RIETI Policy Seminar

Standards and Intellectual Property: Strategies Japan should adopt in light of current global trends (Summary)


  • Time and Date: 13:30-17:30, Friday, December 8, 2017

As social and economic importance of standards is growing due to advances in information and communications technology and the spread of the Internet of Things (IoT), coordination of standards and intellectual property (patents) has become critical. To date, standards developing organizations (SDOs) have worked to coordinate standards and patents by formulating intellectual property rights (IPR) policies and requesting owners of patents which are essential to standards to declare licensing patents on fair, reasonable, and non-discriminatory (FRAND) terms. At the seminar, experts made presentations, considering what is necessary to realize a virtuous cycle in which intellectual property creates innovation and standards diffuse the fruits of such innovation. In Part I, researchers in the fields of economics and law made presentations from the standpoints of empirical research and theories, comparing different approaches to solving disputes. They also discussed the current situation and analyses of licensing negotiations and calculation methods of FRAND royalties. In Part II, experts on Japanese and overseas companies that are proactively engaged in standardization activities as well as policy experts discussed the direction for strategies that governments and companies should take.


Welcoming Remarks

YANO Makoto (President and CRO, RIETI / Professor, Institute of Economic Research, Kyoto University)

Standards and specifications like the metric system have long been created by national governments. Today, however, they are beginning to be formed in a decentralized manner through private organizations.

With advancements in IT, there are many cases where multiple companies or institutions form a pool for cross-licensing relevant patents, within which they establish standards and specifications. Through them, participants can jointly use new technologies while protecting the IPR of these technologies.

Deciding how to contribute patents to a pool represents an important issue for the development strategy and patent utilization strategy of these companies. A pool of patents on the standards and specifications for a particular information technology has important implications in competition policy.

At today's seminar, Professor Masabumi Suzuki of Nagoya University will make a presentation on the research outcomes of RIETI's standardization and IPR project. Also, three prominent guest speakers from the United States—Dr. Anne Layne-Farrar of Charles River Associates, Professor Jorge L. Contreras of the Utah S.J. Quinney College of Law, and Dr. Dina Kallay of Ericsson—will present on recent developments worldwide. Furthermore, a discussion will be held on the future strategy for government policy and business, led by Mr. Tetsuya Mimura of NTT DOCOMO, INC. and Ms. Noriko Fukuoka of Panasonic Intellectual Property Management Co., Ltd., representing corporate Japan, together with Professor Juichi Nagano, former director of International Standards Strategy of the Ministry of Economy, Trade and Industry.

I hope that today's seminar will serve as an opportunity for everyone to deepen your understanding and develop a shared understanding with regard to standardization and IPR.

Presentation 1: Standards and IP–From the viewpoint of economics–

Anne LAYNE-FARRAR (Vice President, Charles River Associates)

Arms' length negotiations vs. litigation

Arm's length negotiations tend to center the royalty discussion on the value that the patents are bringing to the products. They are about the value contributed to the implementer of the standard and seldom touch on whether the patent is essential or not. This value negotiation tends to focus on a handful of representative patents in order to make it a manageable negotiation. Typical terms and conditions are that the license will cover the entire portfolio rights, including standard-essential patents (SEPs) and non-SEPs. You tend to see a single rate for use of the patented portfolio around the world as well as a table with volume discounts.

Litigation on the other hand is quite different. Even in circumstances when the litigation framework is relying on a hypothetical framework, the court does not replicate the circumstances seen in the world. Focus is placed on monetary terms, such as lump sum payments and royalties, and non-pecuniary terms tend to be ignored. This sets up tension between arm's length negotiations and litigation because of the legal limitations imposed on courts. Patents are jurisdiction specific, and efficient and fair court treatment of FRAND requires balance of jurisdictional issues and accounting for the global nature of the portfolio. In addition, SEP portfolios tend to be large. The courts need to think beyond their normal limitations of looking in-depth for a small number of patents, and think more broadly about the special circumstances that arise in FRAND litigation.

