EUROPEAN COMMISSION

Brussels, 27.4.2023

SWD(2023) 124 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT REPORT

Accompanying the document

Proposal for a Regulation of the European Parliament and of the Council on standard essential patents and amending Regulation (EU) 2017/1001

{COM(2023) 232 final} - {SEC(2023) 174 final} - {SWD(2023) 123 final} - {SWD(2023) 125 final}

Table of contents

1.Introduction

2.Problem definition

3.Why should the EU act?

4.Objectives: What is to be achieved?

5.What are the available policy options?

6.What are the impacts of the policy options?

7.How do the options compare?

8.Preferred option

9.How will actual impacts be monitored and evaluated?

Annex 1: Procedural information

Annex 2: Stakeholder consultation (Synopsis report)

Annex 3: Who is affected and how?

Annex 4: Analytical methods

Annex 5: Additional information in support of Problem Section

Annex 6: Additional information on options discarded at an early stage

Annex 7: Additional information in support of impact analysis

Annex 8: SMEs and SEPs

Annex 9: Public consultation – detailed summary

Glossary

1.Introduction

1.1.Political and legal context

Standard Essential Patents (SEPs) are patents that protect technology that is incorporated in a standard. 1  SEPs are “essential” in the sense that implementation of the standard requires use of the inventions covered by SEPs.

SEPs are a combination of two seeming contradictions. Standards ensure compatibility and functionality of complex technology, as well as enable interoperability between devices. The success of a standard depends on its wide implementation and as such every producer should be allowed to use a standard. On the other hand, the goal of a patent is protection of the technology of the patent owner who has the right granted by law to limit, or prevent, the usage of that technology and profit, or not, from it, by choosing how and to whom to license/sell patents, whether or not this is for remuneration. To deal with these contradictions, Standards Development Organisations (SDOs) ask the patent owners who participate in standard development to promise to license their patented technology (i.e., SEPs) on “fair reasonable and non-discriminatory” (FRAND) terms 2 to any implementer 3 that chooses to use the standard. If a patent owner makes this promise (called FRAND commitment) 4 , it cannot refuse to license its SEPs to a party who is willing to agree to FRAND terms.

SEP licensing issues arise primarily in the context of standards where the SEP holders have made a promise to the relevant SDO to license their SEPs on FRAND terms for remuneration. Such standards are, for example, communication standards (e.g. 3G, 4G, 5G, WiFi, NFC), audio/video compression and decompression standards (MPEG, HEVC), as well as standardized technologies for data storage and exchange (CD and DVD), photo formats (JPEG), and Home Audio/Video Interoperability (HAVi). The users of these standards are traditionally producers of telecommunication equipment, mobile phones, computers, tablets, TV sets, etc. With the recent rise of the Internet of Things (IoT) including connected cars, drones, payment terminals, tracking devices, smart meters and other smart devices, the standards listed above are increasingly implemented by a growing number of companies, often SMEs.

Standard setting and the related FRAND obligation are guided by the Horizontal Guidelines 5 and the CJEU landmark ruling from 2015 in Huawei v. ZTE. 6  The summary of the judgment can be found under the link below. 7  The CJEU recognized the right of the SEP holder to seek to enforce its patents in national courts subject to certain conditions that must be fulfilled to prevent an abuse of a dominant position by the SEP holder. Since a patent confers on its holder the exclusive right to prevent any third party from using the invention without the holder’s consent only in the jurisdiction for which it is issued (e.g., Germany, France, the US, China etc.), the patent disputes are governed by national patent laws and civil proceedings/enforcement laws. 8  

In case of unlicensed imports, SEP holders can rely on Regulation (EU) No 608/2013 concerning customs enforcement of intellectual property rights 9  by requesting detention of goods by customs, although that Regulation does not provide specific provisions with regard to applications for customs’ action concerning goods infringing SEPs. 10  Jurisprudence seems to suggest that the parties have to comply with the conditions described under Huawei v ZTE before requesting such seizure or asking for the release of impounded goods on the part of the implementer. 11

In 2017, the Commission’s Communication 12 “Setting out the EU approach to Standard Essential Patents,” called for a comprehensive and balanced approach to SEP licensing to incentivise the contribution of the best technology to global standardisation efforts and foster efficient access to standardised technologies. The Commission acknowledged the need for increased transparency (regarding, among others, existence of SEPs, SEP ownership, and FRAND royalties) and addressed certain aspects of FRAND licensing and SEPs enforcement. The Commission’s views were supported by Council Conclusions 6681/18 13 , with the Council stressing the importance of increased transparency.

Since then, the Commission has conducted several studies 14 and a series of webinars 15 , has established a group of experts on the licensing and valuation of SEPs 16 , and has actively monitored the market situation. On 10 November 2020, by Council conclusions 12339/20 17 , the Council invited the Commission to present proposals for a future EU IP policy. The Council encouraged the Commission to swiftly present the announced IP action plan, with initiatives to make IP protection more effective and more affordable, especially for small and medium-sized EU enterprises (“SMEs”), 18 and to promote effective sharing of IP, in particular critical assets such as SEPs, whilst ensuring adequate and fair compensation to technology developers.

On 25 November 2020, the Commission published its intellectual property action plan (‘the IP action plan’) 19 , where it announced its goals of promoting transparency and predictability in SEP licensing, including by improving the SEP licensing system, for the benefit of EU industry and consumers, and in particular SMEs. The IP Action plan noted increases in SEP licensing disputes in the automotive sector and the potential for other IoT sectors to become subject of such disputes, as they begin using connectivity and other standards. The IP Action plan was supported by Council Conclusions of 18 June 2021 20 and by the European Parliament (EP) in its Resolution 21 . The EP acknowledged the need for a strong, balanced and robust IPR system and agreed with the Commission’s position that the transparency necessary for fair licensing negotiations depends in large part on the availability of information about the existence, scope, essentiality, ownership, and stipulated royalties of SEPs. The EP also asked the Commission to provide more clarity on various aspects of FRAND, and to consider possible incentives for more efficient SEP licensing negotiations and reducing litigation.

In parallel with this initiative, the Commission has updated its Standardisation strategy 22 (EU Strategy) and is revising the Horizontal guidelines 23 . The new EU Strategy, published in February 2022, aims to strengthen the EU's global competitiveness, preserve the high quality of EU standards, and enable a resilient, green and digital economy. The present SEPs initiative is complementary to this EU Strategy and to the ongoing work to revise the Horizontal Guidelines.

Furthermore, this initiative is important in the context of global developments. For example, certain emerging economies are taking a much more assertive approach, not only in promoting home-grown standards, but also by actively supporting participation by domestic companies in global standardisation efforts, providing their industries with a competitive edge in terms of market access and technology roll-out. Moreover, courts in the UK, US and China have, with their own specificities, decided that they have jurisdiction to determine global FRAND terms in specific cases, which may impact the EU industry. 24  Some countries have released 25 or are considering guidelines governing SEP licensing negotiations as well. 26

2.Problem definition

2.1.Market description

SEPs represent approximately 2% of the population of the patents that are currently in force. Standards to which FRAND commitments are made and SEPs are declared include communication standards (e.g. 4G, 5G, WiFi) (90% of declared SEPs), computer technology (5% of declared SEPs), audio/visual field (e.g. codecs) (2% of declared SEPs). The remaining 3% cover machinery, measurement, semiconductors, optics or medical technology. The number of declared SEPs was estimated at around 75 000 (patent families) worldwide in 2021 following a six fold increase over the last decade. They are owned by approximately 260 companies (companies with at least 10 SEPs). 27  One third of all SEPs is owned by Chinese companies, doubling their share in just seven years. Also shares of Korean and Taiwanese SEPs increased over that period to respectively 19% and 1.2%. On the other hand, both the EU and the US shares in SEPs decreased from 22% to 15% and from 26% to 19% respectively. The share of Japanese companies remains stable at around 11% (see figure below). Around 80% of all SEPs held by EU companies are owned by just two companies, Nokia and Ericsson, and a further 10% by Philips and Siemens. The remaining 10% are shared by around 27 firms including telecoms (e.g. Orange, KPN, Deutsche Telekom, Teliasonera, Telecom Italia, T-Mobile, Alcatel), engineering firms (e.g. Bosch), or research institutions (e.g. Fraunhofer Gesellschaft) to name just a few. 28  It has to be noted that the number of declared SEPs may not provide information on the strength of the SEPs portfolio of the different companies. The latter would depend inter alia on which of those declared SEPs are actually essential, which features of the standard they cover and what is their quality.

Figure 1: SEP ownership share by year and country (top 50 owners)

Note: Based on top 50 owners representing from 88% (2014) to 95% (2021) of all SEP families
Source: IPLytics database

The FRAND commitment underlies the SEP owners’ approach to licensing SEPs to implementers, which may be done either free of charge or for royalties. For instance, SEPs in the area of internet (e.g. URL, HTML, XML, TCP, Java Script) are predominantly licensed royalty-free. Cellular standards (e.g. 4G, 5G) are subject to royalty payments, similar to short range communication standards WiFi and NFC, with the exception of Bluetooth, which is royalty-free but requires a small annual membership fee. SEPs related to CD, DVD, digital image formats (JPEG) and most audio/video compression technologies (MPEG, AAC, HEVC) are also subject to royalty payments. 29 Royalty bearing SEPs can be licensed through bilateral arrangements or via joint licensing programmes, including “patent pools” (consortia of patent owners created to license their patents jointly, through a single transaction). SEP holders are expected to license their patents on FRAND terms 30 so as not to abuse their monopoly position.

The largest share of royalty payments for SEP licenses comes from the mobile telecommunications industry, which generates an estimated patent royalty yield of EUR 14 – 18 billion per year 31 with additional EUR 4 billion of non-monetary benefits from cross-licensing. 32 The EU’s share could be estimated at around EUR 3 billion per year. 33

One standard can consist of thousands of technical contributions protected by patents owned by dozens or even hundreds of SEP holders (for example, in the case of 5G, major SEP holders include Ericsson, Huawei, Intel, Nokia, Qualcomm, Samsung, and ZTE, just to name a few). Moreover, a single product (e.g. a car) can implement multiple standards (see figure below).

Figure 2: Standards with FRAND commitments implemented in a modern connected car

Source: Tim Pohlmann, Intellectual Asset Management, May/June 2017, pp. 22-27

Historically, with regard to SEPs in communications technologies, SEP licensing occurred in large part between phone and network equipment companies who licensed their SEPs to each other (i.e., “cross licensing”) 34 , because they all had products using the standards. With the market entry of pure SEP holders and pure SEP users, divestiture and licensing of SEPs for monetization purposes, lead to the rise of patent assertion entities (PAE). 35  More recently, the rise of the Internet of Things (IoT) is resulting in more and more devices using connectivity and the landscape of SEP licensing is shifting even more. Today, (i) some companies incorporate standards into their products while not owning SEPs covering such standards, (ii) others own and license SEPs without using them in any products, and (iii) major SEP holders have significantly reduced their product businesses and focus more on licensing their SEPs. Whereas over the last two decades most high-stakes SEP disputes have centred around mobile communication devices (i.e. smartphones), we are already witnessing more disputes in the automotive sector and expect other IoT sectors to be similarly affected. 

See Annex 5 for more explanations on SEP market and SEP licensing.

The largest SEP licensing market continues to be mobile phones, in which the top ten phone makers are Samsung, Apple, and 8 Chinese companies, including OPPO, Xiaomi and Vivo. SEP licensing, however, is now moving more towards the growing IoT market, including automotive, smart energy (smart meters and smart grids), payment terminals, tracking devices, drones, medical devices, wireless charging stations and other products.

Based on the Orbis 44 database, in 2022 there were around 47 500 potential manufacturing firms worldwide that may implement standards that are subject to FRAND commitments. Around 3 800 (8%) were located in the EU. Among those EU firms, 16% were large and 84% were SMEs (based on employment figures). They employed 2.2 million persons. The combined turnover of these firms for the last available year amounted to around EUR 600 billion. 88 of the companies reported a total R&D spending of EUR 26 billion in the last available year, on average 7% of their turnover. 45  

Furthermore, we estimate that each year around 600 SEP license agreements are signed with approximately 230 EU based implementers (including 80 SMEs) and around 190 non-EU based firms with subsidiaries in the EU. For more information on the implementers’ landscape, please see Annex 5.1.

2.2.What are the problems?

In this section we will present the current problems of SEP licensing, in a situation when a SEP holder makes a FRAND commitment to an SDO, from the point of view of SEP holders and implementers. As SEP licensing concerns commercial deals between two parties and is often subject to non-disclosure agreements, there is only limited public information available (for instance when there is a court dispute). Hence the information below is based on evidence obtained through public consultations, numerous interviews with stakeholders and dedicated studies. The examples provided below are anonymised and aggregated versions of evidence the Commission obtained from testimonies, but which are subject to commercial secrecy. The figure below outlines the argumentation logic of this chapter.

Figure 3: Problem tree

The overarching problematic issues are uncertainty and high transaction costs, which affect differently the behaviour of SEP implementers and owners (this may not be the case when an implementer also owns SEPs as e.g. in case of mobile phone companies, but increasingly with the rise of IoT, the two groups are distinct).

2.2.1.SEP owners’ perspective

In order to participate in standard creation, prospective SEP holders had to invest considerable resources and time in R&D activities to first develop new technology and then to patent it worldwide. CRA (2022) estimates R&D amounts of between USD 2 and 9 billion annually for standards used in a smartphone. 46  All that is done without guarantee that first, the inventor’s patents will be used by the standard, and second, that the standard will be accepted by the market. 47 Even if a standard is accepted, it takes years before it is widely used while an invention is patented for maximum 20 years. As a result, SEP holders have limited time to generate a return for their R&D investments through royalties for the use of the patents.  

A SEP owner, as any patent owner, may license implementers for the use of the protected technology and request remuneration. In contrast to other patents, a SEP holder has to license subject to the FRAND commitment. According to the public consultation 48 , in order to be able to better assess the value that their technology brings to the standard implementations, a SEP holder would wait a few years (around 2 to 4 years) 49  until the standard is implemented in the market and then approach companies in specific markets to offer them licences. This is followed by negotiations, which take on average 3 years 50 , and potentially litigation in case parties cannot reach an agreement (adding another 1 to 2.5 years). 51  

In the public consultation, SEP holders identified their most important problem to be the various reasons used by the implementers to delay the taking of a licence. 57 Examples are provided in Annex 2.

The costs related to SEP licensing are also significant for both SEP holders and implementers. The average bilateral negotiation cost per licence for both the SEP holder and the implementer is estimated at EUR 2 million to EUR 11 million (see Table 1 and examples in Annex 2). Such high cost (both, in terms of time and money) also explains why major SEP holders usually have licensing deals signed with only around 100-200 implementers 58 , that have a sufficiently high volume and/or value of sales that would allow for absorbing these costs. There are also SEP holders whose main source of income is licensing. They constitute a very heterogeneous population, including R&D specialist firms (e.g. InterDigital), public or semi-public research institutes (e.g. Fraunhofer, CSIRO, ETRI), universities, and different types of patent assertion entities (PAE). 59  PAEs are entities whose business model is the licensing and assertion of patents acquired from other parties. They usually conclude below 100 licenses, yielding a modest revenue. Finally, patent pools are platforms through which multiple SEP holders offer a single license to multiple licensees, promoting more efficient licensing. For an overview of licensing practices see table below:

Table 1: Licensing cost and revenue for different SEP holders (worldwide)

Source: -    Baron, J., Arque-Castells, P., Leonard, A., Pohlmann, T., Sergheraert, E., Empirical Assessment of Potential Challenges in SEP Licensing, European Commission, DG GROW, 2023;, Table 16, pp. 138-139

Consequently, SEP owners, in particular those who pursue bilateral licensing, face two main challenges: long negotiations and high cost of SEP licensing. This could potentially reduce their income base and might reduce incentives for participation in development of new standards, especially where strategic licensing markets stagnate and/or move to low-cost business models.

