EU Commission on "FRAND"
New guidance on the “F” in “FRAND” is forthcoming
14 December 2018
The European Commission is seeking to publish a communication giving guidance on what constitute Fair, Reasonable, and Non-Discriminatory (FRAND) licensing terms and practices. FRAND terms are the set of principles with which owners of patents essential to standards (standard-essential patents or SEPs) must generally agree to in return for inclusion of their technology into the relevant standard. Any communication from the Commission on this controversial topic is likely to become extremely influential in any subsequent dispute or negotiation on FRAND licences. This note summarizes key issues on which the draft communication reportedly seeks to offer guidance.
Owners of SEPs must generally agree to give a commitment to license these patents on FRAND terms as a condition for inclusion of their technology into the standard. The FRAND commitment aims to ensure that no SEP holder can unilaterally block access to the standard. Notwithstanding limited guidance from recent United States and United Kingdom court judgments determining FRAND terms in individual cases, there is no precise definition of what constitute FRAND terms. Licensors and licensees of SEPs continue to be divided on how FRAND terms ought to be determined. Some believe that additional guidance from the European Commission is necessary to improve legal certainty.
While a formal communication from the Commission on how to determine FRAND terms could help reduce legal uncertainty, it is also likely to prove controversial. The more favourable the FRAND determination method is to licensees, the lower their likely cost of implementation of standards and the wider the dissemination of the standard. Conversely, under these terms SEP holders may be less able to recover for third party use of their technology and have fewer incentives to contribute technology to new standards.
However, it would be wrong to view the question of what constitute FRAND terms as one dividing "innovators" from "implementers", which would suggest that only the former innovate and that the latter only implement.
First, many companies are both contributors to standards (SEP holders) and implementers of standards. They have balanced interests. Second, and more importantly, innovation is not exclusively the domain of SEP holders. Although SEP owners' contributions to new standards are clearly important to the development of new standards, most innovation is undertaken by firms who build on top of the standard. Indeed, the significance of standards is arguably that they facilitate innovation on top of a common infrastructure. By way of example, the internet is arguably one of the most valuable standards of recent history. Still, the value of the internet as such pales compared to the innovation that has been built on top of the internet.
Thus, any answer to the question of how to define FRAND terms not only determines what SEP holders may demand in return for the use of their technology and how much consumers will pay for their devices, but also to what extent innovators are incentivized to innovate on top of standards. For example, how costly it will be for start-ups to build innovation on top of a standardized infrastructure and how difficult it will be for start-ups to obtain funding for such innovation.
Two main issues
The Commission's draft communication reportedly addresses two key aspects of FRAND terms and conduct, namely (i) refusals to license, and (ii) use-based licensing. The first is concerned with whether the SEP holder is free not to license to particular licensees in the supply chain. For example, to license only to finished product makers rather than component makers. The second is an approach to licensing pursuant to which the licensor charges different rates depending on the end-use of the product incorporating the patented technology, even if that technology is exactly the same. The two concepts appear to be related: the more freedom an SEP holder has to refuse to license at particular levels of the supply chain, the easier it is to choose to license at a level close to the end use, and thus extract value from that end-use.
(i) Refusals to license
SEP owners argue that they should be free to choose at which level of the supply chain they grant licences, and in particular whether to license finished product makers, or manufacturers of components. According to some SEP holders, licensing at the finished product level is more efficient than licensing higher up in the supply chain, as there are fewer finished product manufacturers than there are component makers.
Many (prospective) licensees take a different view. Manufacturers of components incorporating standard-essential technology often prefer to sell licensed products to finished product manufacturers, rather than selling unlicensed products that require customers to negotiate a licence with the SEP holders themselves. Many manufacturers of finished products would also prefer to purchase licensed components, and are concerned that a licence applicable to the finished product in effect forces the device manufacturer to pay royalties that extract value created by the device manufacturer rather than the SEP holder.
Some guidance already exists on the question of whether SEP holders may refuse to license, and departing from this guidance would be a bold move. The Commission's own Horizontal Guidelines provide that SEP holders must license to "all third parties." Recent regulatory investigations by the competition authorities of Korea and Taiwan have also confirmed this principle.
(ii) Use-based licensing
The idea behind used-based licensing is that an SEP holder should be able to charge different rates depending on the end-use made of its SEP, even if the technology covered by the SEP is the same. This would mean that the SEP holder could charge a different rate for his SEP used in a smartphone as compared to the same SEP used in a smartwatch, for example.
