FRAND Guidelines of the 7th civil chamber of the Regional Court of Munich I

These FRAND guidelines are an English translation of the FRAND-Hinweise of the 7th chamber of the Regional Court of Munich I (German original available here) provided by BARDEHLE PAGENBERG. The German FRAND-Hinweise are official and prevail.

Preamble:

These guidelines summarize the chamber’s case law from the proceedings 7 O 5007/25 (ASUS I, GRUR-RR 2026, 215), 7 O 4102/25 (ASUS II, GRUR-RS 2026, 1112), 7 O 7655/25 (Renault, GRUR-RS 2026, 2292), and 7 O 64/25 (ZTE/Samsung, GRUR-RS 2026, 8886) in a concise manner, contain clarifications, and address issues that have arisen in other proceedings where no published decision has been issued. They document the current fundamental understanding of the courts regarding the FRAND framework for standard-essential patents and are intended to assist the parties in focusing their respective factual submissions on the issues that are actually relevant to a decision. The chamber is aware that the specific circumstances of each case require an individual assessment.

The 21st civil chamber has been informed of these guidelines. Due to the large number of issues addressed and the associated level of detail, it is not possible to draft joint guidelines. It is assumed that both chambers share the view that, in many situations, a substantive review of the patent holder’s offers must be conducted.

These guidelines are long. The reason is that much explanation is needed. The situation seems to be that many of the aspects elaborated here are self-evident but need to be put into writing once. Future versions of these guidelines will therefore probably be significantly shorter.

I. Introduction

1 Patent infringement proceedings in which standard-essential patents are asserted continue to be of significant economic importance, and there is no indication that this will change in the near future. It is now the general consensus among the courts actually handling such proceedings in all jurisdictions worldwide that an analysis of the specific licensing offers is required.

2 SEP proceedings are characterized by the fact that the patent holder asserts at least one patent that is essential for the implementation of a standard (e.g., cellular communications, Wi-Fi, streaming, Ethernet …). These proceedings therefore lie at the intersection of patent law and antitrust law. From a simplified perspective, there are two archetypal situations. In the first situation, the plaintiff has a very extensive patent portfolio. In this case, both parties are aware that a license will have to be obtained sooner or later. In such situations, the dispute between the parties essentially centers on the amount and terms of the license fee. In the other situation, holders of smaller portfolios attempt, by aggressive conduct, to compel technology users to pay high –potentially too high – license fees.

3 Based on economic considerations alone, it can be assumed that users of standardized technologies obtain licenses solely because they would otherwise expect disadvantages. Since patent holders and technology users are often based in different countries, administrative measures cannot fulfill the necessary guiding function. Against this backdrop, patent holders must enforce their rights themselves through civil complaints. Due to the global structure of technology users, patent holders have access to the civil justice systems of various countries, which function with sufficient reliability and in which the economic consequences of a court decision have a consequential impact.

4 In this context, initiating and conducting patent infringement proceedings is a legitimate means of negotiation. The Federal Republic of Germany and the Unified Patent Court (UPC) play a significant role in the global system of patent enforcement because the punishment for patent infringement may be, among other things, an injunction within the respective jurisdiction (Germany or member states of the UPCA). In addition, the jurisdiction of the United States of America plays a particularly prominent role, with damages being the primary focus there. Other jurisdictions, such as the United Kingdom and Brazil, play a secondary role, which can be defined primarily as a complement to or reflection of the systems in Europe and the U.S. At present, the role of the judicial system in the People’s Republic of China cannot yet be conclusively assessed. It is expected that the courts there will gain significance over the next few years, not least because the courts have expressed a corresponding self-conception.

5 These guidelines set forth in a detailed and transparent manner the considerations undertaken by the chamber when examining SEP/FRAND matters. In addition to transparency and predictability, the chamber also attaches importance to ensuring that proceedings are not overburdened and can be decided within one year of the filing of the complaint. As a supplement to the established combination of attorneys-at-law and patent attorneys in German patent infringement proceedings, it appears sensible to strengthen the parties’ representation by engaging private experts to calculate license rates in appropriate cases. The chamber sees no reason, even in the future, to appoint court-appointed experts to have FRAND rates calculated. This is because the question of what constitutes a reasonable rate is a legal question and cannot be resolved through expert evidence.

1. SEP/FRAND complaints as the “tip of the iceberg”

6 The European Commission’s amicus curiae briefs in the proceedings that led to the decision by the German Federal Court of Justice’s (“BGH”) “FRAND defense III” (BGH KZR 10/25, GRUR 2026, 383) necessitate an examination of the typical course of negotiations between patent holders and technology users. For a closer examination of the interests of the research institutes and companies involved in the development of standards, as well as those of consumers – who have a right to participate in modern technology – reveals that the criticism expressed by the European Commission ultimately focuses too one-sidedly on supposed consumer interests and fails to do justice to the complexity of the entire issue. This imbalance is evident from the fact that the amicus curiae briefs do not to mention at any point that the German Federal Court of Justice had previously discussed the referenced decision by the Court of Justice of the European Union (C-170 C-170/13 – Huawei v. ZTE, BeckRS 2015, 80933) in two decisions (BGH GRUR 2020, 961 – FRAND defense; GRUR 2021, 585, marginal no. 57 – FRAND defense II). The chamber has already explained that the amicus curiae briefs lack sufficient legitimization (7 O 14174/23, GRUR-RS 2024, 40789, marginal no. 122 et seqq. – Video Encoding). The chamber does cannot assess what exactly prompted the Commission to issue its opinion. What is to be noted is that an important institution has missed an opportunity to enrich the discussion on the topic of SEPs/FRAND by contributing compelling arguments and providing a balanced treatment of the arguments from both sides.

7 (Unlike many higher courts,) the chamber handles approximately 40 to 50 new SEP proceedings each year, which has granted it sufficiently well-founded insights into this quite complex subject matter over the course of several years.

8 To begin with, it can be assumed that the vast majority of negotiations between patent holders and technology users result in the conclusion of a license agreement without the need to bring the matter before the courts.

a. License negotiations

9 In license negotiations, a distinction must be made between negotiations between holders of large patent portfolios and technology users with higher unit quantities on the one hand, and smaller market participants on the other. Both situations have in common that, according to general understanding, it is permissible to begin using the patented technology at a time when no license has yet been granted. In this respect, patent holders make an upfront investment through their development efforts.

10 In the case of larger parties, it can generally be assumed that both parties are aware of the approximate license fee, and that mutual hold-up and hold-out strategies serve to bring their price expectations closer together. When smaller parties are involved, other problems may arise, which can essentially be described as both parties initially – consciously or unconsciously – failing to correctly assess their bargaining position. What these negotiations usually have in common is that they extend over a longer period.

11 Therefore, before a complaint is filed with the Munich Regional Court I, there is typically a period of several years during which license negotiations have taken place. After the complaint is filed, it generally takes between 9 and 12 months to reach an oral hearing, which may result in an injunction. It should therefore be noted that injunctions have a significant lead time and that technology users – the defendants – are not caught off guard by them.

b. Bargaining positions

12 In the chamber’s view, the parties’ bargaining positions change during the process of negotiating the license agreement. As a rule, the technology user initially has a very strong bargaining position vis-à-vis the patent holder because revenue can be realized through the rapid conclusion of a license agreement, while minimal effort is necessary for enforcement: Because of the fact that beginning to use the technology to be licensed was already permitted, its commercialization has already begun. In this regard, it is worth noting the particularity that patent holders, as the actual financiers of the investment in establishing a standard, enter the refinancing phase later than the beneficiaries of the standard enter the commercialization phase.

13 This initially advantageous position changes as negotiations progress – and particularly after the initiation of patent infringement lawsuits – to the detriment of the technology users. Particularly when an injunction is imminent, the patent holder has an advantageous bargaining position. It must be emphasized, however, that this bargaining position generally arose only because the user allowed earlier opportunities to conclude a license agreement to pass. It is only natural that a license concluded under such conditions may be more expensive than one concluded at an earlier stage. At the very least, the patent holder’s additional expenses must be compensated. In the chamber’s view, in addition to mere compensation for the additional expenses, general preventive aspects may also be taken into account when assessing the specific license fee.

14 In actual terms, it can be observed that a certain resistance during negotiations in the initial phase does indeed result in more favorable license rates. In this respect, approaches of hold-out strategies makes sense from the perspective of technology users and should therefore be respected. However, there are limits.

