Despite settlement, Mannheim LD grants first-ever UPC FRAND injunction in Panasonic v. OPPO, disagrees with European Commission

The FRAND part of the decision is the most detailed and most relevant one to study. It makes some clearly reasonable points, but also some highly debatable ones. It is in large part the Mannheim LD’s response to the EC’s amicus curiae brief in VoiceAge EVS v. HMD.

The FRAND approach is materially consistent with the one taken by Presiding Judge Professor Peter Tochtermann at the time when he (not yet a professor or a UPC judge) presided over the Mannheim Regional Court’s Seventh Civil Chamber until the Federal Court of Justice of Germany decided Sisvel v. Haier. Whether the Dutch panel member, Judge Edger Brinkman (The Hague), actually agrees with that stance is not known. It is possible that this was simply a decision by the German majority.

The Mannheim LD seized the opportunity to hand down the UPC’s first-ever SEP injunction over a FRAND affirmative defense. The Munich LD will adjudicate Huawei v. Netgear next month (October 30, 2024 ip fray article). An injunction in that Munich case appears fairly likely, but it will now only be the second UPC SEP injunction. Time will tell which of the two first UPC FRAND decisions will become more influential, but if the UPC as a whole adopted the Mannheim stance, it would draw significant pushback from Brussels.

Huawei v. ZTE step 1: infringement notice deemed sufficient

The EC’s amicus brief in VoiceAge EVS v. HMD took an extremely rigid position on the content of the initial infringement notice. The EC was saying that all of the relevant information had to be included in the notice (such as claim charts). In that case, however, all of the information was available. There simply was a link in the infringement notice to all of the relevant material. ip fray believes that this should be sufficient provided that it was straightforward (after clicking on the link) to find the relevant documents, and even more so when the recipient was such a sophisticated party as HMD.

Therefore, ip fray agrees with the UPC’s Mannheim LD that the requirements placed on an infringement notice under Huawei v. ZTE must be reasonable. Panasonic’s infringement notice to OPPO is not known.

Based on today’s decision, however, it is doubtful that Panasonic’s infringement notice met such reasonable requirements, but possible that OPPO raised its objection belatedly. The decision notes that according to OPPO’s argument, Panasonic provided a claim chart only for a broader Chinese member (Chinese patent ZL201310315589.X) of the family of the patent-in-suit. If that is true (and there is no reason to believe otherwise), Panasonic’s infringement notice fell short of the requirements under Huawei v. ZTE, according to which an injunction request is allowed only if (among other conditions) “prior to bringing that action, the proprietor has, first, alerted the alleged infringer of the infringement complained about by designating that patent and specifying the way in which it has been infringed.”

This means that a claim chart for the actual patent-in-suit is needed. There is no way that the ECJ meant that someone suing over a European patent could satisfy the requirement by providing a claim chart for a non-European (here, Chinese) patent, much less a non-European patent of a different scope. Even without a difference in claim language, jurisdictional differences could make an analysis for a particular non-European patent irrelevant for the purposes of a European infringement litigation. But if the language alone shows that the claim-charted patent has a different scope than the asserted one, there is no question that the claim chart falls short of what Huawei v. ZTE requires.

The judgment says OPPO’s written pleadings merely faulted Panasonic for not giving notice of infringement of that particular patent-in-suit. It was, however, apparently undisputed that the patent-in-suit (EP2568724 on a “radio communication device and radio communication method”) was listed somewhere in that correspondence, possibly even more than once. It was, according to the judgment, only at trial time that OPPO then raised the issue of no claim chart having been provided for that particular patent.

The Mannheim LD’s judgment goes clearly against the EC’s position on strict sequentiality by saying toward the end of para. 206 that a “cooperative license-seeking party” in OPPO’s place could have, and would have had to, point Panasonic to the fact that the Chinese claim chart was based on a broader claim scope than the European equivalent and request clarification.

Nowhere does the ECJ say or even imply that the implementer has any obligations with respect to Huawei v. ZTE step 1, or in between steps 1 and 2 in the form of a step “1bis” by which the implementer has to ask the licensor to redo its infringement notice. This is philosophically somewhere in between the old-fashioned German approach from the Orange-Book-Standard era (when the assumption was that the implementer had to proactively seek a license to all relevant SEPs) and Huawei v. ZTE. But it is not what the ECJ said, and not what the EC says the ECJ said.

