The "Supra-FRAND" Clawback: Is Restitution the Next Frontier in SEP Litigation?

Abstract:

FRAND licensing asks implementers to pay today for a benchmark that courts may only define tomorrow – often in disputes to which they were not a party. When later decisions show that agreed royalties exceeded any plausible FRAND range, a quiet but fundamental question emerges: is overpayment simply the cost of legal certainty, or does private law allow restitution? This post explores whether civil-law doctrines – taking Portuguese law as a reference point – can accommodate clawback of supra-FRAND royalties, and if the theoretical possibility of restitution unsettles the foundations of SEP licensing.

Setting the scene:

The standard-essential patent ecosystem is predicated on a delicate, often precarious equilibrium between the imperative of technological standardization and the proprietary rights of innovators.

SEP licensing is characterized by structural uncertainty at the time many license agreements are concluded, at which time fair, reasonable and non-discriminatory terms have not yet been judicially confirmed.

This uncertainty is not accidental. In much recent case law, FRAND rates have not been regarded as a single, fixed and predetermined rate, but a range - a corridor shaped by comparables, timing, geographic scope, portfolio strength, and risk allocation. In Unwired Planet v Huawei, the UK Supreme Court rejected the notion of a “hard-edged” non-discrimination obligation, emphasising that FRAND permits diversity of outcomes so long as the result is fair and reasonable in the round.

For many implementers – particularly SMEs – the rational response to this uncertainty and the threat of litigation is to accept licenses early on the terms imposed by SEP holders, avoid litigation, and bear the (sometimes high) cost of legal peace. Years later, however, judicial decisions in disputes over the same patent portfolio may reveal that the royalties paid fall outside the FRAND corridor ultimately recognised by courts. At that point, the question arises: is overpayment merely the price of certainty, or does private law offer any route to restitution?

This question matters practically. It affects licensing strategy, litigation timing, and risk modelling on both sides of the market. Yet it also cuts against some of the deepest intuitions of contract law, including pacta sunt servanda and the allocation of valuation risk between more or less sophisticated parties.

FRAND as a normative benchmark:

Courts across jurisdictions increasingly treat FRAND as an objective standard capable of judicial determination through evidence and economic analysis. This enhances legal certainty ex post, but it also transforms FRAND into a normative benchmark against which agreed terms may later be measured.

That transformation is conceptually significant. If FRAND is judicially ascertainable, it becomes at least conceivable to argue that payments materially exceeding that benchmark lack full justification. At the same time, courts are acutely aware that retroactive reassessment of freely negotiated licences risks destabilising settled expectations. Any theory of restitution must therefore confront judicial reluctance head-on.

Four preliminary constraints should be emphasised.

First, mere divergence from a later-determined FRAND rate may not suffice. The FRAND corridor seems to be getting wider, and explanations for variance abound – early-license discounts, litigation avoidance, comparable licenses, geographic scope, cross-licensing, or portfolio composition.

Second, timing matters. A rate may be supra-FRAND ab initio, may become supra-FRAND due to market evolution, or may cease to be so. Private-law remedies are not equally well suited to all scenarios.

Third, FRAND compliance is typically an implicit assumption underlying SEP licences, not an express guarantee. The question is therefore whether the economic exchange it embodies can or even if it should be, in specific cases, be recalibrated.

Fourth, SEP holders assert and make representations that their claimed rates are FRAND.  Where Standard Setting Organisations (SSOs) such as ETSI are involved, the SSO’s IPR policy provides the framework for the FRAND commitment (e.g. French law in ETSI’s case) but does not determine mandatorily the law applicable to the resulting licence agreement. The licence remains a contractual instrument subject to party autonomy, with its governing law typically negotiated between the parties.

In any event, for the purposes of addressing the issue discussed in this article and setting aside competition law discussions, reference is made to principles drawn from Portuguese contract law, which may find counterparts in the laws of several EU Member States.

First Potential Route – Error on the Basis of the Deal:

Generally, from a civil law perspective (as the Portuguese law), one possible – though ambitious – route lies in the doctrine of error on the basis of the transaction (in Portuguese: “erro sobre a base do negócio”, Article 252(2) Portuguese Civil Code; § 313(2) BGB; Article 1135 French Code Civil).

This doctrine applies where the error does not concern a subjective motive, but an objective circumstance regarded by both parties as essential to the meaning of the contract. It may be argued that compatibility with FRAND obligations forms part of that objective basis in SEP licensing: the licence is assumed to discharge the patentee’s competition-law and standard-setting commitments.

If a later judicial decision were to establish that the agreed rate lay outside any plausible FRAND rate (or even a range) at the time it was agreed, the licensee might contend that the contractual equilibrium was built on a false premise. Importantly, the outcome of such a scenario would not necessarily result in the contract being void. Portuguese law allows judicial modification of contracts, which opens the theoretical possibility of a surgical ex tunc adjustment of the price while preserving the licence itself.

That said, courts may be reluctant to accept that FRAND compliance constitutes the “basis” of the deal rather than a valuation risk knowingly assumed. The argument is strongest where information asymmetry is acute, and weaker where both parties are sophisticated and represented. Claims of bad-faith negotiation – such as deliberate non-disclosure of comparable licences, or selective disclosure of existing licenses – may therefore be decisive but will be evidentially demanding.

Second Potential Route – Usury and Partial Invalidity:

A second, more drastic route would frame the issue as one of validity rather than error. Most Civil Codes prohibit usury, defined as the exploitation of another party’s dependency to obtain unjustified benefits (e.g. Article 282 Portuguese Civil Code; § 138 BGB; Article 1143 French Code Civil).

Implementers are structurally locked into SEPs, and information asymmetry is not uncommon. In theory, a licensor who leverages this position to extract royalties far beyond any defensible FRAND benchmark could fall within the scope of usury. In such cases, courts may annul or adjust the price clause alone, replacing it with an equitable rate and ordering restitution of the surplus.

In practice, however, this path is likely confined to extreme scenarios. Courts reserve usury for cases of manifest exploitation, which is hard to demonstrate, and may be particularly hesitant where the parties are large, experienced market actors. As such, while doctrinally available, this route is unlikely to become a routine corrective for supra-FRAND outcomes.

Third Potential Route – Change of circumstances/Rebus sic stantibus and Judicial Caution:

A third possibility lies in doctrines allowing contract adaptation due to changed circumstances (e.g. Article 437 Portuguese Civil Code; § 313(1) BGB; Article 1195 French Code Civil). Here, the argument would be that both parties assumed the rate was FRAND, and that a subsequent judicial determination redefining the FRAND corridor constitutes a supervening change.

This reasoning faces significant resistance. German courts, for example, place strong emphasis on risk allocation and require that continued adherence to the contract be “unbearable” (unzumutbar). The mere fact that a rate later appears excessive is unlikely to meet that threshold. Portuguese law is more flexible, lacking an express “unbearability” requirement, but even there courts may be wary of reopening contracts based on valuation shifts inherent in the market.

Fourth Potential Route – Unjust Enrichment:

Unjust enrichment (literally, “enrichment without cause” in Portuguese law) offers perhaps the most intuitively appealing – but doctrinally constrained – route (e.g. Article 473 Portuguese Civil Code; § 812 BGB; Article 1303 French Code Civil). In systems such as Portugal’s, the concept of “cause” (causa) is understood functionally, leaving some interpretative leeway in the application of the enrichment regime. On that basis, one might argue that the normative cause of SEP royalties lies in compliance with FRAND obligations, such that any payment exceeding FRAND lacks sufficient justification notwithstanding the existence of a valid licence. However, unjust enrichment is of a subsidiary nature in many legal systems, meaning it is only available where other legal remedies do not apply.

One such scenario may arise where royalty rates were initially supra-FRAND but later became FRAND due to an increase in market value. In that situation, other civil-law institutes may prove difficult to apply, as they typically presuppose either a prospective modification of the licence or a declaration of invalidity of a clause (or part thereof) ab initio. If accepted despite the presence of a valid underlying licence – which can only be conceivable in more flexible systems – unjust enrichment could allow for restitution of supra-FRAND payments made in the past without altering the licence or the rates that subsequently became FRAND.

In German law in particular, the existence of a valid licence agreement generally constitutes a sufficient legal ground (Rechtsgrund), foreclosing enrichment claims unless the contract is first invalidated or adjusted. Even in more flexible systems, courts may resist using enrichment to circumvent contractual risk allocation, meaning that the case-specific circumstances are crucial to assess the viability of such a claim.

Accordingly, enrichment claims are more likely to succeed where paired with a prior finding affecting the contract’s validity or content.

Limitation Periods and the Actio Nata Debate:

Perhaps the most formidable obstacle to clawback claims is limitation. Licensors will argue that payments made years earlier are time-barred.

Civil-law systems recognising the actio nata principle offer a partial response. Under this theory, limitation runs not from payment, but from the moment the claimant knew or ought to have known of the right. A licensee may try to argue that knowledge of supra-FRAND overpayment arises only once a specialised court determines the FRAND range.

Courts may be sceptical of this argument, particularly where licensees had earlier opportunities to challenge rates, which will ultimately be a matter of evidence. Nonetheless, the tension between technical knowledge, litigation costs, and legitimate ignorance is one practitioners will recognise from real disputes.

Final thoughts:

Restitution remains the ghost in the machine of FRAND licensing especially in the context of SMEs and smaller licensees: conceptually present, doctrinally elusive, and judicially unsettling. Private law appears capable – at least in theory – of accommodating clawback in exceptional cases. Yet every route encounters resistance grounded in legal certainty and contractual stability.

The more interesting question may not be whether courts can adapt pacta sunt servanda to FRAND’s demands, but whether they will choose to do so. Until then, the most reliable risk-management strategy lies ex ante: careful drafting, transparency, and – where bargaining power allows – mechanisms such as adjustment or most-favoured-licensee clauses.

Whether ex post restitution will ever move from theoretical possibility to practical reality remains an open question – one that future test cases may force courts to confront.