Showing posts with label CEN CENELEC. Show all posts
Showing posts with label CEN CENELEC. Show all posts

Wednesday, June 19, 2019

Qualcomm elaborates on its theories of irreparable harm from immediate enforcement of FTC's antitrust remedies

Originally, Qualcomm asked Judge Lucy H. Koh of the United States District Court for the Northern District of California to set an extremely tight briefing schedule for the San Diego chipmaker's motion to stay the enforcement of the victorious Federal Trade Commission's antitrust remedies pending an appeal to the Ninth Circuit. If the court had adopted that schedule, Qualcomm would even have been prepared to waive its right to file a reply brief in a support of its motion.

But Judge Koh declined to give the FTC only a very few days for its opposition filing--and Qualcomm, though it still could have waived its right to file a reply brief, elected to file a reply brief yesterday (not a single day before the deadline) and counter not only what the FTC wrote but also the amicus curiae briefs submitted by LG Electronics and ACT | The App Association (this post continues below the document):

19-06-18 Qualcomm Reply Iso... by on Scribd

That reply brief is far more interesting--because it is a lot more specific especially on the question of irreparable harm--than the original motion. While I would categorize some of Qualcomm's reply arguments as non sequitur material and consider some others outright smokescreens, that reply brief does contain food for thought.

The first footnote is, however, almost a concession that a stay of the entirety of the FTC's remedies may have been too much to ask for:

"Should the Court determine that the irreparable harm Qualcomm would suffer as a result of provisions (1) and (2) of the injunction does not warrant a stay of the entire Order, Qualcomm respectfully submits that the Court should enter a partial stay of only provisions (1) and (2) of the injunction."

Accordingly, the reply brief focuses on the requirements to (re)negotiate agreements, to sell chips to unlicensed customers, and to extend exhaustive SEP licenses to rival chipset makers. Those requirements flow from the first two provisions of the injunction the FTC secured--the only ones with respect to which Qualcomm, according to the FTC and also in my observation, even attempted to make an irreparable-harm argument in its original motion.

As for any potentially renegotiated license agreements, Qualcomm insists that any harm could not be undone should Qualcomm prevail on appeal. Qualcomm labels as a "red herring" the FTC's argument that Qualcomm would still obtain "fair value" for its SEPs. Qualcomm says the problem is not that those new agreements would be negotiated without the well-known "No License-No Chips" kind of leverage, but "because of the need to negotiate in the shadow of an Order that declares—erroneously, in Qualcomm's view—that Qualcomm's typical licensing terms are unreasonable." The latter is, by the way, what I told CNN within less than 24 hours of the ruling.

Qualcomm is concerned about licensees stopping royalty payments "under valid contracts" (though Judge Koh's order obviously serves to invalidate many such agreements), "even if temporarily," and mentions Huawei as an example.

Then Qualcomm attacks the FTC's suggestion that Qualcomm could, after a successful appeal, seek "damages for any past infringement" against those who (at least temporarily) stopped royalty payments. Qualcomm states something indisputable: a license is a defense to an infringement claim as 35 U.S.C. § 271(a), the statutory definition of infringement of a U.S. patent, applies only to those who practice a patented invention "without authority."

Qualcomm is furthermore worried that the non-discrimination provision of its FRAND licensing commitment or "most favored" provisions in some of its license agreements could ultimately drive Qualcomm's royalties to the "lowest common denominator."

Qualcomm, with a declaration by one of its licensing executives (John Han), alleges that it would be impossible to prevent irreparable harm by agreeing only on "short-term or interim licenses" (as the FTC called them) or (again quoting the FTC) "contractual provisions that would mitigate or eliminate any long-term adverse consequences to Qualcomm").

All of that needs to be considered, but none of it is convincing. The appeal will take time, but we're talking about roughly a year. Negotiations, however, also take a fair amount of time. The FTC v. Qualcomm decision doesn't require Qualcomm to accept a specific set of terms within a short time frame lest the company be held in contempt. I believe it just comes down to how much of a risk Qualcomm is willing to take when betting on a successful appeal. If Qualcomm really was sure that it would prevail, then it could insist on contractual provisions that protect its rights, and there would not be any contempt sanctions before the Ninth Circuit decides. And what's practically totally impossible is that Qualcomm could not only be forced to enter into a license agreement on unfavorable terms with a first licensee and that a second licensee would then have enough time to also get a better deal (or an adjustment under a "most favored" clause) before the Ninth Circuit renders a decision. And even if the appeal surprisingly took longer, anything that happens as a result of a decision that is being appealed would hardly serve as a FRAND benchmark in any other litigation. I would expect any other case to simply be stayed in that event.

What Judge Koh could do now is provide some guidance to Qualcomm on what structural options it does have even if a stay is denied (such as that Qualcomm would not act in contempt of the injunction if it insisted on contractual provisions that protect its rights, including retroactive adjustments, in the event of a successful appeal). Maybe Qualcomm is primarily hoping to obtain such clarification while formally moving for a stay.

