Optis v Apple – UK Court of Appeal awards Optis $700m in blockbuster decision

On 1 May 2025, the Court of Appeal handed down its eagerly anticipated decision in Optis’ long-running SEP licensing dispute with Apple.  The Court of Appeal overturned the first instance decision and awarded Optis a FRAND royalty of $0.15 per device –an order of magnitude higher than the figure found by Marcus Smith J at the first instance.  This equates to a lump sum of $502m, plus interest of approximately $200m, over the term of the licence.

This decision will be welcomed by most standard essential patent (SEP) owners, most of whom have considered both the first instance decision in Optis v Apple and the first instance and appeal judgments in InterDigital v Lenovo as determining very low per unit royalty rates.

The Court of Appeal’s decision

The appeal related to a judgment by Marcus Smith J, handed down in June 2023, in which he decided that Apple should pay an annual lump sum of $5.13m per year for 11 years, totalling $56.43 million.  He reached this figure using a valuation method of his own creation, comprehensively rejecting the evidence of both sides’ experts.

The decision was appealed by Optis on 25 grounds, primarily arguing that the royalty determined at the first instance was far too low.  Optis’ primary case on appeal was that the valuation should be on the basis of the Optis licence with Google, as it considered this was the closest comparable.  Apple, meanwhile, supported the first instance decision (albeit suggesting some modifications to the approach taken by the judge).

The Court of Appeal’s judgment was split into two parts, with Birss LJ dealing with the royalty payment aspects and Arnold LJ dealing with the dispute on the non-financial licence terms and the interplay with a parallel judgment from the Eastern District of Texas, which awarded Optis $300m in damages.

The Valuation Case

Birss LJ identified several critical errors in the approach taken by the judge at the first instance:

1. Rejection of Expert Evidence: The trial judge wrongly dismissed both parties’ economic experts’ evidence, which Birss LJ considered to be “unfair and unjustified.” Birss LJ considered that both experts had taken an appropriate approach and had sought to assist the Court.
2. Approach to Comparable Licences: The trial judge erred in his methodology for evaluating comparable licences. Birss LJ affirmed that the best starting point will be the SEP owner’s own licences, but acknowledged that the implementer’s licences could also assist with the valuation exercise.
3. “Lump Sum” Method: Birss LJ rejected the trial judge’s novel approach of attempting to put the comparable licences on a common scale using a lump sum method. While he considered that the basic premise was entirely logical, Birss LJ concluded that there was a major flaw in the approach adopted by the judge below which he considered “disguises the true relationship” between the comparables.
4. Simple Averaging: Birss LJ found “no precedent or basis” for the judge’s decision to take a simple average of the lump sums he had established using his method, as it was an approach that neither parties’ expert had advocated and could not be justified in principle.
5. Distinction between types of ‘Hold Out’: The judge’s distinction between ‘legitimate’ and ‘illegitimate’ hold out was deemed “unhelpful,” as the job of the Court was to try and unwind any hold up or hold out. In particular, the distinction led to the judge to conclude that all of Apple’s behaviour fell within the scope of what he considered to be legitimate, which then led him to erroneously including almost all of Apple licences, which he then averaged in his calculation.

Having identified these issues, Birss LJ considered whether to order a retrial.  However, he considered that it was unnecessary due to (i) the concessions made by the parties, which considerably simplified the scope of the issues; and (ii) the fact that the “core of the judge’s approach below” could be applied in a way that was not now criticised by either party.  On this basis, Birss LJ considered that it would be possible for the Court to reach a just conclusion as to a FRAND rate and took the following approach:

• Identifying the best comparables. Birss LJ considered these to be the Optis licence with Google, and the four Apple licences that gave the highest implied per unit rates (Ericsson, InterDigital, Nokia, and Sisvel).
• Putting these comparables on a common scale using the implied per unit rates calculated by the experts.
• Recognising that Marcus Smith J’s rate ($0.01-0.02 per device) was too low as there was no justification for a per unit rate below that implied by the four selected Apple licences;
• Concluding that the rate implied by the Optis / Google licence ($0.27) would yield too high a rate, but that the rates implied by the four Apple licences were too low;
• Applying the English law principle of the ‘broad axe’ approach (i.e. taking a reasonable approximation) to select a rate falling between the per unit rates implied by the comparables.  Birss LJ concluded that “realistic options are … $0.20, $0.15 or $0.10”.
• Using a top-down cross-check to confirm which of these options was most appropriate.

Birss LJ ultimately concluded that $0.20 would be too high, while $0.10 would be too low, and set a FRAND rate of $0.15 per device, noting this represented a reasonable balance between the comparables while recognizing “there would appear to be degrees of hold up and hold out involved” in the comparables.

Non-Financial Terms

Arnold LJ criticised the first instance judge for taking an “extraordinary and undesirable” approach towards determining non-financial licence terms. The process had initially followed the usual approach, with parties having prepared a composite draft, in which many of the terms were agreed.  However, the judge then inexplicably provided his own short form ‘Court-Determined Licence’ that ignored most agreed terms and introduced new ones not sought by either party.

Arnold LJ found this procedurally improper, with Apple pragmatically conceding that the Court should set aside the short-form licence and return to the composite draft. Arnold LJ clarified that the first instance judge had misunderstood the Court’s role in relation to non-financial terms and highlighted the importance of allowing parties in a FRAND dispute to seek to agree the commercial terms, as far as possible, between themselves.

A further significant dispute in the appeal concerned the interplay between English proceedings and a $300m damages award Optis secured against Apple in the Eastern District of Texas. Arnold LJ rejected the first instance approach of requiring Optis to repay any excess between the Texas judgment and English award, citing concerns of international comity. Instead, he determined that “the least-worst solution” was to treat the US award as a floor for royalties payable under the English-determined global FRAND licence, should the English Court have determined that the royalty payable by Apple was less than $300m.

Implications of the decision

While this decision is unlikely to be viewed as being groundbreaking, in the way the Unwired Planet case was, it will almost certainly be hugely important.  It adds significantly to the growing body of English case law relating to SEP disputes, and provides possibly the clearest guidance to date on how the English Court should approach FRAND actions.  This should help to reduce the time to judgment and overall cost of these cases going forward.