Updates on SEP royalty rate-setting cases in China
Managing IP is part of Legal Benchmarking Limited, 4 Bouverie Street, London, EC4Y 8AX
Copyright © Legal Benchmarking Limited and its affiliated companies 2024

Accessibility | Terms of Use | Privacy Policy | Modern Slavery Statement
Sponsored content

Updates on SEP royalty rate-setting cases in China

Sponsored by


With the number of Chinese standard-essential patent cases set to increase, Guanyang Yao of Liu Shen & Associates analyses two judgments that are expected to be used as precedents in royalty rate-setting lawsuits

At the end of 2023, two standard-essential patent (SEP) royalty rate-setting judgments were issued in China – the Oppo v Nokia case, by the Chongqing First Intermediate Court, and the ACT v Oppo/Vivo case, by the Supreme People’s Court – which show the legal approach of the Chinese judicial system to handling royalty rate-setting disputes.

Each case has its own characteristics. For example, in the Oppo v Nokia case, a 5G global royalty ‘top’ for telecommunication SEPs was decided and the 5G global royalty rate of Oppo phones based on Nokia’s 5G SEP portfolio was calculated with a top-down approach, while in the ACT v Oppo/Vivo case, several comparable licensing agreements were relied on as the basis to calculate China’s royalty rate of Oppo phones based on ACT’s six audio coding and decoding SEPs.

In view of the significant influence of these cases in China, it is highly likely that the judgments will be regarded as precedents for SEP royalty rate-setting lawsuits. Therefore, it is important to understand the major legal points in these cases.

Trial cycle of SEP royalty rate-setting disputes

In general practice, the trial cycle in China for patent infringement litigation is one to one and a half years for the first instance and another six to ten months for the second instance. An SEP royalty rate-setting lawsuit takes longer, since several complicated legal issues are involved, including:

  • Patent portfolio strength;

  • The comparability of various licensing agreements;

  • The calculation method for royalty rates;

  • Infringement mapping between a standard and a patent; and

  • The evaluation of the behaviours of the parties during the licensing negotiation.

Therefore, two to four court hearings are usually held for such cases, which extends the trial cycle.

The Oppo v Nokia case was formally established on July 12 2021 and the first-instance judgment was issued on November 28 2023. Oppo and Nokia reached a cross-licensing agreement on January 24 2024. This first-instance proceeding took two and a half years.

The ACT v Oppo/Vivo case was formally established on November 27 2018, with the first-instance judgment issued on November 22 2021 and the second-instance judgment issued on December 12 2023, a total duration of five years. Aside from the complexity of the case, the measures to control the COVID pandemic may have been a factor leading to this extended timeframe.

Setting of a 5G SEP global aggregate royalty burden

The highlight of the Oppo v Nokia case is the setting of a global royalty top for 5G telecommunication SEPs. This judgment mainly relies on a report made by the plaintiff's economic expert, Professor Gong Jiong of the Department of Economics at the University of International Business and Economics, Beijing. The report adopted the commonly used ‘feature price model’ in economics to calculate the contribution of 5G telecommunication standards to mobile phone prices relative to 4G telecommunication standards in this case, and then calculated the 5G SEP global aggregate royalty burden.

Firstly, Professor Gong used the feature-based price model method to calculate the degree of consumers’ willingness to pay more for 5G phones than 4G phones; i.e., the 5G coefficient. A sample of mobile phone sales data for regression analysis of the feature price model was used. To estimate the trend of 5G coefficient changes over the next three to five years, Professor Gong used an exponential function to fit the trend of 5G coefficient changes over time. Furthermore, Professor Gong used an exponential function model to predict the period after Q4 2021 and ultimately calculated the average 5G coefficient.

Secondly, Professor Gong calculated the average selling price ratio of 4G and 5G phones over the next three to five years based on an exponential function.

Thirdly, based on the 4G top 6%–8% and 5G coefficient, Professor Gong calculated the 5G SEP global aggregate royalty burden according to the following formula:

5G SEP aggregate royalty burden = (4G SEP aggregate royalty burden + 5G coefficient) * ratio of the average selling price of 4G and 5G mobile phones.

