Licensing of standard essential patents: Hold-up, reverse hold-up, and ex-ante negotiation

Sadao Nagaoka 15 May 2019



Technologies protected by standard essential patents (SEPs) are reflected in an industry standard under the condition that any rights holder makes the commitment to licensing such patents on fair, reasonable and non-discriminatory (FRAND) terms. However, this licensing process has come under scrutiny, as numerous disputes centring on licensing terms have occurred in recent years. Such disputes underlie the publication of the Guide to Licensing Negotiations Involving Standard Essential Patents by the Japan Patent Office on 5 June 2018 (JPO 2018). In this column I offer brief explanations on the following three major issues in the licensing of SEPs – namely, hold-up, reverse hold-up, and ex-ante negotiations – which could help in implementing the licensing negotiations.


Hold-up is a situation in which the existence of an SEP is discovered only after a standard implementer has made investments into a business, with the rights holder demanding exorbitant royalties relative to the contribution of the technology, with which the implementer is forced to comply. The reason why the implementer is forced to accommodate such demands is that it has already incurred significant sunk cost by investing in the business based on the standard, and if the rights holder can threaten the implementer with an injunction on the grounds of patent infringement, the implementer faces the risk of not being able to utilise the investment already made (Farrell et al. 2007, Lemley and Shapiro 2007). For an example, if investments of 10 billion yen were made by an implementer for generating revenues of 11 billion yen, the total royalties would have been limited to the amount of one billion yen (the level of ex-ante profit) or less if licensing negotiations had been conducted prior to the investment. However, conducting negotiations after the investments would only guarantee that the total royalties would be limited to the amount of 11 billion yen (the size of the revenue) or less, which may exceed the business’s profit, thereby exposing the implementer to the risk of not being able to recover its investment.

Standards setting organisations are well aware of such risks of SEPs, therefore they only adopt the technology of the companies with SEPs that have made commitments to licensing them on FRAND terms, as a component of a standard technology.

Nevertheless, unless the price for the entire spectrum of SEPs is predetermined either by a patent pool or by the requirement of a standard organisation of royalty free licensing, the aggregate royalty rates of SEPs are rarely predetermined before the actual licensing negotiations. As a result, it would be possible for SEP holders to demand large royalties of its implementers, if the courts unconditionally agree to issuing an injunction at the request of the patent holder. Thus, an injunction is the necessary legal condition for hold-up. On the other hand, if there is no possibility of an injunction, it gives rise to the problem of ‘reverse hold-up,’ explained below.

Reverse hold-up or hold-out

Reverse hold-up (or hold-out) is a situation in which the rights holder cannot receive royalty income because the implementer refuses to engage in a licensing negotiation. Many experts point out that in recent years, the problem of reverse hold-up by implementers has become an even more serious problem than hold-up by SEP holders, due to the restrictions by courts on injunctions (Epstein and Noroozi 2017). 

To illustrate this problem of reverse hold-up, let us examine the case where a SEP holder is a company specialising in R&D. Under a standard licensing negotiation, this company's R&D investments had already become sunk costs at the time of the licensing negotiations. This is because, normally, negotiations are conducted after the invention is generated and often after the invention is made public, when its patent right is already established. Additionally, because the rights holder has made the commitment to licensing its SEP on FRAND terms, the implementer can use the SEP at any time it wishes.

Under such circumstances, if there is no threat of injunction from the rights holder, there is little incentive for the implementer to promptly conclude the licensing negotiations. The implementer can even enhance its profit by prolonging the negotiation, since it can strengthen its position and create a stronger financial constraint on the rights holder, whose only revenue source is the licensing revenue. Furthermore, if the patent becomes invalidated, the implementer will no longer need to pay the licensing fees. On the other hand, the rights holder cannot take back the right of said implementer to use the technology solely on the grounds of unsuccessful negotiations; moreover, as the R&D investments are already sunk costs and there is no other way of securing returns on these investments, other than concluding the licensing agreement. Consequently, the rights holder would potentially be forced to lower the licensing fees. Under such circumstances, the royalty the rights holder receives can become very low, preventing the patentee from making adequate return on its investment.

In order to address these issues, a penalty on such implementers who refuse to engage in licensing negotiations in good faith would be needed. One measure is to permit the SEP holder to seek an injunction against such firms. Consequently, there are a number of court cases, ruling that injunctions are permitted against unwilling licensees even if they are not permitted against willing licensees.