With a view to LG Electronics, which supports the FTC's opposition to Qualcomm's motion, Qualcomm contradicts LG's description of the state of negotiations between the parties. Qualcomm notes that it "continuously supplied chips to LGE, without any interruption, throughout the negotiations"--but if it intended to do so going forward, the injunction against its "No License-No Chips" tactics wouldn't make a difference. Qualcomm mentions "a written offer to enter into binding FRAND arbitration with LGE to try and resolve the dispute, which included an express guarantee of chip supply during the pendency of the arbitration, but LGE declined that offer." As I've explained on various occasions, including my FTC v. Qualcomm trial coverage, arbitration isn't necessarily fair. Without the right parameters, it can give an extremely unfair advantage to a patent holder making out-of-this-world royalty demands since the licensee can't counterbalance a supra-FRAND royalty demand by proposing a negative royalty.

Qualcomm also argues that a requirement to sell chips to unlicensed OEMs would have profound implications due to patent exhaustion. Qualcomm rejects the FTC's argument that Qualcomm should simply "price[s] its modem chips to reflect the fair value of its patents" (which, by the way, pretty much every other chipset makers does, and even Qualcomm does so in some other business areas than cellular modems). Qualcomm argues that it would either lose chip sales or its ability to fully monetize its SEPs "so long as other chip makers are not licensed and thus do not price into their chip offerings the cost of Qualcomm's patents, this would leave Qualcomm in the untenable position of either charging much more for its chips than do its competitors (and therefore likely losing the sales), or reducing its chip prices so that, once again, they do not reflect the fair value of Qualcomm's patents."

The passage I just quoted may very well be the weakest one in that entire reply brief. If its chipset makers aren't licensed, then this is an apples-to-bananas comparison as device makers (as many of them--and competing chipset makers such as Intel--testified in the FTC litigation) would still have to factor in the cost of licensing Qualcomm's patents.

Finally, and as we all know, Qualcomm dreads the notion of having to extent exhaustive SEP licenses to rival chipset makers. But Qualcomm's argument for a stay of that particular provision comes down to what Judge Koh has already rejected at the merits stage: the question of efficiency of multi-level licensing (licensing cellular SEPs to baseband chipset makers and other patents to device makers) versus component-level licensing. Qualcomm says that neither OEMs nor chipset makers would jump to pay royalties, so there would be disputes over whether certain patents are practiced by particular products or not. Qualcomm describes this as a risk of "obstruction and delay," but that holdout-style argument does not appear to be sufficient to make a case for irreparable harm. Other patent holders face those challenges all the time (as the FTC noted in its opening statement back in January).

In that context, Qualcomm also demands "adjudicative comity" (respect for other jurisdictions) because its settlements with China's NDRC and the Taiwan Fair Trade Commission did not involve a requirement to extend licenses to rival chipset makers. I don't understand the comity issue here: if additional parties get licensed, it doesn't frustrate any policy or decision by other antitrust agencies. If the NDRC or the TFTC settled their cases without such a requirement, why would it hurt them if, say, MediaTek obtained a license as a result of the U.S. FTC case? I can't think of any negative effects on competition in those markets.

Part B of Qualcomm's reply brief addresses the likelihood of success of its appeal, with an emphasis on the need for a court to determine that violations are "likely to reoccur" before the FTC can be granted the injunctive relief it successfully sought here. What is missing here is a proposal for how an antitrust offender in a dynamic market could ever be enjoined from some behavior since the court would always have to set some sort of cutoff date.

Part C, finally, makes a national security-centric public-interest argument, with a particular focus on the decision by the Committee on Foreign Investment in the United States (CFIUS) to block Broadcom's attempted acquisition of Qualcomm. The FTC's position was that the CFIUS decision was unrelated to the behavior at issue in the antitrust case, but Qualcomm points to the CFIUS's position that "[c]hanges to Qualcomm's business model [as a result of an acquisition by Broadcom] would likely negatively impact the core R&D expenditures of national security concern." I tend to agree with Qualcomm that the content of the CFIUS letter is significantly more helpful in the public-interest context of the motion to stay than the FTC acknowledged. However, the CFIUS did not block Broadcom's hostile takeover of Qualcomm in order to enable continuing antitrust violations.

What I found interesting is that, with reference to ACT | The App Association's amicus brief, Qualcomm's reply brief also refers to the European SEP policy debate and, specifically, the two competing CWAs (CEN-CENELEC Workshop Agreements). One of the exhibits attached to Qualcomm's reply brief is CWA1, which Qualcomm and similarly-minded companies support, while there is far broader and stronger support for CWA2.

