Standard-essential patents (SEPs) have been critical to the ICT revolution. SEPs have allowed for the fast rates of innovation diffusion that the world has witnessed in the past 25 years. Yet the SEP system is under pressure. It suffers from a smoldering crisis of confidence as costly legal disputes across several international jurisdictions have caused unpredictable frictions in the markets for standardized technologies. Regulators in several parts of the world are now considering actions that seek to overcome obscurities in the SEP system. Asymmetric information is at the very heart of current problems in the market for SEPs, and all too often resembles a market dominated by a “confusopoly” with little transparency about products, quality and prices. In this paper, we will discuss ideas and concepts for what could be done to maintain a balanced and trusted system that supports technological innovation and at the same time conforms to economic efficiency.
Research assistance by Nicolas Botton and Julie Richert is gratefully acknowledged.
Improving the Governance of SEP Markets
Given that policy ambition, it is natural to ask: what problems, more specifically, need fixing for the SEP-system to remain a central part of new technology creation and innovation? As already mentioned above, due to the nature of most problems of the current system, any improvement will have to start with transparency and ensuring that there is clarification about the actual market: what is the product (license) that is offered and why is it essential to the standard? And there is no point beating about the bush: currently, SEP markets are all-too-often obscure and there is such a degree of market failure that there is no other way than for standard-setting bodies to become far more demanding about the conditions for a patented technology to become part of a standard.
Several court decisions in both Europe and the United States have illustrated the obscurity of the SEP market, notably Huawei v. ZTE, which went into a prescriptive direction for how negotiations between contracting parties should be structured. This obscurity is partly related to the high levels of asymmetric information between the seller and the buyer of SEP licenses. Information about the essentiality of patents as well as the level and differentiation of royalty fees among buyers is often only available to the licensor, while implementers have fragmented information. Consequently, users of SEPs are often unaware of what they have purchased exactly and, as some disputes have shown, whether the license effectively protects them against patent infringement litigation. In a market that is obscure, there is also the risk that users may refuse to purchase licenses because of lack of information or doubts about prices, patent validity or essentiality claims in the first place. This is at the very heart of current problems in the market for SEPs, and all too often resembles a market dominated by a “confusopoly” with lacking transparency about products, quality and prices (see, e.g., Ahmed 2015, Kalaycı 2015; see Figure 2). As always, Dilbert puts the finger on the right problem.
Note: Adams (1997) introduced the word confusopoly. The word is a combination of confusion and monopoly (or rather oligopoly), defining it as “a group of companies with similar products who intentionally confuse customers instead of competing on price”. Examples of industries in which confusopolies exist (according to Adams) include telephone service, insurance, mortgage loans, banking, and financial services. Source of graphic: http://dilbert.com/strip/2010-11-21.
The introduction of FRAND rules intended to improve transparency by mitigating potential market abuse as well as the problems of asymmetric information in SEP markets. That development is still under way and there has been a gradual and evolutionary development in which courts have increasingly interpreted these rules and what they should mean in practice. There is quite some distance to go before court decisions have given clarity to all matters that need to be cleared up – and it is not exactly a linear development either. Nor are all courts necessarily arriving to the same conclusions. Some stakeholders doubt that they can. A study by the Joint Research Centre of the European Commission (2015, p. 11), for example, argues that “the variety of licensing practices has made it more difficult to identify a consensual approach to FRAND licensing.” And, rather than resolving problems, the “involvement in litigation has drawn much attention, fueling controversy on the interpretation and effectiveness of FRAND commitments in the current industry context” (JRC 2015, p. 11).
Amplifying the obscurity is the fragmented information available for parties outside the license contracts. Lack of data regarding the quality or essentiality of an SEP as well as licensing fees are already causing problems for the contracting parties. Consequently, many SEP stakeholders have raised concerns over issues that are difficult to empirically validate – and that is an even bigger challenge for outside observers. The study by the Joint Research Centre (JRC 2015, p. 16) also makes the point that “there is no empirical evidence that clearly breaks the tie between the “hold-up” and “hold-out” arguments.” In that context, a recent report commissioned by the European Commission makes a similar argument, stating that there is “no reliable empirical analysis of hold-up within SSO-based standardization processes.” (Regibeau et al. 2016, p. 12).