Apportionment is the same between arm's length negotiations and litigation. The FRAND rate should cover the value that the specific set of patents conveys to the licensee and its products that are implementing the standard but not more. That said, standards development efforts are like joint ventures, where all of the participants and members in the SDO are contributing to create value to enhance a product or service from which they will all benefit. It is incorrect to understand that SEP holders cannot achieve any value from the standardization because they are helping to create the value, as recognized in the Unwired Planet v. Huawei (U.K.) court case.

Patent counting

The complexities of valuing SEP portfolios, particularly in relation to the other SEPs, as well as the time and money limitations means counting of some sort is inevitable. Patent counting is typically used for the apportionment process. It is not a good use of patent counting to estimate the royalty stack.

Balance between SEP holders and implementers is crucial for maintaining balance, incentives to invest in new technologies, and incentives to continue participating in joint cooperative standard development efforts. This means FRAND requirements impose obligations on both parties. SEP holders need to explain which patents they believe are essential and make offers that are consistent with FRAND before they take any other legal action. Implementers are obligated to respond in good faith to those offers in writing without delay tactics.

FRAND determination methods

The first and my favorite method is use of comparable licenses. It is the most efficient and consistent with FRAND because a comparable license will reflect the market value that the two parties place on the patented technology. However, we cannot simply pull the royalty rates out of the contracts in isolation because non-pecuniary terms can have significant impact on the money that is paid. If you have lump-sum payments as opposed to running royalties, it is tricky to translate lump-sum payments into a monthly running royalty due to biases.

Patent pool licenses tend to fall on the lower end of the FRAND spectrum because lower royalty rates are accepted to encourage widespread adoption. Too few SEP contributors might signal that the pool is being used to suppress FRAND rates. Too few licensees might signal that the rate is above the FRAND level.

The incremental value method is difficult to put into practice. When you have sufficient information, it can at least be a thought exercise.

The "bottom-up" method is the idea of looking at what technologies existed when the standard was adopted. You have to pay close attention to why the alternatives were not chosen. Many times alternatives cannot be rank ordered and instead involve tradeoffs.

The "top-down" method starts with an aggregate value of what it should be, and then it divides that amount among the relevant SEP holders. If you don't have all of the details, then the uncertainty is just moved to an upper level. Using prices or price margins as a measure of the aggregate can also lead to undercounting.

All of these methods try to achieve not only fair remuneration between the two parties in the litigation, but also a message that is going to deter future litigation. If you set the rules in a balanced fashion, you can increase incentives for the licenser and the licensee to enter into good faith bargaining.

We are starting to see some consensus that: 1) global portfolios are most consistent with FRAND as opposed to jurisdiction specific; 2) both licenser and licensee have good faith obligations; and 3) courts and regulators need to keep incentives to maintain balance between the two sides, reduce incentives for litigation, and increase incentives for good faith bargaining and FRAND outcomes.

Presentation 2: Dynamic Effects of Patent Pools: Evidence from inter-generational competition in the optical disc industry

SHIMBO Tomoyuki (Associate Professor, Kanto Gakuin University)

There have been both an increasing need for combining technologies and an increasing fragmentation of patent ownership. They have enhanced the necessity for developing an efficient institutional mechanism for aggregating technologies. A patent pool is one candidate. In the case of DVD patent pools, two pools (3C, 6C) cover almost all bundles of essential patents globally. This means two-stop shopping for global operations. Licensors commit to RAND licensing based on a published price list, and royalty is distributed based on the number of patents by licensors in 6C. Theoretical research shows a well-designed pool makes positive economic contributions to the diffusion of a standard. On the other hand, empirical papers suggest some evidence of negative effects of pools on further innovation by licensors.

Intergenerational competition

A modern pool guided by competition authorities focuses on the integration of complementary patents for a specific standard. Standards often evolve over time, and intergenerational competition is important. Patent pools may affect research and development (R&D) for not only the current generation but also for the next generation. Joshi and Nerkar (2011) is an important study that found the negative effects of pools on R&D by licensors and licensees. However, it has two problems. First, they ignored intergenerational competition. Second, they underestimated the effects of pools because they ignored the effect of the standard agreement before pool formations, and the firms with essential patents would have anticipated the pool formation at the time of the standard agreement. Therefore, in our research, we identified the population of patents in optical disc technologies more correctly. We divided the patents into CD, DVD, and BD/HDDVD.

