaigner, and when software patents are at issue in the cases I blog about, I sometimes mention that fact.

What I'd like to add, however, is that my advocacy of giving FRAND meaning is almost as longstanding as my opposition to software patents. My first write-up promoting FRAND principles was a submission to the European Commission I authored for the world's most famous sports club, Real Madrid, in 2007. The bone of contention was soccer broadcasting rights in the context of wider sports governance principles. Sports bodies are typically monopolists and their tournaments can pratically be to sports teams what one would be more inclined to call an "essential facility" in other areas. Therefore, organizations like UEFA must show that their rules are FRAND (many of them still aren't, but that's another topic).

When this blog was less than a year old, it had already taken the position that, under the law, injunctive relief should not be available over standard-essential patents (SEPs). At the time, there were only a few companies in the industry sharing this view as a matter of policy. Apple was embroiled in its first dispute with Nokia (and shortly thereafter, Motorola, followed by a Samsung complaint); and the likes of Broadcom and Cisco took similar positions, and I believe Intel, too. But even Microsoft (which made FRAND history with its antisuit injunction against Motorola and was already a client of mine at the time, but like all other clients I've ever had, respected my independence) would still submit a position paper to the FTC in 2011 that sought to justify SEP injunctions. Google, when and after acquiring Motorola, wanted SEP injunctions, as did Samsung--an by now, Microsoft, Google and Samsung are all on the FRAND side.

Qualcomm wasn't of great interest to this blog for a long time. From industry players like BlackBerry and Nokia I heard about them even before I started this blog in 2010. But prior to the current wave of antitrust enforcement and litigation, Qualcomm appeared on my radar screen only as a submitter of amicus curiae briefs (including one that bashed Apple as a patent infringer who should be embarassed, and which they withdrew), and because of a restructuring that appeared to be driven by concerns over patent exhaustion through open-source software. It was like lightning from the blue when the Korea Fair Trade Commission (KFTC) slammed Qualcomm with a fine between Christmas 2016 and New Year's Day 2017. Then came the FTC (U.S.) lawsuit and Apple's lawsuit in January 2017.

Since then, various Qualcomm investigations and lawsuits have become the industry's biggest battlefield ever. That said, I also see serious issues in Huawei v. Samsung, just that a Chinese-Korean dispute on U.S. soil gets a lot less attention than U.S.-only matters, though Huawei and Samsung are both known to more consumers than Qualcomm. I may also attend a motion hearing (about whether the jury should hear Samsung's claim that Huawei breached its FRAND obligations) in San Francisco on February 13.

There will be a time when Qualcomm is going to be of less interest to me again. But for the time being it's an interesting mix of issues that are presently unique to Qualcomm ("No License-No Chips"), yet could become a rampant problem if Qualcomm got away with such conduct, and overarching questions the industry cares about--first and foremost, the licensing of rival chipset makers.

Kerrisdale Capital has correctly identified that one--SEP licenses on FRAND terms to the likes of Intel--as the key aspect of the FTC's case, followed by a requirement to renegotiate existing agreements. Just like me, they don't seem to be overly interested in whether or not the FTC has established actual anticompetitive harm from an exclusive deal between Qualcomm and Apple (I view that one as water under the bridge).

The Kerrisdale analysis says that it should be easy for the FTC to prevail post-trial on its pursuit of an injunction requiring Qualcomm to extend FRAND licenses to rival chipset makers because of its summary-judgment win. I'd like to nuance that part a bit:

First, even if the FTC hypothetically (and unrealistically in my opinion) lost all post-trial decisions, the summary judgment on rival chipset licensing would stand unless Qualcomm successfully appealed it. Under its FRAND declarations made to ATIS and TIA (two U.S. standard-setting organizations), the likes of Intel are entitled to a license to all Qualcomm SEPs covered by those ATIS and TIA declarations, period.

Second, what the FTC is seeking now is a broader-based ruling. It would not be specific to particular FRAND declarations and dependent upon their interpretation, but basically say that only FRAND can preserve competition despite the exclusionary nature of standard-setting, and there's no reason why only device makers, but not chipset makers, should be protected that way. So this would involve more standards and more companies.

Three, some of the subissues, such as whether baseband chips materially implement the standards at issue, were decided in the summary-judgment context and Qualcomm's attempts to reargue those subissues at trial were a non-starter with Judge Koh, though Qualcomm obviously needs to preserve its record for an appeal.

I think the FTC is in a pretty strong position with regard to licenses to rival chipset makers, but it's basically seeking a very significant "upgrade" of the summary judgment. Let's not underestimate that hurdle and the significant further effects an FTC victory in this context would (and hopefully will) have.