15 The exact limits of permissible conduct can be determined based on the principles set forth in these guidelines. The chamber is aware that these guidelines are an approximation and that there may always be special circumstances that require a more detailed, case-by-case assessment.

2. Super-FRAND, successful hold-out, and valuation

1 Contrary to persistent rumors, the patent law system described above does not lead to excessive rates. Rather, it is likely that hold-out behavior results in lower licensing rates and, furthermore, that companies that use obvious hold-out tactics are less frequently targeted by other patent holders – meaning that a deterrent effect towards third parties (other patent holders) is consciously intended and desired.

2 It is likely that many of the reservations in this area stem from the very high economies of scale. While the development of standards involves very high costs, these costs are useful for many billions of devices. As a result, even relatively low individual amounts can generate very high total licensing revenues for established standards.

3 This can be illustrated specifically by the ARB for the use of the 5G standard (and all lower-generation mobile communication standards). According to the ZTE/Samsung decision by this chamber, the reasonable amount for use over the entire lifespan of a device is $13.60. At the current exchange rate, this corresponds to €11.78. Currently, an Iced Matcha Latte at Starbucks in Munich costs €6.90. In other words, the use of mobile communication technology costs less than two cold drinks, which can hardly be viewed as unreasonably high.

II. The legal framework for the FRAND review

19 The treatment of the FRAND defense under German law is subject to a fixed assessment framework established by the European Court of Justice’s decision in the case of Huawei v. ZTE (C-170/13, GRUR 2015, 764) and further clarified by the German Federal Court of Justice in its decisions FRAND Defense I through FRAND Defense III.

20 In addition, it should be noted that the Higher Regional Court of Munich, in the original proceedings concerning the FRAND Defense III (GRUR 2025, 738 – Speech Signal Encoder), established further requirements. Although the German Federal Court of Justice did not comment on the concept developed there concerning the amount of the provision of security in its FRAND Defense III decision, it cannot be conclusively assessed whether the Higher Regional Court of Munich continues to require that the technology user provide the amount most recently demanded by the patent holder as security. For the sake of completeness, these guidelines will also discuss the legal opinion expressed there.

1. The (well-known) assessment framework

21 A multi-step assessment framework applies, where the assessment step identified below as Step 5 has been accorded special significance by the Higher Regional Court of Munich.

Step: Notice of infringement by the patent holder
Step: Declaration of willingness to take a license by the patent user
Step: Submission of a license offer by the patent holder
Step: Review of the license offer by the patent user; if not accepted, submission of a counteroffer within a short period (continued willingness to take a license)
Step: If the patent holder rejects the patent user’s offer, the patent user must make a provision of security in accordance with “recognized business practices.”

2. Conventional interpretation

22 The chamber interprets these assessment steps such that the following rules apply (in this regard, reference is made to the chamber’s preliminary decision of July 14, 2025 (GRUR-RS 2025, 19196)):

The first two steps (notification of infringement and notification of willingness to take a license) are subject to a formal review. To date, there is no court decision that had to address the consequences of a lack of these assessment steps.
The issue of a lack of the first two assessment steps could arise, for example, in cases where the parties have been negotiating for a long period of time without a notice of infringement having been sent regarding the specific patent-in-suit. If, in such a case, the focus was placed on a specific notice of infringement regarding a particular patent, this would, under strict interpretation, disregard the negotiation model required as prerequisite by the Court of Justice of the European Union (CJEU). In the chamber’s view, however, this would contradict the spirit and purpose of the decision; thus, it must be assumed that the lack of a specific notice of infringement is compensated for, at least when negotiations of a certain scope have in fact taken place between the parties.
A special situation arises when there was a prior license agreement between the parties (concerning the same standard) or when both sides hold a relevant share in the standard and communicated with each other regarding the taking of a license before the complaint was filed. Even in such situations, the first two assessment steps may be omitted, as requiring formal declarations in these cases would amount to mere formalism.
Only minimal requirements should be imposed on the patent holder’s initial licensing offer. The offer may be vague and contain aspects that require further clarification in many respects, so that it can be tailored to the specific needs of the respective technology users. This is because this initial offer serves to initiate a negotiation process and, as such, does not yet have to meet the FRAND requirements. In particular, it is permissible to demand a fee that exceeds FRAND terms in order to be able to offer discounts during the upcoming negotiations. This understanding regarding the initial license offer follows directly from the assessment framework which is based precisely on the premise of negotiation. If one were to view this differently, one would never reach the fourth step if the patent holder’s offer was already FRAND-compliant. However, the CJEU’s approach cannot be interpreted to mean that the fourth step is always irrelevant.
The line is drawn where the patent holder, by demanding completely excessive licensing fees, clearly attempts to exclude a competitor from using the standard and thereby indicates that it is unwilling to grant a license. In the current economic and competitive environment, this will not be the case simply because the market is so complex that a monopoly position could not actually be implemented. The situation might have been different some 15 to 20 years ago when the first smartphones or tablets entered the market, and participants already active in the market may have had an interest in making it difficult for other market participants to gain access to the market. However, such a situation does not exist at present. The parties generally dispute the amount of the license fee, and patent holders are, in principle, interested in granting licenses – on terms favorable to them.
The focus of the assessment of the FRAND defense in patent infringement proceedings lies on an evaluative assessment of both parties’ willingness to take/grant a license. In this context, it is generally examined whether the lack of an agreement stems from the patent holder’s sphere of influence or from that of the technology user.
The parties’ conduct during negotiations plays a decisive role in this regard. In principle, the patent user is obligated to demonstrate that it has made efforts to conclude a license agreement on FRAND terms. The patent holder is obligated to demonstrate that their offer is FRAND. In this assessment, the speed with which the parties sought to conclude a license agreement plays a significant role. In this regard, the technology user bears an increased obligation to make such efforts because the use of the patent began at a time when no license existed.
In the absence of an agreement (the “sine qua non” prerequisite for the patent infringement proceedings to continue at all), the prevailing view in German case law was that a discretionary determination had to be made as to which party was at fault for the failure to reach an agreement. As a rule, the fault was attributed to the technology user. In its initial decision leading to the German Federal Court of Justice’s decision FRAND Defense III, the Higher Regional Court of Munich formalized this assessment by establishing that a lack of willingness to take a license is presumed whenever the user fails to provide security in the amount of the patent holder’s demand. These requirements can mean that the provision of security in the amount of several hundred million euros must be made, which is economically unrealistic even for solvent companies.

23 Overall, German case law has not, in the past, addressed the substance of the parties’ respective offers. The focus has been on the parties’ conduct – under this framework, the plaintiff’s offer was not examined, except in a few rare exceptions.

24 However, even according to this conventional understanding, it would be consistent to conduct an examination of the plaintiff’s offer if the users either negotiate constructively and/or make a provision of security that meets the requirements of the Higher Regional Court of Munich. According to the conventional understanding, it remains unclear which criteria would be used for a specific examination of the plaintiff’s offer. The decision of the Higher Regional Court of Munich suggests a tendency to appoint an expert in such cases. It remains unclear exactly who could be appointed as an expert and what would constitute an admissible issue of fact.

3. Distinction between external and internal willingness to take a license

25 The conventional understanding described above viewed the parties’ willingness to take/grant a license to generally. For example, if one user clearly held out and another user made a good-faith effort to obtain a license but no agreement was reached, the court assumed in both cases that the user was unwilling to take a license – an assumption typically based on specific failures in the conduct of negotiations. As a result, the judgments based on this reasoning were of limited significance to the parties, since it goes without saying that license negotiations should be conducted expeditiously; this does not need to be confirmed by judgments.

26 In the chamber’s view, a distinction must be made in this regard. Cases of obvious holdouts should be evaluated in simplified proceedings. By contrast, cases in which there is an actual dispute regarding the license fee amount – possibly for good reasons – must be considered in a more differentiated manner. The goal is to address the specific offers in a qualified manner in the reasoning for the decision so that, inter alia, the parties can agree on an appropriate license fee amount subsequent to the proceedings. Furthermore, well-reasoned decisions – which are accessible to the public – always also serve as a guide for similar cases involving other parties.

27 The fact that parties generally dispute the question of how high a license fee must ultimately be has led the chamber to conduct a two-step review as standard practice in FRAND proceedings, in which the external willingness to take a license is examined first and – if necessary – the internal willingness to take a license is assessed subsequently.

28 External willingness to take a license exists when there is no apparent holdout. This is generally the case when the user has paid an amount undisputedly owed and – under certain conditions – has provided qualified security. Only once these conditions are met must the chamber examine the substance of the plaintiff’s offer to determine whether it falls within the permissible FRAND framework. If the offer falls within a permissible framework and is not accepted, the user lacks internal willingness to take a license. The assessment of the internal willingness to take a license thus constitutes a comprehensive examination of the patent holder’s license offer, the scope of which corresponds to the examination regime in license fee determination proceedings in the United Kingdom and the People’s Republic of China. However, since the principles underlying German civil procedural law – specifically the principle of production of evidence – apply, the appointment of an expert by the court is not necessary.

a. Determination of external willingness to take a license

29 The assessment of whether there is an external willingness to take a license is conducted in a strictly formalized procedure. This assessment is omitted only if the technology user has not negotiated at all or has made no offer – or an offer so low that it cannot be regarded as a serious offer under any aspect.

30 Otherwise, the external willingness to take a license (which makes the examination of the patent holder’s offer possible) always exists when the user actually pays the undisputed portion of the license fee (i.e., the technology user’s latest offer) to the patent holder for permanent retention (obligation to make an installment payment) and – under certain conditions – has additionally provided supplementary security. The obligation to make an installment payment is justified by the fact that the beginning of use of the standard constitutes a use of the plaintiff’s portfolio, which triggers an obligation to pay compensation. With regard to the undisputed portion, the defendant has no right of retention.

31 The supplementary security is specifically required if the patent user’s offer amounts to less than 60% of the patent holder’s demand and the difference exceeds US$10 million. Usually, this situation is only relevant if the parties’ negotiations are focused exclusively on concluding an agreement on a lump-sum payment. In such a case, the technology user must provide security equal to the amount payable for one licensing year according to the patent holder’s offer.

Example:
If the patent holder demands 100 for a term of 5 years and the user offers 30, then an amount of 30 must be irretrievably paid as partial payment, and an additional amount of 20 must be deposited as security (assuming the difference exceeds US$10 million).

32 An exception applies if the patent user has obtained or requested a fee determination before another court in a different jurisdiction. This is because a party that has requested a fee determination must, in addition to the partial payment, provide the amount set by the court where the determination was requested as security – or the amount proposed in the request filed – irrespective of whether the defendant in the other jurisdiction has agreed to the proposal of the court there or not.

Example:
The parties’ positions regarding a license with a term of 5 years are 30 and 100, respectively. The decision of the foreign court regarding an interim license stipulates that an amount of 10 must be paid as a fixed deposit and an additional amount of 50 must be provided as security. In the chamber’s opinion, the amount payable for retention is 30, and additional security in the amount of 30 must be provided.

33 Digression: As far as can be ascertained, the concept developed by the chamber in its advisory decision of July 14, 2025, regarding partial payment of the undisputed portion of the license fee has also gained recognition in other jurisdictions and is being implemented. The payment of the undisputed amount does not appear to lead to any insurmountable accounting problems. It is possible to record such payments in accordance with the requirements for proper accounting and to treat them as revenue.

34 To the extent that some criticism has been voiced regarding accounting aspects, it is directed not against the payment of the undisputed portion as such, but against the handling of the (supplementary) provision of security in the United Kingdom. As far as is known, in the United Kingdom, the security is provided directly to the patent holder, who, however, can hardly make use of the money because the payment is subject to the possibility of claw back. In this respect, there is a significant difference from the way in which security is provided in the Federal Republic of Germany. This is because, as a rule, it happens by means of bank guarantees.
b. Overcoming the barrier of external willingness to take a license

35 Once the user has paid the undisputed portion of the license fee and, if applicable, provided security, the plaintiff’s offer is examined. The criterion is whether the plaintiff’s offer falls within the FRAND framework (FRAND range). The exact principles are set forth below.

b. Overcoming the barrier of external willingness to take a license

4. The two aspects of FRAND

36 The patent holder has the right to set the price for its license; however, the amount is limited by regulations under antitrust law. Two aspects are central. First, the price of the license must not be excessive in relation to the contribution made to the development of the standard. Second, the patent holder must not charge disproportionately different license fees to different technology users. This second aspect does not mean that all technology users must pay exactly the same amount. Different circumstances justify different license rates.

37 The question of what amount could be charged in absolute terms is reflected in the terms “fair” and “reasonable” in FRAND. The question of whether licensees are treated differently is expressed in the term “non-discriminatory.” In substance, these are two distinct examination criteria which influence each another. The interplay between these two aspects will be discussed in more detail later.

38 Since this is generally not seriously disputed, it will be mentioned only briefly that the patent holder can determine the scope of the license and, in this respect, only a limited examination takes place. In this regard, it can be assumed that a worldwide license is customary and acceptable. Combinations of multiple standards or of standard-essential and non-standard-essential patents are problematic but may be permissible if at least a single-standard license is also offered as an option and the technology user consciously opts for a more comprehensive license (as in the RENAULT decision, where five different license variants – regarding scope – were offered).

5. Need for a “SAFE HARBOUR” concept?

39 Following the decision by the Higher Regional Court of Munich, the chamber has received numerous inquiries as to whether the court’s reasoning should be understood to mean that providing security would lead to a “SAFE HARBOUR” – that is, an injunction would not be issued if security had been provided.

40 The Higher Regional Court of Munich clarified in a supplementary notification that this is not the case. However, the large number of inquiries prompted the chamber to consider whether there is a legitimate need for a SAFE HARBOUR concept. In this regard, the chamber also took into account that, with the introduction of the Patent Mediation and Arbitration Center (PMAC) in June 2026, the parties have been provided with an additional tool for the resolution of disputes.

41 The chamber decided against introducing a SAFE HARBOUR option. The approach considered was that a SAFE HARBOUR would be possible if:

the defendant had announced to the plaintiff and the chamber, no later than 5 months before the date of the oral hearing, a request for arbitration proceedings and, after one month (4 months before the hearing date), had actually filed such a request on terms acceptable to the patent holder. The objective of such arbitration proceedings would have needed to be to ultimately reach a license agreement that would end the unlicensed use of the plaintiff’s patent portfolio and remain in effect for at least three years beyond the conclusion of the arbitration proceedings.
Neither party would have been allowed to evade the outcome of the initiated arbitration proceedings.
The allocation of costs would have been dependent on the final offers made prior to the initiation of the arbitration proceedings (i.e., the party that opposed the settlement would also have borne the additional costs).

42 Under the proposed concept, a refusal by the plaintiff to participate in arbitration designed in this manner in such a situation might have had implications for the enforcement of the claim for injunctive relief.

43 However, in the chamber’s view, such an approach is contrary to the interests of the plaintiff. The decisive factor in this regard is that even before the introduction of the PMAC, there was a wide range of options for out-of-court dispute resolution, and patent holders often cannot reasonably be expected to wait even longer. The chamber does not see that the introduction of the PMAC has in any way brought about a significant change in the landscape of out-of-court dispute resolution. In this regard, it should be noted that the Regional Court of Munich I has offered mediation services for more than 20 years, with judges specializing in patent law serving as mediators. Furthermore, various arbitration institutions, such as the ICC, offer predictable and reliable procedures. In this respect, the market for out-of-court dispute resolution options has not changed significantly. It has merely been expanded to include an additional option.

44 However, there is also no actual need for a “SAFE HARBOUR”, because the chamber offers an alternative that can lead to swift clarity regarding the FRAND rate: Either party may request a “FRAND FIRST” hearing during proceedings before the chamber. In such a hearing, the discussion is limited solely to FRAND, and the subject matter is restricted to a brief of no more than 25 pages (plus exhibits). It is also possible for such a hearing to take place shortly after the complaint was filed. In that case, both parties would have the opportunity to define their FRAND positions in an orderly discussion led by the chamber. To the extent that a defendant is genuinely willing to discuss unresolved issues, this is a simple way to send a signal. It is therefore not necessary to introduce an official “SAFE HARBOUR” concept. From the chamber’s perspective, the risk of abuse would be too high.