Stakeholders reported that the royalty rates earned by the firms that specialise in technology development have been steady for years in the mobile phone market. 61  It appears that the income uncertainty is currently mitigated by relatively high SEP prices charged to those who are licensed (first example below). And the rate of standard development does not appear to have been affected (second example below).

The ability of the SEP holder to seek royalties is likely to be put under pressure in the future. Some phone manufacturers reported to the Commission that they adopted a low-cost business model to serve lower income consumers, and the royalty demands are not commensurate with their business model. An IoT supplier and an IoT mid-cap 65 reported that the IoT market is very fragmented, competitive and cost sensitive.

These developments, together with the fact that many SEP holders are no longer implementers, or implementers on a much smaller scale (in particular in the IoT), is likely to have an impact on the FRAND royalty negotiations as the disagreements about the FRAND royalty determination are likely to become more pronounced. This in turn will have a negative impact on both the length and the cost of the negotiations.

2.2.2.Implementers’ perspective 

While SEP holders need time to assess the value of the technology in standard implementations and make decisions about who and how to license, and for how much, in particular in new markets, implementers encounter great uncertainty and prospects of much higher than anticipated costs associated with the use of standards, potentially discouraging the implementation of these new technologies.

In principle, goods sold should be free from any right or claim of a third party based on intellectual property, if known to the seller or the seller “could not have been unaware”. 66  If a company discovers that it may be infringing a patent 67 , it has the choice, among other things, to (i) “work around”, using a solution that does not result in patent infringement; or (ii) stop the product development unless the company is able to secure a license from the patent owner for the use of the patented technology. 68 The rationale is to respect the law by ensuring first, that a company does not infringe on another person’s property rights, and second, anyone who buys such a product has legal certainty. The same rules apply to exports. 

50% of the companies who replied to the public consultation claimed that they have proactively sought SEP licences. 69  Not infringing a SEP, having legal certainty over costs and an enhanced ability to plan business activities were mentioned as the top reasons for seeking a license by around three quarters of respondents to the public consultation. 70  Some of those who sought licences did not obtain such licences, mainly because of disagreement on a FRAND royalty (especially in new markets where the SEP holder has not yet decided on its licensing policy and where the SEP holder prefers licensing at end-product level). 17% of the companies who replied to the public consultation reported not to have requested a licence proactively. 33% expressed no opinion. There are various considerations in the decision not to proactively seek a licence.

In practice, it is impossible to sign a SEP licence with all SEP holders whose standardised technology the firm uses. For example, Avanci represents 52 SEP holders of 2G, 3G and 4G but does not include certain large SEP holders such as Huawei, Samsung and others in their membership. SEP holders with whom a company did not sign a licence may ask the firm to enter into a licence at any time (and some may never seek royalties).

Standards are in most cases embedded into components that are used in an end-product sold to the consumer. Suppliers of such components may not have the necessary SEP licenses when they sell such components to the end-product manufacturers. As a result, such end-product manufacturers could be subject to a demand for SEP licenses. For example, in the automotive industry, while component manufacturers were buying their chips licensed for 2G and 3G, they were not licensed for 4G. 71 With the adoption of the 4G standard, SEP holders chose not to license component manufactures but to license the car manufacturers. Component suppliers are not able to pay the royalty requested from the OEMs and pass the cost downstream. Some European suppliers chose to exit the market 72 , others are still litigating 73 .

The likelihood of SEP holders asserting their SEPs in courts is greater in regions with more efficient judicial systems and large markets. During the webinar on enforcement, to which 893 persons registered for participation 74 , 51% of the participants indicated that they prefer to litigate in Europe against 30% who preferred the US. To the question in which jurisdiction they have litigated, 58% of the participants indicated Germany, 7% indicated France and 4% indicated the Netherlands. 31% specified jurisdictions outside of the EU. Using the number of court cases as a proxy, we can estimate that implementers are targets of court actions in the EU primarily in Germany (around 44 court cases per year), France (around 2) and the Netherlands (around 1). 75  Note that not every SEP holder decides to litigate, but only court cases are observable.

As the threat of a court injunction/cessation of production means income loss and loss of market share to competitors and switching to another technology is costly or impossible (because of standard implementation), implementers can engage in court procedures that are cost- and time intensive, attempt to negotiate, or agree to licence agreements at any price (especially SMEs, if they are targeted). This holds true especially for licensing in the IoT, where there is continuous entry of new start-ups that implement standardised technologies in their IoT devices. 82% of SMEs in the targeted survey stated that they do not have resources to negotiate with SEP holders or engage in court proceedings. 77

Start-ups and SMEs lack not only the resources but also the SEP and licensing expertise. 78  Two thirds of SMEs replying to the Commission’s targeted survey stated that there is uncertainty regarding SEPs and that royalty payments affect their competitive position. 79  Currently, there is no public support available to start-ups and SMEs for handling SEP licensing negotiations. So, the majority are de facto not licensed. One third of the SMEs replying to the targeted survey believe that it is unlikely they will be invited to take a licence by a SEP holder (small size and insufficient volume of sales). Another third of the replying SMEs thought otherwise. 80  In principle, nothing prevents a SEP holder from demanding that an SME take a licence.

Whether big or small company, this uncertainty makes business planning more difficult. 82 We will elaborate in the next section on the problem drivers. For more examples of problems that implementers face, see Annex 2.

2.3.What are the problem drivers?

The problems above are caused in large part by the following drivers, which this initiative aims to tackle: insufficient transparency on SEP ownership and essentiality; lack of information about FRAND royalties; and a dispute settlement system not adapted for FRAND determination. There are also other drivers affecting the problem, which this initiative cannot solve directly, such as the use of foreign courts to resolve SEP cases.

2.3.1.Unclear SEP ownership and essentiality

To negotiate a SEP licence, a standard implementer should know which SEP holders license for a fee, how many declared SEPs 83  each has, and how many of those are actually essential (necessary to implement a standard).

Information on patent ownership and essentiality is not readily available. 84  The vast majority of SMEs replying to the Commission’s survey said they did not know who owns SEPs relevant to a standard they implement and stated they did not know if all the patents that a SEP holder presented to them during negotiations were actually essential to the standard. 85  This is consistent with the results from the public consultation. 86  Without such information, the cost and complexity of SEP licensing negotiations increases. With regard to almost all standards, expert third party analysis is required to obtain reasonably reliable estimates of the number of actual SEPs. The cost for implementers of assessing the essentiality of individual patents could be significant from an average cost of EUR 355 to EUR 7 860 per SEP, depending on the evaluation rigorousness. 87

There are three main reasons for the lack of transparency on the number of essential patents for different standards:

First, many Standards Development Organisations (SDO) do not provide comprehensive data on self-declared potential SEPs, as they allow for blanket disclosures. Patent owners may simply declare that they own potential SEPs and are prepared to license on terms compliant with the SDO’s IPR policy, without specifically identifying any of these patents or their relevance to the standard(s) in question. Some estimates indicate that only about 10-20% of all Wi-Fi SEPs are specifically declared at its SDO: IEEE, 88 and approximately 20-30% of all High Efficiency Video Coding (“HEVC”) SEPs are specifically declared at the relevant SDO: ITU-T 89 . Hence, the large majority of potential SEPs are not specifically declared at either SDO. As a result, these SDOs’ databases provide no useful information about the numbers of potential SEPs held by the various owners.

Second, while some SDOs (for example ETSI) require patent owners to identify specific patents and/or patent applications that they believe are or may become essential to a standard, the disclosure (also called declaration) process is complex. This process is geared towards standardisation activities and not adapted for SEP licensing negotiations.

Third, declarations of potential SEPs are often made during the standardisation process (firms wishing to contribute to a new standard identify their patent assets that may cover the standard, depending on what contributions are accepted) and hence do not provide reliable information on the actual SEPs covering the final standard. This is one of the reasons why experts estimate that only between 25% and 40% of the patents found in the ETSI IPR database are in fact essential to the final published standard 90 , as supported by empirical research. 91 In the case of 5G, the essentiality rate can be as low as 15%. 92

Depending on the SDO, SEP holders are not required to regularly update the information originally provided to SDOs, for example by removing non-essential patents or specifying a change in ownership.

Patent pools provide more transparency 93 and conduct essentiality checks. Typically, not all SEPs of an individual SEP holder are checked for essentiality. In addition, the SEP holder chooses which of its SEPs to submit for essentiality check. The likelihood of those SEPs being confirmed essential is, therefore, very high. The essentiality check system of the pools is thus not designed to determine the rate of essentiality in the portfolio of an individual SEP owner. Furthermore, there is no patent pool that has gathered all SEP holders for a given standard. It follows that patent pools are not capable of providing sufficient information on the overall SEP landscape.

More readily usable data is available from commercial providers. The cost varies from EUR 5 000 to EUR 25 000. 94  These private databases are thus available only at high cost, and it is unclear how reliable their assessments of essentiality are. 95

80% of SMEs responding to the SME survey said they did not know who owns SEPs relevant to the standard they use and 90% did not know if patents presented to them during negotiations were essential to the standard. 96

2.3.2.What is a FRAND royalty for SEPs?

FRAND refers to the grant of licences on fair, reasonable and non-discriminatory terms (‘FRAND terms’) in relation to a patent right. Individual patent holders give an undertaking to the SDO that they are prepared to grant licences on FRAND terms. 97

Any SEP holder may demand royalties from an implementer, irrespective of the number of SEPs it owns. Standards such as 5G can have around 200 SEP holders and each of them may own a different number of SEPs covering different patented technologies included in the standard. It is, however, unclear how many truly essential patents each owner has, as explained above. Furthermore, the value of such patented technologies may be different. Therefore, the number of patents contributed by each owner could be seen as an imperfect proxy. Finally, the products of the implementers almost always include many other features and technologies that contribute to the value of their products. This makes it even more difficult to determine the value of the standardised technology. In addition, determining the share of royalties for each SEP holder is also a complex process.

When a SEP holder approaches a standard implementer for royalties, the implementer can choose to engage in in-depth and lengthy negotiations to determine what would be a reasonable royalty (i.e., FRAND) or to accept the demanded rate. Engaging in negotiation requires technical competence, knowledge of patent law and licensing skills, all of which may be less available capabilities in smaller companies as compared to larger firms. Thus, SMEs for example may tend to accept the SEP holder’s offered terms fearing an injunction/stop of production. Moreover, if a company (even a large one) is implementing the standard by using a component supplied by a third party, it will most likely have no knowledge of the relevant technology and must engage external expertise to assist in the assessment of the royalty demand. 98  This comes in addition to the legal resources necessary to negotiate and finalise a licence agreement.

A SEP owner’s refusal to comply with their FRAND licensing obligation may give rise to liability under competition law. However, FRAND may be a range and could mean that the royalty rates differ in different market segments based on the specific use for the standard or parts thereof (e.g. car versus connected vacuum cleaner) and even between producers of the same product, depending on whether they are considered to be “similarly situated” (a FRAND rate should be the same for a company in a comparable situation). SEP holders often disclose their FRAND rate demand at the outset of the licencing negotiations (they may ask for a non-disclosure agreement to be signed, but many do not), however they typically refuse to share information about “comparable licences” or are selective about which comparable SEP deals they will show to the implementer (this kind of information is typically not public). Some SEP holders announce a SEP price in advance, yet in many cases this is a maximum price that can be reduced during negotiations. In most cases this is not a price for all relevant SEPs (of all SEP holders) for a standard (“aggregate royalty”), but a rate the particular SEP owner(s) would like to get for their share of the total SEPs.

Stakeholders consider that the FRAND licensing concept could benefit greatly from some clarification, notably with regard to the determination of an aggregate royalty burden, the level in the value chain at which SEPs should be licensed (e.g. at end product, or component level), or whether an initial offer by a SEP holder must be FRAND (e.g. guidance in addition to that provided in the 2017 Communication). From our analysis and meetings with stakeholders it emerged that especially the level of licencing issue is very controversial with no one-size-fits-all solution, 102  and may not be mandated. It could be recommended, however, that customary industry practices are considered in this respect.

Issues related to SEP licensing are also regularly raised to competition authorities, including the Commission’s department responsible for EU competition enforcement. SEP holders raise concerns regarding implementers unduly delaying negotiations. Implementers complain mostly about SEP holders not following the Huawei v. ZTE process, offers that are not on FRAND terms or about SEP holders refusing to licence at a certain level of the value chain. 103  

Almost all SMEs reported that in negotiations they do not know what a fair price for the SEPs would be. 104

2.3.3.Dispute settlement not adapted to resolve FRAND rate disagreements

SEP holders and implementers often disagree about what constitutes FRAND terms and conditions (primarily royalties) for a SEP licence. If parties cannot agree on a FRAND rate, SEP licensing disputes can turn into patent infringement proceedings before national courts (around 35 per year in the EU) 105 in which SEP holders usually seek injunctions. On average, the annual costs of SEP related court proceedings in the EU are estimated at around EUR 6 million. However, already the costs of a complicated case alone can reach millions of euros. 106  

It is very difficult to properly determine FRAND rates in patent infringement proceedings before a national court, because the SEP holder selects a limited number of SEPs from its national portfolio to adjudicate, but seeks to license all of its SEPs for the particular technology (e.g., 4G or HEVC), including those in other jurisdictions (global FRAND licence).

The CJEU judgment in Huawei v ZTE established that the SEP holder’s offer, and the potential licensee’s counteroffer must be FRAND and left it up to national courts to assess whether the parties observed the process before granting an injunction. National courts in the EU typically focus on establishing whether the good faith negotiation process has been complied with and have so far not been involved in global FRAND determinations. Thus, potential remedies tend to focus primarily on the availability of an injunction, as well as on damages for the past infringement of the asserted patents and only in the relevant jurisdiction. Thus, the parties are still left with little certainty regarding FRAND royalties for future sales.

To deal with this uncertainty, a number of courts outside the EU have declared to have jurisdiction to determine a global FRAND rate, if the parties do not themselves reach agreement on global FRAND royalties (e.g., UK 107 and China 108 ). Courts in the EU have so far refrained from making any FRAND royalties determinations at a national or global level.

While traditional alternative dispute resolution (“ADR”) mechanisms are available to SEP holders and implementers and could potentially result in broader FRAND determinations, ADR is rarely used to settle FRAND-related disputes. 109  Among the key concerns are that (i) implementers may not be able to obtain information regarding comparable licenses and other relevant data from SEP holders (available through discovery proceedings in certain judicial proceedings), and (ii) the proceeding itself as well as the resulting decision are kept confidential and are not publicly available, so they have limited precedential value for future FRAND-related cases. 110  

In response to the public consultation, however, a number of respondents acknowledged the potential for ADR processes to help resolve global SEP disputes, provided that the above concerns are addressed.