The standardization of DVD took place in 1995, and the pool formation of DVD took place in 1998. The number of families of BD/HDDVD starts to increase after the pool formation. In our research, we examined how the agreement and pool formations on the current generation standard (CGS) affect R&D for the next generation standard (NGS). We focused on three types of firms—licensors, licensees, and non-participants—and two events, namely, after the standard agreement and after pool formation.

R&D opportunity for next generation standard (NGS)

When innovation is cumulative and includes multiple generations, R&D for the NGS builds on the technologies for the CGS. For example, there exist 75 families of DVD SEPs in 239 essential families on BD SEPs. If firms developing technologies on the NGS anticipate that the SEPs on the CGS are licensed under RAND conditions ex-post, they can expect that the hold-up problem can be avoided. The standard supported by RAND licensing commitments would have positive effects on R&D for the NGS.

With regard to the experience of licensors when innovation is cumulative and includes multiple generations, the licensors have stronger R&D capability for the NGS. Both events on the CGS would have positive effects on R&D for the NGS by licensors only. Because the licensors made large sunk R&D investment in the CGS, they would have chosen a project for the NGS that would exploit the existing complementary assets. Second, the expected profit from the CGS would make the licensors reluctant to invest in the NGS. Because of these effects, both events on the CGS would have negative effects on R&D for the NGS but only for the licensors. We can propose that both the agreement and pool formations on the CGS will increase R&D for the NGS by licensors relative to nonparticipants if the effect of experience in the CGS dominates the sunk cost and replacement effects (Hypothesis 1).

This study focuses on the standard wars as well. For next-CD, Toshiba and Panasonic developed the SD format, whereas Sony and Philips developed the MMCD format. The DVD standard was announced based on the SD format in 1995. However, two DVD pools were established based on each format. For next-DVD, Toshiba developed the HDDVD format, while Sony, Philips, and Panasonic developed the BD format. Toshiba retreated from HDDVD in 2008, and two patent pools were established such as DVD. Thus, most of the 6C licensors including Toshiba and Panasonic were winners in the competition for the DVD standard. They would have made more sunk R&D investment in the DVD standard than the 3C licensors including Sony and Philips.

Considering these results, we can propose that these R&D promoting effects would be smaller for the 6C licensors with larger sunk R&D investment on the CGS than the 3C licensors (Hypothesis 2).

We have two main findings. First, both the standard agreement and the pool formations on the CGS encouraged the licensors to invest in R&D for the NGS. However, the R&D promoting effects were smaller for the 6C licensors with larger sunk investment on the CGS than the 3C licensors. We can draw the following policy implications from our study. The competition policy and RAND commitment would have contributed to the positive dynamic effects of the DVD pools on R&D. The scope of the pools was narrowly specified, and a clear commitment to RAND licensing for the CGS exists, which is essential for competitive R&D for the NGS.

Presentation 3: FRAND as Private Law and Public Law

Jorge L. CONTRERAS (Professor, University of Utah S.J. Quinney College of Law)

I would like to talk about FRAND royalties from the perspective of public law and private law. Generally in Western economies, standard setting is conducted by private sector organizations as opposed to public organizations. Private ordering allows for collaboration among competitors, focuses on technical, and places priority on optimizing solutions.

Private organizations create policies. These policies contain many types of provisions around due process, openness, balance of interests, and appeals process, and they also relate to patents: the disclosure of patents and the licensing of patents by participants.

In the beginning of the 20th century when there was not much standardization, every city acquired fire hydrants from a different company. When a large fire broke out in Baltimore, hoses from various fire departments could not be attached to the fire hydrants and much of the city burned down. This example shows that interoperability standards have a public character, and create market efficiencies and other benefits that cannot be achieved otherwise.

How does this relate to FRAND? Many standards organizations have an obligation that patent holders license patents to others on reasonable and non-discriminatory terms. This may be considered a patent exchange of value, but there is also a public character in that we want all market participants to be treated equally. This is to ensure that we can create a market that is open to new entrants, does not have high barriers, and avoids excessive royalties. When we have standards subject to FRAND rates, there is a private-public interest in expanding the reach of the standard and also for maximizing the social welfare that arises from having standards broadly adopted.

Bottom-up FRAND rate calculations

One way to determine FRAND rates is bilateral negotiation. They happen all the time, but there is not much transparency in information. They are conducted under confidentiality rules which are reasonable for private party transactions. When we add the public character of FRAND it becomes less reasonable. The negotiation is uncoordinated—pair by pair, one by one. If the parties do not agree on the FRAND rate, they can go to arbitration. This generally is not bad for the parties. However, it has the same confidentiality issues as bilateral negotiation. The results of an arbitration are secret, and thus the next parties negotiating on the same patents and standards do not know what happened. Arbitrations are likewise uncoordinated. You can also have litigation. The results of litigation are made public but are binding only on the parties. Litigation is also expensive. These three methods are what I refer to as bottom-up methodologies for calculating rates.

The bottom-up way does not work very well. A top-down approach creates value for the totality of the patents covering a particular standard that is then divided among the patent holders. In both the Apple Japan v. Samsung case and the Unwired Planet case, the court came up with an imprecise aggregate royalty based on press releases and public statements made by the patent holders. Ways to further improve top-down approaches include involving not only the two parties to the litigation but also all of the patent holders in a single litigation. However, they are quite expensive and long. Another solution is to have both patent holders and implementers negotiate a rate before the standard is adopted. However, due to fears of antitrust liabilities and disagreement by patent holders, this approach has not taken off either. There may be a role for state and possible agency rate setting. The law could make such rates binding. The problem is which state or country gets to do the rate setting. In the Unwired Planet case, while Huawei only wanted a UK license, the judge stated that a FRAND license is a worldwide license. If a UK court can force the worldwide deal between Huawei and Unwired Planet, every other country can do this as well, resulting in a race to the courthouse or a race to the bottom. How can we avoid this from happening with FRAND?

International FRAND alternative dispute resolution (ADR) tribunal?

Suppose we have an international tribunal that determined an aggregate FRAND rate and is recognized in all countries that are signatory to the New York Convention. There are only two ways to encourage countries to do this. One is by treaty. However, treaties take a long time to negotiate, and countries do not always comply after they are negotiated. Another way is through SDO policy. SDOs could require participants to have a FRAND rate determined for a particular standard.

Global adjudication of rates would avoid the race to the courthouse. We can end up with a single adjudication per standard, which would avoid duplicate proceedings. If nothing else, we would remove the threat of royalty stacking. We would know upfront how much the royalty would be. This would also help with compliance with competition law and even help eliminate hold-up and hold-out.

FRAND rate determinations are not merely private agreements. The bilateral and bottom-up methods are not optimal. They create a number of inefficiencies and are also threatening to cause a race to the courthouse around the world as more countries become involved. A more comprehensive system for determining the rates is needed, and I would invite you to think about it.

Presentation 4: Legal Issues Concerning the Enforcement of FRAND-encumbered SEPs: A view from Japan with comparative perspective

SUZUKI Masabumi (Faculty Fellow, RIETI / Professor, Nagoya University Graduate School of Law)

In SEP, there are many legal issues in basically the following three categories: 1) patent infringement disputes; 2) disputes under competition law; and 3) contractual disputes.

Legal grounds for restriction of enforcement

In the first category, there are four legal grounds for restriction of enforcement of SEPs depending on the country: contractual ground; abuse of rights doctrine; theory of patent remedies; and anti-competitive practice. In the contractual ground approach, courts recognize that based on the FRAND declaration between the SSO and SEP holder, the SEP holder has a contractual obligation to potential implementers, which is handled differently from a pure infringement case.