Even though there clearly was a deficiency (no claim chart for the actual patent-in-suit), the question is whether OPPO raised that issue belatedly, in which case the upshot of this part of the analysis would have been the same (even if only for procedural and not substantive reasons):

1. OPPO could have raised that argument earlier. It was about something missing from the original infringement notice that was given before litigation started. They did not need any further information that came to light during the infringement litigation.
2. It is a reasonable assumption that OPPO tactically reserved that argument. They presumably didn’t want Panasonic to have the chance to cure that deficiency at a point that the court would have considered timely. If OPPO could have relied on the court applying the same standard as the EC in its amicus brief, according to which no deficiency can be cured after filing suit, they could have raised the issue as early as in a statement of defense (though Panasonic could then have withdrawn, given notice, and not long thereafter refiled).
3. If one believed that what OPPO made only at trial time was a purely legal argument, yesterday’s decision by the UPC’s Court of Appeal (CoA) in OrthoApnea v. Anonymous (PDF) (to be discussed by ip fray in the next UPC roundup) actually says that they could have presented that alternative theory even on appeal and it would have had to be considered.
4. But one could reasonably say that it was a factual issue. Whether or not a claim chart for the patent-in-suit was included in the original infringement notice is a question of fact. If that fact had been discussed somewhere in the pleadings and the question was just what legal relevance it should have, that would be a different story. But probably OPPO did not discuss the absence of that claim chart.

All in all, the Mannheim LD’s holding that Panasonic satisfied Huawei v. ZTE step 1 may very well be the right result, though the way the court got there (“cooperative license-seeking party” requirement) puts the Mannheim LD fundamentally at odds with the EC’s interpretation of Huawei v. ZTE.

The question is whether the decision as a whole (again, in a case where it appears that the commercial dispute is history) has more to do with forum-selling (attracting SEP complaints to the venue) than anything else.

Huawei v. ZTE step 2: Mannheim LD disagrees with EC by taking subsequent conduct into account

The Mannheim LD’s approach to step 2 is symmetrical to the one in step 1 by rejecting overly rigid requirements on the content of the relevant letter. The court therefore holds in the middle of para. 207 that “the initial declaration [of a willingness to take a license] by the defendant appears sufficient to constitute a serious start of negotiations.” And at the end of the same paragraph, it even says that no circumstances were discernible according to which the defendant’s declaration of willingness “was already at that point in time” insufficiently serious and not a basis to start specific discussions.

But Huawei v. ZTE, as interpreted by the EC in its amicus brief, only requires a declaration of willingness and there is no room for finding based on subsequent conduct that it was not serious in the first place.

Where the Mannheim LD again contradicts the EC and adopts a position somewhere between the German Orange-Book-Standard and Sisvel v. Haier doctrines is that it considers it necessary, in its overall discussion of the legal framework (para. 201), for the implementer to prove the substance of that initial declaration of willingness again and again through continued cooperation.

The Mannheim LD therefore wants to import subsequent behavior into its analysis of step 2 (and, as discussed before, even into step 1, creating an obligation for the implementer to help the SEP holder cure deficiences in the infringement notice).

It stops short of adopting the totality of Sisvel v. Haier only to the following extent:

The Mannheim LD says that the analysis of the parties’ conduct in further negotiations must not skip the assessment of the SEP holder’s offer altogether or reduce it to such a superficial and cursory analysis that it would be devoid of substance.

In other words, the Mannheim LD, contrary to what the EC says (which is not binding, but the EC has the power to ensure that the ECJ will sooner or later adjudicate the question), believes in amalgamation. It subjects step 2 (and even step 1, which may be unprecedented) of the Huawei v. ZTE analysis to an assessment of the parties’ overall conduct during the course of negotiations.

Huawei v. ZTE steps 3 & 4: not even a full-fledged license agreement is required

Not only does the Mannheim LD decline to look at step 3 without also considering step 4 (the implementer’s counteroffer, and beyond what the ECJ said, the implementer’s overall behavior) but it also sets an extremely low standard for the SEP holder to meet.

The part where even the EC appeared somewhat flexible at the recent VoiceAge EVS v. HMD hearing is that the SEP holder’s initial offer does not have to be perfect. There may be more than one iteration. It’s just that the EC reaffirmed at the appellate hearing that whatever the parties propose during the negotiation process must be within the FRAND corridor. But that is not the key issue here. The biggest problem here is a different one:

The Mannheim LD even considers it sufficient for the SEP holder to just outline some general terms as opposed to putting a ready-to-sign license agreement on the table.