There is also no mechanism enabling the use of the customs enforcement regulation 111 by SEP holders because the regulation does not require that the negotiating steps set out in Huawei v ZTE be followed. Customs authorities do not have the competence, resources and expertise to establish whether the negotiating steps set out in Huawei v ZTE were conducted (for more explanation see Annex 5.4)

2.3.4.Drivers out of the scope of this initiative

There are also other factors that may contribute to the uncertainty and high transaction costs for both SEP holders and implementers. These factors include, without limitation, the global nature of standards and products; changes in the business models such as low-cost smart phones; specificities of the IoT markets, which are fragmented, competitive and often a low margin market; and the fact that SEP licensing is sector-specific depending on the nature of the use of the standards. Furthermore, because both SEP holders and implementers typically operate globally, SEP disputes can be adjudicated in different jurisdictions including outside the EU. As such, SEP holders and implementers alike are free to exploit jurisdictional differences to their benefit.

2.4.What are the consequences?

As new standards are developed, additional contributors participate in the standardisation process (particularly from Asia), putting more pressure on existing SEP holders due to further fragmentation of the SEP landscape. Existing SEP holders also face increased challenges to their pricing policies from their main licensing market, mobile devices (primarily for cellular standards). Chinese phone makers are increasing their market share and are focusing on low-cost business models to serve lower income populations. SEP holders need to adapt their licensing model to the realities of these new emerging markets, especially the price pressures.

The IoT market presents additional challenges. It is a market that is growing at a high speed but is fragmented. Volumes for certain applications may be small and profit margins seem tight. 112 Licensing may prove difficult and expensive as those industries are not familiar with SEPs (see Annex 2). The combination of all these factors is likely to deepen the disagreements about FRAND royalties, which is likely to cause more delays in negotiations and to increase the parties’ licensing costs, potentially impacting SEP holders’ revenues. This uncertainty about future revenues could impact the decisions of SEP holders to invest into R&D and may in turn have an impact on participation in standardisation.

On the other hand, the lack of transparency and unpredictability of royalty demands, plus the cost of negotiation and disputes also puts pressure on the costs of products by implementers that are entering the market. This is likely to put pressure on the innovation capacity of implementers as well. CRA(2022) noted that there is a trade-off between high SEP royalties and more innovation upstream (by SEP owners), on the one hand, and less innovation downstream (by implementers) as a result of high SEP royalties, on the other hand.

Both SEP holders and implementers are likely to (i) lose incentives to innovate; (ii) reduce sustainable competitiveness; and (iii) adversely affect supply chain security.

1)Incentives to innovate. Uncertainty regarding SEP licensing, price and conditions, with the risk of litigation may impact investment and R&D decisions as explained above. 

2)Sustainable competitiveness

Global competitiveness

SEP holders often face challenges to the royalty rates offered. Although some implementers may push to obtain lower rates without consideration of what constitutes FRAND, there is also a lack of certainty and transparency with regard to factors that would support a determination of FRAND rates. Uncertainty about which SEPs are essential and the inability of implementers (large and small) to determine the overall essentiality and validity of the SEP owner’s portfolio through technical discussions (which are often limited to one or two dozen SEPs) lead to questions whether the value of the SEP owners’ portfolio is overestimated. The testimonies listed in Annex 2 all refer to a lack of trust in the system and regarding the FRAND-ness of the SEP owners’ offers.

Implementers are typically challenged at courts in the EU first, particularly if they have important sales in the EU. Those that take a licence first face competition from unlicensed competitors.

Furthermore, EU companies implementing more expensive standards face difficulties to effectively compete in markets which have adopted their own competing standards. For example, China promoted the creation of a Chinese patent pool for its audio-visual standards (AVS) and standard contributors accepted to contribute patented technologies for an aggregate fee of one yuan (approximately EUR 0.14) per hardware device. 113  EU implementers of the competing AVC standard may not be able to compete in that market.

The public consultation confirms that if European companies in the IoT sector, such as smart meters, wireless charging stations and payment terminals, are licensed first at rates that their competitors in other jurisdictions would not pay, they are likely to lose competitiveness. 114 More testimonials are available in Annex 2. 

Technology diffusion

There are risks that standards’ take-up will be slower or implementation will be delayed in the EU until applicable SEP royalty rates become more affordable.

Companies may also try to work around the SEP technology by creating their own (not standard) solutions or develop a competing standard on a royalty-free basis, where investment cost permits it. It is also possible that older/less efficient, but lower-royalty or royalty-free technology will be used to avoid high SEP royalty exposure. Half of the SMEs stated they try to use free or older technology in order not to pay for SEPs, while one in five stated that they would try to develop in-house alternatives. 116

Although the study commissioned to support this IA concluded that at least currently there is no systematic evidence of mass scale occurrence of slower than expected or delayed implementation of standards, 118 the public consultation and several testimonies show that the IoT sector is adopting the newest standards slower than the mobile phone market. 119 Many IoT devices can operate on older standards such as 2G (patents expired) and 3G (which is cheaper) but the 2G and 3G infrastructure will be removed in the near future. 120

Both EU SEP holders and implementers are set to lose from this situation.

3)Supply chain security

Delays in the negotiations and increasing costs of licensing, may disincentivise EU SEP holders to continue their participation in standardisation. This would benefit non-EU contributors to the standards. The EU would thus lose its ability to contribute European technology to the global market, also related to critical infrastructure.

There is already proof in the automotive industry that because of SEP royalties, European suppliers are no longer able to compete and exit the automotive supplier market. In a submission to the Commission, the automotive manufacturers’ organisation ACEA reported a shortage of supplier offers and a growing dependence on Chinese suppliers.

2.5.How likely is the problem to persist?

With an increasing demand for connectivity (particularly for IoT) and other standards that promote interoperability, the problem of SEP licensing is set to gain prominence and affect a much larger number of companies. About 70% of surveyed SEP experts 121 as well as conclusions from the Expert Group Contribution confirm that the licensing of SEPs for IoT applications is expected to be more challenging. This is because of current trends in standard development, including the increasing complexity of new standards, the high number of SEP holders with varying interests and business models, a lower volume of products for a larger number of implementers, and increased use of standards across different sectors of the economy. These factors in combination are likely to amplify frictions in the SEP licensing environment, further spotlighting the need for more certainty and transparency.

On the other hand, the level of information on SEP ownership and essentiality is not expected to improve absent a push from outside sources. The Commission has been trying to stimulate voluntary industry solutions since its Communication in 1992 on Intellectual Property Rights and Standardisation 122 , and most recently in 2017 when the Commission issued a Communication on SEPs 123 . Attempts to improve the quality of SEP declarations at SDOs since then show little progress. 124 The polarised responses to the many surveys, webinars, conferences and the Public Consultation have made it clear that the industry will not act on its own.

SEP holders will continue to selectively approach successful SEP implementers for royalty payments (70% of SEP holders in the public consultation answered that they do not contact all known implementers). 125  Although patent pools are likely to emerge for certain IoT applications, such as Sisvel pools for Wi-Fi 6 products, 126  smart meters and tracking devices, 127  such pools will not represent all SEP owners. Consequently, most implementers will not be fully licensed and will operate under the threat of an injunction (particularly if they become successful with higher volumes). Furthermore, due to the increasing number of SEP implementers in an increasingly connected world, SEP holders could choose to outsource SEP licensing (and resulting litigations) to specialized patent assertion entities 128 in the future. 

About 70% of the implementers take a licence without litigation according to the results from the public consultation. 129  There is no indication that traditional alternative dispute resolution mechanisms currently available to SEP holders and implementers will be used more often going forward to settle FRAND-related issues. 130  Respondents to the public consultation considered mediation insufficient as the mediator has no authority to request information and no authority to make a price recommendation. 131  The newly established Unified Patent Court 132 is not yet operational, and it remains unclear how the court will assess and grant injunctions in the context of SEPs and whether it will engage in the assessment of (global) FRAND rates.

No other initiative by the Commission or Member States is going to help provide solutions for these SEP issues. 133  At the same time other jurisdictions already have (Japan, South Korea, Singapore) or are considering SEP related initiatives (including China, India and the UK). If the EU does not take the initiative, it will have to follow developments in other jurisdictions, which will affect EU companies as well as the EU-related businesses of global companies.

3.Why should the EU act? 

3.1.Legal basis

The initiative concerns standards to which a patent holder has contributed a patented technology and which it has committed to an SDO to license on FRAND terms and conditions. Standards for which patent holders make FRAND commitments are applied cross-border among Member States and globally. SEP licensing is also seldom national. Usually, licensing contracts are global and may take into account certain regional aspects. The international standards in question cover technologies such as 4G, 5G, Wi-Fi, HEVC, AVC, DVB and others that ensure interoperability of products worldwide.

Article 114 TFEU constitutes the appropriate legal basis as the objective is to improve the conditions for the establishment and functioning of the internal market. The initiative seeks to ensure the efficiency of SEPs licensing, facilitating lawful access to the standards and promoting wider adoption of standards. There are no specific EU or national rules on SEPs and it has been left to competition law to regulate. 134 In addition, as acknowledged by the CJEU in Huawei v ZTE, apart from common rules relating to the grant of a European patent, a European patent remains governed by the national law of each of the Contracting States for which it has been granted as is also the case of national patents.

The CJEU has confirmed 135 that recourse to Article 114 TFEU is possible, if the aim is to prevent the emergence of obstacles to trade between Member States resulting from the divergent development of national laws. However, the emergence of such obstacles must be likely and the measure in question must be designed to prevent them. Dutch 136 , French 137  and German 138 , 139  courts have been considering FRAND-related issues in national litigation, based on the laws of the Member states and the specificities of the disputes brought before them. Those cases show different approaches (not necessarily different results) with regard to FRAND determinations concerning SEPs covering regional or global standards. It is difficult for EU national courts to handle SEP-related cases and make detailed and consistent FRAND determinations. This is in large part due to the lack of transparency and complexity of the issues that are central to such determinations, such as essentiality of patents, comparable licences and compliance with FRAND requirements. While the initiative will neither interpret the CJEU case-law nor dictate methodologies for FRAND determinations per se, it will establish mechanisms that promote the necessary transparency, increase certainty and reduce the potential for inconsistent rulings. This will be a significant improvement in these courts’ abilities to handle SEP disputes. 

EU-wide rules on transparency regarding SEPs and FRAND terms would have a harmonising effect within the EU, which would facilitate the work of national courts and the future Unified Patent Court. The instrument to implement this initiative should be a regulation. A regulation would be directly applicable, including by empowering an EU agency with the tasks of managing a register of SEPs, and establishing a common FRAND determination procedure that would ensure uniformity across the EU and provide greater legal certainty. These outcomes cannot be achieved by means of a Directive.

3.2. Subsidiarity: Necessity of EU action

Measures taken at national, regional or local level aiming at increasing transparency and facilitating licensing of SEPs may not be efficient for the following reasons. First, instead of one EU-wide solution for SEPs, there might be different national solutions for the SEPs on one specific standard. Second, under an EU-wide approach, it will not be necessary to conduct more than one essentiality check per patent family to find that patents are indeed truly essential to a standard. The check would be done based on a single EU-wide methodology. Third, non-centralized alternative dispute resolution processes may come to different results for the same SEP portfolio, opening the door to “forum shopping” within the EU. An EU-wide approach can help avoid these problems.

3.3. Subsidiarity: Added value of EU action 

Action at EU level is expected to save costs for stakeholders, both SEP holders and implementers, and for Member States. For instance, there would be one register, one essentiality check per patent family, one common methodology for the conduct of such checks, and a streamlined and transparent conciliation (FRAND determination) process. SEP holders and implementers would not have to incur the same costs in each EU Member State which would be the case with national solutions, especially in a situation where most standards are regional or global.

4.Objectives: What is to be achieved? 

4.1.General objectives 

This initiative aims at:

·Ensuring that end users, including small business and EU consumers benefit from products based on the latest standardized technologies at reasonable prices.

·Making the EU an attractive place for innovation and standards development (including for global participants).

·Ensuring that both EU SEP holders and implementers innovate in the EU, make and sell products in the EU and are competitive on global markets.

In particular, the goal is to facilitate SEP licensing negotiations for both SEP holders and implementers, by reducing current uncertainties as to the level of SEP exposure (implementers) and return on research and development activities (SEP owners) thereby also lowering the transaction costs.

4.2.Specific objectives

This initiative has three specific objectives:

·Provide more clarity on who owns SEP and which SEPs are truly essential.

·Provide clarity on FRAND royalty and other terms and conditions

·Facilitate SEP dispute resolution.

Figure 5: Objectives tree

5.What are the available policy options? 

5.1.What is the baseline from which options are assessed? 

The baseline was described in section 2.5 above. In essence, SEP holders will continue to selectively approach implementers, and both will bear the costs of protracted negotiations and court disputes. The information asymmetry between SEP holders and implementers will remain, with the latter having to bear additional costs either by committing resources to long negotiations (including own essentiality checks) or by accepting the SEP owner’s offer and paying potentially higher fees to continue production. With time, patent pools offering more efficient licensing solutions for selected applications may emerge and facilitate licensing, if they gain acceptance in the market. This is not likely to happen because the new emerging pools gather SEP holders who do not implement the technology, for example in the IoT. The FRAND royalty expectations are likely to be higher than the market would suggest. 140  

None of the other patent related initiatives announced in the IP action plan of 2020 (i.e. Supplementary Protection Certificates 141 and Compulsory Licensing 142 ) is tackling nor affecting development of SEP specific problems identified in this impact assessment.

5.2.Description of the policy options

The options concern international standards for which a patent owner has made a FRAND commitment and SEPs are in force in one or more EU Member State(s). The initiative will exclude national standards and standards with royalty free licensing policies. The options are constructed following an incremental approach from the least to the most ambitious. More details on options are provided in Annex 6.

5.2.1.Option 1 (PO1): Voluntary guidance 

This option consists of non-binding guidance that provides further clarification of FRAND concepts and conditions such as, for example, i) the economic value of patented technology should not be conditional on market success of a product that implements it, ii) the value may depend on actual functionalities of the standard used, iii) similarly situated entities using the same functionalities of a standard in similar applications should receive similar terms and conditions; iv) an aggregate royalty must be “reasonable”; v) promote the use of FRAND rate discounts for early standard adopters; vi) differentiation of FRAND terms and conditions between sectors and over time; vii) a recommendation that SEP licensing takes place at the level (component or final product) where intellectual property is customary licensed in an existing industry; viii) an explanation of the legal rules on liability for infringement of third-party intellectual property in commercial contracts and their application in the context of SEPs (e.g. indemnification obligations, passing the cost of the SEP licence downstream; ix) clauses in arbitral agreements that should be avoided (e.g. a prohibition to challenge the validity of a patent). The guidance will also recommend measures to facilitate SEP license uptake by SMEs. Additionally, the Commission may recommend to (or require) SEP holders to adopt existing SEP pool practices addressed to SMEs, such as royalty-free licences for small volume production or discounts. The guidance will be reviewed and updated regularly to take account of legal (e.g. CJEU jurisprudence, other case law) and market developments.