Kerrisdale says this about the potential impact on Qualcomm's business, giving that two thirds of its profits historically came from patent licensing:

"Indeed, in the words of the United States Federal Trade Commission (FTC), quoting an internal Qualcomm assessment from 2015, 'granting a FRAND license to Intel 'would destroy the whole current QTL [licensing] business.''

"If Qualcomm is right about that, then destruction is imminent. [...]"

"Indeed, just last week the FTC confronted one of Qualcomm's witnesses, the company's senior vice president of licensing strategy, with his own past statement, recorded on audio tape, that when “having to choose between licensing chips and licensing at the handset, the handset was humongously more lucrative.' But this 'humongously more lucrative' way of doing business – part and parcel of Qualcomm's overall pattern of monopolistic tactics – will likely collapse when Judge Koh issues her final ruling."

Qualcomm's lawyers and witnesses made a lot of effort to provide legitimate business justifications for device-level licensing. But none of what they said is convincing, and even less so when considering that Qualcomm itself acts as a SEP clearinghouse for the wireless industry by demanding royalty-free, exhaustive cross-licenses that then benefit its customers when they have to defined themselves. Qualcomm's actions contradict its words, and even what it says isn't consistent because internal documents state the real reason for device-level licensing. What those Qualcomm executives said internally is perfectly consistent, however, with what Ericsson once said in public about why it chooses to strike license deals with device makers, not chipset makers.

Again, there is a hurdle involved (going from the interpretation of two contracts to a general FRAND obligation), but none of what Qualcomm has said in its attempts to explain why it shouldn't be obligated to extend FRAND licenses to rival chipset makers is even remotely persuasive--and this is a judge who won't buy things that juries or judges less familiar with these issues might give more weight to.

Kerrisdale's report is the only one apart from my own writings (and it's possible they got the relate hint from this blog) to highlight something important regarding Judge Koh's judicial philosophy in the SEP context: her finding in GPNE Corp v. Apple that the royalty base should be the smallest salable patent-practicing unit (SSPPU), which in a wireless standards context is the modem chip. It's not like GPNE established a bright-line rule, but it showed an approach that is technically, commercially and legally right--but not in Qualcomm's rent-seeking interest.

Kerrisdale points to various situations during the trial where Judge Koh rebuked or disagreed with Qualcomm's lawyers. By contrast, the FTC steered clear of annoying her. However, I don't think (and Kerrisdale doesn't say so either) Judge Koh would penalize Qualcomm for the fact that its lawyers may have gone too far in some of their maneuvering. I believe Judge Koh is simply well aware of the FTC's resource constraints (even regardless of the shutdown) and, on the other hand, sees Qualcomm's enormous legal resources. Those resources are so vast that a lawyer from a major firm, Norton Rose Fulbright, appeared on Tuesday, but that firm doesn't even appear on the filings that already list three other firms. So Judge Koh, like other judges in the U.S. and abroad (another example would be Judge Dr. Zigann in Munich), sometimes doesn't understand why litigants with an army of lawyers on a case can't just provide everything on a timely basis and structured in a way that conserves court resources.

However, it was only in connection with rival chipset licensing that a disagreement between Judge Koh and Qualcomm's lawyers was related to outcome-determinative issues. Everything else was of a purely procedural nature that is not going to play any role in her post-trial decision-making.

In general, I agree with Kerrisdale's high-level perspective that the FTC presented strong evidence (by which I mean testimony from device makers and various documents confirmed by testimony) and that Qualcomm, to put it that way, has been able for an extraordinarily long time to capitalize on its CDMA innovation of several decades ago. By comparison, if Motorola had been able to benefit from its status as the first company to make a functional mobile phone for such a long period, Motorola Mobility wouldn't have been sold to Google earlier this decade and to Lenovo more recently.

I've had to block a whole bunch of "Qualcommies" on Twitter because they just make unreasonable points and don't appear prepared to adjust their position based on facts that are presented to them. But I do respect Qualcomm apart from those FRAND-related issues. Unlike Kerrisdale, I'm not going to take a position on Qualcomm's overall outlook. They do appear to do some interesting things in connection with electric vehicles. In an ideal scenario, I'd like to see Qualcomm successfully modify its business model, continue to innovate, and maybe--though from today's vantage point it sounds like hell freezing over--it will further down the road end up on the good side of FRAND, just like Microsoft, Samsung and Google adopted policy positions I advocated long before them. If anybody sought SEP-based injunctions against them, or demanded supra-FRAND royalties from them, and if I was still active as a blogger (next year this blog will celebrate its tenth anniversary), I'd support them without hesitation. But they now have to go, for the sake of fair competition and innovation, through a process that may be painful.

Share with other professionals via LinkedIn: