Kluwer Patent Blog

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.

Summary

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.