An EU Competence Centre on SEPs will be created within the European Union Intellectual Property Office (EUIPO). Its tasks will include monitoring of the SEP licensing landscape, gathering and providing easy access to SEP-related policies from key jurisdictions globally as well as case-law summaries (including arbitrations) related to FRAND issues, also for foreign jurisdictions, supporting EU SEP policy making (e.g. via studies). The Competence Centre would also seek to sign an agreement with the World Intellectual Property Organization (WIPO) to promote the use of the WIPO Arbitration and Mediation Centre for SEP disputes in the EU and to exchange information. 143

Additionally, the centre will become an SME 144 SEP licensing assistance hub. It will offer European SMEs free SEP advice, both for SME SEP holders and SME implementers, provide trainings on SEP licensing/FRAND negotiations and conduct studies concerning SEP-related issues. These activities will be available at least in English, free of charge and may be funded from the existing SME support mechanisms such as the SME Fund or the IP Helpdesk. 145  

5.2.2.Option 2 (PO2): SEP register with essentiality checks

PO2 builds on PO1 and establishes a SEPs register 146 where SEP holders that seek to license their SEPs in the EU will specify which patents they consider to be essential to a particular standard. Registration will be mandatory for enforcement purposes, such that if a SEP is not registered, the SEP holder will not be able to assert it in court and/or it will not be able to collect royalties or past damages for any use of a SEP prior to the date of registration. This condition will apply to patents for standards (or new versions of existing standards) adopted after the entry into force of the rules implementing this option. 147 A number of standards, however, rely on older technical documents in order to function (for example, a 4G phone should be able to also support 3G) and many IoT applications rely on 4G. Therefore, it will be possible to determine the scope of the standard by listing the relevant technical documents. All SEPs relevant to those documents will have to be registered. SEP holders can register their SEPs anytime following the publication of the relevant standard or the granting of their patents which are declared SEPs. Once a registration is done, any changes, such as ownership, invalidation or other court judgments should be updated within 6 months. 148  

For the registration, SEP holders will be requested to submit: i) full details regarding their patents protected in the EU claimed to be essential, including the patent office that granted the patent; ii) proof of their ownership; iii) identification of the commercial name of the standard for which the patent is essential; iv) information that helps to link the patent to the standard such as the specific section(s) of the standardisation document(s) for which the patent is potentially essential 149  and an identification of which features of the standard are normative or optional; v) evidence of the SEP owner’s FRAND commitment and any other relevant information, such as links to standard licensing terms and conditions as well as contact details. SEP holders will have to update the information in the register to reflect relevant changes (including e.g. validity of a patent).

The essentiality checks on SEPs included in the register 150 will be performed in accordance with a pre-determined EU methodology to be developed by industry experts under the auspices of the Commission. Essentiality checks could concern either: sub-option i) all SEPs in the register or sub-option ii) up to a fixed number (e.g. 50 or 100) 151  of patents selected by a SEP holder (or implementer) and a representative random sample of all SEPs in the register conducted annually to ensure the quality of the register. No SEP holder or other interested party will be able to choose the evaluator for any given essentiality check. The SEP holder will have an opportunity to submit a claim chart (a detailed document that links a standard to the claims of a patent) for each SEP that is checked. The SEP holder will receive the preliminary results of the essentiality checks and have the opportunity to comment before the opinion is issued.

The results of the non-binding essentiality checks will be published in the register (for each essential patent of the “up-to-100” group, and information on the percentage of essential patents for a given SEP holder based on random sampling). After the annual essentiality check process is completed, a SEP holder may request the cancellation of a SEP registration, if the result is negative. A certain number of the essentiality checks will be the subject of a peer evaluation, to be performed annually by a group of evaluators with significant experience. This will ensure the quality and consistency of the essentiality checks. Essentiality checks conducted by an independent person prior to the publication of the standard in the context of pools and essentiality determinations in courts will be reflected in the register. Those SEPs will not be re-checked for essentiality. Any court decision determining the essentiality or lack thereof will be reflected in the register.

Both register and essentiality checks will be managed by the Competence Centre 152 created under PO1. The persons that will conduct the essentiality checks (evaluators) will have to satisfy pre-determined criteria regarding competence and independence and will be designated and remunerated by the EUIPO. The essentiality check methodology and criteria for the selection of evaluators will be adopted as an implementing act under the Regulation.

There will be an administrative fee for SEP holders to enter SEPs into the register and to cover the costs of essentiality checks. Interested stakeholders will be able to access certain basic information in the register for free (for example the list of SEP owners). Fees will be required, however, from implementers and any other interested stakeholder to access detailed information from the register. SMEs (both SEP holders and implementers) will benefit from reduced fees.

5.2.3.Option 3 (PO3): SEP register with essentiality checks and conciliation procedure

This option adds a conciliation (bilateral FRAND determination) process to PO2. Conciliation aims at assisting parties in negotiations of a SEP licence with the FRAND royalty determination designed for their needs.

The Competence Centre will administer the conciliation procedure. It will create a “roster” of conciliators that satisfy the requirements of competence and independence, as well as a repository of conciliation reports, the confidential version of which would be accessible only to the conciliators. The conciliator would be a neutral party with extensive experience in dispute resolution and substantial understanding of the economics of FRAND licensing to be credible. The parties will be given an opportunity to agree on (a) conciliator(s) from that roster and failing to do so will empower a representative of the EU Competence Centre to choose the conciliator(s).

The conciliation procedure would have to be initiated by either the SEP holder or the implementer and concluded within nine months. Neither party would be able to initiate an action in a court of any EU Member State (either patent infringement or FRAND determination) until the conciliator has issued a report 153  (or optional: the procedure has been initiated).

The conciliator would examine the parties’ offers/counteroffers and consider the Huawei v ZTE negotiation steps among other relevant factors. The conciliator would have the power to request and receive information necessary to conduct the task (including access to the SEP register, essentiality checks and past conciliation reports), observing the level of confidentiality attributed to the relevant information in the proceedings. However, the parties would be free to engage in the proceedings and would not be prevented from leaving the process at any time or even not engaging at all. At the conclusion of the procedure, the conciliator would issue a report recommending a FRAND rate. Either party would have the option to accept or reject the conciliator’s report and/or recommendation. The conciliation procedure would be confidential, as well the report the parties would receive at the end, unless they settle beforehand. A non-confidential version of the report containing the FRAND rates and the methodology used (excluding any confidential information) will be provided to the Competence Centre for further publication. The SEP holder would also have to use the report or a notice of termination of the procedure before customs to support a request for customs action with regard to goods infringing SEPs.

The conciliation procedure would be conducted in English, unless parties decide otherwise. All costs of the conciliation process would be in principle shared equally between parties. Each party would bear its own costs in the process. As a support measure for SMEs, the EU Competence Centre would provide free advice to SMEs on SEPs licensing, the conduct of the conciliation and how the SME can best represent itself. SMEs would also be able to benefit from a reduced administration fee for the conciliation and potentially a discount or financial support from e.g. the SME Fund to cover the conciliator’s fees.

5.2.4.Option 4 (PO4): Aggregate royalty for SEP

This option adds to PO3 a process for determining (ex-ante) an estimated maximum aggregate royalty.

Ex-ante i.e., before (or shortly after) the publication of a standard, contributors to a standard should inform the Competence Centre about the maximum aggregate royalty for all SEPs essential to that standard. For complex standards with multiple implementations there may be a need for different aggregate royalties depending on the implementations known at the time of the publication of the standard. 154 It would be possible, however, to reassess the aggregate royalty after a reasonable period of time following the adoption of the standard. Standard contributors would have the right to ask the Competence Centre to appoint a neutral party (a conciliator from the roster set in PO3) to assist them in setting an aggregate royalty. 155  Additionally, a group composed of contributors to a standard and/or (potential) implementers would be able to ask the Competence Centre within a predefined time after the publication of a standard or after a new implementation becomes known, for an expert opinion on the aggregate royalty. An expert opinion would be delivered by a panel of three conciliators, following an open process where any party demonstrating a legitimate interest in the standard (and upon paying a participation fee) will be able to provide information to the panel. There will be a fixed time for the panel to deliver its opinion.

In case an aggregate royalty is not set by any of the methods described above, during the first bilateral FRAND determination (conciliation) procedure related to the relevant standard, the conciliator may make a recommendation for an estimated aggregate royalty (in addition to the determination of a FRAND rate for the parties’ specific dispute). Whether the conciliator makes such a recommendation would depend on the methodology it uses for the determination of the FRAND rate for the specific dispute at hand. Such an aggregate royalty will be specific to the dispute at hand. An aggregate royalty can also be determined or updated during any subsequent conciliation(s).

An aggregate royalty set by any of the above methods will be published in the register and will be non-binding on the negotiating parties. Under this Option there will be no determination of the allocation of the aggregate royalties among different SEP owners.

5.2.5.Option 5 (PO5): SEP clearing house

This option adds to PO4 a one-stop-shop facility for implementers. The aggregate royalty determined under the procedure under PO4 will become a binding aggregate royalty for those using the one-stop-shop. Once the aggregate royalty is determined, any implementer will be able to request a licence from all SEP holders by informing the Competence Centre and depositing in an escrow account either the full amount of the published aggregate royalty (in case of lump sum payment) or the amount corresponding to at least an estimated monthly sales volume of a product concerned (in case of per unit royalty payment). At fixed intervals (at least once a year) the implementer should deposit further royalty payments to the escrow account, provide evidence on real sales volumes (e.g. VAT, financial statements) and correct past royalties to actual sales. Any royalties not collected by SEP holders within a year from the deposit shall be returned to the implementer. The Competence Centre will charge fees to cover the costs of the “clearing” service.

The Competence Centre will notify all SEP holders who have registered, checked and confirmed SEPs in the EU register, so that they may conclude a licensing agreement with an implementer. SEP holders should also inform the Competence Centre how to allocate the aggregate royalty among them. Until they do so, they will not be able to collect royalty payments from the escrow account. Should SEP holders fail to agree on such method, the Competence Centre upon request from any SEP holder involved will appoint a conciliator from the roster set in PO3 to recommend an allocation method (the cost of the facilitator will be borne by all SEP holders involved). Once an allocation is established, any implementer will be expected to deposit in the escrow account the aggregate royalty. Each SEP holder will have a period of time to conclude a licensing agreement. The implementer will be reimbursed the part of those SEP holders that do not conclude licensing agreements. Following notification, SEP holders may not start infringement proceedings in any EU court against an implementer who deposited security for an aggregate royalty in the escrow account.

An implementer may choose not to use the one-stop-shop and to have bilateral licensing with SEP owners.

5.3.Logic of options construction. Could options be standalone?

The options are constructed following an incremental logic proposed already in the Call for evidence. 156 PO1 is voluntary and thus the least ambitious option. PO2 adds to PO1 a SEP register with essentiality checks. PO3 adds conciliation to the register. Conciliation could be a standalone option. However, in that case a conciliator will not be able to use the register and the databases of case-law and other information to complete its analysis. This means that if only one party engages in the conciliation in good faith, the conciliator will be compelled to use only the information provided by that party and any other public information it may find but not on the information available to it under PO1 and PO2. This will significantly reduce the value and credibility of its report. This is why it is important that a conciliator can benefit from the SEP register in making recommendations. This way (s)he can at least check which patents a SEP holder has and what share is essential. PO4 adds aggregate royalty to PO3. SEP holders can benefit from the help of the Competence Centre’s conciliators to propose an aggregate royalty. Stakeholders will also have the right to request a non-binding expert opinion on the aggregate royalty by a panel of three conciliators. If no aggregate royalty is announced or proposed by the panel of conciliators, an aggregate royalty may be determined by any conciliator who decides to use this methodology for FRAND determination in a bilateral process (as explained in PO3). Finally, PO5 can only work when both aggregate royalty (set in PO4) and SEP holders eligible to get a share of that royalty (PO2) are known.

5.4.Options discarded at an early stage

The following sub-options to those presented above were discarded following an initial assessment. For details see Annex 6.

In case of PO1, determination of the level of licensing by the Commission was discarded as currently there are different practices with different standards and there does not seem to be a one size fits all solution. Therefore, such an intervention in the contractual freedom would be disproportionate. 157 In case of PO2, a creation of up to 27 national SEP registers was discarded as a more expensive and cumbersome option. A SEP register without essentiality checks on a sampling basis to control the quality of the register was also discarded because it would have only helped to create some transparency with regard to standards to which blanket declarations are made but would not resolve the issue of over-declaration in existing databases. In case of PO3, a voluntary conciliation was not considered as already existing voluntary alternative dispute resolutions are not frequently used by stakeholders 158 and there are usually no negative consequences for non-participation. In case of PO4, a mandatory aggregate royalty determination after adoption of a standard was rejected as by that time products incorporating the technology are already placed on the market and the price uncertainty problem would not be solved. Moreover, SEP holders would then be in a stronger negotiating position vis-à-vis implementers, shifting the negotiation power to the benefit of the SEP owner. A royalty-free SEP licensing obligation was rejected because from an economic perspective whether or not a standard should be royalty-free depends on the economic interest of standard contributors to be subject to such IPR policy and there cannot be general rules that would determine that. Similar concerns could be raised in case of a uniform SEP royalty per standard irrespective of how the standard is implemented. This applies because a too high price could limit the usage of the standard by certain lower value implementations, while a too low one would reduce incentives to participate in standard setting for SEP holders. For options 1 to 5 a sub-option to cover all past standards (e.g. 3G) was discarded as too costly, because of the large number of patents that would need to be registered, and having the effect of reducing legal certainty for existing contracts. Finally, regarding the choice of an organisation to run the Competence Centre, the EPO was not chosen as changing its mandate would require consent of all 39 member countries, including 12 non-EU countries which are not bound by the EU SEP policy goals. SDOs were not selected as this would mean a dispersed, rather than centralised register, as well as because the EU does not have jurisdiction over non-EU based SDOs.

6.What are the impacts of the policy options?

This chapter presents impact of the options considered as viable. The analysis is qualitative and to the extent possible quantitative. As options build on each other and to avoid double counting we present only quantification of additional features of each option. Quantifications are based on several assumptions and estimations since data on SEP licensing is largely not observable (as explained in the problem definition). The purpose of quantifications is to allow comparison of options and to present relative impacts on affected groups rather than to provide an accurate figure. The fees presented below are just an indication of the fees that would cover the costs of the Competence Centre (based on an initial cost prognosis by the EUIPO). Details on quantitative analysis as well as all the assumptions used for calculations are presented in Annex A7.1.

6.1.PO1: Voluntary guidance

SEP guidelines are necessary to provide clarity on the most problematic and contentious issues in SEP negotiations. Those proposed under PO1 will be voluntary, thus not limiting the negotiating flexibility and customisation of terms and conditions to the specific needs of both parties. At the same time, they will shed light on current best practices and solutions, and can thus serve as a reference point especially for less experienced parties in negotiations (e.g. SMEs). The Guidelines could also be relatively easily and frequently updated to reflect e.g. new court rulings, or market developments.

The Competence Centre will centralise information from diverse public sources (e.g. different licensing policies, published or otherwise revealed (e.g. in court judgements), FRAND terms, conditions and SEP rates/royalties, SEP judgements etc.), produce ad hoc studies to identify SEP holders and their share in different standards, promote WIPO Alternative Dispute Resolution mechanisms, 159 give trainings and advise SMEs on SEP negotiations (e.g. around 80% of SMEs in our survey noted that they do not know strategies to defend themselves in SEP negotiations). 160  Therefore, it may become the first place to be visited by potential SEP implementers, especially SMEs, who are not experienced with SEPs.

In order to limit operational costs, it is proposed that all activities are carried out in English only. The technology firms applying modern standards are most likely already dealing with most of the technical specifications in that language, consequently using only one language is not expected to become an access barrier.

It is estimated that around 400 firms (including up to 80 EU based SMEs), 261 SEP holders and up to 500 other users (e.g. judges, legal counsels) would use services of the Centre each year. The annual costs of operating the centre are estimated at EUR 0.6 million per year. The benefits users could derive from the centre’s activities such as trainings or studies are estimated at EUR 5.9 million (based on cost of buying such services commercially). 161

While guidance will be important and will help to improve transparency and knowledge about SEP licensing, including by providing specific advice to SMEs, it will not be sufficient to help resolve disputes around FRAND royalties.

