The law of innovation must innovate

By Roya Ghafele

Private capacity disclaimer All webinar participants cited here spoke in their private capacities. Their comments reflect personal views and do not represent their employers, organisations or institutions. Affiliations are provided solely for identification.

Intellectual property protects innovation. Its institutions must be prepared to innovate too. We cannot keep assessing new technologies through familiar habits simply because those habits are comfortable. When the businesses, products and markets before a court change, the way we understand them must develop. That was my starting point at OxFirst’s webinar on Munich’s new FRAND Guidelines, and it remains the reason I find the initiative encouraging.

When I read the Guidelines, issued on 13 August 2026 by Munich I Regional Court’s 7th Civil Chamber, I saw a willingness to try. The court had set out its thinking on the economics of patent licensing in unusual detail. Welcoming that initiative does not mean endorsing every calculation. What deserves recognition is the decision to put a method into the open, where others can examine it, challenge it and help improve it.

That willingness matters because patent decisions reach beyond the parties in court. They can affect access to technology and the rewards available to those who develop it. They influence the conditions under which businesses build products and negotiate licences. Courts therefore confront questions with consequences for economic and technology policy. The analytical tools supporting their decisions should match the importance of those responsibilities.

Economics has a clear contribution to make. A court assessing a licence on fair, reasonable and non-discriminatory terms faces a commercial question: what payment has a defensible basis? The answer requires attention to transactions, market conditions and the contribution of the rights being licensed. Negotiating conduct still matters. Munich’s Guidelines make scrutiny of the proposed royalty more explicit, asking parties to bring evidence of value into the argument.

The significance lies in that demand for explanation. A price should come with reasons that others can follow. If an agreement is a useful benchmark, explain why. If its terms require adjustment, show the calculation. If a portfolio deserves a particular share of royalties, identify the evidence supporting that share. This creates a more useful conversation between legal rules and the economic activity those rules govern.

My presentation drew a comparison with Unwired Planet v Huawei. The English proceedings helped establish the importance of comparable licences and used the top-down analysis as a cross-check. Munich is developing related methods within its own legal framework, including its treatment of FRAND as a range. This is one way institutions innovate: they learn from experience elsewhere, adapt the useful elements and expose the adaptation to scrutiny.

The comparison also shows why borrowing a method is only the beginning. Commercial agreements differ. A licence may cover several standards, past use or reciprocal rights. A payment upfront does not immediately tell us the rate per device. Economic analysis must explain how those differences affect the inference drawn from a transaction. Otherwise, a method that looks disciplined can simply reproduce an unexplained price.

Streaming makes the need for new thinking immediate. Transferring a familiar device calculation to a service without testing its assumptions would leave the central economic question unanswered.

This is where criticism becomes productive. A published model lets parties identify the assumptions they dispute and supply better evidence. At the webinar, the Honourable Judge Dr Oliver Schön explained that the streaming illustration invited parties to bring facts before the court. The assumptions must therefore remain open to challenge. The invitation gives practitioners and experts an opportunity to improve the analysis through the circumstances of a real business, rather than arguing only at the level of principle.

Progress also requires a distinction between a court explaining its current approach and declaring the final answer. The Guidelines consolidate recent decisions. Their value will grow if they help future parties identify weaknesses, bring better facts and show where a different treatment is justified by evidence.

Better decisions require knowledge that no single profession holds. Judges bring knowledge of adjudication. Economists examine value and incentives. Licensing practitioners understand how agreements take shape. Their perspectives will not always align. That disagreement can reveal where an apparently clear approach overlooks something important and where research would make the greatest difference.

For OxFirst, bringing economics into intellectual property and competition law is the purpose of our work. A valuation should explain what drives the value of a right and how uncertainty affects the decision. Munich’s initiative creates room to advance that work. Its usefulness will depend on whether stronger evidence changes the analysis when the facts justify it. Innovation in patent law will earn its value through better reasoning about the technologies and businesses it governs.