Proposals for improvements of the SEP system have to take aim at the low degree of market transparency around SEPs, which lies at the heart of the problem of artificial pricing mechanisms in SEP markets. In the past years, there have also been claims that current FRAND requirements are insufficient, and several concerns have focused at the over-declaration of SEPs, no-challenge clauses regarding the standard-essentiality of a patent and confidentiality obligations concerning royalty fees effectively preventing market forces from working towards more transparent, balanced and commonly accepted outcomes. There are also demands for both greater flexibility and limits for how license prices should be determined in the FRAND context. Some suggest that there should be a greater role for price variation in the same way as there are royalty caps. Others argue for pricing models based on the concept of return on investment or pricing according to how the license buyer uses the purchased technology. Let us look closer at some of the main ideas.
Over-Declaration of SEPs
Over-declaration of SEPs is a major determinant of obscurity in SEP markets. The declaration is a decision exclusive to the patent holder and a self-declaratory process. However, several studies show how a significant part of the patents declared as standard-essential are in fact non-essential for the implementation of the standard or not even covered by the standard’s specifications (see for instance Audenrode et al. 2017 and Pohlmann and Blind, 2016).
Problems arise from the fact that neither SSOs nor patent authorities systematically conduct essentiality checks, while at the same time patent holders generally have an incentive to over-declare patents as they aim to protect the enforceability of potential rights. As outlined by van Audenrode et al. (2017), patent holders tend to err towards declaring an SEP even when they doubt the standard-essential characteristics of a particular patent. A recent study done for the European Commission articulated the same observation: “…in a world where there is little objective verification of essentiality, there is an incentive for patent-holders to declare a large number of their IPRs as essential, if only because royalty payments tend to increase with the size of the licensed portfolio” (Regibeau et al. 2017). In addition, SSOs often encourage patent holders to declare all patents that might be essential for a given standard.
The problems associated with over-declaration have been intensively discussed by policymakers in the past, although solutions have not been found. In academia, the problem of invalid SEPs is still “under-recognised” (van Audenrode 2017, p. 3), leading to significant short-comings of empirical analysis. At the same time, policymakers as well as courts have so far struggled to find solutions to the problem of patent essentiality.
None of this is surprising. Essentiality characteristics are indeed a complex matter. Patent claims can be exhausted or amended over time. The scope of protection of a single family or different national patents can vary around the world. In addition, as standards evolve over time, essentiality might change too. For all SEP stakeholders – including patent offices, SSOs and (potential) regulators – thorough essentiality checks can therefore be time consuming and costly.
Markets are generally in the best position to determine the essentiality or the value of a commodity – in this case the value of a patent that has been claimed essential to a certain standard. However, under the current system, a market-based determination of the essentiality and value of a given patent is often curtailed by market participants themselves as well as contractual arrangements. For many companies, including resourceful multinationals, there are often very high transaction costs associated with obtaining the information about the essentiality and value of the SEPs that they are offered.
This is particularly the case when companies are subject to so-called “no challenge clauses” (or no-contest clauses). No-challenge clauses forbid SEP implementers to challenge the validity (or enforceability) of a patent declared as SEP or participate in any challenge or aid any third party challenging an SEP (see, e.g., Collins 2015). As a result, SEP implementers – individually or collectively – are prevented from challenging an SEP, even if they come to the conclusion that an SEP is not essential to a given standard. No-challenge clauses are, therefore, a powerful means to obscure the quality and essentiality of a patent for actors that are less attentive to the needs of SEP implementers.