In this particular case, it says that OPPO would first have had to provide some information on its product sales in order to enable Panasonic to make a suitable proposal. And as long as that was not the case, the court believes it was sufficient for Panasonic to just summarize some key terms.

The ECJ said in Huawei v. ZTE that the SEP holder’s obligation, subsequently to the implementer’s declaration of willingness to take a license, is to “present[] … a specific, written offer for a licence on such terms, specifying, in particular, the royalty and the way in which it is to be calculated.”

Obviously, a full-fledged license agreement along with an explanation of the royalty computation would meet those requirements. But is that a safe harbor or merely a minimum requirement? The Mannheim LD’s implicit interpretation of Huawei v. ZTE is that a lot less than a complete contract may do, at least as long as the implementer does not provide certain information.

Here, like in connection with step 1, the Mannheim LD’s philosophy is that if the SEP holder does not deliver what is required of it under Huawei v. ZTE, it’s up to the implementer to pinpoint any shortcomings and to lend a helping hand so the SEP owner will find it easy to address them.

The judgment as published today is heavily redacted. Several paragraphs relating to the actual offers are redacted in their entirety, and if the unredacted paragraphs are any indication, such paragraphs can be quite long. Is it therefore impossible to say whether there was a sufficient basis for the court to conclude that Panasonic’s offer was FRAND and that OPPO’s was not.

Based on what today’s decision says and what became known as a result of Panasonic’s parallel dispute with Xiaomi, where various facts came to light in the UK and where the appeals court also expressed a certain view of what Panasonic was trying to achieve, it is plausible that neither Panasonic’s royalty demand nor OPPO’s counteroffer were FRAND, and there is room for reasonable doubt about whether Panasonic’s royalty demand was FRAND.

OPPO’s refusal to put sales figures on the table

The judgment says that OPPO wanted Panasonic to base its offers on sales figures reported by third party IDC, a company that estimates such numbers but can also be wrong.

The Mannheim LD really disliked that attitude. It is true that such behavior is neither constructive nor cooperative. Therefore, the Mannheim LD’s negative view of OPPO’s conduct is understandable, but what is simply wrong is to say that a SEP holder is never in a position to make FRAND offer without having access to such data.

There is nothing that stops a SEP owner from making an offer, even in the form of a full-fledged contract, without information from an implementer that has the kind of visibility that OPPO has. If someone made a component sold only in a niche market, and if it was impossible to even take a look at the products or find out at what prices they are sold, the situation may be different. But Panasonic clearly didn’t need anything from OPPO to make a FRAND offer. They could have offered a matrix, even a two-dimensional or multi-dimensional matrix. For instance, they could have specified a percentage of net selling price and a cap. They could have provided brackets. They could have stated different rates for different types of products (for example, distinguishing between smartphones and tablets). And they could have included volume discounts in their offer (e.g., 10% discount for all sales above 50 million units per year, 20% for all sales above 100 million units).

Would it have been more efficient for OPPO to provide the information to Panasonic? Of course.

But Huawei v. ZTE doesn’t talk about an implementer’s obligation to share information, and the EC’s amicus brief in VoiceAge EVS v. HMD requires a strictly sequential analysis, which the Mannheim LD attempts to portray as unworkable by arguing that a SEP holder can’t make a FRAND offer without cooperation from the implementer.

Huawei v. ZTE requires the SEP holder to make a FRAND offer and to explain why it is FRAND. The Mannheim LD now creates some interim steps that are not found in that decision.

To be fair, Huawei v. ZTE would work better if there was a mutual information exchange requirement between steps 2 and 3. In an alternative and arguably better universe, an additional step “2bis” could require them to conclude an NDA and to exchange, subject to that NDA, information on third-party license agreements (to the extent it is possible without a court order) as well as for the implementer to provide sales figures. It’s just that no such interim step exists.

Also, the Mannheim LD does have a point where it says that a SEP holder’s willingness to make major adjustments to a royalty demand during the course of negotiations does not prove that the initial offer was not FRAND.

But where the Mannheim LD simply wants to amalgamete the analysis of the different steps of Huawei v. ZTE, it is at loggerheads with the EC. Regrettably, this comes at a time where there is significant lobbying pressure behind the proposed EU SEP Regulation (especially with a view to the Polish government taking over the rotating EU Council presidency on January 1, 2025). One of many valid arguments against that legislative proposal was that one should first let the UPC develop its SEP case law. And now there is a first UPC SEP injunction, in a case where the implementer may indeed not have been a willing (or least not a perfectly willing) licensee, that in terms of implementer-friendliness is near-identical to Sisvel v. Haier (there is an academic difference, but it is hard to see what practical significance it has) and lightyears away from the EC’s amicus brief.