III. Examination of the plaintiff’s offer

45 The actual examination of the patent holder’s offer for its reasonableness in accordance with the FRAND standards is at the heart of the recent case law of the 7th civil chamber. The starting point is the determination that FRAND represents a range, and the specific assessment of reasonableness is primarily carried out through a comparison with other license agreements that the patent holder has entered into with other licensees regarding the same patent portfolio (priority of comparative license agreement analysis). The top-down approach plays a secondary role. The result obtained through this approach serves primarily to examine the specific offer, but may also be used to determine reasonableness itself in individual cases (e.g., in the absence of comparable agreements or when determining a license fee for the first time).

46 Due to the chamber’s ZTE/Samsung decision of April 30, 2026, the top-down approach has taken center stage. The background is the actual licensing practice in this regard, in which a great many licensors receive lower license fees than would result from the chamber’s top-down calculation approach. Since the top-down approach does not serve to compensate for past negligence, these guidelines are also intended to counteract excessive demands by patent holders.

47 It should be noted that, in the case of an established licensing program, there is no need for a correction if it turns out that a licensing practice results in a higher amount than would result from the chamber’s top-down calculation approach. In this regard, market forces should be given priority, especially since, in performing the calculations underlying the top-down approach, the chamber made assumptions in favor of the technology users – within the available leeway – in several instances.

1. Calculation approach

48 The plaintiff’s offer is examined by comparing the specific offer with one or more license agreements that the patent holder has entered into with respect to its portfolio. In this regard, the patent holder will generally be required to submit several other license agreements. It is then court’s task to decide whether the agreement or agreements are suitable for comparison – that is, whether they actually constitute a comparable license agreement. As a general rule, it is always advantageous for the patent holder to submit as many license agreements as possible. If there are so many license agreements that doing so would entail additional effort, a selection may be made.

49 Essentially, however, it can be assumed that, in principle, even a single license agreement may be sufficient. To properly assess this agreement, however, it is generally advantageous to submit additional license agreements as well. This is illustrated by the following example:

Example:
A submits a comparative license agreement dated May 30, 2026, with a value of 5. Based solely on this agreement, it is not possible to determine how this agreement relates to earlier agreements. If A wishes to increase the value by referring to the 15% rule (described below), this is not possible. This is because it is unclear what applied previously.

If A submits additional agreements dating back (at least) to the year 2024 that also specify a value of 5, then A may demand 5.75 by referring to the 15% rule.

50 Based on the submitted license agreements, which have been deemed comparable, an average price paid per unit is determined. This average price establishes the FRAND framework, which is adjusted based on the parties’ specific conduct during the negotiations. The further the negotiations progress, the more the framework narrows.

2. Criteria for determining a reasonable license fee

51 The license fee does not merely cover the development costs and the costs of enforcing the portfolio. Rather, it also compensates for the risk associated with investing in the development of a standard. Since it is permissible – and often necessary for refinancing – to transfer patents, companies that develop and bring products to market themselves should be treated in the same way as pure patent exploiters (Non-Practicing Entities (NPEs)). Given the actual difficulties in getting technology users to pay license fees, it even appears advantageous – in terms of maintaining the creative potential of developers – for patents to be transferred to specialized exploiters for the purpose of commercialization.

52 In addition to the contribution to the development of the standard, the criteria for determining a reasonable license fee include, above all, the following aspects:

Time of conclusion of the license agreement (absolute/relative)
Scope of use (number of units/single or multi-standard)
(Existence of cross-licensing)

53 When determining what fee is appropriate, it is important to consider whether a license agreement is entered into with a relatively young company at an early stage or at an early stage in a licensing program (absolute point in time). Under such circumstances, very favorable terms are often negotiable and also economically reasonable. If the licensing process drags on and the negotiations consume the patent holder’s time and resources, a higher license fee is justifiable (relative point in time).

54 Another criterion is the scope of use; here, a distinction must be made between the quantity of units and the number of standards utilized. A high quantity of units alone justifies a discount of up to 30%. If multiple standards are licensed, further discounts are possible, which can also be reflected in a surcharge on the most valuable standard.

Example:
If a user with very high quantities takes a license in only the mobile communications standard, a 30% discount may be justified. If Wi-Fi and streaming shall also be licensed under the same license agreement, it appears possible that this could be compensated for by a flat-rate surcharge on the fee payable for the mobile communications standard.
For example, if the relevant base fee is 10, the volume discount can reduce it to 7. The additional standards can be covered by a flat-rate surcharge of 40%. As a result, in such a situation, a multi-standard license can be granted for 9.8.

55 This example shows that the fees paid can vary significantly among individual licensees. This does not constitute discrimination, even though, according to this calculation, a user with a small quantity of units would pay more for a single-standard license than the user with high quantities of units would pay for a multi-standard license. The rationale for this differentiation is that, for users with high quantities of units, the amount of the payment alone is of particular significance. In conclusion, it should also be noted that single-standard licenses are not comparable to multi-standard licenses.

3. What are appropriate license agreements for a comparison?

56 In the chamber’s view, the patent holder is able to determine which of the license agreements it has entered into best reflect the value of its portfolio. In this regard, the patent holder may make a selection, ideally ensuring a high degree of identity between the key terms of the submitted license agreements and the situation of the parties seeking a license. This is because it is impermissible to select license agreements for submission that do not adequately reflect actual licensing practice. Along with the right to submit the agreements to be used for the comparison, comes the obligation to explain which other agreements regarding the standard have been concluded and why those are not being submitted.

57 While this may appear disadvantageous from the perspective of the respective defendant, it does not lead to any actual problems when the principles of onus of presentation and the burden of proof are consistently applied. Rather, an unlimited obligation to produce documents would encourage hold-out strategies. In fact, the chamber is aware of numerous cases in which the plaintiff submitted several license agreements to substantiate the asserted claim, and the respective defendant demanded even more submissions of agreements, while at the same time compiling a non-representative collage of individual clauses from the submitted agreements to justify a low counteroffer.

Digression: regarding the submission of agreements

58 With respect to individual license agreements, a party may request an order to produce documents against itself pursuant to § 142 of the German Code of Civil Procedure in order to be able to produce agreements that are subject to a court order. In such cases, the rule applies that the court will not issue orders against the will of the party in possession of the agreement. This approach is grounded in the system of graduated onus of presentation and burden of proof. Under this system, each party is allowed to determine for itself which facts are submitted to evidence a particular evaluation. It is therefore court’s task to assess whether the facts presented are sufficient. In this regard, the facts presented by the adversary play a decisive role. This concept is incompatible with a requirement, initiated by the adversary, to produce documents that the party subject to the obligation does not intend to produce. In this regard, the chamber cannot rule out the possibility that exceptions may exist in special circumstances, particularly if a party subject to the presentation obligation is prevented from complying with a request for presentation due to contractual obligations toward third parties.

59 If no main proceedings are yet pending, the production of documents can also take place within the framework of independent evidentiary proceedings. This approach may be advisable so that the parties can also rely on agreements subject to a court order during license negotiations. The chamber has not yet had to take a decision on the question whether independent evidentiary proceeding can be conducted in parallel to a main proceeding to make a special confidentiality regime possible. If this were the case, the value in dispute would have to be set at a low level.

Regarding the suitability as a comparable license agreement in detail:

60 As a general rule, license agreements may only be used as comparable license agreements if they are no older than five years (absolute limit). In the case of older agreements, a discretionary assessment must be made to particularly determine whether and to what extent they remain comparable.

61 The specific circumstances under which a license agreement was concluded must be taken into account. For example, entirely different terms may apply to the renewal of a license agreement than to a license agreement concluded in the wake of an injunction. At the very least, upon request, the patent holder must submit the information necessary for the specific classification of a license agreement into the proceedings.

62 Furthermore, the specific scope of other license agreements must be taken into account. On the one hand, this refers to the quantity of units; on the other hand, also to whether a standard was licensed on its own or as part of a package with other standards. As a general rule, a difference in the quantity of units can be easily compensated for by applying a percentage surcharge or discount to the fee. Multi-standard license agreements generally cannot be “unbundled.” Therefore, they generally do not constitute suitable comparable license agreements for a single-standard license agreement.

63 The assessment of license agreements that include cross-licensing poses particular challenges. As a general rule, such license agreements do not constitute suitable comparable license agreements, since the value of the consideration associated with the cross-licensing is difficult to assess. As a result, such submissions are practically never presented as comparable license agreements. If it were relevant, a distinction would need to be made between agreements covering only one standard and multi-standard agreements. In the event that there are no or only a few other agreements covering the same standard exist, single-standard cross-licensing agreements are the most suitable agreements for comparison.