No-challenge clauses underpin the “confusopoly” that characterizes parts of the SEP market. In effect, buyers may have to purchase several licenses for patents that are not essential and, as a consequence, the costs go up for end customers and the allocation of market resources gets distorted. In other words, the over-declaration of patents and non-transparency in matters of essentiality enable and potentially reinforce pricing strategies that stand in opposition to the public desire of promoting innovation and competition.
Essentiality checks are expensive for everyone, including SEP holders and SSOs, which is why there are no mandatory essentiality checks. It is virtually impossible for an SSO to conduct essentiality checks for every declared SEP. There is, however, no way to improve the SEP market without greater information and transparency about the product and its essentiality. Therefore, steps have to be taken to improve transparency. While no system can be made perfect, there are promising ideas for how improvements can be made.
A first step is for SSOs to selectively examine the essentiality of patents, either randomly or by focusing on standards that are more important than others. While SSOs may not have the technical expertise to conduct such reviews themselves, they can commission patent offices for such work. There are also ways for SSOs to make essentiality claims to have a greater effect on the information imbalance between holders and users. For instance, it would be reasonable for SSOs to have an appeal mechanism to which parties could turn for an examination of the essentiality of a patent when it is in doubt. After all, essentiality to the standard is the core of the product that a buyer is purchasing through a license and, given the role of an SSO to develop the standard, there is no way to escape that non-essentiality has to be acknowledged by the SSO for the market to improve. An appeal mechanism would be a powerful institution within the SEP system that gradually would lead to greater clarity about what licenses exactly that a standard implementer needs to acquire and pay for. It would help the market to move towards the patents that technically really matter and make the SEP license a factor only of the relevant SEPs. This should be a critical part of the SEP system. The advantages of a standard are undermined when prices and buyer costs do not reflect the essentiality of a technology to a standard.
A second step could be to incentivize patent holders to be more selective in their declaration of patents by demanding a fee for every declaration. Increasing the cost of SEP declarations, as suggested by Regibeau et al. (2017), might mitigate the problem of over-declaration. However, such a fee would be no panacea. While it could encourage moderation in the declaration process, it would also have to be a high fee in order to be effective. And a high fee is difficult to square with the nature of market and technological discovery – which is evolutionary – and the fact that expected royalty fees are often unknown at the time of the declaration. Therefore, a high fee could also be a deterrent against the declaration of patents that are essential or might become essential in the future, which damages the relevance of the standard. The market reality is that companies patent before and during the standard-setting process, and it is difficult for all actors in the early part of the process to know exactly what patents that really are essential to a standard. For the SEP system to work, it needs to allow the necessary space for such discovery.
An alternative approach is therefore that, while there is no registration fee for the initial declaration, there is a fee introduced on each declared SEP at a certain point after the standardization process when holders have had time to review the essentiality of all their declared SEPs. As a precondition for trust in the system, it is in the interest of competing SEP holders that buyers pay for essential patents only and that resources are allocated in accordance with the standard and market demand, and the purpose of the fee would therefore be to incentivize SEP holders to move in that direction by gradually taking away patents that have proven to be non-essential.
A third option is for SSOs to demand that a patent holder should specify, generally or to a prospective customer, why an SEP is essential to a standard. After all, it should be a standard market practice in any well-regulated market economy that the seller provides the buyer with sufficient information about what they are selling. And it should be normal for SEP markets that patent owners provide sufficient information about why potential implementers need their patents for the implementation of a particular standard. Given the substantial amount of asymmetric information between the seller and buyer, it is reasonable to put significant information responsibilities on the seller. Undoubtedly, this is a responsibility that should become greater when a seller offers a license to a portfolio of patents rather than a single patent as essentiality is even more burdensome to examine for patent portfolios under the current system. In other words, SSOs can complement their FRAND principles by demanding sellers to specify to a consumer how their SEPs read on the standard.
