chevron-down Created with Sketch Beta.

    The Antitrust Source

    The Antitrust Source | October 2024

    Inherent Hold-Up Value, Market Power, and Supra-FRAND Royalty Rates in Standard Essential Patent Licensing

    Yi Cheng and Fei Deng

    Oct 29, 2024

    13 min read

    Antitrust Antitrust Litigation & Enforcement Antitrust Policy & Legislation Exclusionary Conduct Intellectual Property Media & Technology Monopoly Power Pricing Private Antitrust Litigation Unilateral Conduct

    Summary

    • Licenses established through “free market” negotiations between SEP holders and implementers are inherently influenced by the lock-in effect induced by standardization, further compounded by the threat of injunctions.
    • “Free market” license agreements may entail a range of supra-FRAND royalty rates that reflect varied hold-up values.
    • It is crucial for the negotiating parties and courts to recognize the possibility that comparable licenses are affected by hold-up and adjust accordingly in determining FRAND royalties.
    Inherent Hold-Up Value, Market Power, and Supra-FRAND Royalty Rates in Standard Essential Patent Licensing
    Eli Wilson via iStock

    Jump to:

    Hold-up in standard essential patent (SEP) settings arises when an SEP owner leverages the lock-in generated by the standard, and “holds-up” licensees by demanding royalties higher than what could be negotiated in an arm’s-length transaction where the implementer was not locked in. There is a large body of literature on hold-up, including empirical studies of SEP holders’ opportunistic behavior. However, many SEP holders continue to claim that hold-up is merely a theoretical concern, and that there is no evidence of hold-up in practice. 

    The issue is an important one because inflated royalties due to hold-up effectively serve as a tax on new products incorporating the SEP technology, creating a barrier to innovation. This added cost could be passed on to end consumers, leading to elevated prices, limited product variety, and compromised product quality. As noted in economic literature, the excessive royalties driven by hold-up “creates both incentives to get patented technology adopted into standards far in excess of its social value, and disincentives for firms to innovate when their innovations must use the standard, such as by developing goods complementary to the standard.” Various regulatory authorities have also expressed concerns over the effect of hold-up on innovation. For example, the FTC noted: “Patent hold-up can overcompensate patentees, raise prices to consumers who lose the benefits of competition among technologies, and deter innovation by manufacturers facing the risk of hold-up.” In Philips v. Thales, FTC Chair Lina M. Khan and FTC Commissioner Rebecca Kelly Slaughter submitted a written submission to the United States International Trade Commission (USITC), petitioning the USITC against issuing an injunction order on the alleged infringers of certain 3G and 4G SEPs, in which they noted hold-up “harms consumers in the longer run through reduced innovation, competition, quality, and choice if firms choose to reduce investments in standardized products.”

    This article provides a simple economic illustration of the formation and impact of hold-up in SEP licensing. We demonstrate that licenses established through “free market” bilateral negotiations between SEP holders and implementers are inherently influenced by hold-up. As a result, they reflect supra-FRAND royalty rates that greatly exceed the intrinsic economic value of the patented technologies, reflecting hold-up values derived from the market power that the lock-in generated by the standard selection confers on the SEP holder.

    To understand the formation and impact of hold-up in SEP licensing, it is useful to begin with the case of a regular (non-standard essential) patent. Generally speaking, the economic value of a patent depends on the incremental profit a licensee earns (due to higher revenue and/or reduced costs) by using the patented technology over potential alternatives. For a regular patent, if the patent holder demands a royalty higher than its true economic value, the prospective licensee can choose to adopt alternative technologies, work around the patented technology, or exclude the feature from its product.

    The process of standard setting fundamentally changes the negotiation dynamics in favor of the patent holder whose patent gets incorporated into the standard. Once a standard is established and becomes commercially successful, manufacturers invest in creating products that align with the standard. For many standards, the industry will become locked-in to the technologies that were chosen by the standard setting organization (SSO) for inclusion in the mandatory portions of the standard. If left unchecked, SEP holders could leverage the lock-in and hold up licensees by demanding royalties higher than what could be negotiated in an arm’s-length transaction where the implementer was not locked in. This could lead to economically inefficient outcomes by allowing the patent holder to collect “supra-FRAND” royalties that significantly surpass the intrinsic economic value of the patented technology.

