What’s a reasonable percentage of FRAND wins for SEP holders in litigation in a given jurisdiction?

Statistics of litigation outcomes are widely available and often cited. At the same time, the law is not a numbers game. It wouldn’t work to give an individual judge a “quota” to meet. Numbers can be discussed, but one must interpret them carefully.

Last year, when EU internal market commissioner Thierry Breton presented the proposed EU SEP Regulation, he said that only “bad actors” among SEP owners could oppose it. In the firm belief that good and bad actors exist on both sides of this debate, a belief based on granular litigation coverage as well as numerous personal interactions with the whole range from non-practicing entities to 100% defensive operating companies, ip fray rejects that notion.

At the same time, the German judiciary makes it look like (almost precisely) 100% of all implementers of standards who don’t have a license to a patent deemed valid and essential are by definition bad actors that should be enjoined. That is equally absurd.

Is Huawei v. ZTE fair and balanced, or was it at least meant to be?

There is no factual basis for suspecting that the ECJ panel in that case had an agenda favoring one side or the other. Neither are there any utterances nor are there any biographical facts that suggest so.

We’re all imperfect, and so is Huawei v. ZTE. One weakness is due to the way the Dusseldorf Regional Court phrased its questions to the ECJ, and which the ECJ elected not to correct: a conflation of a contract-like FRAND promise and the question of an abuse of market power. Think about it: if someone held a patent enabling them to shut down all 4G/5G networks in the EU and bar the sale of all 4G and 5G devices, but just happens to own it without having made a FRAND promise, would the courts just have to allow such a patent holder to hold the EU hostage? It wouldn’t make sense, and it made even less sense when considering that the Dusseldorf court’s starting point was Germany’s top court’s only SEP decision at the time, Orange-Book-Standard, which was 100% about an abuse of market power and involved no FRAND pledge.

The ECJ then tried to somehow use the existence of a FRAND pledge for an additional reason to justify the Huawei v. ZTE negotiation steps: it referred to expectations having been raised in the marketplace. Under U.S. contract law, promissory estoppel is an interesting argument, but in the EU antitrust context, it can only justify a heftier fine on an abuser, but the absence of a FRAND pledge couldn’t reasonably render hold-up lawful.

The Huawei v. ZTE panel, under a German presiding judge, was aware of Orange-Book-Standard from the pleadings. In retrospect, one can question the wisdom of having created new EU case law by modifying an utterly unbalanced German doctrine only because the originating case came from Germany. They could have looked to what other jurisdictions do, and they could have been more creative.

That said, Huawei v. ZTE is beyond reasonable doubt as far as judicial intent is concerned, and it’s very well-crafted. The idea was to put some pressure on both sides to take reasonable positions prior to litigation, a bit like baseball arbitration where an extreme position can really hurt.

Does it help answer the question of how valuable a standard is? For instance, whether the cellular connectivity that has made such a profound impact on our lives is fairly valued? No, because it is just a procedural framework and not about valuation.

Assuming for the sake of the argument that a court is willing (German courts definitely aren’t, as they wish to focus on swiftly adjudicating cases on the purely technical merits) and able (due to their unwillingness, German courts are “untested” in this regard) to figure out what is in fact FRAND, Huawei v. ZTE would hold a SEP holder to its FRAND obligation to offer a license on lawful terms, but would also penalize an implementer who rejects a FRAND offer without making an alternative offer that is also FRAND.

“Whatever terms are in fact FRAND”

There is some reasonable criticism that advocates of SEP holder interests could raise: based on the strictly sequential approach that the EC says is the correct application of the ruling, an unwilling licensee could come away unscathed because of the SEP owner’s royalty demand having been supra-FRAND (potentially only slightly so). To the UK judiciary’s credit, they addressed that one through Unwired Planet v. Huawei, which doesn’t mean to take any position in this context on the question of whether one country should set a global royalty rate over a party’s objection. It’s just that Unwired requires the implementer to take a license on whatever terms are in fact FRAND.

The first “whatever terms are in fact FRAND” decision, though it wasn’t phrased like that, was Microsoft v. Motorola (Western District of Washington). Judge James Robart made it clear to the parties that he wanted to resolve a dispute on a reliable basis.