The abuse of rights approach was taken by the Japanese IP High Court in the Apple v. Samsung case. The court did not recognize the FRAND declaration as a contract for a third party (contractual ground). However, the court refused to grant injunctions based on the abuse of rights doctrine. The SEP holder had already made the FRAND declaration, and Apple, the implementer, as a willing licensee, had shown intent to obtain a license on FRAND terms. Based on these two facts, injunction was deemed as an abuse of power. The court did not recognize the direct contractual relationship but recognized the good faith negotiation obligations for Samsung as an SEP holder. The issues with the abuse of rights approach are that it is ambiguous and unpredictable. The IP High Court sought to increase the legal predictability. In this particular case, however, the court did not present criteria to judge the willingness on the side of the infringer.

There are legal theories particular to the United States that restrict injunctions even if there is infringement. In the EBay case, the Supreme Court stated that injunctions may be restricted even in cases where patent infringement was affirmed. In Japan and Germany, this line of thinking is not and will probably not be adopted.

The competition law approach developed mainly in the German courts. If patent holders abuse their dominant power, their rights can be restricted based on competition law. Patent law is managed at the country level, but competition law is based on an EU treaty. Therefore, it is unified in its management and operations within the EU. In 2015, the Court of Justice of the European Union (CJEU) issued the decision on the Huawei v. ZTE case which presented a general framework for SEP holders and implementers to follow for the enforcement of SEPs and the use of injunctions. In the Apple v. Samsung case in Japan, the court only gave perfunctory reference to the competition law. The Antimonopoly Act was referenced to confirm that Apple did not have an obligation to compensate for damages exceeding the level of FRAND royalties.

For patent litigation, the other issue is determination of FRAND royalties. As to the issue of apportionment, the IP Court divides by the number of essential patents and comes up with a royalty rate per product. Ideally, it has to reflect the value of the patent. In this case, because of the assertions of the parties and evidence involved, it was inevitable that such a decision was made. There may be differing views on the determination of royalty depending on the court. In the Unwired Planet v. Huawei (2017) case, the decision went into detail about issues related to SEP and raises a significant point that worldwide royalty rates should be assumed.

In relation to competition law in Japan, the issues are unfair trade practices under the Japanese Anti-Monopoly Act as well as private monopolization and discriminatory treatment. The Japan Fair Trade Commission (JFTC) revised guidelines on the use of IP in 2016. A special feature is that they specifically mention refusal of a license or a claim for injunction by the holder of FRAND-encumbered SEPs. The guidelines show various criteria for judging whether the party is willing to take a license on FRAND terms. In Europe, the CJEU and the German courts look at specific procedures of negotiation. Japan has not gone to that extent yet.

Recent developments

U.S. Assistant Attorney General Delrahim gave an address in November in favor of patent holders. It was noted that through the actions of implementers the restrictions become excessive.

On November 29, the European Commission issued a communication on the issue of SEPs. It encourages use of patent pools. It also mentions proportionality for the granting of injunction relief and encourages more discretion. We will have to see how this concept of proportionality will affect Europe.

We will need to monitor and confirm the positioning of competition law in Japan. In Europe, direct violation of competition law is taken up in courts. However, in Japan, IP courts use the abuse of rights approach and only guidelines have been issued on the use of the Anti-Monopoly Act for injunction and denial of licensing. Currently, the Japan Patent Office (JPO) is deliberating on guidelines for licensing negotiations involving SEPs. It is hoped that these guidelines will explain the perspective from which they are presented.

In Japan, the issues of the application of the abuse of rights doctrine and treatment under the Anti-Monopoly Act still remain. Second, we need to establish criteria to judge willingness for licensing on FRAND terms. Third, we need to confirm the legal relationship between SEP transfer and the effects of FRAND declarations based on Japanese law. Furthermore, we need to look at policy issues related to SEPs in view of the development of IoT and the relationship between standardization and open source projects.