    The intrinsic lock-in nature of standards often provides the SEP holders with significantly more bargaining power post-standardization—implementers’ only alternative to reaching an agreement may become to stop using the entire standard and thus lose the benefits of standardization. Any holder of a single valid SEP potentially has the ability to hinder manufacturers from implementing the entire standard, using this leverage to demand excessively high royalties. This increased bargaining power is magnified by the threat of injunction, available in many jurisdictions. Even the mere possibility of an injunction can influence negotiations greatly, as it puts at risk all of the implementer’s profits in a given jurisdiction. These factors establish each SEP holder as a monopolist in SEP licensing negotiations, which in theory could allow them to engage in perfect price discrimination by charging each implementer a royalty rate tailored to their individual maximum willingness to pay.

    Bargaining with and without standardization

    The hold-up problem can be further illustrated by contrasting hypothetical licensing negotiations that would occur without and with standardization. The outcome of any negotiation depends on the gains from trade and the outcome from “outside options” available to each party should the negotiation fail. These factors will therefore determine both the implementer’s maximum willingness to pay and the SEP holder’s minimum willingness to accept.

    Without standardization (i.e., when the patented technology at issue is not integrated into a standard), competition among alternative technologies can drive down the implementer’s maximum willingness to pay for the patent holder’s technology. The patent holder also faces competition and so has lower minimum willingness to accept. Both factors result in a lower royalty rate.

    Consider the scenario in Figure 1. An implementer obtains a profit margin of $40 per unit if it were to use the patented technology at issue, compared to $35 per unit if it were to use the next-best alternative technology. In other words, the patent technology at issue creates a $5 additional value for the implementer compared to the next-best alternative technology—the “intrinsic economic value” of the patent technology is $5.

    Consider the outside options of the implementer and the patent holder in a hypothetical negotiation. Without standardization, the implementer would be willing to pay at most $5 per unit for the patented technology. If the patent holder charged more than $5, the implementer would have “walked away” and adopted its outside option—the next-best alternative technology. On the patent holder side, they would be facing large “sunk” (irrecoverable) R&D and patenting costs upfront, but low marginal costs of licensing. That is, once the technology has been invented and patented, the additional cost (i.e., marginal cost) of granting licenses to any implementer is minimal and is unlikely to increase with the number of licenses issued. The outside option is not licensing the patent and receiving zero royalties. A profit maximizing patent holder, therefore, would be willing to accept a minimal royalty rate that is just at or above its low marginal cost when faced with competition. Therefore, the patent holder’s minimum willingness to accept would be minimal, say at $1 per unit.

    A common assumption in bargaining models is that the two sides have about the same bargaining power and evenly split the gains from trade. As illustrated by the gray bars and the blue star in the left part of Figure 1, this assumption implies a royalty rate at the midpoint ($3 per unit) between the patent holder’s minimum willingness to accept ($1 per unit) and the implementer’s maximum willingness to pay ($5 per unit).

    If the patent was instead essential to a successful standard, the implementer has two alternatives that determine its outside option and maximum willingness to pay. First, it can decide not to license the patent. If it takes this route, this no longer means replacing the SEP with an available alternative technology, but rather that it stops practicing the entire standard. Second, the implementer may choose to enter into litigation with the patent holder to determine a FRAND rate. In this case its outside option would be its expected post-litigation profits, which would depend on litigation costs, the potential risk of exiting countries with injunctions, and other costs injunctions might bring, such as loss of reputation, interruption of normal course of business, and the likely royalties determined by courts. The SEP holder’s minimum willingness to accept becomes the expected court-adjudicated royalties minus its litigation costs.

    For example, consider the same patented technology as the example above but incorporated into a standard, and assume that the implementer expects a $10 per unit court-determined royalty rate and another $10 per unit cost from litigation expenses and loss of profit, reputation, and interruption of normal course of business due to injunctions. Facing a worse outside option, the implementer would be willing to pay at most $20 per unit for the patented technology, instead of at most $5. This leads to an increase of $15 in the implementer’s maximum willingness to pay.

