The SEP state of play in Europe’s major jurisdictions

The European FRAND/SEP landscape is developing dynamically after a recent series of major FRAND-related decisions in Germany, the UK, France and the Netherlands. The question is whether there is a pattern emerging where some jurisdictions are more favourable to SEP owners than others and what actions SEP owners and licensees should be taking as a result. Furthermore: how are things likely to develop?

Germany

German courts have issued by far the highest number of FRAND/SEP related decisions. The result is a wealth of rulings - unmatched in any other jurisdiction - on many detailed FRAND questions. However, German courts have so far shied away from independently calculating a FRAND rate for a given SEP. Instead, their clear focus has been the non-discriminatory (ND) limb of FRAND, essentially reflecting a strong believe that a “fair and reasonable” (FR) licensing rate is not a fixed – mathematically-calculated rate, but rather an acceptable range.

Under German practice, the patentee needs to show the FR-ness of a licence via a sound economic analysis, essentially applying the comparable market/licence concept and/or the top-down approach (some courts combine the two approaches to a cumulative “double approach”). As long as the patentee is able to present conclusive reasons as to why a certain licensing rate qualifies as FR, the court will turn to the analysis of the ND limb and not start its own calculation of FR.

With respect to the ND limb, one of the main open questions is whether a patentee that has acquired the SEP from another owner is bound by any licensing activity of that owner, especially its FRAND commitment made during standardisation. This question has been heavily debated since the early FRAND battles between Nokia and Qualcomm.

Against this background, the decision of the Dusseldorf Court of Appeal of 22nd March 2019 (2 U 31/16) in the case Unwired Planet v Huawei has heavily impacted the debate with a quite radical solution which comes in two parts.

The starting point of this landmark decision is the basic idea that a FRAND commitment made by the patent owner is not merely a contractual act (or even less) towards the respective standard setting organisation, but rather has an effect in rem on the patent itself. Like a partial waiver of rights, the FRAND commitment irrevocably modifies the ability to seek injunctive relief based on the patent in question. This means that no subsequent transaction involving this patent will ever “clean” the patent from its FRAND commitment.

While many people may agree with this first part, the second part of the court’s conclusions has struck the FRAND/SEP landscape like something of a hurricane: the ND limb means that the first licence ever granted involving the patent in question sets the benchmark for non-discrimination, unless any of the a subsequent licences provides an even more favourable rate. Therefore, it is the duty of the current patent owner to produce all prior licences to enable the potential licensee to assess the ND-ness of the licensing offer. If the patentee is unable or unwilling to produce these prior licences, no injunctive relief is available. Within the portfolio of all existing relevant licences, the lowest ever granted rate sets the standard for ND, even if that rate is the result of “sloppy” or unprofessional negotiations. In a nutshell: ND is “super hard-edged”.

The decision of the Dusseldorf Court of Appeal has been appealed to the German Federal Supreme Court. Proceedings before the Federal Supreme Court normally take 18 months. However, given the far-reaching effect of the issue at stake, there is a high likelihood that the Federal Supreme Court will refer these questions to the European Court of Justice for a preliminary ruling.

It should be noted that the decision of the Dusseldorf Court of Appeal is not a binding precedent – ie, the lower courts (in particular the courts of Mannheim and Munich) are free to disagree. While one may expect, as a practical matter, that the Dusseldorf District Court will give deference to its Court of Appeal, the other German courts will most likely not follow this doctrine for the time being, given its extreme and impractical nature.

United Kingdom

Since April 2017, the English courts have surprised the global patent community with their willingness to determine an actual FRAND licensing rate – on a global basis. As regards the FR limb, English courts developed a five-part algorithm to determine one. While the final calculation done by the court looks like an impressive high school math exam, the core of the endeavour is the “double approach” – ie, a combination of the comparable market/licence concept and top-down approach.

Furthermore, very much contrary to the above-mentioned ruling of the Dusseldorf Court of Appeal, the English courts do not view the ND limb as a most-favoured-nation requirement (no “hard-edged” requirement). Instead, ND defines a maximum royalty which shall not be exceeded (a ceiling, in other words). Patentee may grant lower royalty fees than the maximum rate, without being locked into this lower rate for future licences.

The Supreme Court has recently agreed to hear a Huawei appeal against the Court of Appeal's decision in Unwired Planet v Huawei – the case that began it all in the UK. Permission has also been given for Huawei and ZTE to appeal the Court of Appeal's decision in Conversant v Huawei, ZTE.

The Supreme Court is expected towards the end of this year. Unless the UK stays in the EU beyond the end of 2019, it seems unlikely that the Supreme Court will refer these cases to the European Court of Justice for a preliminary ruling.

The Netherlands

Many European scholars regard the patentee’s FRAND obligations as a comprehensive and cumulative list of duties – ie, failure to fulfil one of the multiple obligations results in overall non-FRAND behaviour. In contrast to that, the Hague Court of Appeal very recently held in the case Philips v Asus, Asustek (Case No. 200.221.250/01, 7th May 2019) that the opposite rule applies: if the defendant is not considered to be a “willing licensee” (as was the case with the litigation in question), then it doesn’t matter whether the patentee fulfilled any of its FRAND obligations. In other words, the patentee’s FRAND obligations are not considered a “check list” which needs to be completed before an injunction becomes available.