Panel Discussion


NAGANO Juichi (Professor, Nagoya University / Former Director of International Standards Strategy, Ministry of Economy, Trade and Industry)

The International Telecommunication Union (ITU) Patent Roundtable in 2012 activated discussions on two issues: the availability of injunctive relief in a RAND context; and the meaning of "reasonable" in RAND, not only in ITU but also in the European Telecommunications Standards Institute (ETSI) and the Institute of Electrical and Electronics Engineers (IEEE). I have long participated in such ITU discussions.

Four defenses appear in the context of injunction, reasonable, non-discriminatory, portfolio licensing, and transparency (quality of declarations): these are, namely, (1) essentiality, (2) infringement, (3) validity, and (4) enforceability.

Competition authorities in emerging economies which did not participate in ITU or ETSI discussions have started to define SEPs differently from SDOs, i.e., as a policy tool to perform their national competition policies. This definition, targeted for RAND obligation, even expanded to proprietary technologies much to the worries of some private companies.

There is now a new movement of open source software (OSS). As hardware is increasingly replaced by software, standards bodies are more and more trying to incorporate OSS or OSS-like processes into their standards development procedures. Patents thereby treated as RAND or royalty-free may become a bigger issue here.

A global, birds' eye-view chart describing how the conflict issue occurs between the SEP holders and Implementers, followed by four "trouble shooters" being (1) the Court, (2) Competition Authorities, (3) SDOs, and (4) Patent Offices, against the geographical areas of the United States, Europe, Japan, and emerging economies, presented an overall matrix. This chart included recent trends of FTC/DOJ and JPO moves, OSS, 5G, and transparency as well as EC Communication in the relevant cells of the matrix.

Dina KALLAY (Head of Antitrust (IPR, Americas & Asia-Pacific), Ericsson)

My presentation focuses on Type I antitrust errors ("over-intervention").

Between 2012-2015, the U.S. Department of Justice (DOJ) Antitrust Division engaged in sustained competition advocacy, stating in speeches that hold-up is a systemic problem, when there was no empirical evidence grounding this claim. The DOJ did not purport to offer empirical grounding for these repeated statements, instead, that having the theory of hold-up is sufficient or that "hold-up is like Ebola." Furthermore, the DOJ advocacy during that time strongly encouraged standard development organizations (SDOs) to change their patent policy.

As a result of the DOJ's advocacy, the IEEE decided to change its patent policy, effectively preventing patent holders from using injunction, regardless of how egregious infringers' behavior was. The new policy, that was developed through a closed and highly debated process, included many other elements that devalue standard essential patents. This policy was praised by a DOJ letter issued in February 2015 (despite recognizing procedural flaws in its developments).

Since the new patent policy took effect, almost three-quarters of patent declarations submitted for the WiFi standard have been negative ones. In other words, at IEEE, open standardization secured by FRAND access has broken down as a result of the DOJ's interventionist advocacy.

Now we are seeing the DOJ returning to normal and regretting the three years of one-sided intervention that led to anticompetitive effects. The November 2017 AAG Delrahim speech noted that hold-out is a more serious antitrust risk than hold-up, and recognized that there was a Type I error (over-enforcement or over-intervention by antitrust agencies) by the DOJ during this timeframe. The speech warns of the problem of buyers' cartels, i.e., licensees' collusive behavior in SDOs seeking to lower royalty rates (hinting that the IEEE exercise presented such a buyers' cartel).

The FTC adopts a similar position, recognizing that there has been misguided use of antitrust in the area of SEPs, and calling for empirical grounding for any advocacy or intervention in this area. I ask policymakers to base interventions on empirical findings rather than simply on personal views because the IEEE case serves as a warning natural experiment showing the anticompetitive effects of basing an intervention simply on views.

MIMURA Tetsuya (Senior Manager of Strategic Licensing Group, Intellectual Property Department, NTT DOCOMO, INC.)

I would like to note that most of the views expressed in my presentation are my personal views.

There are gaps between Implementers and SEP holders. For Implementers, since the number of SEPs and SEP holders is large, it is difficult to clearly identify the reasonable rate of royalty. They do not want to pay if they do not have to and competitors are not paying. On the other hand, SEP holders consider that they are charging a valid license fee to implementers. To narrow the gaps, SDOs, industry associations, government institutions, etc. have been holding discussions. The main issues of the discussions include restrictions of injunctive relief, definition of FRAND, and transparency. How to resolve the challenges for implementers seems to be the focus, especially on injunction, and the environment may be slightly in favor of the implementers' standpoint.