Now, German courts cite Mr Justice Colin Birss’s “whatever terms are in fact FRAND” wording all the time. They did so even in Sisvel v. Haier, which is ridiculous at best and cynical at worst, given that Sisvel v. Haier is the exact opposite of “whatever terms are in fact FRAND”: it’s a devious scheme to force virtually every implementer into a license on the patentee’s preferred terms, which will often be supra-FRAND. If all the blame is put on an implementer for allegedly not being prepared to take a license on FRAND terms without properly analyzing the SEP holder’s demand, the result is rarely FRAND.

Huawei v. ZTE is not quite as good as “whatever terms are in fact FRAND,” yet its correct application would have a far higher hit rate than Sisvel v. Haier.

Correct evaluation of Sisvel v. Haier‘s abysmal failure

If serious economists wish to correctly, without any agenda other than seeking the truth, analyze the impact of Sisvel v. Haier, there can be only one measure:

In how many cases did an implementer, after being deemed to infringe a valid and essential patent, prevail on a FRAND defense to injunctive relief?

To clarify this further, three more considerations are key:

• If a patent expires, it’s not a successful FRAND defense but fortunate timing for the implementer.
• It’s all about the net effect. If the lower court enjoins the implementer, that means the SEP holder can likely dictate the terms unless the appeals court quickly stays enforcement, decides the appeal (after the appellate hearing) for the implementer and the injunction never gets reinstated by the Federal Court of Justice (in Sisvel v. Haier, that happened, but the defendant was saved by the bell as the patent had expired, so the injunction decision was academic for that dispute). If the lower court doesn’t hand down an injunction, but the appeals court enters an injunction, there will have been a delay, but after the appeal, where enforcement doesn’t even require providing collateral, the implementer is in serious trouble. So it’s all about whether an injunction issues and can be enforced, whether immediately or with a delay.
• One must focus on independent decisions. If the same judge denies an injunction in three parallel and related cases on the same grounds, it’s just one successful FRAND defense. Likewise, if the same judicial panel, on the same basis (i.e., in the absence of new FRAND facts), enters separate injunctions against a defendant over multiple patents, it’s also just more of the same.

There is also an issue that a major net licensor of SEPs has reasonably raised in conversation with ip fray: if an injunction is enforceable only on the basis of providing an amount of security that a SEP holder can’t practically provide (such as Nokia in the Daimler dispute, where the Mannheim Regional Court required security to the tune of €7 billion), it’s “not really an injunction.” That is true, but even in that Nokia v. Daimler dispute, actual enforcement was on the horizon at the time of the settlement because the patent held up very well on the validity side (parallel proceedings) and the appeals court was prepared to accelerate the appeal to enable enforcement without collateral at the earliest opportunity. That doesn’t mean that this SEP licensor didn’t raise an important policy issue: for instance, hardly anyone can afford provisional (i.e., during the appellate proceedings) enforcement against Apple. That, in turn, enables hold-out. It’s just a question of hold-up vs. hold-out leverage and unrelated to whether Sisvel v. Haier delivers correct results.

With the parameters being clear, what has really happened in Germany since Sisvel v. Haier (May 2020)? It’s dramatic:

In the more than four years since then, all implementers held to practice a valid SEP were enjoined except for a few related and near-simultaneously-decided Dusseldorf cases where a judicial panel disagreed with a patent pool’s duplicative-royalty policy. That was an outlier decision over an extremely rare issue.

Shortly after Sisvel v. Haier, one implementer defended itself, but the appeals court reversed and directly entered an injunction that became then enforceable without collateral.

The Sisvel v. Haier problem actually started a couple of years before the Federal Court of Justice decision with some lower courts in Germany adopting ever less implementer-friendly approaches. If we go back not only four years, but six, there’s just one more incident of a successful FRAND defense (May 20, 2024 ip fray article).

That is a disaster. Even if one assumes that FRAND-unwilling licensees are considerably more likely to be sued than FRAND-willing licensees, there is no way that this legal standard can be right. It’s common sense and also economic logic (homo oeconomicus) that if you can get pretty much everybody enjoined by wielding a deemed-valid and found-essential SEP, there will be some trying to seek supra-FRAND royalties. In fact, they all have a shareholder obligation to try that.

