The European Commission’s recently leaked Proposal for a Regulation on standard-essential patents (SEPs), summarized here by Enrico Bonadio and Dyuti Pandya, establishes a framework for transparent SEP licensing. Some have criticized the proposal, claiming among other things that it is unnecessary, harmful to innovation, and difficult to implement. Much of this critique is unjustified. More transparency is needed in SEP licensing, and the proposal takes a step in the right direction. In this post I explain the issues the Regulation addresses: the current lack of transparency, the question of whether SEP licensing processes work, and why neither the market nor the courts have solved the existing problems. In a following post, I will discuss the various points of critique.

The problem of non-transparency

In a typical licensing scenario, a SEP holder offers an implementer of the standard a license for a portfolio of supposedly standard-essential patents. Though there are exceptions, a patent being standard-essential implies that one needs to use the underlying invention to implement the standard. However, it is challenging to find out whether a single patent is truly standard-essential, let alone all patents in a portfolio including several hundreds. For standards requiring an explicit declaration of potentially essential patents such as cellular standards, the list of declared essential patents offers some guidance, but not much: estimates are that less than half of them are truly essential (IPlytics reports estimates between 20% and 30% for 5G), and the essentiality rate differs significantly across portfolios. Unless the implementer is financially strong and familiar with the standard, and the portfolio is small, it is virtually impossible for the prospective licensee to assess which patents are truly essential. This is particularly relevant for implementers in the Internet of Things (IoT), who are mostly unfamiliar with details of the standardized ICT technologies they use, and to a still greater extent for SMEs as I explain in a recent publication.

It is even more difficult to determine what portion of all SEPs that cover a standard are included in a particular portfolio. This determination, however, is important when the implementer is considering the total royalty burden for using a standard. Determining this portion requires determining the number of all truly essential patents reading on the standard (and ideally, though close to impossible, their validity and technical importance), a lengthy and complicated exercise that no individual implementer can afford to engage in. Commercial reports containing this type of data exist, but are limited to a few standards, carry a high price tag, and their quality is difficult to ascertain. For standards such as Wi-Fi, where no obligation exists to declare individual potential SEPs, not even a list of declared essential patents is available as a starting point.

The current SEP licensing framework does not promote the transparency needed regarding essentiality of patents. In licensing negotiations and court proceedings, prospective licensors typically present a short “proud list” of patents that they consider truly essential and infringed. If this list convinces the implementer or the court, clearly a license must be taken. Determining a FRAND royalty, however, requires information about the overall portfolio and the share of the total stack it constitutes, not about a small amount of proud list patents.

A standard’s aggregate royalty is a further factor lacking transparency. Firms implementing a standard need to know the associated licensing cost early on in order to devise their pricing strategy, especially if the aggregate royalty could be a relevant share of their profit margin. Without a publicly known aggregate royalty this is difficult, especially for fragmented SEP ownership as is the case with many standards. It is unrealistic to expect implementers to have licensing agreements in place with all major SEP holders before implementing a standard, especially device manufacturers who have no knowledge of the standard. Thus, early and reliable public information about a standard’s aggregate royalty is required.

These problems are bound to get worse. SEP numbers are growing with each new generation of a standard, monetization of standards is on the rise, more standards are becoming relevant, and SEP owners increasingly seek to price-differentiate based on the use of the standard. Patent pools can mitigate these problems by reducing the fragmentation of the licensor landscape and, as some do, announcing their royalty demands publicly. However, they are far from a solution as they rarely come close to covering 100 percent of a standard’s SEPs. Furthermore, implementers may for various reasons not seek or not reach an agreement with a pool, thus having to resort to bilateral negotiations.