The public consultation clearly revealed several major issues: e.g. lack of transparency about the FRAND royalties (68% of all respondents, 100% of implementers and 19% of SEP owners), lack of transparency about the SEP landscape (67% of all respondents, 97% of implementers and 13% of SEP owners) or no guidance on FRAND concepts (57% of all, 93% of implementers and 19% of SEP owners). 162

The Competence Centre can gather all public information about pricing but much of it is under non-disclosure agreements. A guidance is not capable of promoting more transparency in this respect. The Competence Centre may also commission SEP landscape studies on a regular basis, but it will need to use existing third parties’ commercial solutions with their limitations. The limitations concern both the information available to these third parties (e.g. missing data and over-declaration mentioned above) and the different methodologies they use. Such methodologies are deemed commercial secrets, have a different level of robustness, and thus may lead to different results.

Finally, while the WIPO Arbitration and Mediation Centre is reporting some positive results, the public consultation revealed that only 35% of all respondents consider mediation useful. Two key concerns are (i) implementers may not be able to obtain information regarding comparable licenses and other relevant data from SEP holders (available through discovery proceedings in certain judicial proceedings), and (ii) the proceeding itself as well as the resulting decision is maintained as confidential and is not publicly available, so it has limited precedential value for future FRAND-related cases.

In order to offer valuable information to interested parties on the SEP landscape, FRAND licensing and royalties, the Competence Centre would need to develop its own data on the basis of independently conducted essentiality assessments and FRAND determination procedures, which this option is unable to achieve.

This option bears similarities with SEP Expert Group Proposal 28 on three licensing principles. 163

6.2.PO2: SEP register with essentiality checks

This option is expected to bring more clarity to the SEP landscape. The register under this option will not replace SDO databases, the purpose of which is to create transparency ex-ante during the standardisation process. It will be created ex-post, after the publication of the standard for the purpose to facilitate licensing negotiations and will be updated regularly. As such it is complementary to the activities of the SDOs.

SEP holders will have to register their SEP(s) in order to be able to enforce their patents and/or collect FRAND royalties for the past. This is reasonable given that on average the SEP holder invites the implementer to conclude a licensing agreement within 2 to 4 years after the publication of a given standard. The registration will be simple (comparable to the declarations at ETSI) and provide information to implementers about the identity and contact details of SEP owners. Moreover, details provided (e.g. link between standard and patent) should help an implementer to judge if the patents cover the functionality/technology it uses (as not all implementers are using the full potential/all functionalities of a standard). 164  

The initial stock of patents that could be registered is estimated at maximum up to 72 000 patent families. 165  Some of the new standards may need to include older technical documents (for example a 5G mobile phone will need to also support 4G). So, all SEPs relating to those older technical documents will also have to be registered. The register will serve a dual purpose. First, the register will be used to prove that a SEP holder has true SEPs. Second, the register will provide additional information regarding the size and value of the SEP owner’s portfolio, if the SEP holder registers all its SEPs.

Since the register may include patents that are not truly essential to the standard, it is necessary to have a mechanism for essentiality checks to ensure the quality of the register. With regard to 4G, experts claim that only between 25% and 40% of SEPs declared to ETSI are truly essential. 166  It is unlikely that the essentiality rate of the new register will be so low. This is because the ETSI database includes patents and patent applications made during the standardisation process. Some of those technologies were not accepted in the standard and some of those patents were not granted or their scope was reduced so that they do no longer fall under the scope of the standard.

The new register will be created after the publication of the standard and will require that the SEP holder indicates the section of the standard to which its patented technology refers. The rate of essentiality would thus be higher. However, it cannot be excluded that patents that are not essential are also included. One should note that no assessment can give 100% assurance that a patent is truly essential until a court has ruled on the essentiality (even then, the court would rely on experts). A recent JRC study found that assessors correctly identify as essential or not around ¾ of patents they check. However, the authors also explain that this share will be higher in practice due to improving routines, and because identifying non-essentiality of the patents in the experiment was particularly difficult (the examiners had a lot of limitations, including no access to support tools and support group). 167

The cost of an essentiality check is estimated at EUR 5 000 per one patent from a patent family. 168 There is no need to check the other members of the family as most European patents have the same claims being examined by the European Patent Office. The cost should also cover the peer evaluation process for some of the patents already checked for essentiality.

A SEP holder will be given the opportunity to submit a claim chart (a document linking the standard to the patent), if a registered SEP is selected to be checked for essentiality. 169 According to experts, all SEP holders who proactively license or cross-license their SEPs have claim charts or internal expertise to create them. The most impacted will be the patent assertion entities that buy the patents from others solely to license for revenue. We estimate that for about half of the patents selected for random essentiality checks, SEP holders will decide to update or produce new claim charts at a cost similar to the one indicated above.

In sub-option (i) all registered SEPs will be checked for essentiality. This will result in the most accurate SEP landscape analysis (subject to the uncertainties mentioned above, including the portion of SEP patents registered by owners). 72 000 patent families are likely to be checked in the first year(s) (registering only those of the family members that are in force in the EU). Following a negative assessment, SEP holders are likely to request re-check for up to 40% of patents. 170  In the following years, the number of patents registered and checked is expected to drop to 10% of the initial numbers.

To implement this option the back-office costs of the Competence Centre are estimated at around EUR 5 million initially and EUR 1.5 million in subsequent years. In order to cover this cost SEP holders will be charged a registration fee of around EUR 140 per patent family registered.

The average total cost to SEP holders of registration and essentiality checks is estimated at EUR 1.9 million and EUR 94.5 million per year respectively. Additional indirect cost of filling the forms and production/update of claim charts is estimated at EUR 2.7 million and EUR 34 million respectively per year. 171 Around 15% of that cost will be borne by SEP holders based in the EU. As indicated in the market description almost all EU based SEP holders are large companies.

The register is expected to give a fairly accurate overview of the number of true SEPs and their ownership.

In sub-option (ii), the same number of patent families will be registered but not all will be checked for essentiality. The checks will be done based on two complementary steps. First, the SEP holders may select up to 100 patent families of which one patent from a family will be the subject of an essentiality check. These are expected to be patents they use often in licencing negotiations (typically up to 15 patents are discussed according to experts, an average of 49 SEPs was indicated by SEP holders in the public consultations). 172 Having checked patents might also facilitate/encourage participation in patent pools as some require essentiality checks. Under this step, it is expected that up to 3 550 patents will be initially submitted with 355 added each year. Following negative assessment around 25% re-check requests are expected (lower percentage than in sub-option i) reflects better quality of SEPs used in negotiations). Second, the Competence Centre will select a random sample from all SEPs registered by each SEP owner, including those designated by the SEP holders under the first step. This is necessary to ensure a statistically valid sample. The Commission will commission a study to develop the sampling methodology that would render the most optimal results. If the sampling includes a patent also selected by the SEP owner, there will be only one essentiality check. Our study 173 finds that the effects of checks conducted based on sampling from the register will be similar to those discussed under sub-option (i) with an acceptably low margin of error. However, sub-option (ii) will limit the number of checks to around 10 000. 

To implement this option the back-office costs of the Competence Centre are estimate at around EUR 3.7 million initially and EUR 1.1 million in subsequent years (figures lower than in i) due to lower workload). In order to cover this cost SEP holders will be charged a registration fee of around EUR 100 per patent family registered.

The average total cost to SEP holders of registration and essentiality checks is estimated at EUR 1.35 million and 13.7 million per year respectively. Additional indirect cost of filling the forms and production/update of claim charts is estimated at EUR 2.7 million and EUR 6.4 million respectively per year. 174 Around 15% of that cost will be borne by SEP holders based in the EU.

The register is expected to give as accurate overview the number of true SEPs as in sub-option i). However, except for patents in the “up to 100” list only a percentage of essential patents in a SEP holder’s portfolio will be indicated. 175 Additionally, the register will reflect any decisions of essentiality by courts and for patents that are examined by an independent examiner of a patent pool prior to the creation of the register. This is however sufficient to allow comparison between portfolios of different SEP owners.

Since sub-option i) is ten times more expensive than sub-option ii) and provides only a marginally more accurate picture, the preferred choice should be sub-option ii).

Implementers and interested stakeholders will have free access to some basic information (for example the list of SEP owners, number of registered SEPs, etc.) but they will be required to pay a fee to access detailed SEP information in the register and the various databases of the Competence Centre. This requirement has two objectives. First, it will enable the development of high-quality data that can be used for aggregate royalty and bilateral FRAND determinations. Second, it will give SEP holders the reassurance that the data they provide is accessed by persons interested in SEP licensing and their representatives, and to specialised professionals. We estimate that at least 380 firms that conduct license negotiation per year will buy access. The fees will cover cost of activities of the Competence Centre described in PO1 and at minimum should amount to around EUR 1 700 for large firm and EUR 850 for an SME for an annual subscription.

SEP holders are likely to withdraw non-essential patents from the register following negative checks and save on patent maintenance fees. This should be considered beneficial to all parties – SEP implementers will have more certainty in negotiations in which only essential patents are presented to them. On the other hand, if the register will be perceived by SEP holders as a means of indicating portfolio strength (and e.g. used in negotiations to determine the share of aggregate royalty applicable to them), they may increase the number of registered patents. 176 Under sub-option i) as all patents are checked, the withdrawals are likely to be higher than new registrations. This is expected to produce net savings for SEP holders of EUR 11.6 million per year (15% applicable to the EU based SEP holders). The savings to SEP holders are losses to the EPO and EU patent offices. In case of sub-option ii) in which a lower number of patents is checked, the opposite is expected. SEP holders are likely to patent more than to withdraw, with a net impact of EUR 29 million in new patent registration/maintenance fees to the benefit of the EPO and EU patent offices. Nevertheless, the impact on the number of patents is uncertain and may not materialise (see also common impacts on patent offices below).

The costs mentioned above should be seen in comparison to an average license revenue a SEP holder gets on its patents. An average worth (net present value) of a positively assessed patent family would be approximately EUR 6 to 10 million (EUR 1 to 1.7 million net present value if the effect of the checks is limited to the EU only). 177

The objective of the register is to have all SEPs in force in the EU registered, irrespective of who the SEP holder is. With the help of essentiality checks, implementers will know approximately what share of the registered SEPs are truly essential. This will determine the scope of the negotiations.

SEP holders will receive a powerful argument when approaching implementers with a licence offer that they have truly essential SEPs and are thus more certain to have the right to ask for royalties. The evidence from the register will allow a SEP holder to justify the FRAND rate demand, and the implementer can better determine whether the offered rate is FRAND. SEP holders with smaller portfolios, who could not afford the costs of licensing until now, would be able to license.

The option also provides efficiency gains as implementers will no longer need to assess essentiality on their own or negotiate without that knowledge. With an estimated around 575 new SEP licensing agreements signed per year (and associated negotiations), the value of information in the register to implementers and SEP holders is estimated at EUR 6.2 million each. 178 These benefits could increase if the number of SEP licensing contracts is going to accelerate due to, among others, a growth in importance of IoT. Almost two thirds of respondents to public consultations (93% implementers and 24% SEP owners) considered that a system of essentiality checks would be useful if assessors are independent. 179  

Moreover, it is expected that the registered SEPs will reflect better the size of the SEP holder portfolio than currently. This is because the SEP holders will have to review their declarations to ETSI or review their portfolio if they made a blanket declaration and register only those SEPs whose claims can be linked to the standard. Since the registration will require them to also indicate the section and features of the standard to which the SEP relates, it will make it easier to determine which SEPs pertain to core features of the standard and which to optional parts of the standard. This will help level the playing field among SEP holders.

Finally, conditioning patent enforcement or licensing on registration (with no possibility to collect royalty for the period before registration) might be considered an encroachment on property rights. This is, however, a justifiable and a very limited restriction. First, the main problem is that either SEP holders do not declare each of their SEPs to the SDO or they over-declare. The choice is made based on the individual business strategy of the SEP owners, but it is detrimental to the SEP licensing system as a whole. Because of the lack of transparency, some SEP owners’ portfolios are devalued, and others are over-valued; furthermore, implementers face business uncertainty. The lack of predictability has a direct impact on both contribution to a standard and standard implementation. Second, the registration is simple and corresponds to what the SEP holders already committed to declare in the ETSI database but do so with a various degree of rigorousness. The registration will normally not impact on any negotiations as it can be done much earlier than the start of the negotiations. Pursuant to the public consultation, negotiations start on average 2 to 4 years after the publication of the standard and last for an average of 3 years.

This option will benefit SEP holders by reducing delays resulting from implementer requests for, or need to obtain their own, essentiality checks and technical documents. This in turn reduces the amount of negotiation time resulting in SEP holders starting to collect royalties earlier.

Around two thirds of all respondents (around 85-89% of implementers, and between 24 and 41% of SEP owners) named the following benefits of a system for essentiality checks: better information regarding the actual SEP exposure of a given product; reduction of the resources spent on licensing of SEPs; it may help to smoothen licensing negotiations; easier negotiations of a fair royalty (preventing over-pricing); trustworthy and reliable overview of the share of each SEP holders’ essential patents. 180

This Option bears similarities with SEP Expert Group Proposals 7, 8, 11, 13 and 14 on introducing essentiality checks by an independent body for those SEPs that SEP holders intend to commercialise. All those proposals received support with 4 out of 5 stars by the experts. It also bears similarities with Proposals 50 and 51 that when a SEP holder asserts its SEP, it should disclose certain information on those SEPs and high-level claim charts. 181 Those proposals were supported with 4.5 out of 5 stars by the experts.

6.3.PO3: SEP register with essentiality checks and conciliations

Conciliation is a process for a FRAND rate determination between a SEP holder and an implementer (prospective licensee). The objective of the conciliation is to facilitate a settlement between the parties before resorting to court. For example, WIPO reports that 70% of its mediations and 33% of arbitrations end up in settlement. 182  

53% of the respondents to the public consultation considered arbitration useful and 35% of the respondents considered mediation useful. Some of them explained that the problem with arbitration and mediation is that the results are not published. For mediation, an additional disadvantage was that the mediator has no authority to request information and no authority to make a price recommendation. 183  

The conciliation thus has all the additional features that have the potential to address those concerns. It provides powers to the conciliator to (i) request relevant information and documents from the parties, (ii) propose a recommendation for a FRAND rate and (iii) to issue a non-confidential report (containing the methodology for the calculation of the FRAND rate).

Since conciliation will likely take place before concluding a licence agreement, and it will be a mandatory step before initiating a patent infringement proceeding or a FRAND assessment before a court, it can be even a more effective tool in limiting the number of court cases. The ability to negotiate a FRAND rate with the assistance of an independent conciliator without the threat of an injunction is important for any implementer. Conciliation can also help SEP holders (e.g. when negotiating with large multinationals). The SEP holder who has made a FRAND offer confirmed by the conciliator will increase its credibility also for future negotiations. Credibility is important as evidenced by this example: The WIPO Centre has seen one party make systemic use of WIPO Mediation – filing numerous requests for WIPO Mediation as one way of demonstrating its willingness to take a licence. 185

Moreover, the conciliation would be more acceptable than arbitration to the parties (where a decision is final) as a conciliator only issues a non-binding suggestion and report on FRAND terms and conditions. The conciliation report will contain a factual summary of the process before the conciliator and include information whether the SEP holder has made a FRAND offer and whether the implementer has responded to that offer. This summary will de facto also cover the information needed to assess whether the SEP holder engaged in the Huawei v. ZTE process with an implementer. The conciliator’s suggested FRAND royalty could also be used by both SEP holders and implementers to determine the appropriate amount of the security that the implementer needs to provide under the Huawei v ZTE process. The non-confidential report (or notice of termination of the procedure) could also be used before customs to support a request for customs action with regard to goods infringing SEPs. 186

Although the conduct or commencement of the conciliation will be obligatory before initiating a court action, it will be up to the parties to decide on their level of engagement (e.g. participation in meetings, providing supporting documents on request by the conciliator). In an extreme case, the conciliator could draw up an opinion based on input from just one party if the other decides to ‘boycott’ the process. This is why it is important that a conciliator has access to the SEP register, the databases of the Competence Centre and confidential reports from other conciliations, so (s)he can receive information on the concerned patents in any event. 187  

During the webinar series on SEP enforcement, it was underlined that parties prefer to agree on FRAND terms rather than having a third independent party determining the FRAND rate for them, even if it would be not binding. Furthermore, the participants in the webinar noted the need of specialised knowledge for FRAND determinations.