A fourth area for improvement concerns no-challenge clauses in license contracts for FRAND-encumbered SEPs. While a prohibition of such clauses would not directly target over-declaration, it would incentivize SEP holders to be more specific in their essentiality claims when they are negotiating with their customers. The invalidation of no-challenge clauses would empower SEP implementers to challenge the validity of declared SEPs before a court, and thereby discipline SEP holders in their declaration practice in the first place. The prohibition of no-challenge clauses might also trigger enhanced cooperation between stakeholders of SSOs, individual corporations and standard consortia to significantly improve the accuracy of essentiality declarations.
These alternative approaches are not mutually exclusive: they can be combined. It is important to recognize that it is difficult, if not impossible, to get improvements in the SEP market by “command-and-control” type of regulations by SSOs that would mandate one specific behavior in all potential situations. Arguably, it is also unlikely that all SEP markets would improve if all SSOs should begin to register all declared SEPs in databases. What SSOs can do is to establish institutions and set the incentives for stakeholders to increase transparency and reduce obscurity about the supply of (and demand for) licenses – that is, provide greater clarity about the product and the market: which patents are truly essential for a standard and, consequently, what a standard implementer should be paying for.
Royalty Rates and the Discovery Process of Markets and Innovation
The concept of FRAND is vague on what is the core aspect of FRAND-encumbered SEPs: setting the royalty rate for a patent that is essential to a standard. There is no particular proscriptive model for how that process should be determined, let alone the price level. It is obvious, however, that the SEP system has to be based on a price commitment by the seller. The combination of a patent and a standard that grants the holder such market power that, without a price commitment, the market would be distorted and the SEP system would be discarded at some point. However, what that actual price should be is a far more complicated issue, and one that can lead to inefficient forms of resource allocation, e.g. money spent or not spent on particular forms of research and development on the side of both SEP holders and implementers. In short, technology creators can be incentivized to develop technologies that are of little relevance to the market if the price mechanism rewards that type of behavior. Getting the prices right is therefore not just an issue of determining what is fair and reasonable; it also informs companies in their decisions about what technologies to invest in and how they strategize to compete in the future.
Obviously, opinions about what FRAND rules actually mean for licensing practices and royalty rates vary among involved stakeholders. Returning to the stylized division of interest between SEP holders and users, we can see that the former group prefers as much pricing discretion as possible, while the latter generally would prefer to reduce pricing flexibility for holders. It has also been argued that neither FRAND rules nor courts give any significant guidance on the actual royalty rate, or even the basis for calculating royalty rates. One study argues that it is unclear whether an SEP holder “could charge no more than the ex-ante ‘incremental value’ of his invention over the next best technical alternative” or “has to set his royalty rate in such a way as to prevent cumulative royalties on the standardized product from exceeding a low percentage of the total sale price of that product” (Geradin 2014, pp. 919-920). At the same time, SSOs like ETSI promise SEP holders a “fair return” after committing to FRAND (ETSI 2017) and, for this reason, Putnam (2016) – along with others – asks whether “FRAND [is] only a royalty rate, a range of rates or a larger set of terms and conditions among which the rate is [only] one variable?”
While it is important to recognize the limits of FRAND rules as far as prices are concerned, and not exaggerate what court decisions in the past have established for the price-setting process, it seems clear that there have been developments through court decisions and competition authorities, which have gradually given more information about how “FRAND rates” should be determined. In a way, one can see developments in many jurisdictions if the form of convergence to specific interpretations of FRAND that limit the space for negotiations of FRAND-encumbered SEPs to deviate from a norm.
Court decisions relevant for royalty pricing and pricing negotiations have acknowledged that FRAND rules intend to resolve market problems caused by asymmetrical information and that they are based on the notion of a price commitment that ensures not just that prices are fair and reasonable, but that they are also non-discriminatory. The sheer presence of a standard gives a patent holder a price leverage and an opportunity to demand a higher license rate for the patented technology than if it had not been adopted into the standard. Courts have therefore argued that it should be the price charged for a patent before the technology was adopted as a standard, and that this “benchmark” should be the relevant metric for determining the “FRAND price.”