No litigation does not mean no problems

Some SEP holders argue, as reported by Adam Houldsworth, that concerns about SEP licensing disputes were exaggerated, since most licensing contracts are closed without litigation. This logic is not convincing. In any negotiation, a party’s negotiating power depends on its options. Beyond concluding a license agreement, the prospective licensee has three options. The first is to stop using the standard. This would significantly reduce product quality, since a standard typically has no direct substitutes, or would mean ceasing a product line entirely – hardly an attractive option. The second option is to continue using the standard without a license, which may trigger legal action by the SEP owner. Even small firms might be sued by SEP holders, e.g. to build a reputation for toughness. The third option is to take the SEP holder to court, claiming it is abusing a dominant market position, or possibly requesting determination of a FRAND rate by the court. In effect, the only viable alternative to concluding a licensing agreement is litigation, and so any SEP licensing negotiation takes place in the shadow of this eventuality. Some potential licensees, however, small implementers in particular, cannot afford litigation due to financial constraints or for reasons such as avoiding uncertainty for investors. Others may refrain from litigation if a negative outcome and even an injunction can be expected, also if they consider the licensing demands excessive. No litigation therefore does not mean SEP licensing runs smoothly.

The market does not fix it

Opponents of regulation maintain that one can rely on the market for determining FRAND rates. This argument relates to the one above and is equally unconvincing. As explained, a firm operating in a certain industry often has no alternative to using the standardized technology, and thus must license the related SEPs. If it rejects a licensing offer, the prospective licensor can turn to the courts. In turn, the licensor is obliged to license its SEPs under FRAND conditions, an obligation that a prospective licensee could try to enforce through litigation. Thus, the situation is not a market in the common sense of the word. It is a bilateral negotiation whereby parties’ fall-back positions are determined by court decisions, actual or potential.

Courts do not solve the problem

Do courts solve the problems of SEP licensing? Hardly. While in various instances courts e.g. in the UK and the US have set FRAND royalties, they neither identify each SEP owner’s share in the various standards, nor systematically set aggregate FRAND royalties for all standards and the wide range of uses that for example the IoT enables. As such, they do not provide the needed transparency for the licensing market to operate efficiently.

Further, the current practice where FRAND litigation cases start with infringement allegations regarding individual patents is problematic. First, if a holder of a large portfolio of supposed SEPs and an implementer of the relevant standard litigate, as Ericsson and Apple did until recently about 5G, it is certain that some patents in the portfolio will be essential, valid, and infringed by the implementer. Thus, the only relevant question is what a FRAND royalty should be. The litigation of individual patents is merely the lever that, if successful, helps the patent owner to force the implementer into licensing. It is superfluous for cases such as Ericsson v Apple, costly and lengthy for the litigants, and clogs the court system. The registry of assessed SEPs in the proposed Regulation offers a shortcut by providing evidence if a portfolio contains truly essential patents. While the Regulation does not provide for validity checks, information on essentiality should nonetheless increase efficiency of SEP licensing.

The second problem are the potentially extreme outcomes of court proceedings. Many litigated patents are found invalid or non-essential, and so a SEP holder who picked the wrong patents for litigation may end up empty-handed. In contrast, a successful plaintiff may obtain an injunction, particularly in the popular German courts. So litigation can be an all-or-nothing gamble, and neither outcome is satisfactory from a policy perspective. A complete loss for the SEP holder would deprive it of the royalties it deserves. And an injunction puts the patent holder in an excessively strong negotiating position, possibly enabling it to extract a royalty above FRAND. Injunctions are particularly unsatisfactory as they make little economic sense in the context of SEP licensing. The question if Huawei/ZTE conditions have been met, and specifically if the defendant is deemed a “willing licensee” hinges often on whether the court considers the implementer’s counteroffer is FRAND. Thus, courts often do decide FRAND royalty levels implicitly. But if a court can determine that an offer is not FRAND, it presumably could also state what royalty is FRAND. The sensible solution from an economic perspective is therefore to order the implementer to pay a royalty set by the court, and only resort to an injunction in the case of noncompliance. The High Court of England and Wales took a comparable approach in its recent decision in the litigation between Interdigital and Lenovo.