As we head towards 5G, patent pools may be a solution, but patent pools for 3G and 4G are not necessarily functioning well.

OSS will probably be used proactively in the 5G network. SEP licensing conditions, whether it is FRAND or royalty free, are an issue and need to be resolved.

FUKUOKA Noriko (Senior Manager of Licensing Department, Panasonic Intellectual Property Management Co., Ltd.)

In the smartphone war among Microsoft, Apple, and Samsung/Google, the FRAND mechanism was attacked and problems were pointed out in 2010-2011. The ITU Patent Roundtable then took up this discussion in 2012. After that, ITU IPR Adhoc took the role of the place for this discussion. Some members, most of whom are located in Europe, proposed the safe harbor approach, while the others, most of whom are located in United States, asserted prohibition of injunction. Manufacturers, semiconductors, handset and base station operators, chip manufactures, carriers, and others belonged to both camps. In the ITU IPR Adhoc, Microsoft, Apple, and Samsung belonged to the same camp. This may indicate a shift from market competition to war on business models.

When you do FRAND licensing, there are factors you have to bear in mind. You should focus on promotion of innovations and industries, balance the benefits of rights holders and implementers, protect good-faith negotiations, establish a level playing field, and secure freedom of negotiation. It must be worldwide and actions should be speedy. We need a long-term perspective and sustainable rules.

NAGAOKA Sadao (Faculty Fellow, RIETI / Professor, Tokyo Keizai University)

I would like to start our panel discussions with the comparison of a top-down vs. bottom-up approach in determining reasonable royalty for SEPs. Given that there exists an individual incentive for excessive pricing as well as difficulty of assessing individual incremental value, which are bound to cause a disagreement between the licensor and the licensee, I tend to concur with Professor Contreras that a top-down approach may be appropriate for achieving reasonable royalties as opposed to a bottom-up approach.

Whichever approach is taken, it should be supported as much as possible by the evidence in the litigation. Both approaches have serious problems in finding real empirical support for valuing the patents. I tend to support comparable license agreements because they are market data. It already takes into account royalty stacking, etc. However, comparable licenses don't always exist. When you have good information on the alternatives, the bottom-up approach could be effective. The choice between the two should depend on the particular litigation at hand and the data that you have.

In the case of DVD or Blu-ray, they can set the price for the standard and divide the pieces. Mr. Mimura, is it easier to determine the price as a whole vs. pieces?

It is difficult to say which is easier. Top-down may be a more general approach. It is not to say that the licensors and licensees have been successful.

Finding and picking these values is not a science. This is somewhat arbitrary. You determine how much the consumer is willing to pay for a DVD and you work backwards.

I would like to move on the second issue of mechanism for implementing RAND licensing. Dr. Kallay reported a very interesting experience of a seemingly failed attempt of the IEEE for clarifying and tightening the RAND obligations. Where might IEEE have failed?

A consensus could have been achieved in IEEE, but we missed that opportunity. At these conferences, there are more implementers than patent holders. That means it tends to go towards the implementers.

If you proceed without consensus, the balanced ecosystem will break down. You need to ensure FRAND access in the future, and consensus of all stakeholders is key to ensuring it.

Prof. Contreras mentioned international ADR. If there is an agreement on this, that will certainly help organize a patent pool. I wonder if a SDO can accommodate such a proposal.

There is already a mechanism in WIPO for RAND disputes-specific arbitration, but it has never been used. There is also a Japanese IP Arbitration Center. I hear that it is again seldom used. We might like to first study why these mechanisms are not being used.