A recent study done for the European Commission argues along the same lines: “[i]t follows immediately that the right conceptual benchmark for the determination of FRAND rates are the rates that independent patent-holders would have been able to charge ex-ante, i.e. before the standard has been chosen and, hence before any patent-holder or user knows whether a given patent will actually end up reading on the standard” (Regibeau et al. 2016, p. 27).
However, while it is important to get the basic theoretical concepts right, and acknowledge that there has to be a price commitment for a system of standardization to work (especially if the technologies in question are under patents), there are many practical issues to consider. An ex-ante price, for instance, may not be possible to determine – and, from a market perspective, it may not be a good price. A price that clearly does not reflect the market valuation of a technology will give companies incentives to allocate resources in ways that do not support technological development and growth. If the current trends of determining the ex-ante FRAND rate locks the system into an artificial type of pricing – setting the price before there has been any real market valuation of it – there are clear risks of the price itself distorting the market. However, what complicates the process of determining prices is that the standard itself will represent such a large part of the value of the license that it becomes difficult to get the market to determine prices in an evolutionary way. Are there alternative ways to approach the issue about what should constitute a price commitment?
Recently, there have been approaches suggested that could work as alternative or complementary models to the current approach of a market-driven ex-ante price determination. For instance, the above-mentioned study by the Joint Research Centre has discussed the introduction of a market-corrective practice that would ensure SEP holders that they get a predicted return on investment (RoI) if the revenues from normal market sales do not cover the cost of developing the technology. The logic is simple: if the creator of the technology does not get adequately rewarded, it follows that the company will not be incentivized to put resources into research and development that could improve technology and lead to more downstream innovation.
This return on investment concept, however, is both principally awkward and impracticable. There is no guaranteed market reward for any investment in a market economy – and if such a guarantee existed, it would distort the market and lead the innovation process in a direction that is unlikely to reflect market demand and consumer preferences. Furthermore, it is difficult, to say the least, to find a way to objectively assign development costs within a firm to one particular patent – and that challenge gets much bigger if the sum also should be considered legitimate by the buyers. There would be incentives for patent holders to exaggerate investment (including research and development) costs, which are difficult to objectively evaluate from the outside, similar to intra-firm transfer pricing arrangements, which constantly give rise for legal disputes. Logically, for such a system to fairly represent the success of a technology creator, any reward in excess of the established RoI metric would have to be redistributed from the patent holder to the users, further reducing the competitive instinct of SEP suppliers. In addition, a return-on-investment concept for determining the “FRAND rate” would likely give companies incentives to declare (artificially inflated) low-valued patents to SSOs while avoid doing it with high-valued patents. The critical task now is to make the SEP market more transparent and lead market participants (sellers and buyers) to contract over technologies that are essential to a standard, and the RoI concept stands in opposition to that quest.
Another idea that has been drafted suggests that the royalty could be determined by the end-use of an SEP, and that the value of that end-product should influence the price that a buyer pays for an SEP. Clearly, this has been part of SEP contracts for quite some time and holders, like any other market participant, have set their prices in accordance with how valuable they think an upstream technology is for a downstream product. Often it has worked smoothly, and as long as a holder is not trying to increase the price when buyers have been locked into using a standard, it should be up to buyers and sellers to determine what the price should be. In a way, inspiration for such an approach can also be drawn from decisions where courts have attempted to determine the value of a patent.
However, there is a context to how this idea is now being promoted which makes it awkward and impractical. While courts have sought to determine a fair value for a patent when other market circumstances have not made that possible, that process has not been an exercise prefaced with the intention of ensuring higher rates for the patent holder. And that is where it differs. The case now for a use-based approach is predominantly about effecting a new form of corporate redistribution where downstream innovators implicitly would be taxed for the purpose of increasing the revenues of the upstream innovator. Importantly, it is partly about trying to leverage the standard itself for the determination of the price and to get the market distribution of income to work more to the benefit of technology contributors.