SEP licensing is riddled with non-transparency, and the problems will only get worse. Registration and essentiality assessment of supposed SEPs as in the proposed Regulation will show whether a prospective licensor owns any truly essential patents, allow parties to consider a SEP portfolio in total, and provide an estimate of the share of the standard it constitutes. A publicly known aggregate royalty removes pricing uncertainty for implementers. Needless to say, the outcome of a conciliation as envisaged in the Regulation may not be accepted by the parties. However, as Thomas Cotter suggests, there could be “considerable pressure to reach agreement speedily once the conciliators made their FRAND determination […]” In cases where the parties settle after conciliation without subsequent involvement of courts, the process will be considerably faster and cheaper than litigation, and even when litigation does ensue the outcome of conciliation may help guide court decisions.


To make sure you do not miss out on regular updates from the Kluwer Patent Blog, please subscribe here.

Kluwer IP Law

The 2022 Future Ready Lawyer survey showed that 79% of lawyers think that the importance of legal technology will increase for next year. With Kluwer IP Law you can navigate the increasingly global practice of IP law with specialized, local and cross-border information and tools from every preferred location. Are you, as an IP professional, ready for the future?

Learn how Kluwer IP Law can support you.

Kluwer IP Law
This page as PDF


  1. Maybe DG GROW has missed a trick with the draft regulation. As well as setting up a register of SEPs, the regulation could also set up a register of users of the technology who commit, in a legally binding manner, to licensing the registered patents under conditions determined by the FRAND determination panel, thereby balancing out the obligations imposed on the SEP owner to license their patent. The absence of such a declaration would then be taken by the courts to be a prima facie indicator that the implementor is not a willing licensee and may be injuncted automatically on a determination of patent infringment.

  2. The EU proposal appears to be the wrong answer to a real problem.

    When determining SEP/FRAND license fees different aspects have to be considered.

    Firstly, which authority decides that a patent is a SEP? It is normally the Standard Setting Organisation (SSO), also sometimes referred to as SDOs. It cannot be the patent holder, although it might have contributed to the setting of the standard.

    Secondly, is the SEP at all valid and hence enforceable? If the patent is invalid, why should ever FRAND, or any license fee, be paid? Transparency should be made so that SEPs can easily be challenged.

    Here we see the whole difficulties with SEP/FRAND licences.

    The SSO cannot decide whether a patent is valid or not. This is normally a civil court in a country in which the granted SEP produces effects. We have thus the need to deal on the one hand with an administrative body and on the other hand with a judicial body.

    That the present system is not satisfactory due to its lack of transparency is a real problem. Transparency is thus a necessity.

    A first step towards transparency would be to forbid mixing SEPs and non-SEPs in a portfolio offered for licensing. This is the main cause of the lack of transparency. Decoupling SEPs from non-SEPs is a bare necessity. A licensee should not be obliged to buy in a package both SEPs and non-SEPs. This is simply an abuse, although it is normal practice.

    An EU directive requiring transparency in matters of SEP would be a good way in the direction of solving the transparency problem. A list of SEPs for a given standard should be held at the SSO. Then every potential licensee would know which patents can be licensed under FRAND conditions and which patents can be licensed under free market conditions. Bundle buying is too easy to conceal useless patents in an effort to monetarise their value. This is also why NPE have got such an importance in the present patent world.

    Going any further from the side of the EU is most questionable. That on top the EU and the EUIPO, which has no competence whatsoever with patents, should have their fingers in the pie of SEP/FRAND licences is not a good solution.

    SSOs should be given a much important role in the system, as thy best know which patents are SEP and which are not. Why could the valuation of true SEPs not be given to SSOs? If there is a transparent list of what is considered a SEP, it would also be easier to challenge their validity in court.

    This is certainly not a proposal which will get an enthusiastic reply from patent holders in general and SEPs holders in particular, but it could be step in the right direction. The thrust should be towards SSOs and not towards another EU construction.

    That UK courts think they can decide on a world fee for licenses in general, and FRAND licenses in particular, should not be tolerated within the EU. After all, with Brexit, the UK has turned its back to the EU and it has to bear the consequences. Wanting to rule in matters of FRAND licenses might be good for the UK legal profession, but it should not be encouraged.

    In how far this new “EU law” is there “to nuke free software” needs explaining. It is an apodictic stance without any content. Lots of SEP might involve software, but it is not software as such and it has a direct technical effect and hence it is patentable.

Comments are closed.