SEP: New EPO Observatory Study Provides Greater Clarity on Determination of FRAND Licensing Rates
During the Euractiv event held on June 24, 2026, key findings on the study (1) “Methodologies for FRAND determination: evidence from global case law” were discussed. The study is a result of a survey of court cases around the world involving the determination and assessment of Fair, Reasonable, and Non-Discriminatory (FRAND) terms and conditions for the licensing of standard essential patents (SEPs). The survey was conducted by the EPO Observatory, a department of the EPO that analyzes market data in relation to patents, in collaboration with the consulting firm BRELA (2).
The Euractiv event (3) was dedicated to Europe’s competitiveness in emerging technologies – and more precisely to the role of standards, markets and quantum innovation. Pere A Castell, from the EPO Observatory, presented the key findings of the extensive 148-page study that aims to offer greater clarity on the determination of FRAND licensing rates for standard essential patents (SEPs).
The goal of the study is said to be empirical and descriptive rather than prescriptive, trying to be as objective as possible. Remarks are illustrated with an extensive list of tables and figures. As such, it is an excellent resource for data on legal decisions relating to FRAND worldwide.
The study reviews a large comprehensive global corpus of related court decisions (65 decisions from 7 jurisdictions: US, JP, IN, CN, UK, DE, NL). The corpus reflects three main approaches through which courts have engaged with FRAND rates across jurisdictions over the last 15 years: (1) FRAND rate determinations, i.e. the court sets a FRAND rate, (2) FRAND rate assessments, i.e. a court decides whether a given rate is FRAND or within a FRAND range, and (3) assessments of the admissibility of FRAND rate determination methods. All decisions are described in section 2 of the study.
Despite operating under different legal frameworks, the study shows that courts have converged on a common purpose for FRAND: allowing a fair reward for the patent owner while ensuring the widest possible adoption of the standard. Section 3 of the study analyses frameworks and principles of FRAND determinations, showing that the framing has evolved and that the center of gravity in the case law has moved from a definitional phase – dominated in 2013–2015 by discussions of FRAND principles – to a methodological phase in which the application of FRAND through comparable-license analysis (and, to a lesser extent, top-down calculations) dominates from 2022 onward.
Methodologies for FRAND determinations are described in Section 4 of the study. The comparable-license analysis emerges as the primary method: it derives a FRAND rate by reference to the rates, royalty bases and terms agreed in prior comparable arm’s-length agreements. The top-down approach is used mainly as a cross check: it calculates an aggregate royalty rate for the whole standard – e.g. for 5G - and then apportions a share of this aggregate royalty to the patent portfolio that is being litigated.
The study analyses how these methods are operationalized in practice and points out different sets of challenges.
Regarding comparable license analysis, courts have developed a structured approach to address recurrent challenges. A first step is selection: courts must assess whether a proposed license covers sufficiently similar technology, parties, scope and time period and whether it is itself untainted by hold-up or hold-out. Second is unpacking: whether the comparable license is a lump-sum agreement, a cross-license, or a license to a larger portfolio, the effective rate is often not directly defined and must be recovered through net-present-value calculations, portfolio-strength ratios and apportionment across standards. Although unpacking is a complex process, courts have generally held that the fact that a license requires unpacking does not mean that the license must be excluded from a comparable licenses analysis. The third step is adjustments: when comparables involving different portfolios exist, courts must adjust for differences in relative portfolio value; when the existing comparable licenses are affected by non-FRAND factors, these factors must be accounted for in order to derive a FRAND rate.
Regarding the top-down approach, the study shows that both steps of establishing the aggregate royalty rate (ARR) and then apportioning a share of that aggregate to the portfolio at issue are contested. For the ARR, courts have relied on different sources, but no single approach governs. As an example, the study points out that the 5% figure for 3G and the 6–10% range for 4G that recur across US, UK, Chinese and German decisions owe their persistence largely to cross-referencing of earlier decisions rather than independent re-derivation, raising questions about the robustness of the apparent convergence. Apportionment of the aggregate royalty rate to the particular portfolio usually relies on patent counting with no evaluation of quality and courts have consistently cautioned that “mere patent counting and dividing is not enough.”
In this regard, the study points out how quantitative patent data plays a central but contested role in all FRAND determinations and this is the object of Section 5. Patent portfolio strength assessments are inevitable in apportionment in the top-down approach but they are also commonly used in several steps of the comparable licenses. The case law illustrates that patent counts are widely used because they are objective and manageable, not because courts consider them fully satisfactory. A spectrum of approaches exists, from purely quantitative declared-patent counts to more refined exercises incorporating essentiality screening, validity adjustments, family de-duplication and technical evaluation. Indicators of individual patent value — forward citations, family size, timing and technical evaluation — have been explored but none has been accepted as a comprehensive or universal measure. This methodological gap is widely recognized but largely unresolved in existing case law.
(1) https://brela-research.com/
(2) https://www.epo.org/en/news-events/events/europes-competitiveness-emerging-technologies-what-role-standards-markets-and
(3) https://link.epo.org/web/publications/studies/en-methodologies-for-frand-determination.pdf