4. Analysis of the original agreement

64 In agreements where a unit fee has been agreed upon, that unit fee serves as the relevant benchmark. An adjustment may be necessary if the comparable license agreement was entered into with a licensee who markets a different product category than the defendant. However, this does not pose a problem in practice.

65 Particular attention must be paid to the analysis of the original agreement if the comparable agreement includes a lump-sum payment, options, cross-licensing, or a combination of these elements.

Regarding the individual elements:

Lump-sum payments:

66 To unpack a lump-sum payment, one must take into account the projected quantity of units at the time the agreement was concluded. Usually, the party presenting a license agreement as a comparable agreement is also able to provide the data that was used as forecast figures at the time the agreement was concluded. To keep the explanations simple, the topic of “past release” (i.e., compensation for the past) will be set aside and addressed later.

67 For a proper submission, it is sufficient for the plaintiff to present these figures. Additional evidence is required only if the defendant can substantiate that the figures may be inaccurate.

68 As a general rule, it can be assumed that a lump-sum payment entails a significant discount compared to regular billing based on a per-unit price. This is because both licensing parties save a considerable amount of effort since billing is eliminated. Furthermore, prompt payment of the agreed-upon amount (if provided for) results in significant interest advantages.

Options:

69 If a license agreement contains options, such as an renewal option at a specific price, these options are generally to be calculated as if they had actually been implemented in the version most favorable to the respective licensee. For renewal options, the projected figures at the time the agreement was concluded are decisive. This is because the parties would have had the option at that time to enter into the license agreement for the longer term right away.

Cross-licensing agreements:

70 Patent holders rarely present cross-licensing agreements as comparable license agreements. In most cases, cross-licensing serves to grant the net payer a discount, so these are often not the best comparable license agreements – from the perspective of the patent holder.

71 However, if such an agreement were to be unpacked, the figures ultimately used for the specific license calculation would also have to be submitted so that the chamber would be able to unpack the agreement as well and accurately as possible. What is significant is that the value of the two portfolios must be determined separately. It is inadmissible to rely solely on the purely quantitative number of patents licensed by each party.

5. FRAND determination within a framework

72 Once the agreements submitted for comparison have been analyzed, the FRAND range must be determined. The decisive factors are the value of the portfolio to be licensed as well as the identity and conduct of the party seeking a license. The starting point is either an existing agreement or a hypothetical comparable agreement. Based on this comparable agreement, an average value must be determined. Based on this average value, a reduction of 50% or an increase of 50% is possible. The decisive criteria for the classification within this range are the quantity of units and the conduct of the parties. In case of comparable facts of the matter, an increase of only 15% is possible. Details are as follows:

a. Determination of the comparable license agreement (calculation of the average value)

73 A suitable comparative license agreement must be selected from the agreements submitted by the patent holder, or a hypothetical comparative license agreement must be determined. In this regard, a distinction must be made between two situations. The first situation relates to a patent holder who has an established and consistent licensing program and uses the same license agreement for all licensees. In such a situation, this license agreement will also apply to the user in the proceedings.

Digression:
Such a situation formed the basis of the ASUS II decision. In that case, the plaintiff was able to demonstrate that it had already entered into more than 90 licensing agreements regarding the same standard, all of which were essentially identical in substance. The per-unit price contained therein then automatically represents the average value. If the value is 5, then the FRAND range lies between 2.5 and 7.5. This is significant in that it determines the maximum amount that can be charged to a technology user whose conduct has caused the patent holder particularly high costs.

74 Then again, situations are conceivable in which the patent holder concluded structurally different agreements with different licensees. In such a situation, the agreements must be analyzed in detail, and a hypothetical comparative license agreement must be determined. In this process, the more recent agreements are generally given greater weight; however, it is precisely by examining the differences between the agreements that conclusions can be drawn about the essence of the license agreements.

Digression:
Such a situation formed the basis of the RENAULT decision. There were several license agreements that differed significantly from each another. In some cases, the agreements were renewed, with major changes between the first and second agreements in each instance. In its decision (where there was no dispute regarding the factual scope of the license), the chamber determined an average value, which then served as the basis for further calculations. Due to this calculation method, it is possible that the specific average value is not explicitly stated in any single agreement, as it represents an abstract calculation parameter.

75 The specific calculation of the FRAND range is performed such that the average value is 2 and the FRAND range lies between 1 and 3. If a submitted agreement is ideally suited for comparison but was concluded with a very large market participant, this can be addressed through an adjustment to the average value.

Example:
The agreement with the large market participant is set at 5. It can be assumed that this market participant’s quantity of units would justify a 30% discount. In that case, 5 corresponds to a value of 1.4 within the aforementioned range of 1 to 3. Then, the average is 7.14. Here, the FRAND range would span between 3.57 and 10.71.

b. Digression on a special problem: compensation for past use (past release)

76 When evaluating comparable license agreements, the question often arises as to how a waiver of claims for past use should be evaluated. This is particularly important during the analysis of license agreements that include a lump-sum payment, because, for the comparison, the chamber apportions the lump-sum payment to the projected unit quantities as of the date the agreement was concluded. Particularly when the agreement is the first license agreement between the parties and periods of use occurred prior to the “official” begin of the license agreement, the question arises as to whether the prior periods were not at least de facto satisfied, too.

Example:
A licenses 4G and 5G to B for the period from 2020 to 2025 for 100. This is the first agreement between these parties. A total of 100 units are applied to these five years. A has had a large 4G portfolio since 2008, and B sold a total of 200 4G units between 2010 and 2020. During negotiations, A initially demands compensation for the period starting in 2010 as well but later limits the claim to the period starting in 2020.

The question arises as to whether the price per unit is 1 or 1/3.

77 In such a situation, the chamber is of the view that strict reliance must be placed on the content of the agreement. If no compensation for past periods is provided for and there are no other indications of abuse, then a per-unit price of 1 must be applied. This is because the patent holder’s decision not to pursue older claims to be able to conclude a current agreement constitutes an economically reasonable decision that must be accepted. Furthermore, the period prior to 2020 would not yet be fully barred by the statute of limitations at the time of the upcoming renewal, so that there would still be an incentive to reach an agreement renewal quickly even in that respect.

78 This assessment does not entail unequal treatment of other technology users either. This is because no technology user can claim that other technology users are not being held liable. The right to equal treatment does not constitute a right to participate in special benefits.

c. Geographic adjustment in comparable license agreements

79 The question whether a geographic adjustment should be made when license agreements submitted for comparison are evaluated typically concerns the question of whether the specific distribution of the patent holder’s patents – based on the production locations or the product distribution area – could result in a lack of comparability. The issue of geographic adjustment when the top-down approach is applied will be addressed elsewhere.

80 Usually, the question of where the patent holder has validated its patents and where the products are sold does not play a significant role in the assessment of comparability. This is because, in the chamber’s view, the license payment compensates the patent holder for its contribution to the development of the standard, as reflected by the number of patent families, provided that these are validated in the world’s most significant countries in economic terms in such a way that safe enforcement can be assumed. Exceptions are possible, however, particularly in cases involving production sites outside the People’s Republic of China and where sales are predominantly in countries with a less developed economy. These considerations are based on the fact that patent holders have typically validated a sufficient number of patents in the People’s Republic of China, as a typical production location, which is not necessarily always the case for other production sites.

81 Here is an example case in which the question arises whether the agreement between A and C submitted by A is suitable as a comparable agreement for the relationship between A and B, or whether a geographic adjustment is necessary:

Example:
A is headquartered in the People’s Republic of China. In total, A has contributed 1,000 patent families to the standard. Approximately 900 patents have been granted in the People’s Republic of China. In the U.S., 400 patents have been granted, and 150 patents in Europe.

B has its production facilities in the Republic of Korea. A holds only 100 granted patents there. The parties are in dispute over whether chipsets manufactured in the People’s Republic of China are used in the production process.

In the license negotiations with B, A presents a license agreement that was concluded with C. C manufactures exclusively in the People’s Republic of China. The sales locations of the products of B and C differ only slightly.

B argues that the license rate from the agreement between A and C should be applied to the relationship between A and B using the following adjustment factor: For the United States: 4/9 (400 patents in the U.S. vs. 900 patents at the production site); for Europe: 1/6 (150 patents in Europe vs. 900 patents at the production site); and for the rest of the world: 1/9 (100 patents in South Korea vs. 900 patents in China).