Sources

Opening remarks and presentation at the OxFirst webinar, 21 October 2026: https://oxfirst.club/webinar-munich-sep-guidelines/#oxf-recording

Accompanying PowerPoint presentation on the Munich FRAND Guidelines and Unwired Planet v Huawei.

Who gets to choose the comparable licences in FRAND rate determination

By Roya Ghafele

Private capacity disclaimer All webinar participants cited here spoke in their private capacities. Their comments reflect personal views and do not represent their employers, organisations or institutions. Affiliations are provided solely for identification.

Comparable licences are central to FRAND rate determination for standard-essential patents. FRAND requires fair, reasonable and non-discriminatory licensing terms. Yet agreement on that starting point leaves a difficult question unanswered: who decides which agreements enter the comparison? At the OxFirst webinar I moderated on 21 October 2026, this became the sharpest point of disagreement. The debate showed that transparency is part of the economic problem itself. A valuation depends on the evidence available, the selection process and the opportunity to challenge what that evidence shows.

In my opening remarks, I argued that the law of innovation must itself be willing to innovate. Munich’s stronger engagement with valuation is encouraging because FRAND rate determination ultimately concerns how much should change hands. The economics of rate setting must therefore address the value behind the proposed payment.

Munich’s new FRAND Guidelines give particular weight to licences covering the same patent portfolio. That makes commercial sense as a starting point: actual transactions can reveal what parties have accepted. But a signed agreement does not explain its own price. Nor does a selection of agreements necessarily establish how representative that price is. Before a benchmark becomes persuasive, someone must assess what the agreement covers, the circumstances surrounding it and the reasons for including it.

Scott Hayden (Amazon), speaking in his private capacity, described the difficulty from an implementer’s perspective. If the patent owner chooses what to disclose, the prospective licensee cannot readily determine whether other, more relevant agreements exist. Hayden favoured early access, potentially restricted to external lawyers. He argued that seeing what comparable competitors had agreed could simplify negotiations. Disclosure, in his account, could help a commercial team identify an acceptable price and conclude a deal.

Nokia’s Clemens-August Heusch, equally speaking in his private capacity, offered a different account of negotiation practice. He questioned whether producing more agreements necessarily brought parties closer together. An implementer might continue finding reasons to distinguish each contract, or combine attractive provisions from several agreements into a package that nobody had actually accepted. His concern was that disclosure could reward delay rather than improve understanding. A limited set of useful agreements, assessed in their entirety, might provide a sounder basis for discussion.

These positions identify distinct risks. A licensor’s selection may leave the other party unable to test whether the sample reflects actual licensing practice. Conversely, a licensee may use a larger evidence set to construct an artificially favourable benchmark. The economics of rate setting must address both. The question is whether the selection and treatment of the agreements support a credible inference about value. Counting disclosed contracts cannot answer that question by itself.

This is why the commercial package matters to portfolio valuation. A payment may include several standards, settlement of past use or reciprocal patent rights. A lower rate may accompany different obligations, a different duration or a different volume of licensed products. Extracting one attractive term can misrepresent the bargain. Equally, relying on an exceptional agreement without explaining its circumstances can overstate what the wider market accepts. The comparison needs a consistent treatment of material differences on both sides.

Freshfields partner Corin Gittinger, speaking in his personal capacity, supported greater transparency early in proceedings. His argument was practical: parties could dispute whether an agreement was comparable more effectively once they knew its contents. From an economic perspective, that sequence is important. Access allows the parties to identify the relevant differences and explain their significance. Without it, they may spend time arguing about the possible existence or relevance of evidence that one side has not seen.

Confidentiality complicates the solution. Agreements can contain sensitive commercial information and restrictions sought by licensees as well as licensors. Heusch and Panasonic’s Andrew Yen, also speaking in his private capacity, both raised this difficulty during the discussion. Confidentiality therefore needs to be addressed as a feature of the licensing relationship. A useful approach must consider who needs access, which information matters to the comparison and how meaningful scrutiny can coexist with protection for unrelated commercial details.

Judge Oliver Schön’s response suggested how Munich’s procedure might develop. Speaking in his personal capacity, Schön maintained that agreements concerning the same portfolio offered the strongest valuation evidence. However, he explained that an implementer’s concerns about the selected comparables, together with an offer of binding arbitration, could prompt the patent owner to produce further agreements. He anticipated that more licences might appear in future cases. This was an explanation of the court’s approach, rather than a commitment to unrestricted disclosure.

My view is that earlier, protected access deserves consideration where it could resolve a material disagreement about the evidence. Its value should be assessed against the questions it enables the parties to answer. Can they identify a representative benchmark? Can they explain exclusions and adjustments? Can they distinguish the economic value of the portfolio from features of a particular transaction? Those are more useful tests than assuming that either maximum disclosure or unilateral selection will reliably produce the right result.

Better visibility may narrow the dispute even when it does not produce agreement. Parties can then concentrate on the assumptions that materially affect price, instead of debating whether they have seen enough evidence.

For economic experts, the task is to make the valuation reasoning traceable. That includes explaining why agreements enter the comparison, how their terms translate into comparable payments and which uncertainties remain. For negotiators, it means engaging with the package as a whole and acknowledging evidence that challenges their preferred position. The webinar demonstrated that credible FRAND rate determination requires a process for testing the evidence. Transparency matters because it gives that process a chance to work. That is fundamental to the economics of rate setting.

Sources

OxFirst webinar discussion on 21 October 2026: Munich SEP Guidelines: The Economics of FRAND

UK Government response on Standard Essential Patents

The UK’s SEP response: no new rate setting track, but a quiet shift in how FRAND gets priced

The UK Government published its much expected response to the 2025 consultation on Standard Essential Patents. Ninety three organisations and individuals responded, from rightsholders and implementers to SMEs, academics and practitioners across the UK, the United States, Europe and Asia.

The headline is restraint. The Government has dropped the idea of a Rate Determination Track in IPEC, left injunctions alone and declined to set up essentiality checks or new ADR services. Read the detail, though, and the response moves the pricing of FRAND licences earlier in a dispute, and partly out of the courtroom.

Key takeaways

  1. No Rate Determination Track in IPEC. The evidence did not show a workable route, and rate setting cannot be cleanly separated from validity, essentiality and the wider dispute.
  2. FRAND stays with the judges, but earlier. The Government will explore with the judiciary an early neutral evaluation of FRAND terms by a judge within existing proceedings in England and Wales, possibly through a pilot Practice Direction with an automatic stay and costs consequences for parties that do not engage.
  3. An IPO FRAND opinion service will be assessed for feasibility. It would be voluntary and not binding, giving an early view on whether an offer falls within a “plausible FRAND range”. The IPO will consult stakeholders this autumn.
  4. Voluntary disclosure of standard related patent information at the IPO, for future standards only and with no verification of essentiality. Mandatory disclosure or sanctions stay in reserve if take up is poor.
  5. FRAND licensing guidelines and guidance on good faith negotiation, plus an assessment of a targeted SEP specific approach before disputes reach court.
  6. No change to the law on injunctions, no IPO essentiality checking service and no new ADR service at this stage.

Five things worth a second read

The courts keep FRAND, but the clock starts earlier. The consultation asked whether a cheaper, faster forum should set rates. The answer is that the courts of England and Wales already have the power, under their case management rules and the Court of Appeal’s decision in Lomax v Lomax, to order early neutral evaluation. What is new is the intention to make that the normal pathway for SEP cases: a claim is issued, proceedings are stayed, a judge gives an early advisory view on FRAND terms, and a party that refuses to engage risks costs consequences. For an economist the practical effect is clear. The valuation evidence must exist at the start of a case, not at the end of a full trial. Comparable licences, top down analysis and portfolio strength will be argued months earlier than they are today.