Finally, the 5G SEP aggregate royalty burden was calculated as 4.341%–5.273%.

The court believed that this 5G top is reasonable, since there is evidence corroborating this 5G top. One is a report from the China Academy of Information and Communications Technology pointing out that the 5G SEP aggregate royalty top should not be higher than the 4G top. Another is empirical research showing that the attractiveness of 5G telecommunication standards to smartphone consumers is relatively limited, and the contribution of 5G technology to the value of smartphones is not significantly improved compared to 4G technology, since we are still in the early stages of the introduction of 5G technology into the smartphone field.

The above reasonings were challenged by economic experts Dr. Jorge Padilla and Associate Professor Tang Mingzhe, who were hired by the defendant, on several bases, including:

  • The adoption of the feature price model;

  • The selection of a Canalys database;

  • The problem of selecting the time period for sales data samples;

  • The issue of not calculating a sales weighted average on the estimated 5G price appreciation data;

  • Not using a Kennedy transformation;

  • A missing variable issue; and

  • Using incorrect methods to predict future 5G price appreciation issues.

The court provided explanations to the above objections one by one and finally rejected all of them and summarised that in the absence of a consensus on the applicability of specific economic methods in the economics field, it is believed that the plaintiff’s calculation methods are rigorous and reasonable.

Application of a regional discount rate

The Oppo v Nokia case appears to be the first judgment in China to accept a regional discount for SEP royalty rates. Both parties provided their own theory for different regional discounts.

The plaintiff created three regions, mainly based on the first and second regions divided by the HEVC (High Efficiency Video Coding) Advance Patent Pool, and combined with the latest available per capita gross domestic product (GDP) data of various countries and regions. The countries and regions with per capita GDP greater than or equal to $20,000 were selected as the first region, Mainland China as the second region, and other countries and regions as the third region. The first region applied a benchmark rate, while the second and third regions offered a 61.42% discount.

The defendant adopted two regions. The first region (countries and regions except Mainland China) has no discount, and the second region (Mainland China) has a certain rate discount.

The judgment held that due to an uneven distribution of licensed patents by patentees in global market regions and economic developments, there are significant differences in regional markets globally. This inevitably causes the value distribution of licensed patents to vary in different regions. Therefore, it is necessary to partition the global market regions to a certain extent and determine relative regional discounts based on existing evidence.

Furthermore, in view of the patent portfolio layout in various countries and regions, the discount provided in licensing negotiations, and country-specific circumstances, the judgment upheld that the regional discount of 61.42% claimed by the plaintiff, compared with the defendant's claim, better reflects the regional economic development status, the patent layout of the parties, and other factors, as an average number calculated based on regional discounts in various judgments, such as Unwired Planet v Huawei, TCL v Ericsson, and Huawei v Conversant.

Evaluation of the strength of Nokia’s 5G SEP

How to determine an SEP’s strength in judicial practice is a key issue when the royalty rate is to be determined. In this case, both parties had their own positions.

The plaintiff calculated the strength of Nokia 5G patents based on the ratio of the number of Nokia patent declarations to the total number of SEP declarations, using the data collected from various third-party companies, such as:

  • A Clarivate report based on SEPs declared to the European Telecommunications Standards Institute (ETSI) from March 1 2016 to December 31 2021;

  • A China Academy of Information and Communications Technology report showing 5G patents declared by each ETSI member from January 1 2017 to December 31 2021;

  • A 2022 IPLytics report showing Nokia's declaration of essential 5G SEP families in Europe, the US, and China; and

  • A Questel report on the actual essentiality analysis of 5G SEPs.