    The SEP holder’s outside option improves and so its minimum willingness to accept increases after standardization. Assume that the SEP holder expects a $10 per unit court-determined royalty rate and a $1 per unit litigation cost, the SEP holder’s minimum willingness to accept would be $9 per unit, $8 more than before. Even more, as illustrated later in this section, standardization and lock-in give SEP holders significantly more bargaining power. The royalty rate from a “free-market” negotiation for an SEP tends to fall on the right end, i.e., close to the implementer’s maximum willingness to pay, at $20, rather than in the middle between that and the SEP holder’s minimum willingness to accept ($9). With standardization, the free-market royalty rate agreed upon by the parties in their license agreements, at close to $20 per unit, significantly exceeds the $3 per unit royalty rate without standardization—the royalty rate that reflects the intrinsic value of the patent technology and equal bargaining power. The difference in these two royalty rates, almost $17 per unit, reflects the hold-up value of the SEP-at-issue, leveraged by the SEP holder based on its market power gained through standardization and the availability of injunctions.

    “Free-Market” Royalty Rate (X) = FRAND Royalty Rate (A) + Hold-up Value (B)

    Figure 1: Illustration of SEP Holder’s Minimum Willingness to Accept, Implementer’s Maximum Willingness to Pay, and Royalty Rate Without and With Stan

    Figure 1: Illustration of SEP Holder’s Minimum Willingness to Accept, Implementer’s Maximum Willingness to Pay, and Royalty Rate Without and With Stan

    Even if the implementer were to expect that after litigation expenses, a court may land on an adjudicated royalty rate that is closer to the intrinsic economic value of the patented technology (say, $5 per unit) or even a $3 per unit FRAND rate based on a pre-standardization “bottom-up” evaluation under the hypothetical negotiation setting, the free-market royalty rate the implementer is willing to pay would still be sufficiently above these rates, due to the elimination of commercially feasible alternatives after the standardization, high litigation expenses, and loss of profit and other costs due to injunctions, etc. Thus, the post-standardization free-market royalty rate would still be supra-FRAND, with the difference between the free-market and real-FRAND royalty rates, i.e., the hold-up value, varying across implementers according to each implementer’s profit margin and willingness to pay.

    Price discrimination and SEP licensing patterns in practice

    It is important to note that unlike in a competitive market where the licensor would be willing to set the royalty rate equal to the licensor’s marginal cost, the lock-in nature of standards and the risk of injunctions provides the licensor with significant market power, which enables the licensor to price discriminate—maximizing its profit by varying the royalty rate charged for each individual licensee based on the expected willingness to pay of each licensee.

    In economic terms, there are three prerequisites for price discrimination: the licensor possesses substantial market power; has the ability to identify or estimate differences in willingness to pay among licensees; and can prevent the resale of the license, i.e., arbitrage. All of these conditions can generally be met in an SEP licensing negotiation. First, SEP holders have substantial market power from each of their SEPs that are truly essential—there is no commercially feasible alternative for the implementer to comply with the industry standard. Second, SEP holders are oftentimes able to estimate the potential licensees’ profit margins, exposure to injunctions, and litigation costs, and thus the potential licensees’ willingness to pay based on public information, such as the licensees’ own financial reports and industry analyses and data published by third parties, combined with their own industry knowledge. Third, SEP licensing agreements often include clauses that restrict resale of the license to other parties, which prevents arbitrage and maintains price discrimination.

    The general trends in SEP licensing practices in the real world are consistent with the theoretical framework of hold-up discussed above. There is evidence that SEP holders strategically sequence their licensing and litigation efforts to exploit the implementers’ varied profit margins and willingness to pay, and maximize their profits by extracting differential hold-up values from the implementers. To be more specific, SEP holders typically structure their licensing and litigation efforts as follows.

    SEP holders may first target major manufacturers with relatively high profit margins, such as Apple and Samsung in the mobile device industry, utilizing injunction threats, to arrive at free-­market license agreements with high per-unit royalty rates and significant amounts of royalties. These manufacturers tend to face significant amounts of potential profit loss from injunctions, including loss from short-term sales bans and long-term market trust issues among business partners and end consumers, which combined with relatively high profit margins drive up their willingness to pay. Therefore, the SEP holders may be able to reach free-market license agreements with this type of licensees at high per-unit royalty rates, extracting high hold-up values and obtaining significant amounts of royalties.

    After the SEP holders successfully obtain license agreements containing high per-unit royalty rates from the previous step, they may use these licenses as “comparable benchmarks” and approach major manufacturers with relatively low profit margins, such as OPPO, vivo, and Xiaomi in the mobile device industry. Since these manufacturers have lower profit margins and therefore lower maximum-attainable willingness to pay, their per-unit royalty rates are generally lower than those of licensees with higher profit margins. But their high exposure to injunctions can still mean that those lower per-unit rates represent a large proportion of their profit margins. Hold-up can therefore still result in excessive per-unit royalty rates compared to truly FRAND royalty rates.