It may have worked as intended by the presiding judge who kept defending it over the years instead of issuing a public mea culpa. From a short-term perspective, many German SEP litigators may have liked it (and may still like it). But it’s unsustainable. It wasn’t going to stay that way.

German judges’ lack of FRAND balance is harming German consumers and workers: less choice, fewer jobs

The outlier position that Germany’s SEP case law has on a worldwide basis is also evidenced by the negative economic impact of the unbalanced and irresponsible misapplication of EU competition law in Germany:

• OPPO and vivo left the market. Those companies are major patent holders themselves and proud of their innovative capacity, but it made no sense for them to bow to Germany’s dictate of unreasonable global terms. If they sell 50 or 100 times more phones in other jurisdictions, an incremental global royalty cost of, say, €1 per device (the actual damage between multiple patent holders is greater than that!) as a result of German extremism has the same effect as paying an extra €100 per device sold in Germany. They laid off a number of people.
• A company like Transsion will probably never make an effort to put itself on the map in Germany (July 25, 2024 ip fray article).
• Lenovo was enjoined a few months ago, the appeals court (despite already having received the EC’s amicus brief in another case at the time) inexplicably declined to stay enforcement, and is currently unable to ship products to Germany.
• The next one to leave the market will come unless some people wake up now and start to think hard about the real world outside their courtrooms.
• In 2012, well before Sisvel v. Haier but in response to the same (one might even call it neofascist, figuratively speaking) SEP enforcement ideology (Orange-Book-Standard at the time), Microsoft moved its European warehouse out of Germany and over to the Netherlands. That fact was confirmed and it was widely reported in the press. Microsoft is one of the world’s leading patent holders and innovators. It values IP, but had to make a decision.

All of that is a disaster and a tragedy for consumer choice, competition, innovation, foreign investment and employment. Now that DG COMP has explained where the German courts are wrong (and has thereby vindicated those who were the voices of reason (August 2024 LinkedIn post by ip fray) until the Federal Court of Justice gutted Huawei v. ZTE, something must give.

Now get it right, but don’t throw out the baby with the bathwater

The UPC gets the EC’s guidance just in time. In Panasonic v. OPPO, the Mannheim Local Division’s Presiding Judge Dr. Peter Tochtermann raised, in a non-suggestive manner, the question of whether an unwilling-licensee holding would dispose of OPPO’s FRAND counterclaim (June 27, 2024 ip fray article). He was one of the voices of reason on how to apply Huawei v. ZTE. He has that chance again now, sitting on the UPC and being supported by EU institution that has the constitutional role of the Guardian of the Treaties.

The Munich Higher Regional Court should have lifted the injunction in InterDigital v. Lenovo, where there is a FRAND determination from the UK (now also from the appeals court) that shouldn’t just be ignored (at least unless German courts decline to make that kind of effort themselves). In VoiceAge EVS v. HMD, the circumstances are different and even after the EC’s amicus brief, affirmance of the injunction could still be justified. The amicus brief discusses differences between the Munich and Mannheim rulings, but Mannheim enjoined HMD as well (even twice as often as Munich did).

Sisvel v. Haier is a now-discredited decision by a court of (apart from rare circumstances) final appeal that made bad law every time it decided on SEPs. Fortunately, that court’s opinions are not binding beyond a given case.

So what should happen now?

Huawei v. ZTE should again be applied sequentially.

The bottom line should not be that all or almost all SEP holders are denied injunctive relief because that (while some would advocate such policy) is not what the ECJ said in Huawei v. ZTE.

If instead of virtually 100% “only” 75% of all SEP holders overcame an implementer’s FRAND defense, one would have to look at the reasons decision by decision, but at least the win rate would then have entered the realm of prima facie plausibility.

It would oversimplify the question to assume that a 50-50 split is the most correct outcome. There will be a higher percentage of bad actors on one side than the other. Bad actors will sue more often if they’re SEP holders, and will get sued more often if they’re implementers. There are too many factors in play that one could reasonably assume there must be as many FRAND wins as losses.

After the EC’s amicus brief, there will likely and hopefully be more than just one successful FRAND defense in four years (again, an outlier not just statistically but also for substantive reasons) and more than two such defenses in six years. And that’s not because of the EU SEP Regulation, which faces a doubtful future now as it doesn’t tackle the real issue, which is now likely to be addressed, hopefully in a way that treats both sides fairly.