In May, 2019, the U.S. District Court for the Northern District of California ruled that Qualcomm, Inc., the dominant supplier of wireless mobile telecommunications chips to the global smartphone industry, violated its obligation to license patents covering key communications standards to smartphone manufacturers on terms that are fair, reasonable and non-discriminatory (FRAND). The court held that both the rate that Qualcomm charged (approximately 5% of the price of a smartphone) and its method of calculating royalties (based on the total price of a smartphone rather than the chip embodying the standard) violated this commitment. As a result, the court ordered Qualcomm to renegotiate close to three hundred existing license agreements with manufacturers around the world. Whether or not this decision is upheld on appeal, it clearly illustrates the large number of FRAND licenses in the market and the enormous potential impact that they have.
Background: Standards, Patents and FRAND Licensing
Technical interoperability standards like Wi-Fi, USB and 4G/5G are critical to the modern networked economy, yet patent litigation concerning these standards continues to plague courts, competition authorities and technology firms around the world. One of the key disputes is the amount that can lawfully be charged to use patents that cover such standards (standards-essential patents or SEPs). While the holders of patents are generally free to charge what they wish to users of their patents (licensees), the developers of technical standards, working within trade associations called standards development organizations (SDOs), typically agree to license their SEPs to manufacturers of standardized products either royalty-free (e.g., standards like USB, Bluetooth, XML and IPv6) or at royalty rates that are fair, reasonable and non-discriminatory (FRAND) (e.g., standards like Wi-Fi, LTE and MP3).
Despite this general agreement, which is intended to ensure the broad adoption and deployment of standards in products around the world, SDOs do not specify the magnitude of FRAND royalties applicable to their standards, or even how such FRAND royalties should be determined. Thus, SEP holders and product manufacturers are left to negotiate FRAND rates themselves, usually in confidential, two-party (bilateral) transactions. As such, it is difficult for manufacturers to know whether the rates that are offered to them are actually reasonable, or whether they are the same or different than the rates offered to other manufacturers. This lack of transparency has led to a lack of trust among parties and protracted negotiations, often resulting in an impasse without agreement.
FRAND Litigation - A Lack of Transparency, Consistency and Comprehensiveness
When parties disagree over FRAND royalty rates, they may resolve their disagreement in court or through binding arbitration. Arbitration is generally confidential, so while it can resolve disputes between particular parties, this lack of transparency does little to inform the rest of the industry. Recently, an increasing number of cases regarding FRAND royalty rates have been heard by courts around the world. In such cases, the manufacturer wishing to sell a standardized product typically alleges that the SEP holder failed to comply with its commitment to offer a license on FRAND terms, and the SEP holder argues that the manufacturer was not willing to pay a FRAND royalty rate. Recent cases involving the assessment of FRAND royalties, or the methods for calculating them, have included In re. Innovatio (U.S. C.D. Ill., 2013), Ericsson v. D-Link (U.S. Fed. Cir. 2014), Samsung v. Apple Japan (Japan, IP High Court 2014), Microsoft v. Motorola (U.S. 9th Cir. 2015), Unwired Planet v. Huawei (U.K. EWCA 2018), TCL v. Ericsson (U.S. C.D. Cal. 2018), and FTC v. Qualcomm (U.S. N.D. Cal. 2019).
As the number of courts worldwide deciding FRAND-related cases has increased, so has disagreement over the interpretation of FRAND commitments. There are at least a dozen points of serious divergence among courts, both across national borders and within the same country, with respect to the methods of calculating FRAND royalty rates. These methods can vary dramatically from court to court and case to case, even for patents covering the same standard, leading to a profound lack of consistency among judicial FRAND determinations.
