Richard Vary Partner Bird & Bird London, United KingdomIn recent years, UK first-instance trials have elevated their approach to calculating a FRAND royalty, adopting increasing levels of sophistication. This shift is partly due to the creativity of economic experts, who introduce adjustments that real-world negotiators might never consider. Their goal? To refine the figures in favor of their clients. While this ingenuity often leads to more precise results, it comes at a significant cost. UK FRAND trials now run each side an eye-watering €15 to €20 million, a stark contrast to the relatively modest expenses of German or Chinese FRAND trials, which cost a fraction of this amount.
The irony lies in the aftermath: both recent UK FRAND trials, despite their meticulous analysis, were appealed. At the appellate level, however, the process takes on a markedly different character. The Court of Appeal adopts a more pragmatic, albeit rough, approach. At first instance in InterDigital v. Lenovo, experts on both sides deployed an array of adjustments and corrections—either to enhance their methodologies or, cynically viewed, to obfuscate matters for the judge. The Court of Appeal felt that the first instance judge had indeed become confused in some areas. Yet, unlike the first instance trial, the appeal court does not have the luxury of directly engaging with experts to rework their analysis. Instead, the appellate judges imposed an arbitrary increase in royalties, openly acknowledging the lack of precision.
A similar approach seems imminent in the Apple v. Optis appeal. Paradoxically, this appellate method is closer to the commercial realism of deal negotiations than the intricate first-instance trials. Nevertheless, the disconnect between the detailed first-instance process and the more expedient appellate rulings can be unsettling for industry observers.
Adding to this irony is the criticism frequently leveled by English courts at German FRAND determinations. The appellate process that ultimately sets the rate in England exhibits no greater degree of accuracy than the German first-instance courts that the English courts are so ready to criticise.
This article delves into the increasingly detailed methodologies embraced by the English first-instance courts in their recent decisions.
When arguing FRAND cases before the English courts, both parties rely on expert economists or accountants to present evidence on portfolio valuation. The court is tasked with evaluating and choosing between the experts’ arguments. However, since valuation is a multi-step process, the court might reject one step of an expert’s analysis while accepting others. This can leave the court in uncharted territory, attempting to apply the expert’s methodology with altered data, often risking errors in the process.
In InterDigital v. Oppo, Mrs. Justice Joanna Smith utilized “hot-tubbing” for the experts—a method where both experts present together and respond simultaneously to the judge’s questions. For example, the judge might ask, “If I rule against you on step X, how would that affect your respective calculations?” This approach has also been embraced in FRAND arbitrations.
Additionally, Mrs. Justice Smith required parties to produce a “calculator” and a “roadmap.” The road-map outlined disputed points in the FRAND calculation, such as whether to use sales forecast data from the license date or actual sales data available at trial. The calculator, a spreadsheet, enabled the judge to input variables at each decision point and immediately see how they influenced the final figure—removing the need for manual calculations.
The English courts favor analyzing comparable licenses, which may include licenses granted by the SEP owner or those issued by other SEP holders. During disclosure, both parties typically gain access to relevant agreements, such as out-licenses and in-licenses.
While one-way licenses are easier to compare, many agreements involve cross-licenses with terms that vary: ad valorem or per-unit rates, running royalties, or lump sums. The licensed portfolios also differ in size and value since portfolios evolve over time. Each of these variables must be adjusted for during analysis.
To control for differences in patent portfolio strength, economists assess each portfolio’s value at the relevant date. When using comparable out-licenses, this means tracking the SEP owner’s portfolio strength over time. If the comparison involves licenses granted by other cellular SEP holders, the court evaluates the strength of their portfolios at the time the license was issued.
An alternative method is the top-down approach. Here, a party argues that the SEP owner holds a specific share of all SEPs in a standard and deserves a corresponding share of the value the standard creates.
The total value of a standard can be gauged by comparing customer willingness to pay for products with and without the standard. For instance, how much more will consumers pay for streaming movies in HD versus SD, or for a 5G device compared to a 4G one? Crucially, this value must be fairly divided between patent holders and implementers. An imbalance—where either party takes all the value—undermines incentives for innovation or adoption.
While some courts worldwide use the top-down approach as a cross-check, one English judge recently dismissed it as unhelpful. In InterDigital, the judge found that the approach yielded a figure exceeding those derived from comparable licenses. Additionally, the judge questioned whether this methodology constituted an “experiment” under English civil procedure, requiring extensive validation. Consequently, the top-down approach has fallen out of favor in England, though it remains influential in other jurisdictions.
Since portfolio value cannot be directly measured, economists rely on proxies. The simplest of these is counting unique SEP families in a portfolio.
Even this basic measure involves technical complexities, such as identifying declared SEPs, removing expired patents, assigning ownership, and categorizing patents by standard generation. The ETSI database alone is insufficient—it lacks detailed bibliographic and expiry data needed for precise calculations. Tools, like the patent portfolio analysis tool developed by Bird & Bird, help account for these variables and produce accurate family counts for any industry player at any given time.
Recognizing that not all SEP families hold equal value, economists also use proxies like forward citations. This method evaluates how often a patent is cited by others, adjusting for factors such as:
After establishing the patent portfolio strength for both portfolios in the cross-license at the time of entry, and how these have changed today, it becomes possible to adjust the comparable license’s rate to determine today’s FRAND royalty rate. To deconstruct a cross-license into two one-way licenses and derive the one-way rate, historical sales expectations for both parties at the time the cross-license was issued must be known.
Historic sales data for handsets and tablets are available from sources like Strategy Analytics or IDC. However, the analysis must rely on estimated sales expectations at the date the comparable license was entered, rather than actual sales informed by hindsight. These expectations can often be inferred from industry forecasts and internal communications from the time. For instance, Counterpoint has published company-specific sales forecasts for mobile handsets over the past decade.