82 In general, when determining whether a product is to be licensed, one must consider the place of manufacture and/or the place of distribution. In this regard, it is simplest if there is sufficient coverage of the portfolio at the typical manufacturing locations. However, for the patent holder, this entails the risk that relocating the production site for essential components to another country could trigger a different assessment. What matters here is the production location of the components essential to implementing the standard – most often, if not always, the chipsets. In this regard, a mere relocation of the assembly location would not alter the assessment under patent law.

83 Typically, patent applications (based on the priority application) are filed in such a way that they cover not only the most important production locations but also the most important markets. These are primarily the U.S. and the major countries in the European Union. These countries significant in economic terms have a judicial system that makes the enforcement of patents possible in practice. Even in countries with large populations, such as India, Malaysia, Nigeria, Egypt, and Brazil, the number of patent applications has traditionally been very low.

84 In the chamber’s view, these geographic considerations may generally be taken into account when analyzing license agreements, provided, however, that an adjustment will be warranted only in exceptional cases. The background is as follows: If patent applicant A in the latter example were to seek absolute protection in every country, all 1,000 patent families would have to be validated in virtually all countries. At the very least, all potential production sites would have to become filing countries to prevent the licensing obligation from lapsing due to a change in the manufacturing location. However, this is unreasonable because it would ultimately result in the need to maintain large patent portfolios in emerging economies of minor economic significance. This contradicts the principles of economic conduct and would divert resources from the patent system that can be better used for research and development.

85 However, with regard to the example described above, there are good reasons to question whether a coverage of 100 patents at the alleged place of manufacture is sufficient to assume adequate coverage. This is because, should serious disputes actually arise between the parties, a portfolio of this size (100 patents) would in all likelihood be significantly reduced by targeted nullity complaints. And then the criterion required by the chamber – sufficient protection of the contribution to the prior art – might no longer be met. In this regard, in the situation described above, it might indeed be necessary to actually examine the plaintiffs’ argument and discuss more thoroughly whether the chipsets are, after all, manufactured at least partially in the People’s Republic of China. Whether this is the case is a question of fact and remains open. In this respect, the agreement between A and C does not constitute an agreement particularly suitable for comparison.

86 If the plaintiff cannot present a more comparable agreement, a geographic adjustment would indeed be necessary. However, this would not be carried out as proposed by the defendant. This is because such an approach would contradict the aforementioned purpose of the licensing agreement, under which the patent holder’s contribution to the standard is compensated. Rather, a percentage discount would have to be applied to the unit price specified in the agreement between A and C. In the present case, such a discount could range between 20 and 30 percent. Such a discount would be justified because the patent holder failed to fulfill its obligation to validate the patent.

d. Determination of framework

87 The question whether the plaintiff’s offer falls within the FRAND range is determined based on the final offer made at the conclusion of the oral hearing. Generally, the specific course of the negotiations – and particularly the question of the reasonableness of interim offers – no longer carries any special significance. In this respect, it is, for example, inconsequential if the plaintiff’s penultimate offer was still outside the FRAND framework. Otherwise, the chamber would have to consider the dynamics of the negotiations in its analysis, which makes no sense if the parties have since made new offers. It can also be assumed that the parties are sufficiently professional to limit themselves to a purely numerical assessment.

Specifically, the following applies:

88 Once the average value – which constitutes the value of the plaintiff’s portfolio – has been determined based on the submitted agreements, the FRAND range can be established. In principle, this FRAND range encompasses a 50% upward and a 50% downward adjustment. It should be noted at this point already that this very broad range is generally not available, as the actual circumstances narrow it. The broad range is necessary to be able to capture the complexity of the various licensing scenarios. To illustrate this, it can be assumed that the lowest amount corresponds to a license renewal by a very large user, for example; the highest amount, however, is appropriate, for example, for a smaller customer who has resisted obtaining a license for years by all means possible and is ultimately faced with a final and thus legally binding injunction.

89 The amount a patent holder can choose within a range is determined by the specific circumstances of the negotiations. Since it is common for both parties to reconcile their positions more and more during negotiations, the patent holder can start with a high figure (even above the FRAND range) in order to have leeway for compromise during the negotiations. The key point is that the patent holder must tailor their respective offer to the user’s conduct or specific offer.

e. Distinction between the FRAND range and the 15% limit

90 A conceptual distinction must be made between the actual FRAND range and the possibility of a 15% increase (15% limit). The FRAND range is the range in which, in abstract terms, an offer for a specific portfolio may fall. Where within this range the reasonable fee for a specific license seeker lies is determined by the specific circumstances of the individual case (the assessment criteria scope, conduct, etc. have already been discussed). If all factors are the same between a comparable agreement and the license seeker, then a 15% increase is still possible. The reason for this is the possibility of adjusting the fee.

Important:
This does not mean that chain increases are possible. If an agreement with A dated May 1, 2026, is comparable for B1 and specifies an amount of 10, then the patent holder can increase this agreement by 15% and demand 11.5 from B1. If, on August 1, 2026, the patent holder asserts its rights against any other comparable technology user, B2, then the agreement with B1 is not the comparable agreement; rather, it is still the agreement with A. If the patent holder now consistently charges 11.5 in comparable situations, then as of the 1, 2026, an average can be calculated from the old agreements and the agreements with B1, B2, B3, etc., so that an average of 10.5 could then be assumed. If the patent holder establishes 11.5 over a period of 3 years, then as of May 1, 2029, an average of 11.5 could be applied (which, in turn, could be increased by 15% if necessary).

91 The general concept is illustrated by the following numerical example (which is taken from the RENAULT proceedings but has no connection to the specific facts described there):

If the average value is 5, then the FRAND range is between 2.5 and 7.5. This means that it can be permissible for a very inexpensive license to be set at 2.5 and a very expensive license to be set at 7.5. However, there must be good reasons for such a distinction (otherwise, it would constitute discrimination). Typical reasons for such differentiation include the scope of the license (quantity of units/duration) and the defendant’s conduct. This is because a party that quickly enters into a license agreement causes less burden for the plaintiff, and this can be taken into account when determining the license fee.

92 Particular attention must be paid to determining the average value. This determination is based on an evaluating decision. In case of doubt, the unit rate used for comparison should be taken as the average value. A different approach may be warranted, for example, if additional agreements exist. This is illustrated by the following example:

A fundamentally comparable license agreement specifies a rate of 5. However, there are other agreements that are less suitable for comparison, for example, because the size of the companies or the products differs. In those cases, however, a rate of 6 is specified for products with a certain similarity.

In such a situation, there are indications that the value 5 is on the low side and therefore does not represent an average. The specific calculation would be such that the rate within the range of 1 to 3 is not 2 (the average) but could be set at 1.8. The range would then be between 2.78 and 8.33 (5 equal to 1.8; the average is then 5.55).

93 This range is further narrowed when specific comparable license agreements are available. This is because the principle that a licensor may not discriminate then applies.

94 In concrete terms, this means the following figures:

If, in the first example, there were a license agreement with a competitor (similar size/similar products) set at 4.5, then the generally applicable range would be narrowed (provided that the parties seeking a license behave similarly (see below)). In this regard, the chamber has already stated in its preliminary ruling of July 14, 2025, that only a 15% increase would be possible in that case. The FRAND range would thus end at 5.175 (115% of 4.5).

95 Based on the reactions to the Renault decision, the chamber assumes that the concept has been understood in principle by those applying the law. In this regard, no further explanation is necessary here.

f. Digression: the expert in license calculations

96 In SEP/FRAND proceedings, private expert opinions are regularly submitted to have the license rate deemed reasonable from the perspective of the respective party calculated. These experts are often present in person during the hearing. The chamber has already taken advantage of the opportunity to personally hear the experts present on several occasions, and this has almost always provided clarity.

97 The written expert opinions alone contribute to a better understanding of the parties’ arguments. Since the expert opinions are more detailed than the briefs, individual arguments can be more easily followed and understood. It is particularly helpful when experts present figures from independent analysis companies that are not publicly available and that, because they come from independent third parties, are often accepted by both sides. This can provide a basis for further calculations.

98 It goes without saying that an expert’s actual submission is deemed to have been admitted by the party that appointed the expert. If a party does not wish this to be the case, it must expressly state so; however, this could then raise the question of whether the expert opinion is useful.