Proponents of use-based licensing argue that the value of SEPs cannot be determined in a vacuum but is demonstrated principally by reference to particular use cases. Therefore SEP holders should be able to collect royalties that depend on the nature of that use. In addition, it has been argued that such price differentiation would allow for lower prices for applications that do not use the patented technology as intensively as others, thereby potentially lowering barriers to entry. SEP holders have also expressed concerns that inability to pursue a use-based licensing approach would endanger their ability to be fairly remunerated for the use of their technology in standards, which in turn would reduce incentives to innovate and contribute to new standards.
Opponents of use-based licensing disagree that the value of an SEP depends on how it is used. Their view is that the technology covered by the SEP fulfils exactly the same role in any standard-compliant product regardless of its end-use, as indeed the function of the technology covered by the SEP is defined by the standard. For example, the SEP used in the smartphone and the smartwatch is exactly the same, and so is the function fulfilled by the technology covered by the SEP. Opponents of use-based licensing thus argue that the differences between products incorporating the same standard (and thus the same SEPs) lies in the innovation others have added to the product. For example, even if the same standards (and thus SEPs) may be used in a smartphone and a smartwatch, the smartphone generally contains various valuable components and inventions that may not be found in the smartwatch. It has thus been argued that, if the SEP holder could charge a higher royalty for the smartphone than for the smartwatch under a use-based licensing model, the SEP holder would effectively be permitted to charge royalties on technology invented by others. This could be regarded as a tax on innovation, and which would not be consistent with the aims of standardization. Opponents of use-based licensing also argue that, conversely, in order to ensure adequate incentives to innovate on the part of SEP holders, it should be sufficient to reward SEP holders for their own innovation. According to them, enabling SEP holders to extract the value of the follow-on innovation added by third parties may disincentivize third parties from creating follow-on innovation. In addition to this, they argue that SEP holders cannot credibly claim value extracted from follow-on innovation as nobody knows what things others will do with the standard (consider the internet example).
Concerns have also been expressed about whether charging different rates for similar (or identical) situations might violate the "ND" in FRAND.
Questions also arise as to the compatibility of use-based licensing with existing guidance on standardization agreements, and in particular the European Commission's Horizontal Guidelines. The Horizontal Guidelines do not appear to recognize use-based licensing as a legitimate form of FRAND licensing. They suggest that FRAND valuation should not be based on any strategic or hold-up value, and suggest methods of evaluating the ex-ante value of the SEP (incremental value in relation to alternatives). If the SEP holder is permitted to extract part of the value of follow-on innovators by charging a royalty based on the technology's end use, this could in effect mean that he would be allowed to extract part of the hold-up value - the value the SEP holder can only extract by virtue of his technology having been adopted as part of the standard. It could be said that use-based licensing thus confuses (i) the value of the SEP holder's contribution to the standard, and (ii) the contribution of the standard to the end-product.
How would principles co-exist? Does one undermine the other?
According to some reports, the Commission may be minded to endorse both a prohibition on refusals to license and the principle of use-based licensing. If correct, this raises the additional question of whether these two principles could in fact co-exist, or whether one would undermine the other. In particular, some have pointed out that endorsement of use-based licensing could effectively undermine a prohibition on refusals to license. For example, even if a prospective licensee operating upstream, such as a chip manufacturer, may be entitled to request a licence from the SEP holder under a prohibition on refusals to license, the SEP holder on its part may be entitled to insist on a use-based licensing model, requiring the chip manufacturer to inform the SEP holder what the end-use of the licensed product (the chip) is. The prohibition on refusals to license could be said to be at odds with the use-based licensing approach where the chip manufacturer is unable to answer this question. Will use-based licensing trump the prohibition on refusals to licence, meaning that the chip manufacturer may be unable to negotiate a licence without knowing what his chips are used for? Or does the prohibition on refusals to license prevail, meaning that the SEP holder must offer a licence even if it cannot be one following the use-based licensing principle?
Given the sharply divided views on refusals to license and use-based licensing, the question arises whether there may be too many unresolved questions to provide formal guidance on these issues. Even if not legally binding, any such guidance will likely prove very influential in any subsequent FRAND dispute, and, by extension, on incentives to innovate in Internet-of-Things technologies. The Commission need not become part of the controversy, however, and may prefer to choose a middle road. For example, in relation to use-based licensing in particular, the Commission may choose to stop short of an absolute endorsement and instead recognize that, in appropriate circumstances, use-based licensing may be compatible with FRAND principles.
Key Take-Away Points
- The European Commission is expected to give guidance on the controversial topic of what constitutes FRAND licensing terms for Standard-Essential Patents.
- The guidance may endorse the principle of use-based licensing, pursuant to which licensors could charge different royalties depending on the end use of their patented standard-essential technology.
- Even if not legally binding on licensors or licensees, guidance from the European Commission is likely to become very influential in any subsequent dispute on FRAND terms.