    Below are a few examples of SEP holders securing license agreements with major manufacturers who have relatively high profit margins first, and then with those who have relatively low profits margins, based on public information:

    • InterDigital, a major cellular standard SEP holder, first secured licenses with major mobile device manufacturers who had relatively high profit margins—Apple in 2016 and Samsung in 2014, before aggressively targeting major mobile device manufactures with lower profit margins, such as Xiaomi (with litigation started in 2020 and license agreement reached in 2021), OPPO (with litigation started in 2021 and in the process of litigation now), TCL (with license agreement reached in 2023), and Lenovo (with litigation started in 2019 and in the process of litigation now).
    • The licensing and litigation history of ZTE, who is another major cellular standard SEP holder, demonstrates a similar pattern. According to public information, ZTE secured licenses or initiated litigations against major mobile device manufacturers who had relatively high profit margins first—Apple (license agreement signed in 2020) and Samsung (license agreement signed in 2021 followed by litigation brought by a related party G+ in 2024), before approaching major mobile device manufacturers with lower profit margins, such as Tinno (patent dispute litigation started in 2021), vivo (cross-license agreement signed in 2024), and OPPO (patent dispute litigation started in 2022 and settlement reached in 2024).
    • In earlier years, Nokia secured license agreements with major mobile device manufacturers at the time—Research In Motion (“RIM”, the manufacturer of BlackBerry), Apple, and Samsung—who had relatively high profit margins. After that, Nokia reached license agreements with major mobile device manufacturers that had lower profit margins, such as with Xiaomi and Huawei in 2017, with OPPO in 2018, and with vivo in 2019. It appears that upon the roll-out of 5G technology, Nokia started another round of licensing negotiations with major manufacturers, including Samsung, Apple, OPPO, and vivo.

    Due to the lack of publicly available information regarding license term details, a comprehensive and precise quantitative evaluation cannot be performed to calculate and compare the royalty rates each licensor imposed across different licenses, or to assess the extent to which hold-up value is embedded in the agreed-upon royalties. Nevertheless, for demonstration purposes, it can be roughly estimated, for example, that the per unit royalty rate Nokia charged to Apple, an implementer who has a relatively high profit margin, is about twice that of OPPO, an implementer who has a much lower profit margin. This is consistent with the theoretical framework of hold-up discussed above, indicating that SEP holders have pricing power in the licensing market and thus are able to extract differentiated royalties from implementers based on their profit margins and willingness to pay.

    Notably, Nokia was able to extract a relatively high portion of the profit margins as royalties from manufacturers like OPPO—while the average per-unit profit margin of OPPO, vivo, and Xiaomi is less than 3% of Apple’s per-unit profit margin, the per-unit royalty rate paid by OPPO is about half of that paid by Apple. On average, then, OPPO, vivo, and Xiaomi paid roughly seventeen-times as much as Apple in royalties as a proportion of their profit margin. This suggests that low profit margin manufacturers may have been even more held-up by SEP holders, paying even higher portions of their profit margins as royalties.

    Conclusion

    Licenses established through “free market” negotiations between SEP holders and implementers are inherently influenced by the lock-in effect induced by standardization, further compounded by the threat of injunctions. Consequently, these agreements may entail a range of supra-FRAND royalty rates that reflect varied hold-up values. To the extent that these licenses are used as comparable licenses in subsequent license negotiations or court rulings, royalties calculated based on such comparable licenses would exceed the intrinsic value of the patented technology, exacerbating the issue of hold-up. Thus, it is crucial for the negotiating parties and courts to recognize the possibility that comparable licenses are affected by hold-up and adjust accordingly in determining FRAND royalties. For example, consider that comparable licenses that could be used as a benchmark may be at a range of royalty rate, using the lowest rather than the average or median royalty rate among them is less likely to reflect the impact of hold-up. Evaluating the patent-at-issue based on a pre-standardization “bottom-up” approach under a hypothetical negotiation setting would also yield a royalty measure that is more consistent with the FRAND principle.

    The views expressed herein are those of the authors and do not necessarily represent those of the authors’ firm or any of their clients.

      Authors