Mandatory conciliation should benefit both SEP holders and implementers in reaching a licence agreement faster without costly court proceedings. Since the procedure would be limited to maximum nine months, it would expedite the negotiation process thus making delaying tactics on either side less attractive. This should be especially useful for SMEs (according to WIPO in 94% of SEP ADR at least one party was an SME). 188  

On the other hand, the procedure may potentially lead to delays in licence payments. However, past payments are either resolved in the license agreement or the SEP holder can request damages for past use before the court, albeit depending on national laws requiring commencements of national proceedings before Member States’ courts of choice.

An indicative hourly wage of a conciliator is about EUR 500. 189  We estimate that a conciliator would spend on average around 40 hours per case (e.g. as a conciliator may need to come to a conclusion and draw up a report in case the parties do not settle). With estimated around 70 conciliation cases per year (35 based on the current number of FRAND related court cases in the EU and another 35 not related to court proceedings), the cost of back-office support to conciliation is estimated at EUR 800 per case. Thus, the total cost of a conciliation should amount to around EUR 20 800. This is eight times lower that the average SEP court cost in the EU of EUR 170 000. 190 Based on WIPO experience (70% settlement rate) we estimate that up to 24 court cases could be avoided. Additionally parties could receive help on FRAND terms determination in at least 35 cases per year. Altogether the additional benefits of this option are conservatively estimated at EUR 6 million and could increase further as the mechanism becomes more known and used.

This Option bears similarities with the Expert Group Proposals 68 and 70 on establishing an independent board of experts for determining FRAND royalty upon request and a specialised mediation institute for FRAND licensing disputes. 191  Proposal 68 was supported with 4.5 out of 5 stars and Proposal 70 was supported with 4 out of 5 stars by the experts.

6.4.PO4: Aggregate royalty for SEPs

Aggregate royalty can be considered a total potential cost of licensing a standard. 192 Economically, an aggregate royalty makes sense because implementers value the standard as a whole, not a collection of inventions represented by an incomplete portfolio of SEPs. It may also help to overcome problems of royalty stacking, a modern version of Cournot’s (1838) 193 well-known problem that the independent pricing of complementary goods – here: licenses to different SEP portfolios on the same standard – leads to excessive prices, above those that a single party offering all those goods jointly would charge. As Contreras (2017) 194 notes, “[t]op-down approaches [i.e. aggregate royalty] avoid many drawbacks associated with bottom-up approaches in which royalties for individual SEPs are assessed, often in an inconsistent and piecemeal manner, without regard for the other SEPs that cover the standard.”

70% of all respondents to the public consultation and 100% of implementers considered that it is important to know the aggregate royalty. Views of SEP holders were mixed, with more (40%) considering it not important, than those who thought it was (20%). Moreover, 70% of all, 93% of implementers, and 25% of SEP holders considered it important to have a transparent process for aggregate royalty determination (38% of SEP holders disagreed).  195

Knowing it for different applications would simplify cost planning for SEP implementers and assessment, if their business model is profitable. As Henkel (2022, Section 4.1) shows, these are very real concerns for SMEs implementing SEPs: “During the development of an IoT device, it is unknown which demands for SEP royalties will surface later, so innovators cannot make reliable cost calculations.” For SEP holders it will allow to decide if they wish to contribute their technology to a standardisation process and help estimate potential future revenues.

Under this option standard contributors may announce jointly an aggregate royalty to the Competence Centre. Finding an agreement among contributors might however be a challenging task due to often opposing interests of contributors. Some may also be implementers and thus are likely to favour lower royalties, while others are “pure” SEP holders whose primary source of income is licensing and who are interested in higher aggregate royalty levels. To help contributors in reaching an agreement, the Competence Centre upon request may appoint a conciliator to facilitate their discussions. Finally once royalty is set, it will not be for ever, there will be a mechanism to adjust the aggregate royalty to reflect changing market conditions (e.g. the fact that a standard ages with time).

The aggregate royalty could be irrespective of the level of licensing 196 , and it should be possible to pass the cost of the licence downstream without a mark-up (the commission usually charged by the various levels in the value chain on top of the price). As per PO1 guidelines, the industry is expected to figure out the appropriate level of licensing to avoid the so called “double dipping” (collecting royalties twice for the same SEPs).

From the point of view of an implementer, one risk could be that the aggregate royalty announced by the standard contributors is too high. SEP holders might set it high strategically in order to have room for negotiation.

This may, however, dissuade implementers from taking a licence, which should have a self-limiting effect on SEP holders, as wide implementation of the standard is of interest to the SEP owners. Nevertheless, in order to cater for such situations, a group composed of standard contributors and/or implementers may ask the Competence Centre for a non-binding expert opinion on the aggregate royalty. The process will be inclusive where any interested party would be able to submit its views. We estimate that around three such requests will be coming each year and that the cost of delivering one expert opinion (consisting of approximately 240 working hours of experts’ time and Centre’s back-office support) is estimated at around EUR 135 800. The costs would be shared by requesting parties.

Besides predictability, this option is also expected to facilitate SEP license negotiations. Asked about problems in negotiating with SEP owners, 90% of respondents to the SME survey rated a fair price for SEPs as either a “very important” or an “important” issue. 197  In fact, SEP licensing negotiations are mainly about royalty rates. By combining an aggregate royalty with a SEP owner’s share in all essential patents for a standard (from PO2), one obtains a proxy (albeit imperfect) of a royalty level to be expected. This could establish a reference point to help negotiating parties in reaching an agreement. It might be especially useful for smaller and SME implementers (who have limited resources to conduct own research or gather evidence), but also for (smaller) SEP holders who could underpin their FRAND offer with additional, publicly available evidence.

With an estimated 575 SEP licensing negotiations taking place in the EU each year, we conservatively estimate the value of publicly known aggregate royalty rates at approximately EUR25.3 million per year.

In case licensing negotiations fail, an aggregate royalty would also facilitate the work of conciliators and judges by providing a clear reference maximum price for using a standard. This will not prejudice the use of other valuation methods in bilateral FRAND determination processes. For instance, during any conciliation process, the conciliator may choose the methodology (s)he uses to propose a FRAND rate and may also choose to look at the aggregate royalty. The non-confidential version of the report will be published and will inform any future discussion on an aggregate royalty.

The publication of aggregate royalties is not expected to impact on innovation by SEP owners, as exemplified by the following research:

For a more developed analysis of economic rationale for an aggregate royalty please consult Annex A7.2.

This Option bears similarities with SEP Expert Group Proposals 38 and 42 on encouraging SEP holders to announce ex-ante most restrictive licensing terms and determine a reasonable aggregate royalty. 200 Both proposals received support with 4 out of 5 stars by the experts.

6.5.PO5: SEP clearing house

This option depends on an aggregate royalty established above, and adds a clearing house facility, or one-stop-shop for implementers to acquire SEP licenses with a single (annual) payment and without any negotiations.

PO5 would be most helpful for mid-caps and SMEs developing solutions in the IoT as it will enable them to simply deposit the aggregate royalty in an escrow account and then just sign standard deals with SEP holders. 201  One third of respondents to the SME survey quoted too much efforts to enter license agreements with all SEP holders as a reason for not obtaining a license; and around 80% complained they did not have sufficient resources to negotiate with SEP holders. 202  However, any implementer may always choose not to use the clearing house and negotiate with SEP holders bilaterally instead, to receive better conditions.

The initial escrow account would be based on the estimated volume and value of sales by the implementers. Once SEP holders are informed by the Centre, those that are interested in a licence will reach out to the implementers concerned to conclude a licensing agreement on standard terms. This agreement will detail the sales and the monitoring mechanism and will regulate the relationship between the SEP holder and the implementers. The amount allocated to SEP holders that chose not to conclude a licensing agreement will be returned to the implementer within one year of the deposit.

Given the fragmentation of the IoT and the large number of potential licensees this option should also benefit SEP holders. In particular, those SEP holders with small portfolios (e.g. SMEs) who have limited resources to engage in bilateral negotiations with many (or larger) implementers. The clearing house would allow them to reach a wider audience of implementers at limited cost. All money transfers will be dealt with by the Competence Centre.

There are two main challenges regarding this option. First, the aggregate royalty from PO4 would have to become binding on the parties. Consequently, a mechanism should be established to allow both updates and challenges of this royalty (e.g. in case it is set at too high level, or when the standard ages). However, changes to the aggregate royalty levels should not occur too often, as this would be rather disruptive to the licensing process, especially as regards predictability.

Second, the problem is how to incentivise SEP holders to agree on the allocation of the aggregate royalty between them as they will be able to access escrow deposits only after doing so. In practice, it may not be possible for the SEP holders to agree on allocation methods because not all SEP holders license their SEPs for remuneration, and they may have diverging commercial interests. For some licensing is a business model, for others this is just a gateway for building business relationships. To get to an agreement SEP holders may use services of Centre’s facilitator. We estimate that on average there would be three such request per two years. Remuneration of facilitator(s) is estimated at EUR 120 000 per case and will be borne by the requesting parties. In any case however, it is likely that at least some SEP holders will be dissatisfied either with the level of binding royalty or allocation method. This can result in numerous legal challenges against both the Commission and the Competence Centre.

Nevertheless, if the allocation is agreed upon, the transfer of royalties should be automatic without the need for individual negotiations. We estimate that the majority of implementers will want to use the clearing house for at least some of the standards they use. We estimate that this option has the potential to eliminate the need for up to 500 negotiations per year, producing net benefits of up to EUR 76 million annually. Fees for using the service and covering cost of the Centre are estimated at just below EUR 100 per payment. 203

A potential drawback of this option from the point of view of implementers is that more SEP holders (also those who did not actively seek royalties in the past) might register to get their share of royalties. So, it is likely that more SEP holders will decide to license their SEPs, e.g. once Avanci became known and used by vehicle manufacturers, the number of SEP holders grew from an initial five in 2016 to around 20 in 2018). 204 In practice this could mean that the implementers will pay (almost) the full aggregate royalty. Currently not all SEP holders seek royalties, so the implementers are likely to pay only to those who proactively request royalties. Second, it may happen that due to ease of licensing under the new system, standards that thus far were in practice royalty free could become royalty bearing.

This option bears similarity (but is different from) Proposal 74 from SEP Expert Group “On demand collective licensing agencies”. 205  This proposal received 2 of 5 stars support from the experts. To the extent that it bears similarities, the objections to that proposal would apply. Those include that mandatory patent agencies would create strong disincentives for companies to participate in standard setting and mandatory patent agencies would result in that SEP holders only licensing their SEPs defensively, would collect licence fees through the agency, contrary to their policy not to seek licence fees from other users of a standard.

6.6.Common impacts

The general objectives are to ensure that end users, including small businesses and EU consumers benefit from products based on the latest standardized technologies at reasonable prices; and to make the EU an attractive place for innovation and standards development. Ultimately, the initiative wants to ensure that both EU SEP holders and implementers innovate in the EU, make and sell products in the EU and are competitive in non-EU global markets.

Specifically, the initiative aims at providing the tools for the parties to negotiate and conclude a FRAND license agreement successfully. All options aim at facilitating SEP licence take-up in the EU by reducing information asymmetries and redundant transaction costs. The initiative is thus likely to balance the negotiating power between SEP holders and implementers. It should provide SEP implementers with predictability as to their SEP exposure, allowing for proper business planning and usage of the latest technology in the EU. It should also ensure sustainable licensing income for SEP holders thus enabling them to continue participating in standard development. It would promote a level playing field for SEP holders with good quality SEP portfolios.

The precise impact of the options on SEP holders and on costs of implementers is difficult to predict. As transaction costs go down (progressively with each option) it will be easier for implementers to take a licence (options mainly reduce information asymmetry that implementers currently face) and thus, not to infringe on patents. SEP holders would also benefit from such wider market increasing their licensing income base.

Impact on fundamental rights

This initiative may impact the right to intellectual property of patent holders (article 17(2) of EU Charter), if it is viewed as a restriction on the ability to enforce individual patents within a standard for which a FRAND commitment was made, but that has not been registered. The same may apply to the requirement to conduct a conciliation procedure before enforcing individual patents. IP rights are not absolute rights and limitations to the exercise of these rights are allowed under the EU Charter, provided that the proportionality principle is respected. According to settled case-law, fundamental rights can be restricted, provided that those restrictions correspond to objectives of general interest pursued by the European Union and do not constitute, with regard to the aim pursued, a disproportionate and intolerable interference, which infringes the very essence of the rights guaranteed. 206  In that respect, this initiative is in the public interest in that it provides uniform, open and predictable information and outcome on SEPs for the benefit of SEP holders, implementers and end users, at EU-wide level, and in that it aims at promoting technological innovation and the dissemination of technology to the mutual advantage of the SEP holder and implementers. Furthermore, the rules concerning the conciliation procedure and the determination of the aggregate royalty are aimed at improving and streamlining the process, but are not ultimately binding.  207

The initiative should also improve the conduct of business for both SEP holders and implementers, and ultimately other businesses downstream (Article 16 of the Charter).

The initiative is also entirely consistent with the right to an effective remedy and to access to justice (Article 47 of the Charter), as the implementer and the SEP holder fully retain that right. In case of failure to register, the exclusion of the right to effective enforcement is limited and necessary, and genuinely meets objectives of general interest, described above (see the first paragraph of this Section). 208  

Impact on innovation

Reducing uncertainty about costs and legal exposure will have a positive effect on implementer-level innovation. Reasonable SEP licensing costs and a reduction of negotiation delays will have a positive impact on SEP holder innovation. Both SEP holders and implementers may experience negative impact if a balance is not achieved in the determination of the FRAND terms.

A more sustainable standardisation system and the availability of open standards to a wider range of implementers will have a positive effect on innovation that builds up on those standards.

SEP licensing is expected to remain a profitable business after the options considered are implemented. For instance, the total costs (including indirect costs and excluding benefits) for SEP holders of PO2 of EUR54 million annually account only for 0.3% of the estimated global royalty payments on cellular standards alone (EUR 18 billion). 209 Thus, incentives to innovate are likely to continue. 210  Two thirds of respondents to the public consultations considered that efficient SEP licensing would foster innovations by implementers, including start-ups and SMEs. 211

Social and environmental impacts

The options have no direct social impacts. Indirectly by facilitating take-up of new technologies they can help in the digital transition such as e.g. smart cities, connected cars, telework, etc. More jobs could emerge in the new technology sectors.

The options have no direct environmental impacts. They comply with the “do no significant harm” to the environment principle. Indirectly, by facilitating application of new technologies they can help in the green transition e.g. by contributing to a reduction of energy usage (via smart grid, smart meters) or air pollution (e.g. connected cars avoiding traffic jams).