UPC SEP injunctions in cases where even the EC’s proposed standard is met would not do any harm. In fact, they might even show that the UPC is going to make a positive impact on Europan SEP case law. But the same cannot be said of today’s Mannheim judgment.

Which end of the FRAND range?

Another respect in which the Mannheim LD’s ruling differs fundamentally from the EC’s position in VoiceAge EVS v. HMD is that the judgment says a SEP holder does not have to grant a license at the lower end of the FRAND range. The EC, however, clearly takes the position that an implementer who makes an offer within the FRAND range (albeit at the lower boundary) must not be enjoined.

Past damages not necessarily capped by FRAND

Judge Gilstrap held in a case in the Eastern District of Texas that an implementer may lose the entitlement to a FRAND license. The UPC’s Mannheim LD, at minimum, doesn’t assume that damages for past infringement necessarily amount to (and cannot exceed) a FRAND royalty.

UPC jurisdiction over FRAND counterclaims

It is interesting that Panasonic’s counsel in this action said the opposite in a public LinkedIn discussion from what it told the UPC about FRAND jurisdiction. Last month, after ip fray said that U.S. and UK (and, by the way, also Chinese) courts have further-reaching powers in the FRAND context (particularly when it comes to obligating SEP holders to extend licenses), one of the two founders of the firm representing Panasonic here (an excellent lawyer and one of the leading UPC litigators) wrote:

“Thanks for the article, but why should the UPC’s competence to determine global FRAND rates be more limited than of other courts like the ones in UK? They reasoned that they are competent because the defence against an SEP injunction is a global license that is to be determined by national courts. So the UPC can do as well.”

Today’s judgment reveals in para. 67 that they argued on Panasonic’s behalf that OPPO’s FRAND determination counterclaim was impermissible with respect to any patent other than the particular patent-in-suit. That would be an extremely limited kind of FRAND jurisdiction for the UPC.

Also, para. 243 of the decision said that the UPC did not have the power to force Panasonic to accept OPPO’s offer to take a license. That is different from courts in certain other jurisdictions.

It is definitely to be welcomed that the UPC’s Mannheim LD does, in principle, assume jurisdiction over FRAND (counter)claims. In the case at hand, OPPO complicated everything, however, by gerrymandering the world’s major markets. The UPC was asked to set a rate for Europe, Japan and the United States. But OPPO wanted a Chinese court to set a FRAND rate for the markets where most of its volume is. Then, at a late stage of proceeding and only after the court notified them of the deficiency of their gerrymandered approach, they apparently opened up to a further-reaching FRAND determination (last sentence of para. 250).

It is understandable that the Mannheim LD threw out the FRAND counterclaim. There were some issues. But repeatedly that part of the decision refers to OPPO’s unwillingness to provide sales data, which has no support in Huawei v. ZTE and reinforces the impression of forum-selling.

No ECJ referral

The Mannheim LD is not a court of final jurisdiction and, therefore, any preliminary reference to the ECJ would be in its sole discretion. Its decision to deny OPPO’s request for a ECJ referral was debatable on substance (given that differences between the court’s and the EC’s interpretations of Huawei v. ZTE were indeed outcome-determinative here), but from a pragmatic point of view, it was the right decision, given that this is apparently a dead lawsuit walking and the ECJ might never actually get to decide.

How influential is this decision going to be?

It will probably serve its suspected forum-selling purpose, but it appears unlikely that the partly radical (even if disguised as compromise) positions taken by the Mannheim LD will be controlling law for years to come. It would be surprising if those extreme positions had been a unanimous decision even by that panel.

It is unlikely that this decision will influence the Munich Higher Regional Court while it is working on its decision, scheduled for early February, in VoiceAge EVS v. HMD, but should the Munich appeals court read that decision, it would presumably agree-in-part and disagree-in-part.

Panel and counsel

Panel: Presiding Judge Professor Peter Tochtermann, Judge Dirk Boettcher (“Böttcher”) Judge Edger Brinkman (The Hague) and Technically Qualified Judge Klaus Loibner.

Counsel for Panasonic: Kather Augenstein’s Christopher Weber.

Counsel for OPPO: Vossius & Partner’s (Vossius & Brinkhof UPC Litigators’) Dr. Andreas Kramer.