99 Expert opinions submitted by a party must generally be addressed in the judgment. Specifically, at least a brief substantive explanation is required as to why the opinion is not followed or is not followed in its entirety.

100 Based on expert opinions that are always easily verifiable through calculation, the chamber can readily understand why the parties arrived at their offers – which in some cases are very far apart. The topics of “past release”, “geographic adjustment”, and “patent pools” are particularly frequently decisive in this context, providing a reason to address how these topics are handled in expert opinions.

In detail:

aa. Past release

101 For reasons of contractual autonomy, the question of how past use is taken into account when concluding a license agreement is within the discretion of the contracting parties. This discretion must be taken into account when unpacking license agreements for the purpose of comparison. Often, the reasons for particularly low rates in expert opinions stem from the fact that the experts did not follow this rule.

102 Generally, the patent holder is under no obligation to waive compensation for past use. However, there is also no obligation to collect compensation for past use either. There may be good reasons for both (claiming compensation and waiving it). On the one hand, a technology user should not benefit from the fact that claims against them are enforced late (which is why they cannot demand that compensation be waived). On the other hand, the parties will probably focus on the future. To achieve this, it may sometimes be necessary to set aside the past.

Digression:
To the extent that the chamber held the view that compensation for the past could be claimed only to a limited extent only under certain conditions – because the user lacks the ability to pass on the costs of obtaining a license to the buyers – this view is no longer upheld, at least for the period beginning in 2015. By then at the latest, the requirement to obtain a license was likely well known to all users of telecommunications technology. From that point onward, corresponding reserves could have been set aside. For the period prior to this, a case-by-case assessment must be conducted.

103 However, if past use is waived, a review must be conducted to ensure there is no misuse. It is impermissible for a waiver of past use to be offset by a significantly higher rate for the future if such an agreement is intended solely to allow the license agreement to be used as a comparable license agreement with respect to other technology users in order to justify and enforce higher rates.

Digression:
To establish a suspicion of misuse, the defendant would have to present figures showing that the waiver of past use, in terms of unit volume, exceeds the numbers licensed for the future in terms of quantity of units. If such a submission existed, it would be the patent holder’s task to explain how the negotiations for the comparable license agreement unfolded and why past use was waived.

104 Economic considerations alone play a role in the decision whether to waive past use. It is not apparent that considerations under antitrust law are affected to any significant extent. Therefore, third-party technology users have no (equal treatment) claim to being favored by the patent holder to a comparable extent through a waiver.

bb. Geographic adjustment

105 Another popular area of activity in the preparation of expert opinions is the performance of geographic adjustment. Generally, this involves recalculating the plaintiff’s patent portfolio to a smaller size based on actual circumstances. The possibilities are diverse, and special attention must be paid to ensuring that – as in fraction calculations – any adjustments are made not only to the numerator but also to the denominator.

106 There are good reasons to examine the geographic distribution of the patent holder’s portfolio when analyzing comparative license agreements. However, it must be taken into account that, in the chamber’s view, a license covering a larger patent portfolio does not constitute a license covering the entirety of individual patents (and patent applications) but rather compensates the patent holder for their contribution to the development of the standard. For this reason, it is not necessary for all patents to be filed and validated in all countries. Rather, it is sufficient for the patent holder to strategically distribute their patent applications in such a way that the intellectual property rights can almost certainly be enforced. This is the case if a sufficient number of patents are in force in all relevant manufacturing and sales markets. In this respect, a geographic adjustment is likely to result in only minor adjustments, at least for owners of very large patent portfolios. An adjustment exceeding 30% seems difficult to imagine.

107 The situation is different in case of small and medium-sized portfolios. If, for example, there is no patent protection in the technology user’s country of manufacture, no claim for compensation can arise from this (e.g., for the rest of the world, where there are no intellectual property rights either). If there are only a few intellectual property rights in the country of manufacture, this may justify significant discounts.

108 The chamber addressed the issue of geographic adjustment in the RENAULT and ZTE/Samsung decisions:

In RENAULT, the defendant argued that a geographic adjustment should have been made because the group did not sell vehicles in the United States of America. Therefore, the defendant argued, a discount compared to other automakers should be applied. One of the defendant’s arguments in that case was that the Renault group would therefore not face the risks of U.S. litigation. Consequently, the defendant argued, the patent holders’ potential to threaten litigation would be significantly lower, justifying a lower license fee.
In the ZTE/Samsung case, the issue of geographic adjustment was addressed within the context of calculating the top-down approach. The defendant argued that a geographic adjustment should be made because ZTE’s portfolio (approximately 3,800 patent families) was primarily validated in the People’s Republic of China. However, the defendant argued, Samsung manufactured the essential components in the Republic of Korea, where only approximately 400 patents had been granted to ZTE. This, too, justified a discount, in the defendant’s opinion.

109 Common to both lines of argument was that the respective defendant attempted to diminish the strength of the plaintiff’s patent portfolio by focusing on individual regions or countries with respect to the number of the plaintiff’s patents, while citing the number of patents granted worldwide as the relevant reference. Specifically, this was done as follows:

110 Renault (rounded figures):

Other automakers pay 10 for 20,000 patents (worldwide). Of these, 10,000 patents have been granted in the U.S. The defendant argues that – irrespective of all other arguments – it should in no case be required to pay more than 5 because it does not use the 10,000 patents in the U.S.

That submission failed to take into account the global distribution of the patents. If a geographic adjustment were to be made at all in that situation, it would have to be done by comparing the portion of the patent holder’s patents in a region to the total number of patents granted in that region.

111 ZTE/Samsung(with rounded figures):

There are 60,000 patent families. The plaintiff has 4,000 patent families. This corresponds to a share of 6.66%. In Korea, there are 400 patents from these patent families, and in the Federal Republic of Germany, there are 1,000 patents (all patents granted in Korea originate from patent families that have also been granted in Germany). There are no patents in Liechtenstein. According to the defendant’s submission, a share of 1.67% should therefore be applied for Germany and 0.67% for Liechtenstein (where only the country of manufacture is taken into account).

The chamber did not follow this approach and instead based its calculation on the patent holder’s share of the entire standard, because the patent holder had ensured sufficient coverage in all relevant markets.

cc. Patent pools

112 The analysis of patent pools and the calculation of the often very cheap fee are also frequently found in expert opinions. The value determined in this way is then usually used to demonstrate that the patent holder’s offer is significantly higher and that the technology user’s offer is based solely on the pool fee.

113 These arguments are generally ineffective, as patent pools do not, in principle, constitute suitable comparable license agreements. The lack of comparability stems from the different structures of a bilateral license agreement and a pool license agreement.

114 There are numerous reasons why a patent holder might join a pool. Often, large manufacturers of end devices participate in pools either to establish a standard or to effectively introduce a lower actual licensing burden. Another important aspect is that holders of small and/or weak patent portfolios monetize their patents through the pool, knowing that they would achieve little or no success if they were to enforce their patents in court on their own (an example of a patent pool is the Access Advance Pool for HEVC discussed in the ASUS II decision, which has 90% coverage and brings together both large and small providers, but charges a very low rate). In this respect, a single licensor not participating in a pool – whose interest is usually focused solely on the commercial exploitation of the patents through licensing and who is not afraid of disputes with potential licensees – typically differs from the majority of patent holders gathered in a pool.

115 It is not overlooked that there are also pool operators (such as SISVEL) that do indeed fully exploit the potential of the licensed patents. However, these – more expensive – pools are rarely used to reduce the requested license fees. This will be discussed in greater detail in the context of a discussion of the top-down approach.

Example:
A has a portfolio of 100 patents and demands 10. B is a patent pool with 1,000 patents and demands 5. User C believes that the 10 demanded by A are significantly excessive compared to the pool. C argues that at most, 0.5 would be reasonable, because this would correspond to 100 patents in the patent portfolio.

Usually, this line of reasoning will not be accepted. The pricing of patent pools and individual license agreements is subject to different underlying criteria. Even the relatively large difference in price and value does not justify a different assessment, because the reasons for a pool’s pricing are often not obvious, and the patent holder is under no obligation to justify the pool rate.

g. The defendant’s own license agreements

116 By its very nature, the plaintiff’s patent portfolio can be evaluated only by reference to the plaintiff’s other license agreements. However, since defendants repeatedly refer to license agreements they themselves have entered into, this argument will be addressed at the end of this section.