A patent office opinion on price. The UK IPO has issued non-binding opinions on validity and infringement since 2005. Extending that model to whether a licence offer sits within a plausible FRAND range is a different kind of question, because it is about value rather than legal status. The response is candid that such a service would not decide liability, validity or infringement, and would not impose a global rate. It also floats something more significant: publication of anonymised outputs or benchmarks, subject to safeguards. If that happens, the UK will gain a public reference point for FRAND ranges that does not exist today.

FRAND is officially a range. The response cites WIPO’s June 2026 report on FRAND economics and accepts that a licensing scenario can yield a range of reasonable outcomes rather than one precise rate, and that comparables and top-down methods are complementary. That is a quiet but important statement from a government. It aligns policy with how valuation is done, and it means the useful question in a dispute is where the edges of the range sit and which evidence moves them. A range that is too wide guides nobody, so the methodology behind any opinion service will matter more than its existence.

Transparency through a register, not verification. The disclosure measure asks rightsholders to provide standard related patent information at grant, and for EP(UK) patents at renewal, through the IPO’s search service. Nothing will be checked for essentiality, and it applies to future standards only. The design is deliberately light, and it comes with a visible stick: if evaluation shows significant non-compliance, the Government will consider mandatory disclosure or sanctions. It will make SEPs easier to find. It will not tell a licensee which of them are truly essential, which remains the open issue.

The road not taken. Compare the European Commission’s proposed SEP Regulation, with mandatory essentiality checks, aggregate royalty determinations and a FRAND procedure at the EUIPO, which the Commission withdrew in 2025 for lack of foreseeable agreement. The UK has chosen the opposite architecture: courts rather than an agency, voluntary information rather than mandatory registration, and no change to the position settled in Unwired Planet, where an injunction follows only once an implementer refuses court determined FRAND terms. The Government describes the current framework as broadly adequate and the UK’s approach to injunctions as respected internationally.

 

What to watch this autumn

The IPO will consult on the opinion service this autumn. Work on the court anchored pathway has started, and a Practice Direction could move faster than anything that needs legislation. Voluntary disclosure may require legislation and depends on a legislative vehicle, so the Government itself expects a longer timeframe. In the meantime, the interesting work is in the detail: what evidence an early evaluation or an IPO panel will accept, how confidential licence data is handled, and whether a plausible range can be made narrow enough to be useful.

We plan to take these questions up with judges, industry and economists at the 15th IP and Competition Forum in Oxford on 13 to 14 January 2027.

The full response: https://www.gov.uk/government/consultations/consultation-on-standard-essential-patents-seps/outcome/government-response-to-the-consultation-on-standard-essential-patents

#SEPs #FRAND #patents #IPpolicy

Tesla v InterDigital: FRAND Valuation and the $32 Question

The UK Supreme Court’s decision in Tesla v InterDigital brings an important question in standard-essential patent licensing into sharper focus:

How should a court determine whether a rate offered through a licensing platform is actually FRAND?

Tesla challenged the terms available through Avanci’s 5G Vehicle platform. The High Court rejected Tesla’s licensing claims, and the majority of the Court of Appeal reached the same conclusion, with Lord Justice Arnold dissenting. The Supreme Court has now allowed Tesla’s appeal.

Importantly, the Court did not determine that Avanci’s $32-per-vehicle rate is non-FRAND. Nor did it finally decide that Avanci itself is directly bound by the ETSI FRAND undertaking.

Instead, it held that there is a serious issue to be tried as to whether the FRAND obligations of SEP owners remain relevant when their patents are licensed collectively through a platform. That is where the economic analysis becomes particularly interesting.

Platforms and FRAND

Licensing platforms can generate considerable efficiencies. Instead of negotiating separately with numerous patent owners, an implementer can obtain access to a large portfolio through a single transaction. Avanci’s aggregate portfolio covers a substantial proportion of the cellular SEPs relevant to connected vehicles and therefore offers a genuine one-stop licensing solution.