The defendant believed that Nokia's patent strength should be comprehensively determined based on the number of granted patents, the number of proposals approved, and the actual essentiality rate of patents, by mainly citing:

  • Updated research results on the essentiality of declaring 5G SEPs, released by ReyB Company and Amplified Company;

  • The 5G Standard Essential Patent Report, Version B2, released by PA Consulting; and

  • The Essential Patent Analysis Report for LTE and SAE (4G Wireless Standard) Declaration Standards as of June 30, 2009, released by Fairfield International Resources.

Although, as shown in those reports, there are multiple dimensions in determining patent strength (such as the number of standard contributions, the number of contributions approved, the number of technical project leaders, and the number of chairpersons of relevant working groups), the judgment focuses more on counting the number of patent declarations and maintains that other dimensions cannot be directly analysed from a qualitative and quantitative perspective. The judgment asserts that those reports have many problems, such as:

  • Production time that is too long to reference;

  • Relatively small sample sizes and questionable data rationality; and

  • A lack of exposure and verification of technical analysis processes and conclusions.

At the same time, the judgment accepts that the prerequisite for using declaration numbers for strength evaluation is that all SEPs have the same quality. The judgment does not rule out that if evidence can be provided to prove the quality of the specific core patents, there is room for adjustment on the strength evaluation.

However, in this case, the defendant did not provide sufficient evidence to prove that its patent quality is significantly higher than the industry’s average quality. Furthermore, the judgment stated that as time goes by, it is necessary to consider the changes in patent strength, which is based on the number of patent declarations.

Calculation of royalty rates for 5G multimode mobile phones

The calculation of a 5G royalty rate is at the core of the Oppo v Nokia case, with a comparable agreement approach and the top-down approach both applied.

The formula of the top-down approach is as follows:

Nokia 5G multimode phone royalty rate = (5G SEP global aggregate royalty burden * Nokia's 5G SEP strength * 5G SEP value in a multimode phone) + (4G SEP global aggregate royalty burden * Nokia's 4G SEP strength * 4G SEP value in a multimode phone) + (3G SEP global aggregate royalty burden * Nokia's 3G SEP strength * 3G SEP value in a multimode phone) + (2G SEP global aggregate royalty burden * Nokia's 2G SEP strength * 2G SEP value in a multimode phone).

The formula of a comparable licence approach is as follows:

Nokia 5G multimode phone royalty rate = (5G SEP aggregate royalty burden * Nokia 5G patent strength proportion during the agreement term * 5G technology value proportion 50%) + (4G multimode royalty rate (calculated through the 2018 Oppo agreement unpacking) * 4G multimode technology value proportion of 50%).

The judgment held that both approaches are reasonable – with small differences in numerical ranges, and even crossing over each other – thus both can adequately perform cross-verification. Considering that the plaintiff's final claim was calculated using the comparable agreement approach, which is more favourable to the defendant, the court determined the 5G multimode mobile phone licensing fee range and a single unit licensing fee of $1.151 based on the plaintiff's claim of using the comparable agreement approach.

In the ACT v Oppo/Vivo case, a royalty rate of six Adaptive Multi-Rate Wideband SEPs was calculated based on comparable licensing agreements. The judgment provided guidance on how to select comparable agreements, with consideration of the following factors:

  • The environment for licence negotiation, such as transaction background and the condition of both parties, and whether the agreement is voluntarily reached or under psychological coercion (such as ongoing litigation or threatening to sue, or an injunction threat);

  • The similarity of licensors, judged from various aspects, such as the business model, the business scope, and the relationship between the licensor and the licensee;

  • The similarity of licensed patents, such as whether patents in comparable agreements are consistent with, or at least cover, the patents in the dispute, and whether they have the same or similar quantity and quality; and

  • The similarity of licence terms, including the calculation of licence rates, the licence scope, the licence term, the licence type, and the licence fee payment method.

In view of this, the court found a comparable licence agreement generally complying with the above factors, which was reached in a normal licensing negotiation atmosphere and that objectively and reasonably reflects the market value of the six patents involved in this licensing agreement. The royalty rate for the six patents was calculated by dividing $6 million by sales amounts of 760 million phones, resulting in $0.008 per unit. Of course, the prerequisite is the assumption that all SEPs have the same quality.

