In an earlier post I explained the issues that the proposed EU Regulation on SEPs intends to address, and why neither the market nor the courts solve them. Here, I discuss three points of critique brought against the Proposal: that it would reduce innovation incentives; that a family-level essentiality assessment would be too imprecise; and that the EUIPO is not prepared for the tasks for which it has been designated in the Proposal. I also comment on differences between the leaked draft of the Regulation and the final version.
SEP royalty income
Critics have expressed concerns that the proposed Regulation will reduce SEP royalty income, and consequently harm SEP holders’ innovation incentives and even European competitiveness in ICT. Mallinson writes that the Regulation would do “economic harm to licensors, including leading EU innovators Ericsson and Nokia — that are highly dependent on SEP income to fund R&D […]” Wild perceives a “[t]echnology security threat” for Europe, arguing that a reduced licensing income of Ericsson and Nokia, which the Regulation would presumably entail, would jeopardize the viability of their network businesses.
These are important and relevant concerns. Yet, it is worth deconstructing the underlying arguments and assumptions. The first claim is that the Regulation would reduce SEP licensing income, when it is designed to promote more transparency and consistency regarding the aggregate royalty and does not address the level of royalty. The underlying assumption seems to be that the expert opinion provided for in the Regulation – as a third potential step after determination of an aggregate royalty by the contributors alone, and with support by a conciliator – yields a royalty below current levels. That may be so, and it may be justified on economic grounds (see below), but it is not certain. In fact, Müller expresses concerns that the expert opinion will yield high royalty rates. Another criticism is that the cost of the proposed registration and assessment system will reduce net licensing income. However, licensors will benefit from savings due to simplified licensing and reduced litigation, at least partly offsetting the additional costs. Owners of SEP portfolios with an above-average share of truly essential patents will benefit since, based on the assessment, they can claim a larger part of the aggregate royalty. Furthermore, revenues might increase if licenses are concluded with more implementers due to the added transparency imposed by the Regulation.
SEP holders’ innovation incentives
The second claim is that SEP holders’ innovation incentives would diminish significantly, or even disappear, if their net royalty income decreased. Currently, SEP royalties for cellular and other important standards obviously provide sufficient incentives for standard-related R&D (comprising inventive activities yielding potential standard contributions, participation in the standard development process, and patenting). Will EU standards developers still be incentivized if the Regulation reduces net licensing income for SEP holders?
The answer seems to be clearly yes. Pentheroudakis and Baron conclude, in a report prepared for the European Commission, that “the remuneration of SEPs […] appears to be attractive” and “SEPs can generate higher economic returns for their owners than other patents.” For cellular SEPs, there are several indications that licensing revenues are considerably higher than the development cost of the standards. Estimates by Mallinson, Sidak, and Galetovic, Haber, and Zaretzki for 2014 and 2016 based on analysing licensors’ royalty income put the annual cumulative royalty revenue from cellular SEPs, including benefits from cross-licensing, at around USD 20 billion. From 2013 to 2019, Mallinson estimates annual licensing revenues of five major Western SEP holders to be stable at USD 10-11 billion. Considering the rise of Asian SEP holders in this period, one can assume that the total across all SEP holders, accounting for the value of cross-licenses, has increased compared to the earlier estimates of around USD 20 billion. Benefits from the use of an implementer’s SEPs in its own products must also be considered. It seems hardly plausible that the aggregate annual spending on R&D related to cellular standard development comes anywhere close to this value. According to a recently published study by Charles River Associates (CRA), prepared for the Fair Standards Alliance, “smartphone-related upstream R&D spend of SEP holders in 2020 amounted, roughly, to only $2 billion.” Even if this estimate should be overly conservative, standard-related R&D costs are only a fraction of aggregate benefits from cellular SEP ownership. Also the increase in the number of SEP-declaring parties from around a dozen for 2G to several hundred for 5G suggests that developing cellular SEPs is highly attractive. Thus, incentives for cellular standard-related R&D are very likely to persist, even if the assumption that the proposed Regulation will reduce net royalties turns out to be correct.
Relocation of standard-related R&D outside the EU?
Could it be that the proposed Regulation will cause standard contributors to relocate standard-related R&D and standards development outside the EU, as Müller warns? It is not clear why this would happen. The Regulation applies to SEP licensing for EU patents (and FRAND determinations of global SEP portfolios), irrespective of where the underlying R&D is conducted, and applies to standards developed by non-EU SDOs such as IEEE just as much as to ETSI standards. If SEP holders wanted to avoid being subject to the regulation, they would have to stop licensing SEPs in the EU.