Two thirds of respondents to the public consultation considered that efficient SEP licensing would i) increase employment and keep a high level of competence in the EU and ii) foster the EU’s transition to the green economy. 212

Impact on SEP holders’ royalty income and availability and cost of products to customers

Many implementers currently use standards without having a license. Thus, the effect on availability of SEP embedding products is unknown. We expect that lower transaction costs will promote license-based applications and more implementers will take a license. It may be just a shift from unlicensed products to licensed products, or real increase in technology take-up resulting in more SEP embedding products on the market (also due to expected increase in IoT applications).

The impact on SEP prices (royalties paid by implementers) is also unknown (see Annex A7.1). Announcements of aggregate royalties and FRAND determination processes may (but do not have to) contribute to lowering the royalties paid by implementers.

Consequently, we have two effects working in different directions: i) potentially more firms taking a license (increasing implementers costs, and income of SEP holders) and ii) potentially lower royalties paid (decreasing implementers cost and income of SEP holders). Finally, the impact on prices for customers will depend on the competition on a given product market. Any change in royalties paid by producers may be internalized by a firm or passed on to final customers.

Impact on national patent proceedings

The additional information on essentiality shares, on aggregate royalty as well as information from conciliator reports are likely to facilitate judgements in SEP related cases. Moreover, mandatory pre-trial conciliations are expected to reduce the number of SEP litigation by up to 70%. The benefits for the national judiciary are estimated at EUR 2.8 million in terms of access to information, and at EUR 4 million in terms of reduction of the number of court cases. In the analysis these benefits were attributed to SEP holders and implementers.

Impact on patent offices

Options starting from PO2 may create additional demand for European patents, in case the share of SEPs in the register will become a reference point for aggregate royalty allocation. Especially larger stakeholders may take advantage of the fact that there is certain probability that a non-essential patent will be considered essential by evaluators. 213  This effect however is likely to diminish as evaluators gain experience. Thus, an increasing number of SEPs in the register may improve SEP holders position vis-à-vis other holders. We estimate that in case this effect occurs, it may result in additional income to EPO/national patent offices of around EUR 29 million annually (in case PO2 sub-option ii) is chosen).

Impact on SEP patent pools

All options will have positive impacts on patent pool formation and patent pool licensing.

First, all patent pools conduct essentiality checks for some or all patents included in the pool. In the future, SEP holders may include in the pool SEPs that were found essential under PO2. 214 Merges and Mattioli (2016) e.g. estimate that the cost of these essentiality checks related to the MPEG Audio pool represented USD 5 250 000, approximately two thirds of the total set-up cost of this patent pool. Patent pools may continue to do their own essentiality checks.

Second, the requirement to announce an aggregate royalty is likely to promote patent pool formation as independent patent pool administrators have significant experience in pool formation. They also have internal revenue allocation mechanisms, which may increase transparency with regard to the share of the various SEP holders and facilitate their business planning.

Finally, if a patent pool comprising all SEP holders relating to a given standard would be formed, it would replace the use of PO5.

Impact on commercial database providers

Providers of commercial solutions include companies such as LexisNexis Patentsight, Clarivate Derwent Innovation, Questel Orbit, Patsnap, Minesoft Patbase and many others. Recently such companies have integrated SEP declaration information in their databases; i.e. patent datasets include a flag whether or not a patent has been declared. Some solution providers such as IPlytics, PatentCloud Inquartik, Patently or Unified Patents have created designated patent declaration analytics solutions that allow searching across the full text of patent and standard documents. 215

These providers may have access to the register, use it in their commercial offers and build services on top of that data. There may be new users and thus enhanced competition. Providers that are currently focused on collecting and curating the currently missing or imperfect data available are likely to lose that part of their business. The companies above, however, all offer additional services. While users may no longer need to purchase commercially produced reports on the SEP landscape, they may benefit from the price decrease and quality increase in new analytical commercial reports induced by the free availability of the Competence Centre’s information on essentiality rates or aggregate royalty.

Digital impacts

The SEP database may be used for research purposes, and/or to train machine learning (ML) algorithms to recognize “true” SEPs among larger populations of potential SEPs. These methods are expected to significantly improve over the next years, and the policy options may further contribute to this evolution. The causal effects of improved ML approaches to essentiality determinations are complex, but potentially significant. Among others, it may limit significantly the costs of essentiality checks.

7.How do the options compare?

Guidelines and support measures created under PO1 will support SEP license negotiations and improve the SEP licensing environment mainly by increasing transparency and reducing some costs. Their impact, however, may be limited. The assistance package for SMEs can be seen as the biggest advantage of this option. As explained under PO1, much of the public information about pricing is under non-disclosure agreements. A guidance is not capable of promoting more transparency in this respect. Any SEP landscape study will suffer from natural limitations such as missing data and over-declaration, and the use of different methodologies to produce them, which may lead to different results. WIPO mediation will not address the key concern about transparency of the results of the mediation that are to benefit future comparable situations. In order to offer valuable information to interested parties, the Competence Centre would need to gather its own quality data through unbiased and transparent processes. This option is unable to achieve this.

By making registration in the EU register mandatory, and enforcement and/or collection of past royalties conditional on registration, PO2 will create transparency about the SEP landscape (improving on objective 1). In addition, regular essentiality checks will guarantee the quality of the register. Essentiality checks will also give an indication on the number of truly essential patents per SEP holder. With time this might lead to creating a register of ‘truly’ essential patents. Implementers will have a (comparative) proxy of the shares that each SEP holder has in the respective technology. There will be benefits for both the SEP holders and the implementers. The SEP holders that register will demonstrate the legitimacy of their claims and see reduction of some costs related to the technical discussions plus shorter negotiation periods. The implementers will be able to see who the SEP holders are who may have legitimate claims on SEPs, what the potential strength of those claims is and how those may impact on its business. 216  They will economise on their own essentiality assessments which should aid in negotiations – improvement on objective 2. Nevertheless, implementers will still have to negotiate licence and royalties individually with each SEP holder with, in most cases, no aggregate royalty available as a reference. PO2 also includes all impacts of PO1.

By adding an obligatory pre-litigation conciliation, PO3 is expected to speed up dispute resolution and lower its cost as compared to court proceedings, thus significantly improving on objective 3. A conciliator would also propose FRAND conditions including royalty, thus further helping especially implementers and improving on objective 2.

PO4 is going to solve one of the key issues in SEP licensing negotiations by adding an aggregate royalty to the SEP register and providing a reference price for using a standard – thus significantly improving on objective 2. It should speed up license negotiations and dispute resolution, thus further improving on objective 3. It will especially help smaller SEP implementers, facilitate production cost planning and profit calculations. It will also restrict room for exploitation by SEP holders by setting reasonable maximum royalty rates.

By creating a one-stop-shop for obtaining SEP licences, PO5 eliminates the business uncertainty about the aggregate royalty. It gives the opportunity to implementers to sign licence deals with many SEP holders. For SEP holders it gives clarity about their share of the aggregate royalty and facilitates licensing with implementers who choose to use the one-stop-shop. This option will reduce the licensing costs in a similar way as for pools. Given that all SEP holders will be covered by this option, it would represent an improvement in comparison to a pool. This option radically improves on objective 2 and could eliminate the majority of disputes, thus improving on objective 3. However, the SEP burden on implementers might increase as transaction costs get very low, more SEP holders than currently might want to participate in SEP licensing (including those that were not licensing thus far, or even those that had a royalty free policy might change to royalty bearing licensing). PO5 would eliminate the issue of licensing in the value chain as any implementer may use the one-stop-shop and its customers will be licensed. Finally, for PO5 to work, SEP holders have to be incentivised to agree on a mechanism for royalty allocation. This is a very complex exercise because of the many diverging interests of SEP holders (those whose main business is using the standard, those whose main business is SEP licensing and those who do both). It is very likely that they will fail to agree, and the issue will be referred to the FRAND determination process. This option has an impact on the freedom to conduct business.

Table 2: Comparison of policy options against the effectiveness and efficiency criteria

Note: Only additional quantifiable costs and benefits of each option presented. Options build on each other thus for instance impact of PO3 consist of impact of PO1, PO2 and PO3.
 Quantifications are only indicative based on several assumptions. Not all impacts are quantifiable, notably impact on SEP licensing revenue. See annex A7.1 for details.

Table 3: Comparison of the impacts of policy options on stakeholders (including quantifiable costs and benefits)

* added in equal shares to benefits of SEP implementers and holders
Note: Only additional costs and benefits of each option presented. Options build on each other thus for instance impact of PO3 consist of impact of PO1, PO2 and PO3.
Quantifications are only indicative based on several assumptions. Not all impacts are quantifiable, notably impact on SEP licensing revenue. See annex A7.1 for details.

7.1.Compliance with the proportionality principle 

None of the options goes beyond what is necessary to achieve the identified objectives. The options are characterized by progressively increasing market intervention, which is justified to balance the distorted balance of power and information asymmetry, and to reduce unnecessarily high transaction costs in licensing negotiations between SEP holders and implementers, especially with smaller (including SMEs) implementers and ensure that the principle of FRAND licensing is implemented.

7.2.Coherence with other EU legislation

The options are coherent with the application of EU competition policy as regards the enforcement of the FRAND obligation arising from Article 102 TFEU. Other instruments in the area of enforcement of intellectual property rights are not impacted by this initiative.

Possible legal instruments for implementing the policy options are in case of PO1 an EU recommendation for setting the guidelines and a regulation to create the Competence Centre (e.g. adding new tasks to the EUIPO). In case of PO2 to PO5, an EU regulation would be required, supplemented by delegated or implementing acts (e.g. establishing the methodology for essentiality checks, criteria for essentiality assessors, criteria for aggregate royalty determination, rules for conciliation).

7.3.Coherence with international obligations (WTO rights)

Certain proposed limitations on the rights of a SEP holder, including requirements to (i) register its patents in a designated register, and to (ii) conduct a specified FRAND determination process (conciliation) in order to be able to enforce their patents and/or collect FRAND royalties for the past, could be considered a legitimate limited exception as stipulated in Article 30 TRIPS to the exclusive rights conferred on SEP holders under Article 28 TRIPS (a more detailed assessment is provided in Annex 8).  217  There is only limited guidance on Article 30 TRIPS, which on the one hand seems to allow only a narrow curtailment of the legal rights of a patent holder, but on the other hand suggests that public interest objectives stated in Articles 7 and 8(1) TRIPS, are legitimate objectives permitting encumbrance on the use of intellectual property. 218

The temporary exclusion of the right to effective enforcement of exclusive rights may be considered to be limited. The inclusion of a patented technology in a standard gives the SEP holder dominant market power over an implementer of the standard, as they cannot design their products around SEPs because the technology is essential for implementing the standard. SEPs holders benefit from the adoption of the standard and therefore usually tolerate the infringement of their patents to ensure the standard is widely used in products. When SEP holders commit to FRAND licences in order to promote adoption of the standard, their objective is not to stop the sale of infringing products but to collect royalties from such sales. The exceptions to the exclusive rights of SEP holders are thus consistent with the objectives of the TRIPs agreement to promote technological innovation and the dissemination of technology to the mutual advantage of the SEP holder and the user of the technology (Article 7). It would also be consistent with its principles of preventing the abuse of intellectual property rights and adopting measures for public interest reasons (Article 8).  219   

The exception to the exclusive right would not conflict with the normal exploitation of the patent. The normal exploitation of the patent in the context of standard-compliant products is to be able to collect FRAND royalties. Thus, exploitation rights should be more strictly defined or limited because of concerns regarding potential restrictions to fair competition and discrimination.

Furthermore, the legitimate interests of the patent holder would not be unreasonably prejudiced either since enforcement would be possible following registration and would also include royalties due for the conciliation period. Conciliation would only temporarily suspend the enforcement and neither the conciliation result nor the published aggregate royalty would be binding upon the parties. Moreover, the proposed initiative takes full account of the legitimate interests of third parties, notably implementers and end users.

Article 40 of the TRIPS Agreement on the Control of Anti-competitive Practices in Contractual Licences provides that the TRIPS Agreement does not prevent Members from specifying in their legislation licensing practices or conditions that may in particular cases constitute an abuse of intellectual property rights having an adverse effect on competition in the relevant market. The proposed FRAND determination process (conciliation and aggregate royalty setting) is intended to address, among other issues, concerns about whether the demanded royalty is truly FRAND, which may have potential anti-competitive effects. Such anti-competitive effects may impede the adoption of the standardised technology mainly by new entrants and SMEs that lack the resources to deal with such demands or pay potentially non-FRAND royalties. Any potentially abusive practices in the licensing of IP rights may result in harm to the consumer and public interest. Therefore, the proposed registration of SEPs in a designated register prior to enforcement and the FRAND determination process are likely to fall under Article 40 of the TRIPS Agreement.

Furthermore, the introduction of a conciliation is not unreasonably expensive or burdensome, given the specific SEP practices, and is thus in line with the obligation in Article 41(1), (2) TRIPS, which requires WTO members to ensure that the different types of proceedings relating to IP enforcement shall be available under the respective national laws. SEP holders benefit from the adoption of the standard and therefore usually tolerate the infringement of their patents to ensure that the standard is widely used in products. When SEP holders commit to FRAND licences in order to promote adoption of the standard, their objective is not to stop the sale of infringing products, but to collect royalties from such sales. As such, the FRAND determination (conciliation) process cannot be considered a restriction of the SEP holder’s right to enforce its patent against an implementer 220 .

8.Preferred option

Option 4 is the preferred option – a mandatory register for SEPs with essentiality checks, a process for determining an aggregate royalty, and a mandatory pre-trial conciliation, combined with voluntary guidance on SEP licensing and a central Competence Centre offering, among other services, also assistance to SMEs.

This option is expected to strike the right balance between negotiating powers of SEP holders and implementers and at the same time allow sufficient room for private negotiations between them, with more relevant information at hand. The option reduces information asymmetry between a SEP holder and an implementer by providing the latter with information who the relevant SEP holders are, how many patents they have registered in the register and what their essentiality rate is (derived from a representative random sample of all registered SEPs) and what the potential [or maximum] total cost of using a standardised technology (aggregate royalty) is. A pre-trial obligatory conciliation is likely to reduce SEP dispute settlement costs to about 1/8 as the conciliator will assist both parties in reaching an agreement. A Competence Centre will provide objective information, guidance, trainings and support to SMEs in relation to SEPs and SEP licensing. Moreover, SMEs will benefit from reduced administrative fees and/or financial support from existing facilities such as the SME Fund to cover e.g. conciliation costs.

The option is likely to reduce the time for negotiation and the cost of the negotiation for both parties. Transparency will inject trust into the system. SEP holders are likely to reach a wider set of implementers therefore securing adequate remuneration on their innovation, and incentivising participation in future standard setting. Implementers (especially smaller ones) will be able to plan their business costs and make their risk assessments.

Finally, the initiative will cover future standards (or generations of standards) and there may be a possibility to request the inclusion of older versions of the standards, such as LTE, of particular importance to the roll-out of the IoT. The option will complement and not replace the Huawei v ZTE process. It is expected not to impact licences that are already in force.