Obviously, this approach intends to correct the changing fortunes of businesses, where margins often are better in downstream markets and increasingly squeezed in upstream markets. It is thus not a surprising development: given market developments, that no one wants to end up in a market position where other companies in the value chain are better at capturing potential margins. Creating a way of redistributing revenues backwards through the value chain is an appealing idea if the downstream users are considered to benefit unfairly from the upstream contributor.
However, it is difficult to see how a use-based approach could work in practice when the task remains to set the norms for determining the FRAND price – the ex-ante price that discounts from the market value that comes from being part of a standard. A use-based approach could, in the first place, be discriminatory as users could be charged a differentiated royalty for the same technology. Given the general market development, this could have effects on the competitive relationship between downstream users of the SEP. A key point of FRAND still is to minimize such effects and ensure that downstream competition is not determined by the contract relation that different parties have with an upstream contributor of technology. Hence, an end-use approach would run the risk of undermining the current system of standardization because users would not be able to trust that it ensures stable conditions for competitive neutrality. This judgment holds even if the discrimination takes place at the level of product classes and not just between companies that are competing head-to-head with similar products.
Finally, an end-use approach to establishing a policy-based FRAND rate (policy set by SSO, courts, antitrust authorities or market regulators) would face the same practical difficulties as determining a FRAND rate based on the return on investment for the patent holder. It is hardly easier to design a credible system for determining the value of an input for the final product than it is crafting a trusted estimate over the costs incurred for developing a patent. It has been attempted by courts in patent litigations but, even if it is a practice that is necessary when all other alternatives have failed, it is not exactly a straight-forward process. Nor are the outcomes necessarily more advantageous for companies who think they are poorly rewarded for their upstream technologies because, as expressed by opinions by the U.S. Federal Circuit, the valuation of royalties should be based on the “smallest salable patent-practicing unit” with the purpose “to calculate damages based on the product incorporating the patent, not a downstream product with a value much less related to the patent’s value” (Muris 2017, p. 6). In a similar vein, the European Commission’s “Horizontal Guidelines” express the view that the preferred method of patent valuation is the “licensing fees charged by the company in question for the relevant patents in a competitive environment before the industry has been locked into the standard (ex ante)…” (European Commission, 2011, para. 288-89).
There is a market logic behind the skepticism to a use-based approach. The value of a connectivity standard in an automobile, for instance, differ between brands and models, and over time as products, consumer preferences and market competition change. Markets can be fickle and the whole point of market-driven innovation is that the hedonic outcome for customers, not the inputs and the costs incurred, determines the value of a good or service. Consequently, a use-based approach would risk resulting in greater costs for those companies that are good at understanding what customers are willing to pay for.
What the consequences would be in practice is actually harder to predict. In sectors where SEPs are important – predominantly ICT – technological development forces rapid changes in products and markets, and what is relevant for the valuation of an SEP at one point could be irrelevant soon after. Furthermore, when standards are developed and/or when the prices are first determined, the actual shape and structure of potential downstream markets are often unknown.
There is much talk now about the future shape of the IoT market, and expectations about growing connectivity between various products have fueled controversy about the entire SEP system. Yet this discussion shows that the policy aspects of the SEP system are sometimes distant from market realities. For instance, today there are greater connectivity opportunities for a refrigerator, and those are likely to increase in the future. But what are consumers prepared to pay for the opportunity to get an automatic update when they are about to run out of milk and butter? Can automated retail delivery be connected to the technology, and assist households in making sure there is food in the refrigerator when they need it? If so, how much would consumers be willing to pay for that service? All these aspects of the market are unknown. And, therefore, there is no objective way of singling out the value of one particular input technology that would not be prone to a very big scope of interpretation, negotiation and manipulation. Consequently, the use-based approach does not reduce the artificiality associated with determining an ex-ante price for the FRAND royalty.
