With this article, I report from a very interesting panel on “Patents & Antitrust – Key takeaways for IP attorneys” which took place at John Marshall Law School’s 61st Annual Intellectual Property Conference. Panel speakers were Dina Kallay (Ericsson), Dirk Weiler (ETSI/Nokia), Tim Syrett (WilmerHale), Carlos Aboim (Licks), and my humble self. The panel focused on standard essential patents (SEPs) and FRAND (fair, reasonable and non-discriminatory) licensing commitments regarding such patents. Instead of just reproducing what was said on the panel, I am going to combine a couple of key takeaways with additional information on the legal situation in the EU.
As a core topic, the panel discussed the relation between SEP ownership and market power. Antitrust agency communications, such as the EU Commission’s Horizontal Guidelines and the FTC/DOJ Licensing Guidelines underline that market power does not necessarily result from patent ownership as such. They contain, however, no specific language on standard-essential patents which are – if they are valid and truly standard-essential – different from other patents in that they must, by definition, be used in order to operate on the respective standard-based market. In Europe at least, it seems to be increasingly accepted that SEPs can convey market power but that they do not necessarily always do so. Advocate General Wathelet’s proposition (para. 57 et seq. of his opinion in the Huawei/ZTE case) to establish a rebuttable presumption that SEP ownership generates market power has not been taken up by the CJEU’s Huawei/ZTE-decision, probably because the parties already agreed that Huawei held a dominant position (para. 43). But court decisions from the UK (for instance Unwired Planet/Huawei, a summary of the case is provided here) and Germany (for instance LG Düsseldorf, 26.3.2015, 4b O 140/13) have taken a case-sensitive approach, looking not only at the leverage generated by a SEP but also at circumstances which may limit its holder’s power. Factors limiting market power may be, inter alia, a high number of SEPs in the standard, the need for the SEP owner to engage in cross-licensing with implementers, a strong market position of implementers in general, or even the FRAND commitment in and of itself. Furthermore, typical ICT standards consist of many different elements, not all of which are equally relevant to all implementers. While most implementers are likely to use the “core” of a standard, a handheld producer will employ additional parts of the standard which are different from the additional parts implemented by – say – a producer of base stations. And some parts of the standard may even be “dead language”, not implemented at all, for instance because the features encoded there never appeared sufficiently attractive to be put to the market. Against this background, it makes an evident difference whether your standard-essential patent reads on the core of the standard; whether it reads on an additional part that is key to the implementer you are negotiating with; or whether it reads on a part of the standard which is of little interest to that particular implementer.
Pointing to a recent publication by FTC’s Acting Chairman Maureen Ohlhausen, the audience raised the question whether a FRAND commitment eliminated the potential for market power altogether by taking away the patentee’s ability to leverage on the standard-essentiality of its patent. Personally, I have my doubts whether this is true since – to give just one reason – most FRAND licenses are not formulated or controlled by courts or independent agencies and at least in such cases SEP-based leverage can impact licenses negotiated between the parties. EU courts, in any event, feel free to apply EU competition law to SEP cases in spite of the existence of a FRAND commitment.
On a different, but related note: It was stated in the discussion that, in order for a FRAND commitment to eliminate SEP-based market power, the FRAND commitment should ensure royalty rates reflecting the value of the respective patent before it was integrated into the standard. Assuming that the definition of a technical standard and the integration of a SEP into the standard adds value to the SEP, I personally doubt this often-repeated credo according to which the patentee is not entitled to (any of) this added value under a FRAND regime. Certainly, royalties should not fork over all of the added value to the patentee. Given that the patentee’s willingness to standardize its patent-protected technology and to accept a FRAND commitment helped to create the added value in the first place, though, I don’t see why implementers and/or consumers should reap all of the added value and the patent holder none.
In any case, most court decisions in Europe (and apparently also in Brazil, as Carlos Aboim pointed out) are very short on the question of market power via SEP-ownership and focus more on whether the parties’ conduct complied with the requirement established by the CJEU in Huawei/ZTE. This may provoke the critique that the rules of EU competition law on unilateral abuses (mainly Art. 102 TFEU) are applied without a key prerequisite – namely dominance, conveyed by the SEP and the other circumstances of the case – being properly established. SEP holders often argue that the difficulty of getting implementers to license SEPs at all contradicts the narrative of SEP-based dominance. Reliable empirical research on this point would be much appreciated, just as further court decisions on the relation between SEP ownership and market power.
As a second main issue, the panel discussed where the law stands regarding the right of a SEP holder to seek an injunction in spite of the existence of a FRAND commitment. While the eBay/MercExchange factors and the Noerr-Pennington Doctrine loom large under US law, the EU lacks – so far – a harmonized patent law and Member State law can be pretty different from the US framework. Germany – being one of the most important patent jurisdictions in the EU and, for that matter, a notoriously patentee-friendly one – has, for instance, no such thing as an eBay/MercExchange rule in its patent law. Instead, an injunction is traditionally the almost automatic result of the finding of an infringement. In this world, the creation of a far-reaching, competition law (Art. 102 TFEU)-based FRAND defense which permits licensees to bar injunctions in case they are willing to take a FRAND-license is a very momentous development. Arguably it forms, at least for the ICT sector, the most significant limitation on injunctions ever established in German patent law. The existence of this new barrier is, however, not to mean that injunctions are no longer issued in SEP/FRAND cases. In a number of cases, courts in EU Member States have, in fact, issued injunctions because they found an SEP to be infringed and the implementer not to be sufficiently willing to take a FRAND license. Some of these injunctions have been stayed by second instance courts, but not all (find an extensive – German – summary of the post-Huawei case-law in the EU here).
Finally, the panel tackled the question of whether a SEP holder is free to choose on which level in the chain of production and distribution it grants a license. Licensing the producer of the end product (e.g. a smartphone) versus licensing the producer of a patent-implementing component of the end product (e.g. a chipset) are two prominent options in this respect. Panelists disagreed on whether the IP policies of important standard-setting organizations (SSOs), such as ITU or ETSI, require the granting of FRAND licenses to interested implementers on all levels. Dirk Weiler from ETSI/Nokia reported (without naming parties) a case in which ETSI and ITU were asked to treat a patentee’s refusal to license on a particular level as the complete refusal of a FRAND commitment, but apparently the SSOs declined to do so. Some of the (first instance) post-Huawei case-law, by the way, seems to establish a formula according to which the SEP holder has to license all willing implementers, regardless of the level they operate on, while he is, in principle, at the same time free to choose the level on which he sues for infringement if no license is in place (cf., for instance, LG Düsseldorf, 31.3.2016, 4a O 73/14). It will be very interesting to see whether appellate courts confirm this formula.
The panel seemed to be in agreement, however, that efficiency ought to be a key criterion for choosing the licensing level and that, hitherto, licenses were usually granted on the level of the end product. For patentees, this is advantageous inter alia because it avoids early patent exhaustion: Once a licensed component is sold to the next level in the chain of production patent rights regarding this particular component expire under the doctrine of exhaustion. If, however, the license is given on the end product level exhaustion usually also kicks in only on this level. It remains to be seen whether new technologies and new business models – such as “smart cars” and IoT devices more generally – call for a change of the established end product level-licensing as component producers may now be in a relatively better, more informed and more efficient position to negotiate licenses than end producers.