Ex-ante negotiations and incentives

The problems of hold-up and reverse hold-up (or hold-out) stem from the fact that royalties set in ex-post negotiations are subject to the variations in bargaining power. One step to prevent this is to adopt the framework of ex-ante negotiations. In an ex-ante negotiation, royalties, as appropriate incentives for innovation investments, are set in advance, before the sunk investments. This concept of ex-ante negotiation is being widely used in the US as hypothetical negotiations between the willing licensee and the willing licensor, to determine the amount of damages.

We can confirm that if we assume that the licensing fee will be determined through negotiations by the implementer prior to its investment (hypothetical ex-ante negotiations), then it will generate appropriate incentives for the implementer's investment. If the implementer had conducted licensing negotiations prior to executing its investments, the royalties will be set within the amount of expected profits to be generated from the use of the standard. The implementer can choose not to execute the investment if the higher amount is presented. Consequently, if there is a clearly established rule that the royalty fulfilling reasonable licensing terms is the royalty resulting from such an ex-ante negotiation, then it would generate an appropriate incentive for investment by the implementer to use the standards and alleviate the problem of hold-up.

Some believe that ex-ante negotiations should be set, assuming that such negotiations would have taken place when the alternatives for the standard existed, that is, even earlier than the point when the standard had been agreed. This is because, if the value of the standard were determined by network externality through its wide use (but not the standard technology per se), licensing negotiations conducted after the standard had been widely used would overvalue the contribution of the standard technology. Moreover, such licensing negotiations conducted after the establishment of the standard would not reflect the competition among the technical alternatives for the agreed standard. Thus, the Fair Trade Commission (2011) argues for the application of the hypothetical negotiation framework and recommends that “[c]ourts should cap the royalty at the incremental value of the patented technology over alternatives available at the time the standard was chosen.”

However, if the ex-ante negotiation were set earlier to the point when competition for the standard exists, it would be essential for the negotiating rule to secure appropriate incentive for the R&D of the standard technology. if the incremental value (i.e. the value of the proposed technology adopted minus the value of the proposed technology that is next in line) is used as the reasonable royalty value of the adopted standard technology, then when the two proposed technologies are similarly superior to the current standard, the value of not only the technology that was not adopted but also the technology that was adopted would become zero, thereby minimising the ex-ante incentive for R&D. Additionally, if the new standard technology were an important source for the scale of network externality, it would be critical to reflect the value of new technology in the royalty in order to ensure appropriate incentives for developing superior standard technology. 

Concluding remarks

Under the patent system, if several companies almost simultaneously generate similar inventions, the first filing company acquires the patent (exclusive right). As for the second company, if its contribution to further significant progress is recognised compared to the initial patent, then a further patent is awarded for that portion. If no progress is recognised, no patent is awarded. In other words, even if identical inventions are made, the patent system provides an ex-ante incentive for R&D, as the exclusive right is awarded to one party. While the patent race under such incentives is not necessarily the best system (Scotchmer 2006), it has played a major role in motivating R&D. In the case of standards, setting royalties for a new standard in line with its additional technical value compared to the existing standard is consistent with the basic concept of the patent system that motivates technological progress. However, the concept of paying royalties in line with the difference in value between the most superior proposals for the new standard and the next superior proposal seems to contradict the basic concept of the patent system, thus undermining the ex-ante incentive for an R&D.

Therefore, if we were to adopt the framework that sets ex-ante negotiations earlier than the point where the standard is set, we would also need to think about the incentives for R&D of standard technology. That is, a standard based innovation requires complementary investments by both the firms developing the standard technology and the firms using the standards. Thus, royalties will need to be determined through ex-ante negotiations that do not contain the elements of hold-ups, and at the same time generate appropriate ex-ante incentive for the R&D of a new standard.


Epstein, R A and K B Noroozi (2018), “Why incentives for “patent holdout” threaten TO dismantle FRAND, and why it matters,” Berkeley Technology Law Journal 32: 1381–1432.

Fair Trade Commission (2011), Evolving IP market place report: Aligning patent notice and remedies with competition, Federal Trade Commission, report.

Farrell, J, J Hayes, C Shapiro and T Sullivan (2007), “Standard setting, patents, and hold-up,” Antitrust Law Journal 603: 616–18.

Japan Patent Office (2018), Guide to Licensing Negotiations Involving Standard Essential Patents.

Scotchmer, S (2006), Innovation and Incentives, MIT press.

Lemley, M A and C Shapiro (2007), “Patent holdup and royalty stacking,” Texas Law Review: 1991–2049.



Topics:  Competition policy Productivity and Innovation

Tags:  patents, standard essential patents, hold up, reverse hold up, ex ante negotiation, licensing, technology

Professor of Economics, Tokyo Keizai University and Program Director, RIETI


CEPR Policy Research