With sales figures in hand, it’s possible to generate a rate chart showing a range of possible gross rate pairs that correspond to the net rate. For example, if the net rate is €2 per device, one party might pay €2.50 per device, while the other pays €1.50; or one could pay €3, while the other pays €1.
To solve for the correct rates, knowledge of each party’s patent portfolio strength at the relevant date allows calculation of the portfolio’s attributed value at the time the license was issued. Assuming equal bargaining power between parties during negotiations, the following equations help derive the one-way royalty rate:
R is the gross rate; S is sales and P is portfolio strength. As the sales expectations and the balancing payment will be known, this leaves a single unknown, the royalty rate of the first party, which can be solved for.
Not all patents notified as potentially essential to a standards-setting organization like ETSI turn out to be essential. Implementers may fear they will have no chance to challenge a patent owner’s portfolio during a rate-setting exercise, but in portfolio rate-setting arbitrations, implementers present technical arguments regarding the portfolio.
However, assessing the share of truly essential and valid patents requires consistent scrutiny across both the patent owner’s portfolio and the broader industry. Courts, as seen in Unwired Planet and Ericsson v. TCL, have attempted to account for “true essentiality” rates. These cases involved sampling studies by experts to estimate the industry-wide number of essential patents. Critics argued these studies overestimated the total, artificially diminishing Ericsson’s share.
A notable data source in this context is PA Consulting’s database, which evaluates the essentiality of patent families declared to standards like CDMA2000, UMTS, and LTE. This data supports two approaches:
While PA’s survey lacks the rigor of courtroom scrutiny, its consistent application across all patents ensures fairness. More intensive studies would be ideal but impractical outside of the highest-value cases. For technologies like 5G, PA’s reports rely on sampling rather than comprehensive analysis. Other surveys also have limitations. For example, in rapidly growing patent landscapes, time-to-grant disparities can skew results: companies with faster patent grants may appear to hold a higher share of a standard at a given point in time. Robustly addressing such biases, as required by English first-instance courts, adds further complexity and increases costs.
Recent FRAND case law has seen discounts applied for sales in specific regions, notably (i) China and (ii) emerging markets, with discounts ranging from 35 percent to 50 percent. Key examples include:
Such discounts have practical roots in industry practices. For instance, China’s NDRC decision against Qualcomm led to royalties being calculated on a reduced selling price. Similarly, Nokia’s agreement with China’s SAMR limited royalties to a percentage of the device sales price for Chinese customers.
In InterDigital v. Oppo, Oppo argued for discounts across all emerging markets, citing lower customer willingness to pay. However, a counterargument emphasized that equivalent Oppo phone models are often sold at similar prices in both developed and emerging markets. Granting a discount would thus provide an unjustified windfall for Oppo, rather than reflecting true market conditions. Oppo also sought discounts in regions with lower InterDigital patent coverage. This was contested by pointing out that manufacturing in China constitutes infringement, creating a “floor” for discounts tied to patent coverage. Given the strength of Chinese portfolios among most SEP holders, such arguments are unlikely to achieve major discounts.
Collecting royalties on past sales is not easy. Licensors face challenges due to limitation periods, and the practical impossibility of bringing damages claims on all of the patents in all of the countries where damages have accrued. This incentivizes implementers to delay agreements, to increase the number of their sales that become past sales. This leads licensors to offer significant discounts on past sales to secure agreements.
In IDG v. Lenovo (CA), the English Court of Appeal ruled that discounts on past sales are not FRAND. It considered such discounts as “sub-FRAND” factors that push overall payments below the FRAND range. Consequently, rates unpacked from earlier licenses (e.g., pre- IDG v. Lenovo decisions) can be uplifted to remove the sub-FRAND factor. However, it remains unclear how courts outside England would approach this issue or how the English courts would handle licenses agreed upon after the IDG v. Lenovo ruling. In theory, licensees knowing that they may be on the hook for full past royalties will stop seeking past discounts. In practice past discounts may be too embedded in industry practice to be overturned by one decision from one court.
In IDG v. Oppo (HC), both parties’ licensing experts acknowledged discounts on past sales but disagreed on their level and nature. In IDG v. Lenovo (CA), the Court of Appeal increased the unpacked rate from $0.24 to $0.30 to reflect discounts previously applied to past sales.
Licenses to one group of SEPs often include “standstill” provisions, deferring negotiations on other patent groups to a later date. This reflects the commercial reality that convincing a licensee to accept multiple licenses at once—for example, cellular SEP and video codec patents—can prolong negotiations and complicate internal approvals. A standstill offers breathing room, often until the next renewal date.
However, standstills raise questions during license unpacking. While a true standstill has little intrinsic value—delaying payments without affecting the total owed—it can shift more sales into the “past sales” category, which provides practical value to the implementer. The court must determine whether such standstills were included to artificially inflate the perceived value of a license to one group of patents or simply reflect a pragmatic strategy to avoid pushing licensees beyond their limits.
Ad valorem licenses (based on a percentage of the selling price) result in higher royalties for high-end devices, while per-unit licenses assign a uniform royalty regardless of price. This disparity often sparks debate:
In InterDigital v. Lenovo, Justice Mellor leaned toward the high-end sellers, stating: “I find it difficult to understand why the royalty paid for each of those phones should differ significantly or, for that matter, at all.”
The truth likely lies between these extremes. While high-end devices should contribute more on a per-unit basis, the relationship probably does not scale directly with ASP.
As FRAND litigation evolves, economic experts will continue to develop increasingly sophisticated methods for adjusting comparable licenses, adding layers of complexity to courtroom battles, and flow through into negotiation tables. The process of negotiating FRAND is not getting any easier.
Latest Articles Of The Month
Login
