FRAND Is Not a Number: The Prior Question About Standard-Essential Patents

In contemporary SEP litigation, debates about FRAND almost always open with the same questions. What is the correct royalty rate? Which licences should count as comparables? Should the rate be global or national? Should the methodology be top-down, comparable-based, or anchored in a hypothetical negotiation?1

These are not minor questions. They explain why standard-essential patent disputes have become one of the most sophisticated laboratories of modern patent law.

They are, however, downstream questions. Before any of them can be meaningfully answered, a more basic problem has to be addressed: why does a FRAND obligation exist at all, and what is its legal nature?

That prior question is too often crowded out by the technical and economic vocabulary of SEP litigation — hold-up, hold-out, royalty stacking, portfolio valuation, global rates, implementer willingness. Each of those notions matters. None of them is intelligible without something more elementary: an enforceable undertaking, identifiable beneficiaries, and a legitimate basis for judicial intervention in giving that undertaking concrete effect.

The central difficulty of FRAND, in other words, is not primarily economic. It is juridical.

Standardization as legal infrastructure

Technical standards are routinely described as instruments of interoperability — devices, infrastructures, networks and platforms made to speak the same language. The description is accurate, but incomplete.

Once adopted and widely implemented, a standard does more than solve a coordination problem. It structures an entire technological ecosystem: it directs investment, conditions market entry, determines compatibility, and creates path dependency. For many actors, participation in the market becomes practically impossible without implementation of the relevant standard. A handset that cannot run cellular standards is not technologically deficient; it is commercially out of the market.

That gives standardization an infrastructural dimension with significant legal consequences. It generates structured dependence and produces legitimate expectations that the standardized technology will remain accessible. Those expectations are real, but they are not, of themselves, legal rights. A technical standard does not automatically create a duty to license, and essentiality does not, by itself, convert exclusivity into a general access regime.

This is where FRAND becomes decisive. The FRAND undertaking is the legal mechanism by which expectations generated by standardization are transformed into enforceable obligations. The standard creates dependence; the SSO’s intellectual property framework organizes institutional expectations; the FRAND declaration gives those expectations juridical form.

These layers should not be collapsed into one another. The technical standard is not the FRAND undertaking. The SSO policy is not the downstream licence. The undertaking itself is not yet the licence through which performance will ordinarily occur. Much of the conceptual confusion around FRAND litigation comes from running these distinct normative levels together.

Beyond competition law: a contractual source

European FRAND litigation has been shaped, naturally enough, by competition law — most prominently since Huawei v ZTE2. The procedural framework articulated by the Court of Justice on injunctive relief by dominant SEP holders remains central to current practice.

Competition law does not, on its own, account for the FRAND undertaking. It may discipline the abusive exercise of exclusionary remedies, address hold-up, and set the conditions under which an injunction becomes disproportionate. What it does not do is identify the private-law source of the obligation to make access available on FRAND terms.

Patent law has a parallel limitation. It defines the exclusive right and governs validity and infringement. It does not explain why a patent holder who has participated in a standardization process becomes bound to license — on FRAND terms — implementers who were never parties to that process.

The legal source of the FRAND obligation must therefore be sought elsewhere.

Within the ETSI framework, that source lies in the contractual architecture of standardization itself: the ETSI IPR Policy3, the written licensing declaration, the governing-law clause, and the undertaking assumed by the SEP holder. Read together, they point toward a contractual analysis. Under French law — which governs the ETSI undertaking — the most coherent characterization is that of a stipulation pour autrui4.

This is not doctrinal ornament. It is structurally explanatory. It accounts for why the undertaking is formally made to ETSI yet remains invocable by implementers; why the class of beneficiaries can be indeterminate at the moment of formation and become determinable at the moment of performance; and why implementers hold something more than a mere factual expectation but less than an automatically perfected licence.

Implementing the standard does not, of itself, give rise to a licence. It gives rise to a direct right to demand performance of the FRAND undertaking. That distinction does most of the analytical work.

Undertaking and licence are not the same thing

A recurrent error in FRAND discourse is moving too quickly from the undertaking to the licence itself.

The two are connected; they are not the same legal object. The undertaking is upstream — the commitment assumed within the institutional framework of standardization. The licence is downstream — the bilateral instrument by which that commitment will ordinarily be performed.

The undertaking does not generate a licence automatically, but it is not reducible to a vague duty to negotiate politely either. Its content is more precise: the SEP holder undertakes to make access available on FRAND terms, ordinarily through a licence, and to perform that undertaking consistently with its own normative logic.

This clarifies the role of courts. When a court determines FRAND terms, it is not acting as a standalone price regulator, nor is it inventing obligations ex nihilo. It is giving operational content to a pre-existing undertaking whose performance has become contested.

Whether one believes that FRAND identifies a single point or a range of acceptable outcomes, the underlying premise is the same: some terms satisfy the undertaking, others do not. The undertaking has substantive content, even if that content can only be specified contextually.

Substance and behaviour

The FRAND undertaking operates along two complementary dimensions.

The first is substantive. FRAND terms must remain fair, reasonable, and non-discriminatory. It is doubtful that each adjective functions as an autonomous legal test; FRAND is better understood as a composite benchmark. Each element nonetheless illuminates a distinct aspect of the undertaking. Reasonableness disciplines the attribution of value: the SEP holder is to be remunerated for the technological contribution of the patent itself, not for the lock-in effects produced by the act of standardization. Fairness prevents the licensing framework from reintroducing extraction through ancillary mechanisms, even where the headline royalty appears acceptable. Non-discrimination addresses unjustified disparities of treatment between similarly situated implementers, without collapsing FRAND into a rigid most-favoured-licensee rule.

The second dimension is behavioural. Because the undertaking is open-textured and ordinarily requires concretization through negotiation, its performance presupposes cooperation, transparency, responsiveness, and good faith.

That is where the procedural sequence articulated in Huawei v ZTE becomes relevant — notice, willingness, FRAND offer, counter-offer, security, diligent negotiation. These are important indicators.

They are not a mechanical algorithm. A party may comply formally with each step and still behave opportunistically; conversely, a justified departure from the sequence is not, by itself, incompatible with genuine FRAND performance.

The true standard is not ritualistic compliance. It is good faith — intensified — in the performance of a pre-existing juridical obligation.

That intensified good faith operates symmetrically. The SEP holder may not exploit standard-induced dependence to extract supra-FRAND terms. The implementer cannot invoke FRAND to justify indefinite delay or strategic disengagement. Hold-up and hold-out are not slogans. They become legally relevant when, and to the extent that, they are anchored in concrete conduct.

Juridifying private technical normativity

The significance of FRAND ultimately runs beyond SEP litigation.

FRAND illustrates one of the major transformations of contemporary legal systems: the juridification of privately produced technical infrastructures. Standards are not enacted by parliaments. They emerge from private or hybrid bodies, often organized around expert-driven governance. They nonetheless shape markets, technologies, and access to whole sectors of economic activity.

Law cannot simply ratify those structures as though they were neutral facts, nor can it ignore their practical normative force. It must qualify them juridically.

That is the function FRAND performs. It does not dissolve patent exclusivity into compulsory access. It does not turn courts into general regulators of technology markets. It does not reduce standardization to a subsidiary branch of antitrust. It provides the juridical form through which expectations generated by standardization become enforceable, while preserving the legitimacy of patent exclusivity itself.

A contractual theory of FRAND offers a coherent framework for that complexity. It identifies the source of the undertaking. It locates implementers within it. It distinguishes the undertaking from the downstream licence, and the substantive content of FRAND from the behavioural obligations governing its performance. And it allows patent law, competition law, contract law, and remedies to function at their proper conceptual levels rather than collapse into a single undifferentiated regulatory discourse.

This matters in practice. FRAND litigation will continue to expand across jurisdictions, institutions, and technologies. Global portfolio disputes, the growing role of European courts, and the development of alternative mechanisms for determining FRAND terms — including arbitration — point in the same direction: the future of SEP disputes will not be settled by valuation methodology alone.

These broader questions will be developed further in forthcoming work devoted to the contractual structure of FRAND commitments and the juridification of technical standardization.

The prior question remains a legal one. What is the undertaking? Who may invoke it? What does it require? How should it be performed? And how should courts and arbitral tribunals give it effect without becoming regulators of technology markets?

A coherent answer begins by recognizing FRAND for what it is: not a number, not a defence, not a competition-law constraint, but a contractual mechanism through which private technical normativity becomes legal obligation.

• 1See Matthieu Dhenne, Calculation of FRAND Royalties: An Overview of Practices Around the World, 41 European Intellectual Property Review 754 (2019).
• 2Case C-170/13, Huawei Techs. Co. v. ZTE Corp., EU:C:2015:477 (CJEU July 16, 2015).
• 3ETSI Rules of Procedure, Annex 6: ETSI Intellectual Property Rights Policy, cl. 6.1 (FRAND undertaking).
• 4See, e.g., C. civ. arts. 1205–1209 (as amended by Ordonnance No. 2016-131 of Feb. 10, 2016, portant réforme du droit des contrats, du régime général et de la preuve des obligations); formerly C. civ. art. 1121.