Downstream innovation incentives
Regarding incentives to innovate, it is surprising that much of the public debate is about the standards developing firms, not implementers. In the same vein, the 2021 SEP Expert Group’s report on the licensing and valuation of SEPs repeatedly addresses innovation by SEP owners, but does not mention innovation by downstream firms at all. A notable exception is the study by CRA mentioned above, which discusses the effect of royalty levels on downstream innovation. My own research on licensing SEPs in the IoT space, with a focus on SMEs, suggests that the lack of transparency in SEP licensing and ensuing uncertainty harm downstream innovation. It bears repeating that downstream innovation is important in the EU, especially in the IoT space. One of the Regulation’s explicit goals is to support innovation by implementers, and increased transparency will be helpful in this regard.
Precision vs. cost of essentiality assessments
The Regulation seeks to establish estimates of the share of truly essential patents in the relevant portfolios. In this regard, patent litigators Müller and Müller-Stoy criticize that “[i]f only one patent per family is reviewed for essentiality as planned, the register will not provide any transparency”, and “[s]imilar problems apply with respect to different patent claims in the same patent.” It is of course true that family members may differ regarding essentiality, and that only some of a patent’s claims may be essential. Yet, this critique imposes an overly strict standard. The Regulation is not attempting to establish full transparency, but significantly improve the current state of transparency (or lack thereof) regarding essentiality of patents that are sought to be licensed. To achieve this goal, the Regulation must make pragmatic simplifications, just like commercial providers of essentiality assessments do. It is an empirical question in how many families (DOCDB simple patent families, containing patents that have exactly the same priorities) of alleged SEPs on a standard the individual members of that family differ with respect to essentiality. Will this really be a significant share? One should also bear in mind that in real-life SEP licensing, most patent families in a portfolio of alleged SEPs are not at all assessed for essentiality, let alone every family member or even every claim. There is necessarily a trade-off between an assessment’s precision and cost. Usually, the best option is somewhere in the middle ground. Regarding potential differences between family members, the essentiality assessment of a family will be based on an EP family member, which reduces the options (possibly to one) and thus the variability of the essentiality characteristics.
The role of the EUIPO
Several commentators have raised the point that the EUIPO has no experience with patents. This critique misses the point. The EUIPO has been selected in the Regulation because unlike the EPO it is an EU agency, and because of the organizational framework it offers. Also an institution like the EPO that does have experience with patents would need time to create a unit for SEP assessment, and would have no advantage over the EUIPO regarding expertise in conciliation and FRAND determination. The main challenge for any institution will be finding qualified assessors, conciliators, and FRAND experts. While the EPO has examiners experienced with standard-related inventions, it needs these experts for examination and would hardly delegate them for SEP assessment. And the fact that no EU agency currently has experience with SEPs cannot mean this needs to remain the case forever. It will certainly take time to build the required capabilities, but standards and SEPs are here to stay, and so a long-term investment makes sense.
Well-established licensing practices?
The final version of the proposed Regulation, published on April 27, differs in several respects from the draft that was leaked a month earlier. Importantly, recital 4 stipulates that “the aggregate royalty determination and the compulsory FRAND determination prior to litigation, should not be applied to identified use cases of certain standards or parts thereof for which there is sufficient evidence that SEP licensing negotiations on FRAND terms do not give rise to significant difficulties or inefficiencies.” This change is surprising. While it is true that SEP licensing for smartphones is more established than e.g. in the IoT context, it is anything but smooth. Recent global court battles, such as Ericsson v Apple and Nokia v Oppo, show that considerable inefficiencies remain. Oppo, for instance, has halted its sales in Germany since August 2022 after injunctions issued by the Munich Regional Court. These are significant difficulties. Furthermore, as I explained in my earlier post, the fact that most SEP licensing contracts for smartphones are closed without litigation does not mean that licensing runs without problems. And finally, the exemption of certain use cases from the Regulation according to Article 1.4 will open the door for lobbying regarding the identification of such use cases by the Commission.
The Proposal for Regulation of SEPs is facing critique on various grounds. The discussion above and in my earlier post refutes several important points of critique. Others are valid and should be considered by the Commission. Contreras, for example, notes a number of drawbacks, but states that “[o]verall, I believe that the EU Proposal makes a positive contribution to the efficient and fair resolution of FRAND disputes”. SEP licensing is not a well-functioning market – it is a game of bargaining power between unequal parties, driven by deep pockets and court decisions. Given the ever increasing importance of ICT standards in our societies, SEP licensing needs to become more transparent. To achieve this goal, regulation is needed.