As far as quantification is possible, and under several assumptions, the net benefits of the preferred option are estimated at around EUR 24.5 million per annum. In terms of distribution of costs and benefits, the majority of net benefits (estimated at around EUR 24 million) will be directed to SEP implementers localised in the EU (split almost equally between firms with headquarters in the EU or non-EU firms with subsidiaries on the territory of the EU). With an approximate number of less than 600 new SEP licensing contracts signed per year in the EU, net benefits are estimated on average at around EUR 40 000 per negotiation. SEP holders are expected to bear the brunt of costs of this initiative mainly due to the cost of SEP registration and essentiality checks. The costs for this group amounts to EUR54 million, with 85% of these costs applicable to non-EU based SEP holders. The estimated benefits to this group account for around half of costs, so the net effect is negative (EUR 29 million). It should also be noted that a large part of SEP holders’ costs (EUR 29 million) is due to an expected increase in patent fees due to an expected increase in number of patents. This impact is uncertain and thus may not materialise, in which case the overall impact on SEP holders would be neutral. However, if it were to happen, it would constitute a new income to the European and National Patent Offices. In terms of geographical distribution of impacts in the EU, it is expected that firms from all but three (FI, SE and LU) Member states are likely to experience net benefits, with DE, IT and FR accounting for 75% of benefits. Finally, it should be noted that certain impacts, such as e.g. a change in SEP licensing revenue are difficult to predict, thus are not included in the cost/benefits calculation. For discussion of those impacts as well as for details on all calculations see Annex A7.1.

The table below summarises costs and benefits of the preferred options:

Table 4: Ten-year-average total approximated annual costs and benefits of the preferred option per affected party and localisation (EUR million)

* concerns non-EU implementers with subsidiaries in the EU
Note: numbers rounded which may affect totals
Source: Own assessment based on studies and assumptions. See annex A7.1 for more explanations

8.1.REFIT and One-in-one-out

This initiative is not part of the REFIT simplification effort as there are currently no EU rules on SEPs that could be simplified or made more efficient. The administrative costs of filing the forms necessary for SEP registration, estimated at EUR 200 per patent or around EUR 2.7 million annually over a decade are considered eligible for one in one out calculations.

9.How will actual impacts be monitored and evaluated?

The legislation to be proposed would include a provision requiring evaluation every five years.

Table 5: Monitoring indicators

Annex 1: Procedural information

1)Lead DG, Decide Planning/CWP references

DG for Internal Market, Industry, Entrepreneurship and SMEs (DG GROW).

2)Organisation and timing

This proposal is scheduled for adoption by the Commission in the second quarter of 2023.

Inter-service group meetings took place on 23.09.2021, 12.10.2021, 23.11.2021, 09.12.2021, 20.06.2022, 19.09.2022, 11.10.2022 and 27.01.2023.

The following Commission services were invited to participate: DG AGRI, CNECT, COMP, ENER, JRC, MOVE, RTD, SJ, SG, TAXUD, TRADE, and GROW. DG JUST was consulted on specific issue with regard to mediation. The EUIPO was also involved with regard to the institutional aspects. Those services were actively consulted throughout the project, including in discussions with consultants.

3)Consultation of the RSB

A meeting with the RSB took place on 15.03.2023. On 17.03.2023 the RSB delivered a positive opinion. The table below shows RSB comments and how they were addressed in the revised text.

4)Evidence, sources and quality

The analysis presented in this impact assessment is based on the following key sources:

-Baron, J., Arque-Castells, P., Leonard, A., Pohlmann, T., Sergheraert, E., Empirical Assessment of Potential Challenges in SEP Licensing, European Commission, DG GROW, 2023;

-Baron, J., Essentiality Checks for Potential SEPs – Framework for Assessing the Impact of Different Policy Options, European Commission, DG GROW, 2023;

-Charles River Associates, Transparency, Predictability, and Efficiency of SSO-based Standardization and SEP Licensing, European Commission, DG GROW, 2016, https://ec.europa.eu/docsroom/documents/48794;

- Commission webinar series on Standard Essential Patents (2020 – 2021);

-Baron, J., Geradin, D., Granata, S., et. al., Group of Experts on Licensing and Valuation of Standard Essential Patents ‘SEPs Expert Group’ (E03600): Contribution to the Debate on SEPs, 2021, https://ec.europa.eu/docsroom/documents/45217 ;

-European Commission, Joint Research Centre, Bekkers, R., Henkel, J., Tur, E. M., et al., Pilot study for essentiality assessment of standard essential patents, Publications Office of the European Union, 2020, https://data.europa.eu/doi/10.2760/68906 ;

-Number of other studies and research papers are used and references are provided in the footnotes;

-Submissions to the call for evidence, the public consultation, the survey of SMEs and numerous interviews with stakeholders and papers submitted in confidential and non-confidential format to the Commission;

-The Options considered all contributions of the Expert Group, the suggestions made during the webinar series and contributions by practitioners, academia and judiciary.

Additionally, the following data sources were used in order to perform in-house analysis:

·IPLytics Platform – data extracted in 2022;

·Experts analyses based on contracts with the Commission.

The remaining sources are provided in the footnotes, whenever they are referred to in the text.

Annex 2: Stakeholder consultation (Synopsis report)

The public consultation took place between 14 February 2022 and 09 May 2022. During that period 74 replies were submitted using the EUSurvey tool.

This annex presents summary of results, a detailed question by question analysis is presented in Annex 9. All received responses are published. 221

1)Respondents’ characteristics

Respondents were asked to provide basic information about themselves, such as country of origin, area of activity, and registration in the transparency register.

Two thirds of the replies came from the European Union. The highest number of EU replies came from Germany. The remaining one third of the replies came from third countries, with the highest number coming from the USA. The graph below (Figure 6) presents responses per country of origin.

Six out of ten replies came from companies (39 replies) or business associations (7). Followed by seven replies from academia, two from public authorities, one from a non-governmental organisation (NGO) and one from trade union. Six replies came from EU and non-EU citizens (see Figure 7).

Among the 39 companies, 77% were large entities and 23% were small and medium-sized entities (SMEs).

30% of the respondents identified themselves as both SEP holders and implementers, 13% as only SEP implementers and 4% as only SEP owners. Around 50% identified themselves as “other” (which consists of for instance attorneys, advisers, academia…).

Around half of the respondents (35) are registered in the transparency register. 222 The transparency register is a tool to allow European citizens to see what interests and on whose behalf the respondent represents.

Figure 6: Distribution of responses per country of origin of respondent

2)Summary of replies

Participants to this consultation were invited to provide answers on the following blocks of questions: general questions, questions on the licensing process and related problems, questions on transparency of SEP licensing, on essentiality of SEPs, on fair, reasonable and non-discriminatory (FRAND) licensing terms and finally questions on SEP enforcement. Below is a short quantitative summary of the responses based on selected questions per block.

Replies to general questions.

Around 60% of the respondents considered that the current legal framework sufficiently protects against implementers “hold-out” (e.g. unreasonably delaying the conclusion of a licence by an implementer). Around 30% had opposing views. The responses showed a similar pattern across different stakeholders’ groups (companies, associations, academia, respondents from the EU and non-EU countries).

Respondents were divided on the issue whether the current legal framework provides sufficient protection against SEP holders “hold-up” (broadly opportunistic behaviour by SEP holders, such as using their market power to extract excessive rents or terms from implementers) with 43% agreeing and 48% disagreeing. Companies, citizens, and non-EU respondents mainly disagreed. While academia/authorities/NGOs/others, and EU based respondents tended to agree with that the current legal framework provides sufficient protection against SEP holders “hold-up”.

The final question in this section asked about the impact of the current SEP licensing framework on SMEs and start-ups. Around half of all respondents assessed the impact as negative, a third thought there is no impact, and around 5% deemed it positive. Responses showed similar patterns for all analysed groups, except for associations, implementers and SEP holders. Around half of the associations and SEP holders considered that there was no impact, while around 80% of implementers thought that the impact is negative.

Questions on the licensing process

SEP implementer perspective: Around half of the respondents reported seeking a licence before a SEP holder approaches them. SEP holders usually contacted the respondents around 3 years after the first implementation of a standard in a product. Licence negotiations with a large SEP holder on average concerned around 60 patents.

Almost three quarters of the respondents request a licence in order not to infringe a SEP and 60% to be able to plan production and costs.

Reported costs of estimating SEP exposure per product amounted to on average around EUR 230 000.

In terms of consequences of the current SEP licensing practices: Two thirds of all respondents and a majority of implementers try to share SEP cost/risks with their suppliers. Around 40% of all respondents, and two thirds of the implementers will try to settle with the SEP holder as quickly as possible to avoid litigation or will search for an alternative to SEP (another technology or royalty free standard). Around a third of the respondents (and half of the implementers) will increase prices and may become less competitive.

SEP holder perspective: The main reasons for licensing/having SEP are securing the return on investment (70% of answers), followed by use of SEP for defensive/bargaining purposes (60%) and participation in the standardisation process in the future (40%).

On average SEP holders said they start contacting implementers two years after the publication of a standard. Around 70% do not reach out to all SEP implementers though. On average around 60% of the contacted implementers reply within one year. Around 70% take a licence without any litigation. It takes on average 3 years and 3 months to conclude a licence.

Questions on the problems related to SEP licensing

Lack of transparency on the FRAND royalty rate, on the SEP landscape (who owns SEP) and divergent court rulings were named as the key problems by three quarters of all respondents (and almost all implementers). From the SEP owners’ point of view the main problems were hold out and anti-suit injunctions.

Questions on SEP transparency

Respondents asked for more public information on SEP as regards “patent and application number” (88% of all responses), “relevant standard, version, section of the standard” (80%), “contact details of SEP owner” (80%), “transfer of ownership” (77%), “licensing programs” (76%) and “standard FRAND terms and conditions” (72%).

Patent pools should disclose “standards subject to pool licensing” (100%), “product royalties per programme” (94%) and “list of SEP owners” (87%).

Around 70% of the respondents considered that a confidential repository of licensing agreements could help judges and arbitrators to determine a FRAND rate. Such repository should contain information on “licensed SEPs” (96%), “royalties” (96%) and “methodology used to calculate the royalty” (94%).

Questions on essentiality

Around 60% of all respondents and 90% of the implementers supported third-party essentiality checks as long as independent experts do them. Only 24% of the SEP holders supported such a solution. A third of all respondents considered that essentiality checks should serve only an advisory role with no legal consequences.

Around two thirds of all respondents and around 80% of the implementers thought that essentiality assessment might help in assessing SEP exposure of a product and deciding whom to negotiate with, smoothen licensing negotiation and prevent over pricing. More than half of the SEP holders disagreed with these impacts but agreed that checks might provide a reliable overview of the share of each SEP owners’ essential patents.

As regards practical implementation, the respondents preferred that the European Patent Office (EPO) conducts the checks (63% of all replies) on just “one SEP per SEP family” (63%).

Questions on FRAND

Between 55% and 75% of all respondents and between 85% to 100% of the implementers, consider that SEP holders cannot refuse a licence following a request from an implementer. The majority of SEP holders were of the opposite view.

60% of all respondents and 93% of the implementers consider that licensing could take place at any level of the value chain. Around 70% of the SEP holders consider that it should be at one level only (level allowing for the best monitoring of applications).

Around three quarters of respondents (93% of SEP owners) agreed that fair and reasonable terms and conditions might depend on functionalities of the standard implemented in a product. Around 70% thought these terms could depend on the level of licensing.

For non-discrimination assessment, it matters if companies that use the same functionalities of the standard in similar applications are put at a competitive disadvantage (around 75% of all answers, and 94% of holders’ answers).

SEP holders considered that discounts between 28% and 40% do not cause discrimination. For implementers reasonable discounts amount to between 5% and 10%.

70% of all respondents and 100% of the implementers argued that it is important to know the reasonable aggregate royalty rate for a product. Only 20% of SEP holders shared that view.

Questions on enforcement

Respondents estimated that court costs could range from approximately EUR 2.1 million for essentiality, EUR 6.6 million for injunction and EUR 7.1 million for FRAND disputes.

Arbitration (53% of all answers) was deemed more useful than mediation (35%) for FRAND assessment, especially by SEP holders and academia/authorities/NGOs.

Two thirds of the respondents were of the opinion that efficient SEP licensing would foster innovation by implementers, increase employment and allow for keeping high level of competence in the EU as well as foster transition to the green economy.

3)SMEs survey

Internet survey targeting SMEs was active from 25/10/2022 until 27/12/2022. It used EUSurvey tool. It is not representative but gives an impression of the issues start-ups and SMEs face in the context of SEP licensing. A summary of replies is presented below, a detailed question by question analysis can be found in Annex A8.3 – SME survey.

The sample is quite diverse with respect to geography, sectors, and standards used.

In total, 39 firms responded to the survey. They are mostly start-ups and SMEs (37), from 17 different countries, mostly from the EU (30). They are active in ICT (24) and various other ICT- and IoT-using industries. In the sample, cellular standards are the most widely used or developed (by 24 firms), followed by Wi-Fi (21), LPWAN (15), NFC (12), and video/audio codecs (10).

Implementers dominate the sample; standards are very important for them. 

Most firms (35) are implementers of existing standards, e.g. by producing or buying components that incorporate a standard. 12 stated that they participate in standard setting. Asked about the importance of the respective standard, 24 out of 30 responding considered the standard very important for their business. Seven firms are SEP licensees, for cellular standards (4), video and audio codecs (4), Wi-Fi (3), and other technologies.

SEP licensing is difficult for SMEs, due mostly to resource constraints, lack of expertise, uncertainty about infringement, and the fragmentation of SEP ownership.

Most respondents have, when buying or importing components that incorporate standardized technologies, a clear preference for components with all IP rights fully licensed (25 out of 34 prefer this solution “in most cases”). One respondent described it as “ridiculous that patent pools should come after each company doing some modern radio technology, rather than settle cost per module with manufacturer.” The main stated reasons not to take a SEP license are: (i) the assumption that purchased components are fully licensed (15); (ii) difficulties to find out the SEP holders (10); and (iii) an excessive effort to enter license agreements with all SEP holders (9). Related, the main stated problems in negotiating with SEP holders are: (i) lack of knowledge what would be a fair (FRAND) royalty (15); (ii) lack of resources to negotiate or engage in court proceedings (14); (iii) uncertainty if own product infringed the patents at hand (13); and (iv) the firm’s ignorance how to defend against SEP holders (13). Few (8) have in-house patent specialists.

For SMEs, device-level licensing was acceptable only if it was efficient (ideally one-stop), at publicly known royalties, and consistent across all implementers (which given their fragmentation seems difficult to achieve). Respondents consider the best solution for SMEs to be upstream SEP licensing, to module or component makers.

Respondents stated that the uncertainty about SEP licensing would affect their business in various ways. Most frequently, they stated: Royalty payments will affect the firm’s competitive position on the market (18); potential future SEP payments may make the firm unable to compete (13) – however, if competitors pay the same royalties, licensing will not affect the business significantly (13); the firm tries to use alternatives available on the market in order not to pay for SEPs (e.g. older or free technology) (12). In open comments, respondents expressed wishes among others for licensing to component makers. Failing this, efficient licensing through pools, ideally one-stop, rather than having multiple licensors per standard and even portfolio split-ups to several NPEs. Publicly known royalty rates that are paid consistently by all and protection from abusive licensors.

About 40 percent of firms in the sample contribute in some way to standard development, though very few own SEPs. The majority does not contribute, mostly for lack of resources. 

Of the respondents, 15 participated in standard setting, though mostly not to complex ICT standards, such as cellular and Wi-Fi. Only 2 firms own SEPs. Motives to contribute to standard development comprise improving the standard, influencing the direction of standard development, facilitating adoption of the firm’s own technology, learning, and networking. Reasons not to contribute to standard development comprise a lack of resources and expertise, and the domination of SDOs by large firms.

4)Interviews with stakeholders

During the course of 2021-2022, the Commission services met with various SEP holders and SEP implementers in order to assess the problems they face with SEP licensing. As these meetings concerned confidential information based on licensing negotiations, below we present selected stylised summaries of issues raised that combine the experience of several parties we interviewed.

Problems faced by SEP implementers