But aggregation does not make the FRAND question disappear. If anything, the greater the proportion of relevant patents brought together within a single licensing mechanism, the more important the aggregate price becomes. The next question is therefore straightforward:

Is $32 per vehicle FRAND?

Answering that question is considerably less straightforward. Several valuation approaches may contribute to the analysis.

Comparable licences

Comparable licence agreements are an obvious starting point. But comparability is rarely simple. Licensing agreements may differ. A platform licence raises an additional issue: part of its value may result from the reduction in transaction costs created by aggregation. There is also a broader economic question about the extent to which negotiated SEP licences can be treated as conventional market prices. SEP licensing does not necessarily take place in a transparent competitive market. An implementer may negotiate against the prospect of an injunction if no agreement is reached, potentially creating significant bargaining pressure. At the same time, many licensing agreements remain confidential, meaning that courts and market participants may have only a partial view of the terms agreed elsewhere. Comparable licences can therefore provide important evidence of negotiated outcomes.

Top-down analysis

A top-down approach provides another potential reference point. It starts by considering an aggregate royalty burden associated with a standard and then examines the share attributable to a particular portfolio.This approach may be particularly relevant where a platform itself aggregates a large proportion of the patents required to implement the technology.

The question is not simply what an individual portfolio is worth in isolation, but whether the overall royalty stack remains economically sustainable for downstream innovators.

Top-down analysis is not free from assumptions either. Much depends on how the appropriate aggregate royalty burden is established and how the relative strength and contribution of individual portfolios are assessed. It may nevertheless provide an important cross-check on other valuation approaches.

Does downstream use matter?

Another possibility is a use-based approach. The same standardised technology may make different economic contributions across different downstream applications. Connectivity in a smartphone is not necessarily economically identical to connectivity in a vehicle or an industrial device.

The argument underlying a use-based approach is that, where a technology is used more intensively or contributes more strongly to the value of a particular application, this may be relevant to the valuation exercise.

This does not mean that royalty rates should automatically differ across industries. It does, however, raise the question of whether the economic context in which a technology is embedded can simply be ignored.

Could hedonic price analysis help?

Econometric techniques such as hedonic price regression may provide another source of evidence. Hedonic analysis seeks to estimate how individual product characteristics contribute to observed market prices while controlling for other features.

Applied carefully, such an approach may provide insights into how consumers value functionality associated with connected technologies.

But econometric analysis also has important limitations. It cannot directly reveal the value of an SEP or mechanically determine a FRAND royalty.

What does FRAND mean?

Ultimately, the choice of methodology depends partly on how FRAND itself is understood. There is no methodology that is free from assumptions or limitations. A robust FRAND analysis may therefore need to draw on several approaches. The Supreme Court has not yet told us whether $32 per vehicle is FRAND. What it has done is open the door to a much deeper examination of how collective SEP licensing should be valued.

That may ultimately prove to be one of the most significant aspects of Tesla v InterDigital.

Watch the OxFirst discussion

OxFirst recently discussed these questions with Tim Powell of Morgan Lewis and Le Chen of Xiaomi in the webinar:

Putting a Price Tag on Global Patents: UK FRAND after Tesla

The discussion examines the implications of the decision for global SEP licensing, platform licensing and the determination of FRAND rates.

Watch the webinar:
https://oxfirst.club/webinar-tesla-frand/


OxFirst and FRAND valuation

OxFirst works at the intersection of intellectual property, economics and competition policy.

Our work includes the economic valuation of intellectual property, royalty-rate determination, licensing strategy and FRAND analysis.

Learn more about OxFirst’s work on IP and FRAND valuation:
https://oxfirst.com/ip-valuation-services/

Discuss a valuation or licensing matter with OxFirst:
https://oxfirst.com/law-economic-analysis/


Continue the conversation with OxFirst Club

Join us in Oxford and Munich

Two flagship Fora for the international patent, FRAND, licensing and competition community in 2027.

