I'll admit that, over the past few weeks, I haven't kept as close a watch on the Qualcomm antitrust litigation as I should have, but I have now read Judge Koh's opinion in FTC v. Qualcomm that I mentioned on the blog earlier today. Perhaps there's something (or quite a lot) that I'm just not understanding yet, and I'm not saying Judge Koh was wrong to deny the motion to dismiss--this litigation is at a very early stage--but there are two matters that I'm finding perplexing about the FTC's case.
First, according to the opinion the complaint alleges that Qualcomm is charging an above-FRAND rate for its SEPs. That may well be what the complaint says, but I'm wondering how the FTC plans to prove this allegation if the matter proceeds to trial. One possibility is that a FRAND royalty is a commitment to charge less than the market otherwise would bear, that is, to voluntarily forgo some revenue in return for participating in the standard setting process. (If correct, this would suggest that in a patent infringement case involving a FRAND-committed SEP, the amount awarded as a FRAND royalty would be less than what would have been awarded as a reasonable royalty, absent the FRAND commitment.) And maybe this is the right way to think about FRAND royalties, but as I've stated in the past I'm skeptical; by what criteria would a court measure the different between a FRAND royalty and a freely-negotiated (reasonable) royalty? Of course, one might challenge my view by asking what the point of a FRAND commitment is, but I think the answer is that a FRAND commitment is just that--a commitment that otherwise wouldn't exist to license one's patents, one that might be enforceable in a breach of contract or promissory estoppel lawsuit.
Alternatively, maybe the idea here is that Qualcomm is charging a royalty that exceeds the value of its technology because it includes "holdup" value--some amount, in addition to the value of Qualcomm's SEPs in comparison with alternatives, that reflects the implementer's sunk costs or differential switching costs ex post. (See my blog post with Norman Siebrasse on this issue, here.) But while I'm sympathetic to that theory when the facts are consistent with it (e.g., when the implementer has invested sunk costs already, and ex post the patentee uses the threat of an injunction to demand a royalty reflecting in part those sunk costs), if I'm reading the opinion here correctly it sounds as if Qualcomm negotiated the licenses at issue ex ante. It certainly sounds like Qualcomm drove a hard bargain, too, but that fact alone doesn't in my view mean that Qualcomm is demanding a supra-FRAND royalty; rather, it's charging what the market would bear ex ante. Am I missing something here?
The other issue I don't quite get is the theory that Qualcomm is putting its competitors in the market for modem chips at a disadvantage by (1) not licensing its SEPs to them, and (2) instead requiring the OEMs to license Qualcomm's SEPs, regardless of whether the chips the OEM is using come from Qualcomm or from one of Qualcomm's competitors. The reason I don't get it is this: if Qualcomm's patents are indeed standard-essential, then wouldn't the chips produced and sold by the competitors necessarily practice Qualcomm's technology? If so, then Qualcomm is entitled to extract a royalty for the use of its patents in these chips, but I don't see why it has to extract it directly from the competitors instead of from the OEMs. (Moreover, as I mentioned earlier today, U.S. antitrust law doesn't condemn monopoly pricing as such, but rather the willful acquisition or maintenance of monopoly power. There has to be something more than just excessive pricing, which is why the FTC alleges and Judge Koh discusses whether Qualcomm is using its monopoly power to exclude competitors.) Of course, if the proponents of the SSPPU concept are right, Qualcomm will be able to extract a higher royalty from the OEMs than it could from the competitors, because the OEMs' royalty base is higher; but as readers of this blog and of my scholarship are aware, I'm somewhat skeptical of that theory too. I suppose if the SSPPU theory is correct, though, then this strategy may enable Qualcomm to extract a higher royalty overall than it otherwise could, and this potentially puts its competitors at a disadvantage, since they could otherwise underbid Qualcomm (which is what Judge Koh talks about). Even so, I'm not sure I'd buy into the proposition that a FRAND commitment (or antitrust law) necessarily requires a firm to license its patents to the component manufacturer instead of to the end product manufacturer.
Then again, maybe I'm just missing something--maybe the competitors' chips don't practice Qualcomm's SEPs? But if so, how could those chips be competitive in the first place? I realize that overdeclaration may be a problem generally, but surely some of Qualcomm's asserted SEPs must be essential in fact, no?
Readers, enlighten me.