France

While French courts have been silent on FRAND/SEP for a while, the recent decision in Conversant v LG (Cour d’Appel de Paris, Decision No. 061/2019, 16th April 2019) touches an important side-aspect of the debate, namely the confidentially regime protecting the disclosure of existing licences to show the FRAND-ness of the offer made to the defendant. To protect confidentiality, the court applied the EU Trade Secrets Directive and its implementation into French law, resulting access to the courtroom was restricted to the parties’ counsel and some representatives of the parties bound by strict confidentiality obligations.

Is there a pattern – or “gold standard” – emerging? How are things likely to develop?

There seems to be an emerging consensus in Europe that the FR limb of FRAND should be determined based on the “double approach” – ie, a combination of the comparable market/licence concept and top-down approach. However, there is a quite fundamental disagreement over whether this approach can be programmed into some kind of a mathematical algorithm, which – at least in theory – enables a court to “calculate” a FRAND licensing rate.

Some market players view the absence of number acrobatics in German FRAND/SEP proceedings as problematic, so decreasing the attractiveness of German proceedings. On the other hand,  many concede that the German approach results in very high efficiency, both in terms of cost and time. In particular, in cases where the SEP owner publishes and consistently maintains its FRAND licensing terms, Germany remains the most patentee friendly jurisdiction, as the track record of, for example, the MPEG LA cases shows.

One may regard the identification of a “comparable” licence agreement entered into by the owner of the SEPs or third parties as mathematically precise as the Hogwarts’ world of science, so that the overall FRAND “calculation” at least contains a very subjective element. However, one must concede that the English bar has done a great job in promoting the FRAND “calculation” offering of the English courts as a go-to solution for critical FRAND disputes, albeit coming with a heavy price tag and a serious risk of losing the patent in an invalidity counterclaim.

On balance, it seems high time for German courts to find the courage to rule that a given licensing rate is positively FR, applying the double approach, in order to catch up with the developing FRAND debate. However, courts (in particular English ones) should concede that this is not a mathematically precise exercise. Judges are trained to professionally exercise discretion as their statutory task and there is nothing wrong with also applying this skill when it comes to FRAND. Therefore, there is no reason to crouch behind a monument of numbers when judges rule on a FRAND rate – based on discretion.

As regards the ND limb, case law is currently drifting apart. Reaching a consensus would require clear directions either from the ECJ or by a legislative act of the EU. However, given that Brexit is unfortunately a very realistic scenario, Europe may end up with an English approach and a rest-of-Europe approach, unless the UK Supreme Court injects some wisdom into the pending appeals in the Unwired Planet and Conversant cases.

Under the “soft-edged” ND approach taken by the English courts, powerful licensees are typically able to get better licensing deals than small and medium sized companies that have a weaker market position. Preventing such discrimination is the core of the FRAND doctrine under Article 102 TFEU, and the English approach seems to be a toothless tiger in that respect.

The ND approach taken by the Dusseldorf Court of Appeal, however, gives every impression of being conceived in the Ivory Tower of theoretical antitrust law, as it puts a quite impractical burden on the patentee, and/or seriously hampers any asset transaction involving SEPs.

From a legal perspective, the court converted the “abuse of market dominance” requirement of Article 102 TFEU into an abstract requirement totally detached from the actual behaviour and abilities of the current SEP owner, which seems to be a quite wrong understanding of the law. Given the patentee’s burden envisaged by the Dusseldorf Court, the value of SEP portfolios would be likely to drop significantly, making larger restructuring projects (such as the carve-out of Osram from Siemens) very difficult or even impossible, and therefore destroying valuable assets instead of promoting competition.

The recent French decision might actually set a new European trend. As the procedural instrument of “protective orders” does not exist in most EU civil law jurisdictions, while the Trade Secret Directive is now implemented in almost all EU member states, the quite creative approach taken by the French court might be copied in other jurisdiction, in particular Germany.

The recent Dutch decision admittedly creates a very favourable forum for SEP holders in the Netherlands, possibly bringing more SEP litigation activity to the country. However, looking into the crystal ball, this “party” may come to a sudden end pretty soon, when a Dutch court files a references for a preliminary ruling with the ECJ, and that court aligns Dutch practice with the “check list” approach taken by the other European jurisdictions.

Actions SEP owners and licensees should be taking right now

Given the general importance of the “double approach” and the (low to medium) risk that the ND doctrine of the Dusseldorf Court of Appeal might become the European rule of play, access to existing (prior) licensing terms should become a key provision in future asset purchase agreements involving SEPs. SEP owners should also try to revise or re-negotiate existing asset purchase agreements to ensure compliance with upcoming disclosure obligations.

As regards litigation strategy, the recent developments in the Netherlands add an interesting option when dealing with notoriously “unwilling” licensees, as the court will not expect the plaintiff to argue FRAND as long as the defendant remains unwilling. This might save significant costs. However, as explained above, it remains to be seen how long this new option will be available.

In contrast, the ruling of the Dusseldorf Court of Appeal has suddenly turned that extremely popular jurisdiction into a no-go area for SEP owners. Those currently engaged in litigation before the Dusseldorf District Court should expect that deference will be given to the doctrine of the Court of Appeal, although it is not technically binding. This might boost the popularity of the equally competent courts of Mannheim and Munich for the time being.

From the licensee’s perspective, the ruling of the Dusseldorf Court of Appeal is great news - and it is clear that every implementer currently involved in litigation and/or negotiations should vigorously plead this doctrine. However, given how extreme it is, this doctrine might not have a long lifespan. Hopefully, it will be soon replaced by a more balanced and practical approach.