Importantly, neither the RoI nor the end-use approach takes us any closer to a market which is more transparent and predictable, and less like the “confusopoly” that too often is allowed to set the tone of the SEPs market. Both approaches would create more problems than they would solve. They would introduce greater uncertainties about the terms of pricing and more frictions between contracting parties that cannot find alternative arrangements that are reasonably reliable. Furthermore, they do not make the system more reflective of market conditions and how market participants’ valuations evolve over time.
An alternative approach is to change the price setting approach and allow for much greater price variation – or price ranges – in contracts of FRAND-encumbered SEPs. One argument behind this approach is that the devil of FRAND is always in the details – and that there is no way to settle on one “objective” FRAND price that could steer markets infinitely. It has been argued that FRAND royalty rates or licensing fees that are “deemed” to conform to FRAND always constitute a system of artificial pricing, in which prices do not properly reflect supply and demand characteristics, which eventually causes a misallocation of resources, including distorted investments in technological R&D and business model adaptation. Furthermore, as illustrated by the high number of FRAND litigations, it is often left to courts to determine appropriate FRAND rates and negotiation practices, although courts are not in the best position to determine an “objective” market price.
This view is thought-provoking – for reasons that we will return to soon. It is difficult, however, to get a laissez-faire or price-range approach to work in a system whose credibility ultimately has to be based on a price commitment by SEP holders. The process of standardization is very likely to lose trust if an upstream input supplier is given big pricing power and when that is already known at the time of the actual setting of the standard. Rather, such approaches take us back to the idea that FRAND rules are actually supposed to mitigate the granting of additional market power to companies that already enjoy exclusivity rights through patents. The market distortions that are likely to follow such market power would provide incentives for users to develop own solutions rather than accepting a standard, and consequently there would be increased technological fragmentation and greater problems of interoperability between products and platforms.
What makes the “pricing flexibility” view interesting, however, is that it accounts for what indeed constitutes a (perhaps the) major problem of the current system: correcting a royalty determined artificially before or after the standard had been developed and decided. The FRAND concept of fair and reasonable prices may by all means be necessary, but it does not ensure that a royalty is set at a level that reflects market conditions and what, beyond the premium that is caused by a patent being essential to a standard, is a relevant market valuation. While there is a degree of flexibility in negotiations, the point with FRAND rules is that there should not be any significant ex-post changes in royalties. Ex-post prices that deviate from the ex-ante valuation should in principle not be part of the market. In reality, however, many holders and implementers are accepting that practice, but developments through courts and authorities – moving in the direction of the ex-ante valuation – may be about to reduce that flexibility.
Locking the market to the ex-ante price is awkward. Courts and other institutions respond to litigation claims and give guidance on what general principles like fair and reasonable mean in practice, but their task is not to determine if other developments than the standard have given reason for the royalty to change. And such changes can go in both directions, as there can be situations when a patent holder is motivated to either cut or increase a royalty. Prices would then more adequately signal how market participants should allocate resources, notably R&D investment, for economic payoffs to improve. Hence, the merit of this approach is to avoid that markets, and not just individual companies, get locked into using a sub-optimal technology because of failing incentives to encourage new technology and more competition.
The pricing flexibility view is also interesting because, despite FRAND rules and court decisions, the current regime all too often leaves participants in the dark about the FRAND royalties charged by SEP holders to other firms, including firms operating in different positions along the same value chain or in other sectors. Hence, adequately improving price transparency would have to take precedence over attempts to amend or detail the framework for how FRAND-encumbered SEPs should be negotiated.
That latter part follows directly from recent court cases, most notably the decision by the Court of Justice of the European Union in Huawei v. ZTE where judges established the basic pillars of the negotiation framework (see for instance Jacob and Milner 2016). While it has become obvious through court disclosures over the past years that there is indeed a degree of discrimination as far as royalties are concerned, few proposals that have been suggested so far to address concerns about discriminatory practices.