Rhodes House, University of Oxford

Oxford

15th IP and Competition Forum

13–14 January 2027
University of Oxford

Max-Joseph-Saal, Residenz Munich

Munich

16th IP and Competition Forum

2–3 June 2027
Residenz Palace, Munich

A senior international community across industry, law, economics and policy – continuing the conversation on IP, standards, licensing and competition.

Munich venue photo: Rainer Herrmann

For further analysis on FRAND, SEP licensing, patent valuation and competition economics:

Follow OxFirst on LinkedIn:
https://www.linkedin.com/company/oxfirstltd/

MUNICH COURTS GO ECONOMICS

Munich’s new FRAND Guidelines put economic evidence at the heart of comparable-licence analysis and top-down valuation.
The practical message for SEP litigants is simple: hire an economist early—well before the FRAND hearing.

German SEP litigation has historically placed considerable emphasis on the parties’ negotiating conduct—particularly whether the implementer was a “willing licensee.” Many cases could therefore be decided at that threshold without a detailed examination of the royalty. This is increasingly changing.

On 13 August 2026, the 7th Civil Chamber of the Munich I Regional Court completed and issued its FRAND Guidelines. Although the Guidelines are not binding law—and represent the position of that Chamber rather than the Munich court as a whole—they explain how the Chamber has approached, and intends to approach, FRAND disputes.

The Guidelines emphasise economic analysis and the use of party-retained experts in appropriate cases. At a high level, Munich’s methodological hierarchy resembles that adopted by the English courts in Unwired Planet v Huawei: comparable licences provide the primary evidence, while a top-down calculation is used mainly as a cross-check. Where suitable comparables or an established licensing practice do not exist, Munich may use the top-down approach to establish the royalty benchmark directly.

The institutional setting is nevertheless different. English courts can determine the terms of a global FRAND licence. Munich generally asks whether the SEP holder’s offer falls within a permissible FRAND range in the context of an infringement action.

Munich also takes a materially different approach to the production of comparable agreements. It permits the SEP holder to select the agreements it presents as the best comparables. However, that selection must adequately represent the holder’s actual licensing practice: the SEP holder must identify other agreements covering the standard and explain why they were not submitted. The Chamber generally will not compel production of a licence against the holder’s wishes. The Guidelines nevertheless provide unusually detailed guidance on what makes an agreement economically comparable.

Core comparability criteria

CriterionMunich treatment
PortfolioSame licensor and materially the same patent portfolio
RecencyNormally no more than five years old; older agreements require particular scrutiny
Standard scopeSingle-standard licences should ordinarily be compared with single-standard licences
Product categorySame or economically comparable products; relevant differences may sometimes be adjusted
VolumeDifferences may justify adjustments, including discounts of up to approximately 30% for exceptionally large volumes
Contract timingA first licence, a renewal and an agreement concluded under the threat of an injunction represent different economic circumstances
DurationLicence duration must be considered when determining and normalising the effective rate
Litigation contextRelevant, but litigation or injunction pressure does not automatically disqualify an agreement as a comparator
Cross-licenceUsually unsuitable unless the respective portfolios and reciprocal consideration can be valued independently
Geographic coverageGeographic adjustments are permitted only exceptionally
Party conductMay affect where the appropriate rate falls within the FRAND range

The message is clear: FRAND litigation before Munich’s 7th Civil Chamber increasingly involves economics, not merely formal negotiating conduct. Party experts can assist with unpacking lump-sum agreements, normalising rates, valuing contractual options, analysing past releases, assessing patent quality and testing top-down assumptions.

The Chamber does not intend to appoint its own experts to determine FRAND royalties, because it regards the ultimate reasonableness of the rate as a legal question. Economic evidence will instead be introduced and tested through the parties. For litigants appearing before Judges Dr Schön, Tözsér and Dr Schweyer, the practical implication is straightforward: prepare the economics early—and have your economist ready for the next FRAND hearing.