Evaluation of FRAND terms regarding behaviour during licensing negotiation

In the ACT v Oppo/Vivo case, an important issue is to evaluate which parties comply with their fair, reasonable, and non-discriminatory (FRAND) obligations based on their behaviour during the negotiation process. The judgment provides considerations for the licensor and the licensee.

With regard to the licensor’s behaviours, the following main factors should be considered:

  • Whether it directly files a royalty rate-setting litigation without issuing a written notice of infringement to the implementer;

  • Whether the implementer’s licensing request has been explicitly rejected;

  • During the negotiation process, whether the implementer is repeatedly threatened by the filing of lawsuits or injunction applications;

  • Whether it interrupts negotiations without justifiable reasons;

  • Whether it refuses to disclose necessary patent information, such as the number of SEPs and exemplary claim charts; and

  • Whether it refuses to disclose to the implementer the calculation basis or method concerning royalty rates.

With regard to the licensee’s behaviour, the following main factors should be considered:

  • Whether it responds within a reasonable time after receiving a written notice of infringement from the patentee;

  • Whether it provides a positive response on acceptance or rejection within a reasonable time after receiving the patentee’s offer;

  • If the offer made by the patentee is deemed unreasonable, whether the licensee submits a counter-offer or a licensing proposal to the patentee in a timely manner, or whether a corresponding licensing fee for the counter-offer has not been deposited in a timely fashion;

  • Whether it delays or interrupts negotiations without justifiable reasons;

  • Whether it refuses to sign a confidentiality agreement with the patentee; and

  • Proposing obviously unreasonable licensing conditions to the patentee.

In view of the above, the judgment concluded that ACT and Oppo had faults that rendered the licensing agreement not concluded, therefore each party shall bear 50% responsibility for the losses caused thereby. Compared with previous cases, this judgment provided more protocols for both parties to comply with, thus creating relatively strict requirements for the parties in patent licensing negotiations.

In the 4G era, Chinese companies were less involved in the standard-setting process, while in the 5G era, the gap between Chinese companies and international telecommunication companies has significantly narrowed. Therefore, an increase in SEP disputes reflects not only the developments in 5G technology in China, but also the importance that Chinese manufacturers attach to their legitimate rights and interests.

In this context, Chinese courts are actively participating in the handling of international SEP disputes, demonstrating extensive experience of Chinese courts accumulated in recent years, starting from the IDC v Huawei case in 2011. On the basis of a short trial cycle, the professional knowledge of intellectual property judges, and an active attitude of courts, it is believed that more SEP cases will be initiated in Chinese jurisdictions in the future, and Chinese courts will gradually showcase their capabilities and confidence in resolving SEP disputes to the world.

more from across site and ros bottom lb

More from across our site

Four sources reveal which tools they have been using – or building – to help them with a range of tasks from invention generation to claim sufficiency
Managing IP reveals Wednesday's highlights, including a discussion on how AI is helping lawyers improve their "gut instinct" trademark decisions
Managing IP reveals Tuesday’s highlights, including an illuminating discussion celebrating women in the workplace and the challenges that remain
Dana Northcott, INTA’s 2024 president and associate general counsel for Amazon's IP team, talks about her work for the association
Managing IP reveals highlights from the INTA Annual Meeting, including law firms’ diversity and ESG concerns and a new beginning for a Chinese firm
Firms with a broad geographic reach are more likely to win work, especially from global companies with high turnovers, according to survey data of nearly 29,000 corporate counsel
IP STARS, Managing IP’s accreditation title, reveals its latest rankings for trademark work today, including which firms are on the up
Highlights from Sunday included judicial insight from across the globe and a keynote address from Martin Luther King Jr’s daughter
Managing IP’s senior reporter Rani Mehta interviewed attendees at the INTA Annual Meeting in Atlanta about how they made the most out of their first day
A team of lawyers who joined Norton Rose Fulbright from Polsinelli say they were drawn to the firm's global platform
Gift this article