Share with other professionals via LinkedIn:

Wednesday, June 12, 2019

Broadbased IoT industry support for CWA2 set of standard-essential patent (SEP) licensing guidelines and best practices

At times, courts are asked to clarify what FRAND (fair, reasonable and non-discriminatory) means for standard-essential patent (SEP) licensing terms. The two most fruitful U.S. cases in this regard have been Microsoft v. Motorola (Judge James L. Robart, Western District of Washington) and FTC v. Qualcomm (Judge Lucy H. Koh, Northern District of California). In the EU, Huawei v. ZTE proved reasonably helpful, though definitely not at a level with the aforementioned U.S. cases.

But besides litigation, policy initiatives as well as efforts by standardization bodies can and do provide orientation. In the U.S., the IEEE has become a particularly important discussion forum, and the arguably most important European guideline-development effort, a workshop under the umbrella of European standard-setting organizations CEN and CENELEC with support from Germany's DIN, has just concluded with the publication of a 51-page "CEN Workshop Agreement" on the "Core Principles and Approaches for Licensing of Standard-Essential Patents" after more than a year of discussions. The official dcoument number is "CWA 95000." The shorter name is "CWA2." Why the number 2? Because there's another set of recommendations by another group of companies, and it advocates pretty much the opposite approach, but it has considerably less support from industry than CWA2.

This is an important time for the publication of such proposed guidelines for different reasons. The industry is transitioning to 5G, and in an Internet of Things world, ever more--and an ever greater diversity of--devices will implement wireless industry standards. Also, a new EU Commission will be appointed this year, and the incoming Commission will have to address SEP licensing principles in different fields of policy-making as well as in competition enforcement.

The Fair Standards Alliance issued a press release on CWA2 this morning. The FSA partnered with ACT | The App Association. ACT already pointed to CWA2 yesterday evening in an amicus brief supporting the FTC against Qualcomm's motion for an enforcement stay. FSA and ACT teamed up with the leading German standard-development body, Deutsches Institut für Normung (DIN), and DIN served as the secretariat for this workshop.

All in all, 56 organizations have already expressed their support for the document, even including some who were not involved in the process but agree with the conclusions. Europe's largest automotive association, ACEA, and IP2Innovate, an industry group whose membership significantly overlaps with that of the FSA and which promotes reasonable patent enforcement policies (unlike the FSA, without focusing on SEPs), also agree with the CWA2 guide. Here are some of the household names among the companies listed in the CWA2 document: Apple, BMW, Cisco, Deutsche Telekom, Renault, Honda, Juniper, Volkswagen, Daimler, Ford, Hitachi, HP, Lenovo, and Toyota. If your company or other organization would also like to throw its weight behind the CWA2 guide, please get in touch with the FSA.

The supporters of CWA2 own and generate many patents, and some of them have actively enforced their intellectual property rights. In other words, they're the opposite of "infringers." But they do understand both sides of the licensing equation, and that's why the recommendations made in the CWA2 document appear reasonably balanced. They don't seek to devalue SEPs, or to enable "holdout" (the act of refusing for an extended period of time to take a license). At the same time, the CWA2 guide promotes principles that greatly reduce the risk of patent holdup (the following summary is consistent with my terminology on this blog, not always identical to the terms used in the CWA2 document):

  1. access to injunctions only as a last resort

  2. availability of licenses to all comers including chipset makers

  3. FRAND valuation based on technical contribution at time of standard-setting, without capturing post-standardization increases in value or collecting percentages of components beyond the smallest salable patent-practicing unit (SSPPU)

  4. separate availability of SEP license on FRAND terms without tying this to a license to non-SEPs

  5. no overbroad NDAs that would complicate an implementer's evaluation of assertions of essentiality or infringement in the negotiation process

  6. no circumvention of FRAND licensing obligations through tranfers ("privateering"); if SEPs are transferred, implementers shouldn't be worse off

Obviously, some patent holders with a particularly strong interest in aggressive monetization don't want any of the above. That's why there's also "CWA1" as I mentioned before. The FSA and ACT expressed their disagreement with the CWA1 approach in an open letter earlier this year. While that letter apparently didn't have the desired effect of creating more balance, it shouldn't be too hard for policy makers (such as the incoming EU Commission) to figure out which of the two sets of recommendations is conducive (CWA2), and which one is detrimental (CWA1), to innovation and competition. That's easy to tell based not only content but also by simply looking at the backers:

  • CWA2 already has 56 backers, including some of the most significant technology companies in the worldand especially in Europe, versus the 17 organizations behind CWA1.

  • While CWA2's supporters have a rather reasonable track record in patent enforcement, almost a third of CWA1's backers have been on the receiving end of antitrust complaints or antitrust lawsuits: Qualcomm (around the globe), Dolby (see this blog post by a Korean IP law firm), Ericsson (in China), Nokia, and InterDigital, which I sometimes refer to as "InterDigitroll" (they were investigated in China and sued in the U.S.).

Thanks to CWA2, there now is a comprehensive proposal for a FRAND licensing framework on the table that has the support of many large players from the automotive and information and communications technology industries. That document will be referenced on many occasions going forward, in Europe and beyond.

Share with other professionals via LinkedIn: