2 Mutually beneficial terms – the role of the FRAND undertaking and valuation methodologies

Topics covered in Part 2:

  • The FRAND undertaking – legal basis and key principles

  • Common types of FRAND disputes

  • Why reliable valuation is central to resolving FRAND disputes

Building on Part 1, which has outlined the importance of the FRAND concept in balancing incentives for innovators and implementers in the context of standardization, this part of the Report sets out the legal basis and key principles of the FRAND undertaking, identifies common types of disputes surrounding FRAND and shows why these disagreements can often be mitigated by reliable valuation.

As discussed in Part 1, the inclusion of the FRAND undertaking in SDO IPR policies may contribute to balancing the interests of standard innovators and implementers and mitigating competition law (1)See Section 1.3.2 above. concerns. Although the legal foundation, interpretation and enforcement of FRAND undertakings differ across SDOs and jurisdictions, the central objective – and the economic rationale that underpins it – is broadly consistent. The FRAND undertaking seeks to safeguard incentives: with respect to innovators, to develop and contribute valuable technologies for inclusion in the standard; and with regard to implementers, to adopt the standard and develop products and complementary assets that users value. In this way, FRAND aims to align licensing terms with the value created when innovation and implementation are combined, increasing the quality and value of both the functionalities that standards enable, and the products that implement them.

In practical terms, the FRAND undertaking provides legal assurance to both sides. It gives implementers confidence that they can obtain access to the essential patented technology needed to produce standard-compliant products and services. At the same time, it preserves innovators’ rights over successful technologies, supporting continued investment in R&D and participation in the development and evolution of standards. In settings where lock-in arises from standard adoption, the FRAND undertaking provides a framework that can reduce uncertainty associated with ex post opportunism and support licensing, investment and market participation.

The FRAND undertaking is a legal obligation. Its interpretation therefore depends on the relevant SDO policy, applicable laws and the approach taken in the specific jurisdiction. This Report focuses on valuation and the economics of FRAND, which broadly do not vary with legal interpretation. It will therefore not go into detail concerning the legal interpretation of FRAND in different jurisdictions and potential different outcomes.

However, as a basis for the subsequent economic discussions, this section briefly summarizes several of the legal issues that may arise in FRAND cases, the interpretation of which can vary between jurisdictions.

As noted above, the inclusion of FRAND undertakings in SDO IPR policies may be motivated, in part, by competition law considerations. In some jurisdictions, consequently, the interpretation of FRAND has been approached primarily through the lens of competition law, while in others it is treated predominantly as a matter of contract law (for example, as a third-party beneficiary arrangement). In practice, both perspectives may be relevant. A detailed analysis of these approaches is beyond the scope of this Report. It is sufficient to note that these two foundations serve different functions: the contractual undertaking supports standardization by providing assurance that innovators and implementers can make relationship-specific investments before detailed licensing terms are agreed, with a reduced risk of subsequent appropriation. Competition law, by contrast, focuses on protecting the competitive process and is directed at constraining certain forms of SEP licensing conduct, particularly where companies hold a position of dominant market power.

A practical implication is that a breach of a contractual FRAND undertaking does not necessarily imply a competition law violation, and vice versa. Contract and competition law apply different tests and provide different remedies. As a result, a royalty may be considered to be outside the FRAND range under an undertaking made at an SDO if understood contractually, without meeting the separate – and often higher – threshold for an antitrust infringement, which typically requires additional elements, such as dominance and competitive harm. (2)Specifically in the context of FRAND and SEPs, see Optis v. Apple [2025] EWCA Civ 552 (link), para. 119: “A finding that a degree of hold-up or hold-out was involved in a given real negotiation is not a finding that either party has acted in an unlawful manner. It is simply a finding that that outcome cannot be taken as the FRAND rate. It may be close to it or far away, and if a view can be taken about the degree of hold-up or hold-out involved that might shed some light, but these are different issues.”

Naturally, differences in applicable competition laws, as well as differences in the wording and in the applicable laws governing different IPR policies, affect the outcome of what is considered to be FRAND. (3)For example, in the UK Supreme Court’s assessment of non-discrimination, it was relevant that ETSI had previously considered and rejected the imposition of a “most-favourable licence” clause in the undertaking, which requires that the lowest rate agreed with one party is offered to all other parties. See Unwired Planet v. Huawei [2020] UKSC 37 (link) , para. 116 et seq. However, some differences in terminology can be superficial. For instance, some policies use “FRAND,” as is the case in ETSI’s policy for the governance of cellular standards. Others use “RAND,” such as the ITU’s policy for video codec standards. Generally, these terms are treated as substantively the same undertaking in practice. (4)This report uses the term FRAND.

2.1.2 Key principles of FRAND

Without going into the details and differences between jurisdictions, some general key principles in the interpretation of FRAND can be observed.

Committing to “fair and reasonable” terms is typically framed as the price a willing licensor and a willing licensee would agree, both acting in good faith and with appropriate information, taking account of the value of the patent rights. (5)See, for example, Georgia-Pacific Corp. v. United States Plywood Corp., 318 F Supp. 1116, 6 USPQ 235 (SD NY 1970), May 28, 1970. See also Optis v. Apple [2025] EWCA Civ 552 (link) , para. 119: “The relevant context is the court’s task of identifying a rate which is FRAND. As a matter of principle that is a rate which a willing licensee and willing licensor would agree. This is an idealised legal standard.”; and Intellectual Property Office, Patents Act 1977, United Kingdom (amended up to 2025). Courts may use valuation to determine fair and reasonable royalty terms that are FRAND, either to directly award for terms, or to assess whether a party’s offer or counter-offer is FRAND. (6)For example, in litigation between Samsung and ZTE, the High Court (England & Wales) and Chongqing Court both determined a royalty they considered FRAND for a global license between the parties; their valuation differed. The Munich Regional Court did not set a FRAND rate, but compared ZTE’s offer to its own FRAND assessment. It considered ZTE’s offer was FRAND. See Samsung v. ZTE [2026] EWHC 999 (Pat),  ZTE v. Samsung, The People’s Republic of China Chongqing First Intermediate Court Civil Judgment, Yu 01 Min Chu No. 905; B, Munich Regional Court, 30.04.2026, 7 O 64/25 – Bestimmung der Angemesenheit von FRAND-Lizenzrate (link) 

The concept of a willing licensor and a willing licensee is often assessed against the parties’ behavior during the negotiations. It permits good faith disagreement over terms but seeks to distinguish between parties that genuinely seek to reach an agreement within a reasonable range and those that strategically refuse to engage, delaying negotiations or insisting on unreasonable positions. Parties are generally expected to engage with each other in good faith, and their conduct – including the exchange of offers, the provision of information and the use of legal remedies – may influence whether outcomes are consistent with a reasonable range. The assessment of such conduct is context specific. Jurisdictions differ in how far they treat negotiation conduct as a distinct requirement. Courts and government authorities in a number of jurisdictions have developed practical guidance on negotiation steps and information exchanges that may be relevant to assessing good-faith behavior and the appropriateness of particular remedies. (7)For instance, see discussions in Huawei v. ZTE Deutschland GmbH, CJEU [2015] Case C-170/13 (link) ; Unwired Planet v. Huawei [2020] UKSC 37 (link) ; and Federal Court of Justice, 05.05.2020, KZR 36/17 – FRAND-Einwand ; Japan Patent Office, Guide to Licensing Negotiations Involving Standard Essential Patents ; METI, Good Faith Negotiation Guidelines for Standard Essential Patent Licenses (31 March 2022 ); Tokyo District Court IP Divisions, Guidelines for Standard Essential Patents Judicial Mediation (SEPJM) (January 2026) .

It is generally recognized that there may be a range of reasonable prices for a given portfolio and set of circumstances, rather than a single rate. Commercial negotiations do not normally converge on a single point estimate that reflects “the” FRAND rate in all cases. (8)This does not mean that the range is unbounded, and a “range” may take on different meanings depending on the context; e.g., a FRAND range may have a different meaning in the context of offers (where there may be limited or asymmetric information) from the range that is considered in a FRAND rate determination (which may be due to factors such as measurement issues). Terms may also be structured in different ways, while providing broadly the same economic value. For example, licenses with terms that provide the same economic value may involve a lump-sum payment, a running royalty per unit or a cross-license; provided the overall value exchanged is broadly comparable once the differences in their scope, timing, risk allocation and other commercial terms are taken into account. (9)See Section 3.2 below.

The “non-discriminatory” element seeks to ensure that similarly situated implementers are not placed at a competitive disadvantage. For example, it prevents a vertically integrated licensor from using market power to impose disadvantageous terms on downstream rivals. (10)Swanson and Baumol offer a principle for determining non-discriminatory royalties when the licensor is vertically integrated, known as the efficient component pricing rule (ECPR). The ECPR provides a pricing rule that is competitively neutral in downstream markets. The ECPR states that the price of the bottleneck input (here, the license fee for the SEP) should equal the price that the vertically integrated firm charges for the final product using the intellectual property (IP), net of the incremental cost to that firm of all other inputs used to produce the final good. Swanson, D.G. and W.J. Baumol (2005). Reasonable and non-discriminatory (RAND) royalties, standards selection, and control of market power. Antitrust Law Journal, 73(1), 1–58. It generally does not require identical licensing terms and differences may be justified where they reflect relevant commercial factors (for example, where one license covers more licensed units, or those units are sold in countries with materially different conditions, or the license lasts longer, includes more patents or the payment structure allocates risk differently between the parties)

FRAND undertakings typically do not specify a predetermined royalty rate or a fixed set of payment terms. Neither, typically, do they specify a predetermined process for negotiating or calculating FRAND terms. Instead, they operate as a principles-based undertaking that licensing terms will satisfy the core concepts that underpin FRAND terms and negotiations. This allows flexibility in licensing arrangements but also leaves scope for disagreement over what constitutes FRAND in practice.

2.2 Common types of FRAND disputes

FRAND disputes arise when parties disagree, in the context of a particular negotiation, about what the FRAND undertaking requires. Such disagreements may concern different dimensions of the undertaking and may, depending on the applicable law, procedural framework and the parties’ own choices, arise separately or together in the same proceedings. Against that background, four recurring categories of dispute can be identified. Each is connected, directly or indirectly, to the value of the license agreement to the parties, and each may be informed, clarified or narrowed by reliable valuation and economic analysis.

2.2.1 Disputes about price and payment structure

These disputes concern whether proposed payment terms fall within a FRAND range. Parties may disagree about the royalty level, the royalty structure (for example, a running royalty per unit, a percentage of a price or a lump sum) and the treatment of cross-licenses or other forms of non-monetary consideration.

These disagreements often reflect different views about (i) the value of the SEP portfolio, (ii) the appropriate benchmarks for reasonableness and (iii) the adjustments needed to compare alternative structures on a cash-equivalent basis.

Disputes about price are often resolved through mediation and arbitration. The WIPO Arbitration and Mediation Center (WIPO AMC) regularly manages disputes concerning FRAND royalty rates. Under the WIPO Rules, parties and neutrals are free to determine applicable valuation methodologies. See also WIPO ADR for SEP/FRAND Disputes and WIPO ADR Options for FRAND Dispute Management and Resolution . Court determinations of FRAND terms are not equally common across jurisdictions, reflecting differences in legal frameworks, procedural design and the claims or remedies pursued by the parties. Box 2.1 provides examples of approaches in different jurisdictions. In some jurisdictions, most notably the United Kingdom and China, courts have, in a number of cases, determined FRAND rates. In the United States, FRAND rates have also been determined by courts, but such determinations remain relatively uncommon. In continental Europe, by contrast, courts have generally approached FRAND disputes according to the applicable legal framework, including the Huawei v. ZTE line of case law in the European Union (EU), focusing in particular on the parties’ conduct and on whether proposed terms are FRAND-compliant or evidently non-FRAND, rather than themselves determining a complete FRAND rate or royalty schedule. These differences are also relevant when considering the judicial sources available on valuation methodologies. (11)Judicial decisions from jurisdictions in which courts regularly determine FRAND terms, including global portfolio rates, may provide more detailed discussion of valuation methodologies than decisions from jurisdictions in which courts more often focus on conduct and FRAND compliance without themselves setting a complete rate or royalty schedule.

Box 2.1 Disputes about royalty level and structure (case examples)

United Kingdom: Global portfolio rates and pricing of past sales. In Unwired Planet v. Huawei, the UK Supreme Court confirmed that an English court may determine FRAND terms for a global portfolio license, backed by an injunction on UK patents if the implementer refuses that license. In InterDigital v. Lenovo, the Court of Appeal addressed historic unlicensed sales, holding that using heavy discounts and limitation periods to reduce the price for past use was not FRAND, as it risks rewarding delay and weakening negotiation incentives.

China: Increasing use of court-set rates, including global rates. In OPPO v. Nokia (Chongqing), the court set global FRAND rates for Nokia’s cellular SEP portfolio. There are also domestic FRAND determinations, such as the Supreme People’s Court judgment in the Advanced Codec Technologies v. OPPO dispute.

United States: Court-set FRAND rates are relatively uncommon. Most SEP disputes in the United States are resolved without court FRAND determinations. However, US courts ruled on some of the first and most cited FRAND determinations – Microsoft v. Motorola, Innovatio and TCL v. Ericsson.

Germany and the Unified Patent Court (UPC): In application of the Huawei v. ZTE framework, both the German national courts and the German divisions of the UPC have generally assessed FRAND as a defence in infringement proceedings, rather than as an independent rate-setting exercise. Under German FRAND-Einwand case law, courts focus on the parties’ negotiation conduct and whether the SEP holder’s offer is so clearly outside the acceptable FRAND range that it cannot be regarded as a serious FRAND offer, in which case injunctive relief may be unavailable. The UPC’s early Panasonic case law similarly applies Huawei v. ZTE in a conduct-based manner. Although implementers have sought FRAND-rate or FRAND-terms determinations by counterclaim before the UPC, the Mannheim Local Division dismissed those counterclaims and did not itself set a FRAND royalty rate.

Interaction between jurisdictions with divergent views on the global FRAND rate. An SEP licensing dispute between Samsung and ZTE was heard in several jurisdictions. Three of them opined on the FRAND rate, reaching different conclusions. In the UK, the High Court awarded a global FRAND license valued at USD 392 million for a five-year term. In China, the Chongqing Court valued a six-year global license at USD 731 million. The Munich Regional Court did not set a rate, but concluded the upper bound of a FRAND offer was USD 798.6 million for a five-year term. The reasons reflect differences in the evidence presented, how its relevance was assessed, and the interpretation of the previous license between the parties.

Sources: Unwired Planet v. Huawei [2020] UKSC 37 (link) ; InterDigital v. Lenovo [2024] EWCA Civ 743 (link) ; OPPO v. Nokia [2021] The People’s Republic of China Chongqing First Intermediate Court Civil Judgment, Yu 01 Min Chu No. 1232 ; Advanced Codec Technologies v. OPPO [2022] ZGFZMZ Nos. 907, 910, 911, 916, 917 and 918, SPC (link) ; Microsoft Corp. v. Motorola, Inc. No. C10-1823JLR (W.D. Wash., 2013) (link) ; In re Innovatio IP Ventures, LLC, Patent Litigation,   921 F. Supp. 2d 903 (N.D. Ill. 2013)(link) ; TCL Communication Technology Holdings Ltd. v. Telefonaktiebolaget LM Ericsson, No. 8:14-cv-00341, (C.D. Cal. Dec. 21, 2017), amended and superseded (C.D. Cal. Sept. 14, 2018) (link) , rev’d in part, vacated in part 943 F.3d 1360 (Fed. Cir. 2019)  CJEU, Huawei Technologies Co. Ltd v. ZTE Corp. and ZTE Deutschland GmbH, Case C-170/13, judgment of 16 July 2015 (link) ; Federal Court of Justice ,  05.05.2020, KZR 36/17 – FRAND-Einwand ;  Federal Court of Justice,  24.11.2020, KZR 35/17FRAND-Einwand I, Federal Court of Justice , 27.01.2026, KZR 10/25FRAND-Einwand III, especially headnotes and paras. 50–53, 60–63; UPC Local Division Mannheim, Panasonic v. OPPO, UPC_CFI_210/2023, decision of 22 November 2024 (link) ; the judgment states that OPPO filed a counterclaim for determination of a FRAND royalty for the EP territory. Same judgment, Panasonic v. OPPO, UPC_CFI_210/2023, decision of 22 November 2024 (link) , dismissing the FRAND counterclaim / FRAND-rate determination request. Samsung v. ZTE, [2026] EWHC 999 (Pat) (link) ; ZTE v. Samsung, The People’s Republic of China Chongqing First Intermediate Court Civil Judgment, Yu 01 Min Chu No. 905; Munich Regional Court, 30.04.2026, 7 O 64/25 – Bestimmung der Angemessenheit von FRAND-Lizenzraten (link) .

2.2.2 Disputes about discrimination

Non-discrimination disputes concern whether similarly situated implementers are being offered sufficiently comparable terms.

As outlined in Part 1, FRAND does not necessarily require identical terms for all licensees. Differences in circumstances can justify differences in terms, particularly where licensees operate in distinct product markets or where the same technology creates different value in different uses. The dispute typically turns on (i) whether the implementers are comparable for the relevant purpose and (ii) whether differences in terms are objectively justified and consistent with lawful, non-exclusionary competition.

These disputes therefore rely on reliable valuation and economic analysis, not a simple comparison of whether the terms are “broadly similar”.

Whether two firms are “similarly situated” often depends on the economic facts: whether products compete and, if so, how intensely, how the standardized functionality contributes to value in each use, and whether differences in license scope, timing or risk allocation and so forth change the cash-equivalent value of the proposed terms.

Some argue that differential treatment is discriminatory treatment if royalty rates differ substantially among similarly situated licensees after controlling for other factors. Others have argued for a higher standard, based on whether such rates would distort downstream competition – for example, by raising a rival’s costs relative to a vertically integrated licensor, or by conferring a cost advantage on one implementer that cannot be explained by objective commercial differences. Economic analysis can help to assess (i) the relevant product market and competitive conditions, (ii) which differences between implementers are economically material, and (iii) whether differences in terms are likely to affect competitive outcomes rather than merely reflect normal commercial variation.

In practice, non-discrimination disputes revolve around three questions, as set out in Box 2.2.

Box 2.2 Discrimination disputes (case examples)

Are firms similarly situated? The similarity of the parties’ situations should be assessed by reference to economic facts, such as product type, geographic footprint and sales scope. For instance, in TCL v. Ericsson (United States), major global smartphone manufacturers were treated as relevant comparators to TCL, whereas firms with a narrow geographic focus on China (CoolPad) and India (Karbon) were not.

Must similarly situated firms agree the same rate? In both the United Kingdom and Germany, courts have generally rejected a hard-edged approach under which any lower rate offered to another licensee automatically makes an offered rate discriminatory. The analysis instead focuses on whether differences are objectively justified and, in particular, whether they could distort competition. The jurisdictions may differ in emphasis – for example, in the weight given to competition-law framing and to information and procedural expectations – but not in requiring identical rates in all cases.

Does structure matter, or only the value of the terms? Courts often focus on whether different payment structures deliver broadly comparable cash-equivalent value, rather than requiring identical structures (for example, English FRAND decisions including Optis v. Apple). Some decisions, however, treat structure and transparency as more significant in particular contexts. The Karlsruhe Higher Regional Court in Datenpaketverarbeitung considered it relevant that the licensor did not adequately explain why it accepted a lump sum from one party but offered a running royalty to another In India, competition proceedings have also treated aspects of royalty structuring as potentially discriminatory, particularly where they may affect downstream competitive conditions.

Sources: TCL Communication Technology Holdings Ltd. v. Telefonaktiebolaget LM Ericsson, No. 8:14-cv-00341, (C.D. Cal. Dec. 21, 2017), amended and superseded (C.D. Cal. Sept. 14, 2018) (link) , rev’d in part, vacated in part 943 F.3d 1360 (Fed. Cir. 2019) , cert. denied 141 S.Ct. 239 (2020); Unwired Planet v. Huawei [2020] UKSC 37 (link) ; Optis v. Apple [2025] EWCA Civ 552 (link) ; Federal Court of Justice, 05.05.2020, KZR 36/17 – FRAND-Einwand (link) ; Higher Regional Court Karlsruhe, 30.10.2019, 6 U 183/16 – Datenpaketverarbeitung (link) ; Higher Regional Court Düsseldorf ,  22.03.2019 ,  (link)  2 U 31/16  – Improving Handovers (link) , paras 413–414 and 420. For Indian competition proceedings, see Micromax v Ericsson, Case No. 50/2013 , and Intex v. Ericsson , Case No. 76/2013 (link) .

2.2.3 Disputes about negotiation conduct

Conduct disputes focus on how the parties negotiate – for example, whether offers and counteroffers were made in a timely way, whether information was shared sufficiently to evaluate proposals or whether either side delayed unreasonably. While these disputes are formally about process, they often matter because negotiation conduct can affect bargaining leverage and therefore influence the terms ultimately agreed or imposed.

Reliable valuation and economic analysis can help to resolve these disputes, even though they are about “conduct” not “prices.”

In many cases, whether a party is acting as a willing licensor or willing licensee depends in part on whether its proposals and responses were anchored to plausible FRAND terms. For example, an offer that is materially outside a reasonable valuation range may support an inference that a party is using process strategically rather than seeking agreement. Conversely, a refusal to engage, prolonged delay or insistence on terms that remain unsupported by any credible valuation may indicate unwillingness, even if the party asserts that it is “still negotiating.” Such issues, however, can be addressed by valuation, which contextualizes parties’ conduct, providing information on whether particular positions or remedies are proportionate in light of the value and terms at stake.

A recent development in conduct disputes is the strategic use of parallel proceedings in multiple jurisdictions. This is particularly evident where parties litigate in – or seek to avoid – courts that are prepared, in some circumstances, to determine FRAND terms for global portfolio licenses (notably the United Kingdom and China and, in more limited procedural contexts, the United States). Box 2.3 highlights disputes where courts assessed whether the parties’ global litigation strategy – including seeking injunctions in parallel proceedings – was consistent with negotiating a FRAND license in good faith.

Box 2.3 Global litigation strategy and good-faith FRAND negotiations (case examples)

United Kingdom/United States; United Kingdom/UPC/Germany – cross-border interim measures in global SEP disputes (Ericsson v. Lenovo; Amazon v. InterDigital). In the United Kingdom, courts have considered interim measures in global SEP disputes. For example, in Ericsson v. Lenovo, the Court of Appeal granted an interim license declaration   holding that, pending a final FRAND determination, a willing SEP holder would offer a temporary license so the dispute can be resolved without the prospect of an injunction distorting negotiations. In the United States, the Federal Circuit sent the case back to the district court after holding that it was wrong to refuse, at the threshold stage, to consider an anti-suit injunction that would stop enforcement of foreign SEP injunctions while the US court considered alleged FRAND-related breaches. Further, in Amazon v. InterDigital, the UK proceedings seeking interim or adjustable RAND licensing relief led to parallel applications for anti-interim license relief before the UPC Mannheim Local Devision and the Munich Regional Court.  Both courts granted ex parte anti-interim-license injunctions restraining Amazon from seeking or continuing UK interim license relief, or equivalent measures, that would effectively prevent InterDigital from pursuing or continuing patent infringement proceedings before the UPC or German courts, or from enforcing any resulting judgments or measures. The UK High Court  subsequently granted an ex parte anti-anti-suit injunctive relief (AASI) intended to prevent InterDigital from seeking further anti-suit relief before the UPC Mannheim Local Division, the Munich Regional Court, or any other court to restrain Amazon from carrying on its claim to substantive, final RAND relief in the UK proceedings.

China/India – anti-suit and anti-anti-suit injunctions (InterDigital v. Xiaomi). In China, the Wuhan court granted an anti-suit injunction ordering InterDigital to withdraw or suspend the Indian proceedings and restraining it from seeking injunctions or FRAND determinations elsewhere while the Chinese case proceeded. In India, the Delhi High Court granted relief restraining Xiaomi from enforcing the Wuhan anti-suit injunction against InterDigital in relation to the Indian proceedings (an anti-anti-suit/anti-enforcement order), so that the Indian case could continue.

Sources: Lenovo v. Ericsson [2025] EWCA Civ 182 (link) ; Ericsson v. Lenovo., No. 24-1515 (Fed. Cir. Oct. 24, 2024)  (link)  Amazon. v InterDigital [2025] EWHC 3170 (Pat) (link) ; InterDigital v Amazon, UPC_CFI_936/2025, Local Division Mannheim, order of 30 September 2025; Munich Regional Court, 24.09.2025, 21 O 12112/25, clarified by order of October 1, 2025 [Klarstellungsbeschluss ] pursuant to sections 319, 938 (1) German Civil Procedure Code (ZPO); Amazon v. Interdigital [2025] EWHC 3170 (Pat) (link) ; InterDigital v. Xiaomi [2021] Delhi High Court CS(COMM) 295/2020 (link) .

2.2.4 Disputes about the appropriate licensing level

Parties may also disagree about the appropriate point in the supply chain at which SEP licenses should be taken. For example, disputes can arise over whether licensing should occur at the component level (for example, where a Wi-Fi chipset manufacturer takes a license) or at the end-product level (where a laptop manufacturer takes a Wi-Fi SEP license).

In simplified terms, two positions are often described (12)On SEP valuation and the level of licensing, see, for example, Heiden, B., J. Padilla and R. Peters (2021). “The value of standard essential patents (SEPs) and the level of licensing. American Intellectual Property Law Association Quarterly Journal, 49(1). Llobet, G. and D. Neven (2023). Investment and patent licensing in the value chain. Journal of Competition Law & Economics, 19(4), 527–555.:

  • License-to-all” approaches emphasize the availability of licenses on FRAND terms at different levels of the supply chain, including to upstream component suppliers. Proponents often argue that this can reduce transaction costs, provide clearer pricing signals earlier in the supply chain and support downstream access.

  • Access-to-all” approaches emphasize access to the standard on FRAND terms for all implementers, even where licensing takes place at a particular level of the supply chain, such as the end-product level. Proponents often argue that licensing at the level where the standard is implemented and sold can be administratively simpler and more closely aligned with how value is realized in products and services.

These issues raise several economic questions, including the costs of negotiating licenses, as transaction costs can differ depending on the number of firms at each level in the supply chain, whether the licensing level and potentially different royalty base affects the reasonable price that parties agree to, the risk of double-dipping, and how the benefits and costs of a license are passed on to firms and consumers downstream. Box 2.4 provides examples.

Box 2.4 Disputes about licensing level across jurisdictions (case examples)

India – upstream component versus downstream device (Micromax v. Ericsson; CCI proceedings). In 2014, the Competition Commission of India (CCI) ordered an investigation into Ericsson’s SEP licensing practices following complaints including Micromax v. Ericsson. Among other issues, implementers argued that royalties should be linked more closely to the SEP-practising component, rather than calculated on the selling price of the end device. In 2023, the Delhi High Court held that, in this context, the Patents Act prevailed over the Competition Act and quashed the CCI proceedings. In 2025, the Supreme Court dismissed the CCI’s Special Leave Petition against that judgment, leaving the Delhi High Court’s position undisturbed.

United States – no antitrust duty to license at the component level (FTC v. Qualcomm). In 2020, the Ninth Circuit reversed the district court’s ruling and held that Qualcomm’s refusal to license rival chip suppliers did not, on the facts, amount to an antitrust violation. The court further indicated that disagreements about licensing level under the FRAND undertaking were better addressed, in that case, through contract and patent law rather than under the antitrust framework.

Germany/EU – connected vehicles and tiered supply chains. In a series of connected car disputes, the vehicle manufacturer argued that licenses should be taken by upstream suppliers, while the SEP holder pursued licensing at the vehicle manufacturer level. In Germany, one first-instance court found infringement and rejected the car manufacturer’s FRAND defence; another first-instance court referred questions to the CJEU concerning, among other matters, the level of licensing. The litigation and related proceedings illustrate how licensing level questions can become central in multi-tier industries, where components are supplied by specialist firms but value is realized at the end-product level. The parties settled in 2021.

Sources: Federal Trade Commission v. Qualcomm Inc., 969 F.3d 974 (9th Cir. 2020) (link) . For German connected car disputes, see Regional Court Mannheim,18.08.2020, 2 O 34/19Lizenz in der Wertschöpfungskette, as to the preliminary stay of enforcement, see Higher Regional Court Karlsruhe,12.02.2021, 6 U 130/20Wurzelsequenzordnung, as to the CJEU referral, see Regional Court Düsseldorf, 26.11.2020, 4c O 17/19 ; for the Indian competition proceedings, see Case Competition Commission of India v Monsanto Holdings Private Limited, Telefonaktiebolaget LM Ericsson & Micromax Informatics Limited. (link) 

2.3 Why reliable valuation is central to resolving FRAND disputes

Valuation methodologies can be used to establish both the economic value that is generated when innovators and implementers of technology standards combine their contributions and the prices that parties would negotiate in light of that value and those specific circumstances. (13)How parties negotiate terms that divide the value created by their cooperation is underpinned by the economics literature on bargaining. In particular, on bargaining power and complementary values, see Nash, J.F. (1950). The bargaining problem. Econometrica, 18(2), 155–162; Shapley, L.S. (1953). A value for n-person games. In Kuhn, H.W. and A.W. Tucker (eds), Contributions to the Theory of Games (Vol. II), Annals of Mathematics Studies, 28. Princeton University Press, 307–318.

2.3.1 The role of valuation

Ensuring that license negotiations and FRAND determination are informed by reliable valuation helps both to resolve disputes and to deter them.

2.3.1.1 Resolving FRAND disputes

Reliable valuation is most obviously relevant to disputes about royalty level and structure. In essence, these disputes arise because the parties have incompatible views about what constitutes a reasonable range of terms. To the extent that robust valuation methods can identify, or at least narrow, a plausible FRAND range, they can reduce disagreement and help parties reach agreement.

In practice, however, valuation evidence in SEP disputes has often been contested. Opposing parties sometimes present estimates that diverge widely, even where each party claims to have applied recognized methodologies. This divergence does not necessarily imply bad faith: even when everyone is acting honestly and using the same data, results can still differ significantly because analysts must make many legitimate choices along the way, and those choices change the answer. (14)The uncertainty created by the variance in these choices is referred to as “nonstandard errors.” Standard errors capture randomness from the data-generating process (sampling noise): if a new sample is drawn, the estimate would alter. Nonstandard errors capture randomness from the evidence-generating process: different analysts, working in good faith, will make different defensible choices – what to include/exclude, how to define variables, which specification to use, which adjustments are “reasonable.” Those choices can change estimates substantially. Menkveld, A.J. et al. (2024). Nonstandard Errors. Journal of Finance, 79(3), 2339–2390. This is exacerbated by the intrinsic features that make the value of SEP licenses uncertain: (15)See Section 1.3.3 above. complexity, incomplete information and uncertainty about the strength and relevance of particular patents, about comparability of benchmarks and about how value should be apportioned.

Divergence is exacerbated when analyses are difficult to interpret or verify. If key assumptions and modelling choices are not transparent, it becomes challenging for opposing parties – and for courts, arbitrators and mediators – to identify the source of disagreement, test sensitivity to inputs or determine which conclusions are robust. In such circumstances, decision-makers may face a practical difficulty: even if they find weaknesses in both sides’ analyses, they must still determine terms. Where valuation evidence does not provide clear, verifiable anchors, outcomes can become more dependent on judgment and discretion, which can in turn increase uncertainty and the possibility of subsequent disputes. The impact of uncertainty and ambiguity is exemplified by the dispute between Samsung and ZTE. Considering broadly similar evidence, three courts – the High Court (England & Wales), Chongqing Intermediate Court, and Munich Regional Court – reached different conclusion about the FRAND terms. Those differences relate to the evidence presented, their assessment of it, and interactions with other issues regarding the scope of the license. (16) Samsung v ZTE [2026] EWHC 999 (Pat) ZTE v Samsung, The People’s Republic of China Chongqing First Intermediate Court Civil Judgment, Yu 01 Min Chu No. 905; and Munich Regional Court, 30.04.2026, 7 O 64/25 – Bestimmung der Angemessenheit von FRAND-Lizenzrate (link)  n.

For these reasons, reliable valuation should be understood not only as producing a plausible numerical estimate, but also as using methods that are transparent, reproducible and appropriately grounded in evidence. Valuation that meets these standards reduces the scope for incompatible expectations and supports more efficient resolution of disputes.

2.3.1.2 Deterring disputes

Valuation also has a bearing on whether disputes arise in the first place. Good valuation not only helps to resolve disputes, it also makes disputes less likely to occur.

Parties typically form their negotiating positions by considering the range of outcomes that a court or arbitrator might impose if negotiations fail. If adjudicated outcomes are reasonably consistent and predictable, parties are more likely to agree terms commercially. Conversely, if outcomes are highly uncertain, parties may be encouraged to “wait and see", to litigate strategically or to delay in the hope of obtaining more favorable terms from a particular forum. (17)Economic models of bargaining and settlement predict that, when parties have limited outside options and disagree about expected adjudicated outcomes – especially under asymmetric information – negotiations are more likely to involve delay or impasse; conversely, more predictable and transparent valuation anchors narrow the range of plausible outcomes and increase the likelihood of agreement. See Rubinstein, A. (1982). Perfect equilibrium in a bargaining model. Econometrica, 50(1), 97–109; Fudenberg, D. and J. Tirole (1983). Sequential bargaining with incomplete information. The Review of Economic Studies, 50(2), 221–247; Admati, A.R. and M. Perry (1987). Strategic delay in bargaining. The Review of Economic Studies, 54(3), 345–364; and Bebchuk, L.A. (1984). Litigation and settlement under imperfect information. The RAND Journal of Economics, 15(3), 404–415.

Uncertainty can therefore increase the incentive to dispute even where the parties privately recognize that an agreement exists within a reasonable range. More predictable valuation frameworks can reduce these incentives by narrowing the plausible range of outcomes and making the implications of key assumptions easier to test. They can also support better investment decisions by both innovators and implementers, because expected licensing costs and revenues become less uncertain.

2.3.2 An overview of valuation approaches

There are three broad approaches that are commonly discussed – two of which can be applied to SEPs. Each offers a different way to anchor the assessment of FRAND terms, and each has strengths and limitations.

2.3.2.1 Market-price-based approaches

Market-price-based approaches assess the reasonableness of proposed terms by reference to prices that arm’s-length parties have agreed in comparable license agreements. The aim is to identify a “going rate” for similar rights in similar circumstances and to adjust for differences in scope, geography, timing, portfolio strength or other relevant factors.

This approach aligns with valuation practice in many markets, where the price of an asset is benchmarked against prices for similar assets. In SEP licensing, however, comparables can be difficult to apply effectively. Licenses are often confidential, may bundle multiple rights and may reflect settlement dynamics, cross-license value or litigation risk. In addition, comparable agreements are not necessarily FRAND – as the comparable contracts may also have been agreed by parties without credible outside options to apply competitive pressure on prices.

Nonetheless, valuation by reference to comparable contracts is widely used in practice and frequently features in litigation and arbitration. See Part 3 for a detailed discussion of this approach.

2.3.2.2 Value-based approaches

Value-based approaches assess the reasonableness of licensing terms by reference to the economic value of the benefits enabled by the technology. In practical terms, this means focusing on forward-looking expectations: the future benefits, revenues, cost savings or avoided losses that access to the standardized technology is expected to deliver and how that expected value should be shared between innovators and implementers.

In principle, value can be described in two related ways:

  • absolute economic value: this is the value of the total benefits enjoyed by users and implementers (for example, improved performance or new functionality); and

  • incremental value: this is the value of the technology relative to the next-best alternative that would have been available absent standardization, recognizing that, once adopted, standardization eliminates alternatives.

Identifying both the economic value and the incremental value that depends on licensed technology helps to clarify the stakes of negotiation. The economic value establishes the scale of the benefits to which both parties have contributed. The incremental value over alternatives is one of the factors that influences the specific terms – or price – that the parties agree, which determines how the total value will be divided between them.

Value-based approaches are particularly relevant in innovation-driven sectors because development costs and economic value are not closely linked. A technology can be expensive to develop and yet create limited value, or relatively inexpensive to develop and yet create very significant value. A valuation framework that focuses only on cost risks rewarding effort rather than impact.

At the same time, value-based approaches require evidence and assumptions that can be challenging to verify. In this Report, two value-based approaches are discussed in detail: the bottom-up method (Part 4) and the top-down framework (Part 5).

2.3.2.3 Cost-based valuation approaches

Cost-based approaches assess reasonableness by reference to costs and a margin – often described as “cost-plus". In regulated sectors with high fixed costs and stable demand, cost-based approaches can provide administrable benchmarks for ensuring cost recovery and incentivizing investment.

In SEP licensing, however, cost-based approaches are generally less informative. Research and development is inherently risky: many projects fail and returns are earned from the subset that succeeds. A benchmark that focuses on the costs of successful inventions alone can understate the level of returns required to sustain overall R&D effort. Consequently, the appropriate “plus” – the margin required to reward risk, reflect opportunity cost and incentivize ongoing improvement – often becomes the central question, and it cannot be determined from costs alone.

Cost evidence can nevertheless play a role in some settings. It may be relevant as part of a broader assessment of commercial reasonableness; for example, by offering insight into whether proposed terms are clearly inconsistent with sustainable participation or by providing a cross-check for extreme outcomes. However, cost-based methods rarely provide a sufficient standalone basis for determining FRAND terms in innovation-led standardized technology markets.

2.4 Summary

FRAND disputes have several recurring dimensions: disagreement over royalty level and payment structure; disagreement over non-discrimination and comparability; disputes about whether parties negotiated in good faith; and disputes about where, within a supply chain, licenses should be offered.

Across these categories, the same practical challenge recurs. FRAND is expressed at a high level, but settlement and adjudication turn on concrete issues: the scope and strength of the portfolio, the value contributed by standardized technology, the relevance of market benchmarks, and whether particular offers and counter-offers are consistent with that evidence.

The concept of FRAND has an important economic rationale. It is intended to preserve incentives on both sides of standardization: for innovators to contribute patented technologies for inclusion in a standard, and for implementers to invest in developing, manufacturing and bringing standard-compliant products and services to market. In practice, that objective requires an assessment of the value that standards help to create, the contributions that innovators and implementers make to that value, and how specific license terms divide it between licensor and licensee. Carrying out this assessment in a disciplined way depends on reliable valuation.

Reliable valuation helps to bridge the gap between an undertaking to the principle that licenses are FRAND and the agreement to terms that are FRAND in practice. Where parties can anchor their positions with robust methodologies and verifiable inputs, the range of plausible FRAND outcomes narrows and negotiations are less likely to be distorted by strategic delaying tactics or entrenched positions. Where clear valuation anchors are lacking, courts, tribunals and mediators are left to exercise broader evaluative judgment when settling FRAND disputes. In such circumstances, the prospect that a resolution may provide a “mechanical” middle-ground outcome would not only risk failing to serve the markets that rely on FRAND license terms, but can also encourage parties to adopt more extreme positions, rather than converging on terms that balance their interests.

The next three parts of this Report address the principal valuation methodologies put forward in SEP license negotiations, whereby the order follows an economic logic, which does not reflect any preference for, or assessment of, the practical utility of the methodologies. Part 3 examines the comparable-licenses approach in detail. Parts 4 and 5 then turn to value-based frameworks – bottom-up and top-down approaches, respectively – and explain how they can be applied, and tested, to better inform negotiations and resolve disputes more effectively.