This is an area where further guidance is needed because protection against discrimination is critical to ensure the competitive neutrality of standards and SEPs. One approach would be to have SSOs giving clarification for how parties should negotiate in good faith and what that entails in practice. Another approach, currently under consideration in Europe, is to have guidelines issued by regulators on the same matter. A third approach is to establish the equivalent of a license registry or repository where parties to a contract of a FRAND-encumbered SEP would disclose the material conditions of a contract.
Such a change would increase transparency and trust. Importantly, it would also increase the ability of both parties to negotiate legitimate ex-post changes of royalties. Many SEP implementers accept the notion of an ex-post change to the price as long as changes are reasonable and do not discriminate between various buyers. The problem now, however, is that the obscurity of the SEP market stands in the way for ex-post royalty corrections that reflect other changes than the essentiality to a standard and its lock-in effect on the industry, i.e. supply and demand dynamics. Furthermore, improved transparency in other aspects, and greater demands on all parties to ensure that the SEP contracts concern technologies that are essential to a standard, would additionally support a greater role for the evolutionary market discovery as far as prices are concerned. If it is regularly the case that SEP implementers lack the information to judge whether all the SEPs they are buying licenses for are essential or not, they are hardly in a position where they can engage in an informed dialogue about allowing greater variation of prices for reasons unrelated to the essentiality of the contracted technology to a standard. Naturally, if the market is considered distorted by multiple layers of non-transparency, the first order of priority is to defend against any misuse of the market power that a standard gives to an SEP holder.
Market obscurity reinforces potential problems associated with how prices are determined for FRAND-encumbered patents. While there is no universal formula for how the initial, ex-ante price is determined, the reality is that prices and standard-setting often evolve simultaneously. During a standard-setting process, many technology creators have incentives to patent as much as they can in the territory of the future standard. It is impossible for everyone at this point to anticipate the market utility and valuation of each patent, and the only thing that seems safe to predict is that a patent that will be essential to the standard will increase revenues. But there is often no real market where a valuation can be discovered. Consequently, there are risks that the standard-setting process fails in “getting the prices right” and that markets will not develop as well as they could – and should – if prices were better at reflecting market valuations. Undoubtedly, this is particularly a concern for those companies that signaled an ex-ante price commitment that is lower than what subsequently would be the real market valuation irrespective of the patent being included in the standard or not.
One conclusion is that more time in the standard-setting process should be given to establishing the conditions for how to set the price – and, thus, to avoid as much as possible the problem of getting the prices wrong. Discussing the price before a standard is defined is an awkward process in a market-based economy, but the standard itself and the FRAND rules that should govern essential patents change the price mechanism to such an extent that corrections are difficult to achieve at a later point.
Many companies participating in the standard-setting process already consider the market aspect of the future standard and have quite an advanced idea about which companies and patents that will be central for the standard to be implemented. It would therefore be natural for the parties involved to engage in a dialogue ex-ante about what the establishment of the standard should mean for the royalty and what factors that could constitute legitimate reasons for ex-post price corrections. In other words, when price and market valuation are such a central part of the market for FRAND-encumbered SEPs to function efficiently, the standard-setting process could improve the evolution of future SEP markets by combining the technical decisions about the standard with transparent discussions about what should constitute the initial price and on what grounds these prices could be amended in the future.
Arguably, an initiative by SSOs to facilitate a discussion between companies about prices follows logically from the opinion by the Court of European Justice in the Huawei v. ZTE case. In essence, the court’s attempt to define the key steps of a market negotiation process is about building a normal market and helping the SEP system better conform to normal market practices, like being clear about what product is on offer and why it is essential to a standard. Under standardization, the lion share of the value of a product will be its essentiality to a standard: technologies that are not essential will be valued completely differently. It follows, therefore, that after a decade of SEP disputes and controversy that have made the SEP system unpredictable for many companies, it is time for SSOs to better facilitate the market for SEPs and not just their technical properties. SSOs are effectively creating a market place for technology – and the task now should be to ensure that good markets norms govern the behavior in this market.