Moreover, in litigation, the royalty due to a SEP holder is determined individually, without reference to other patents covering the same standard or product. In most cases, only one patent holder out of dozens presents arguments to the court, generally over-emphasizing the value of its own patents above the others. Thus, even if such a royalty might meet some test of reasonableness if considered in isolation, it is not likely to be reasonable when combined with many other independently-calculated royalties applied to the same product. This is the well-known issue of royalty "stacking." In reality, despite the inherently bilateral, adversarial nature of litigation, the determination of a FRAND royalty is not strictly a bilateral matter. Rather, it must involve consideration of the other patented and unpatented technologies embodied in the standardized technology, as the royalty determination made in one case necessarily affects the level of royalties in many other cases. Thus, FRAND royalty determinations, particularly when they relate to standards covered by diversified patent holdings, should be comprehensive, taking into account the totality of patents and technologies involved in the relevant standard or product.
These three deficiencies in the FRAND licensing system—transparency, consistency, and comprehensiveness—impact the market in two principal ways. First, they reduce the overall fairness and accuracy of the system by producing results that could favor some market participants over others. Uncertainty regarding the level of FRAND royalty rates makes it more difficult for manufacturers to predict the cost of entering the market. Second, the system imposes excess transaction costs through duplicative negotiation and litigation both among different parties disputing rates over the same patents and among the same parties litigating in different jurisdictions (just consider the 300 licenses that Qualcomm alone has entered). As reported in recent cases, negotiations over FRAND licenses routinely take years and often result in an impasse that can be resolved only through litigation. Such transaction costs could deter or prevent small and medium enterprises (SMEs) from entering the market and increase costs for larger firms, which in turn could lead to reduced consumer choice and higher consumer prices. The result could undermine the benefits that industry-wide, non-discriminatory licensing of SEPs is intended to achieve. These concerns are particularly worthy of attention given the emergence of important new technologies, such as broadband 5G/6G data transmission and the Internet of Things that will be heavily dependent on standardization.
Toward a Global FRAND Tribunal
To address the three issues of transparency, consistency and comprehensiveness in FRAND licensing, I have proposed the creation of a global, non-governmental FRAND rate-setting tribunal (FRAND Tribunal) (details of this proposal can be found here). SDO policies should then be modified either to mandate or make available the resolution of FRAND royalty disputes through this FRAND Tribunal. Like existing rate-setting bodies such as the U.S. Copyright Royalty Board, the FRAND Tribunal will collect all available evidence regarding the patented and unpatented technology involved in a particular standard, determine an aggregate (top-down) royalty rate for the standard as a whole, and appropriately apportion royalties among all SEP holders. Beyond these rate-setting matters, the FRAND Tribunal will adjudicate no other issues, and related claims for breach of contract and antitrust and competition law violations would continue to be subject to adjudication by national courts.
In order to eliminate inappropriate leverage by SEP holders in these rate-setting proceedings, all SDO participants should be required to refrain from seeking injunctive relief against potential licensees during the proceedings, but once FRAND royalty rates are determined for a specific standard, a SEP holder should be permitted to pursue injunctions against product manufacturers that refuse to accept a license at the designated rate.
Though the proposed rate-setting mechanism will not result in allocations to individual SEP holders that precisely account for the variable value of each SEP, this degree of precision is neither a realistic expectation nor a necessary condition to the functioning of the market. Rather, it is hoped that the proposed FRAND Tribunal will make the FRAND licensing marketplace operate more efficiently by eliminating excessive transaction costs and enabling broad dissemination of standardized technology around the world, while still achieving a fair distribution of proceeds among affected stakeholders. While further details regarding the implementation and operation of the FRAND Tribunal must be considered and debated, this general approach has the potential to bring greater predictability and stability to the technology development ecosystem while reducing costly and disruptive litigation.
The determination of FRAND royalties today suffers from a lack of transparency, consistency and comprehensiveness. The proposed FRAND Tribunal, which addresses each of these issues, is designed to reduce negotiation, litigation, and other excess transaction costs while bringing greater predictability and stability to the technology development ecosystem and achieving a fair allocation of resources among affected stakeholders.