For IPRs, you have the WIPO system, the Paris Convention, and the WTO Trade-Related Aspects of Intellectual Property Rights (TRIPS) agreement, so harmonization within the IP system is happening. For standards, there is no international convention that will harmonize all of the SDO entities concerned. It is relatively a select number of international SDOs that are affected, but they differ in legal status, private or governmental, and furthermore there are hundreds of other forum and consortium standards bodies. Some of these bodies are influential, so we might have a challenge in persuading SDOs to participate in this kind of worldwide harmonization.

I think international ADR can be good. The scope should be narrow, focusing only on the rate of royalties. If the scope is broader, it can invoke the issue of ability to arbitrate patent disputes.

Can SDOs force participants to abide by the rate set by ADR?

If there should be such a discussion, it should be done in the ISO Technical Management Board (TMB) which I served for eight years. But I have never heard of this kind of discussion; the closest being just once when we discussed anti-trust guidelines for standards-development participants so they don't discuss illegal issues such as pricing or production. I guess there are similar situations for other SDOs.

I agree that the scope should be limited to contractual monetary rate, not to issues of patent validity and competition law. WIPO and other tribunals are not used because the lawyers like to stick with the arbitral institutions they know.

The SEP holders tend to press for arbitration and implementers tend to resist it. Both parties have to agree to go to arbitration.

The new EU FRAND consultation document, issued recently on November 29, 2017, is commendable because it suggests that if parties cannot agree on FRAND terms and conditions, then their agreement to a third party adjudication that will determine the FRAND terms will be considered bona fide behavior. Conversely, refusal to such adjudication that determines the FRAND terms for all disputed patents should be seen as unwillingness to negotiate in good faith.

My empirical research on patent pools suggests two scenarios. When there is considerable symmetry among the parties that own the key patents, this allows for consensus building on the terms and conditions for the pool. The other condition is when a sufficient number of SDO members are worried that the standard will fail commercially without one.

I think pools also tend to succeed when there is a critical mass. I am somewhat optimistic with the recent Avanci licensing pool announced for IoT that includes Panasonic, Sony, and Sharp, because most of the major patent holders are members of Avanci. Last Friday, December 1, 2017, Avanci announced a license with BMW.

It is critical to find potential big players as first licensees so that the next party finds it easier to join. Another question is how to mobilize big patent holders.

Finally, could we discuss the new issue of open source software and SEPs. What is the fundamental source of conflict of OSS and SEPs and how can we address the issues?

The Internet Engineering Task Force (IETF) decided that open source implementation of its standards is desirable. The system for licensing IETF standards has a track for the text and another track for software code. FRAND licenses are incompatible with the Open Source Initiative (OSI)—defined forms of open source code licenses. It is not just the payment terms. Other terms like reciprocity are also problematic.

OSI open source definition No. 7 says the rights attached to the program must apply to all to whom the program is redistributed without the need of execution of an additional license. This additional license includes patent licenses. The implementers' side is inclined towards this because their business model is that of "open and closed," combining as free as possible technology and their own propriety technology as opposed to the RAND business model. There are two parties: OSI/OSS side and non-OSI/OSS side, corresponding to the already-seen conflict between Implementers and Right-holders. By the way, I see the BMW-Avanci agreement might be a first promising step for much worried IoT-related SEP issues.

I would now like to open the floor.

Q1. You have horizontal cooperation to create the standard and horizontal cooperation to implement the standard and no competition. Is there a concern?

I think there would be competition among software developers.

Q2. In telecommunications, there is now this kind of open source community that is creating open source hardware, not only software.

There is also another competition issue. Unlike standard development that is like an open, peer review process to pick the best technology, open source typically has two or three maintainers. They decide what goes in and what goes out, and that has a huge potential for exclusion.

I also hadn't heard about this. If you did have a small proprietary software vendor who is excluded because some get together and form an open source version, that may be a risk.

Q3. We have a lot of new IoT implementers who are unfamiliar with SEP licensing. In Europe the CEN-CENELEC process is trying to create an IoT licensing portal for implementers and SEP holders. Maybe this can be considered in Japan.

A JPO opinion paper states that there should be common recognition on an information sharing database. Such a database with quantitative information like FRAND rates will promote smoother bilateral negotiations.

Thank you very much for the lively discussion.