Example:
A has a portfolio of 100 patents and demands 10. User B has an agreement with C covering 200 patents for 10 and an agreement with D covering 100 patents for 3. B believes that A’s demand is excessive. B argues that, with regard to the agreement with C, 5 would be reasonable, and with regard to the agreement with D, 3 would be reasonable. They continue that, the fact that there are two significantly lower agreements, this proves that A is making excessive demands.

117 Generally, this line of reasoning cannot be accepted. This is because there is no presumption that the individual agreements cover the same contribution to the standard. Furthermore, patent holders may have differing interests when it comes to licensing. For example, patent holders who manufacture and distribute end devices themselves tend to have an interest in lower general license rates. Pure development or enforcement companies, in contrast, are interested in a particularly high licensing rate.

118 From a procedural standpoint, it should be noted that the chamber is not required to address the substance of such a submission until the defendant has provided a substantiated argument regarding its own comparable license agreements. In addition to submitting the relevant agreements, this always includes a written submission containing an analysis of the agreement. In particular, the figures on which the agreement was based at the time of its conclusion must generally be submitted. Only then is it possible to analyze the agreement submitted for counter-comparison. Essentially, the rules governing the submission of comparable license agreements apply accordingly.

6. Top-down approach

119 The application of the top-down approach in determining license rates for standard-essential patents represents one of the most contentious aspects of the FRAND debate. This is because assumptions must be made to determine what total burden is reasonable for a specific product or service. This opens the door to debate.

120 In reality, however, this is at most a matter of a lack of transparency. This is because in various technology sectors – particularly in the field of mobile communications – there is a substantial body of empirical data. The major market participants have already renewed existing agreements on multiple occasions. In this respect, the figures from this sector serve as a benchmark for other standards. Due to the large number of proceedings before the chamber, its members are therefore particularly well-versed in the overall situation on the licensing market.

121 In the chamber’s view, the primary purpose of the top-down approach is to make an examination of the appropriateness of the result derived from comparable license agreements possible. This is because, by its very nature, the amount determined through the top-down approach is a notional value that indicates which fee would be readily permissible for the use of a specific standard.

122 Based on this approach, a result derived from comparable license agreements is always considered reasonable if it matches or is lower than the result obtained using the top-down approach. If the result derived from comparable license agreements is significantly higher than the result of the top-down approach, then a qualified review of the comparable license agreements used is required. Essentially, the purpose of the reasonableness review is to prevent a patent holder from using a small number of non-representative comparable license agreements to attempt to enforce license rates that do not correspond to actually established licensing practices.

123 The following example illustrates this:

A demands 10, citing comparable license agreements, and the top-down approach yields a result of 10. In this case, the reasonableness of the demand is indicated. If, however, A demands 50 based on comparable license agreements and the top-down approach yields 10, then special attention must be paid to the comparable license agreements submitted. If they are freely negotiated agreements between contracting parties comparable to the user in question, then the application of 50 may be FRAND.

a. Implementation of the top-down approach

124 In the top-down approach, a discretionary assessment is used to determine how high the total burden for the use of a standard for a specific product should be (the so-called Aggregate Royalty Burden – ARB). In this regard, two factors are primarily at play: the price to be applied and the portion of that price to be paid for the use of one or all standards.

125 In licensing practice, it is common to assume a unit price with a cap (e.g., 400 euros for mobile phones). A portion of that is then allocated for the use of the standard. Based on the chamber’s experience, the percentage for the use of the mobile communications standard is generally set between 4% and 8%. The exact percentage is often dependent on the reference value. In other words: the higher the sales, the lower the percentages.

126 For the sake of simplicity and in the best interests of technology users, the chamber has decided to base its assessment not on the actual price of the users’ products, but on the price of a sufficiently equipped product of average type and quality. The chamber is aware that this approach tends to favor manufacturers of very expensive products and disadvantage manufacturers of low-cost products. However, the chamber considers this justifiable because the use of the standard provides the same options for both expensive and lower-cost products. If, as a consequence of this approach, the price of lower-cost products were to rise slightly, this would be acceptable. The limit should be set at the point where license fees would result in a situation in which parts of the world’s population are excluded from using certain technologies. However, given the amounts in question, this is unlikely to be the case.

127 Although it has been argued in some cases that lower-priced products do not take full advantage of all the possibilities offered by a standard, these claims made in previous proceedings have always been general assertions, and thus this argument has not convinced the chamber. It remains open, however, whether a different assessment would be possible if the facts were presented in a sufficiently concrete manner.

128 In conclusion, this means that a specific ARB must be applied to a product in a given category. In the ZTE/Samsung decision, the chamber ruled that a price of US$13.60 is appropriate for the use of the 5G mobile communications standard (backward compatible, i.e., including all predecessor standards) in a mobile phone. This figure from the decision of April 30, 2026, remains current as of the date of these guidelines.

129 Before delving into the details, it must be noted that this figure is purely hypothetical and says nothing about how much is actually paid for the use of the mobile communications standard in a mobile phone. There are clear indications that the amount actually paid is considerably lower, because not all patents are actually licensed, because pool offers are lower, and because even single licensors are unable to price agreements in a way that allows them to actually enforce their ARB share.

In detail:

aa. Standardized per-unit price

130 When applying the top-down approach, the chamber uses a standardized per-unit price for certain product categories. As set forth in the ASUS I decision, this standard price falls within the following range. Although the decision dealt with the Wi-Fi standard, these values are also applicable to other standards:

- Mobile phones: US$150 – US$200 (specified in ZTE/Samsung as $170)
- Tablets: US$150 – US$200
- Laptops/PCs: US$500 – US$550
- Routers: US$130 – US$150

131 In the ZTE/Samsung decision, the chamber specified the standardized per-unit price for mobile phones as US$170. No determination has yet been made regarding the other categories. Consequently, the lowest value will always be applied.

132 Building on the ASUS I decision, it should be clarified that these product groups are groups with the characteristics typical of their respective device type. This means that for tablets and laptops/PCs with mobile communication capability, there is a different price range, which is likely to be approximately US$100 higher.

bb. Aggregate royalty burden

(1.) Single standard ARB

133 In determining the aggregate royalty burden, the chamber, in the ASUS I decision, relied on the amounts that had been cited as reasonable by the respective parties in various proceedings. The lower figure reflected the technology user’s statement, while the higher figure reflected the patent holder’s statement. These figures related to the use of Wi-Fi in the respective product category. Taking into account the submissions presented in the proceedings since the ASUS I decision, the chamber is of the opinion that, for device types that do not use the mobile communications standard, a higher ARB must be applied with respect to the use of Wi-Fi. This is made clear by the additions in parentheses.

- Mobile phones: 2 – 6%
- Tablets: 3 – 7% (corrected: 4 to 10%)
- Laptops/PCs: 2 – 7% (corrected: 3 to 8%)
- Router: 6 – 10%

134 Determining the exact percentage within this range requires a discretionary assessment. In the ZTE/Samsung decision, the chamber noted regarding the mobile communications standard that the parties had cited figures ranging from 4% to 8%. The chamber opted for a rate of 8%, i.e. at the upper end of the range, while taking particular account of the fact that the per-unit price of US$170 is set low. This particularly applies to the specific ZTE/Samsung case because Samsung sells a high proportion of high-end mobile phones; however, the application of this price is also justified for less expensive mobile phones. This is because – as far as can be seen – expensive and inexpensive mobile phones do not differ in terms of their standard-essential mobile communication capabilities, but rather in terms of storage capacity, display quality, camera performance, and external factors such as design or brand prestige.

135 As a result, the chamber ruled that the ARB for the use of the 5G mobile communications standard (and all backward-compatible standards) for a mobile phone is US$13.60.

(2.) Total ARB

136 In order for a device to be used effectively, it must often be capable of implementing multiple commercially exploited standards. In the case of a mobile phone, this applies not only to the mobile communications standard but also to Wi-Fi and the essential streaming protocols, assuming backward compatibility in each case. Overall, the following standards should be assumed:

- Mobile phones: mobile communications, Wi-Fi, streaming
- Tablets: Wi-Fi, streaming
- Laptops/PCs: Wi-Fi, streaming

137 Applying the individual standards to these devices yields the following values, with the proportion for streaming being higher for tablets and laptops than for mobile phones because their specific usage differs:

- Mobile phones: Mobile communications: 8%; Wi-Fi: 2 – 6%; Streaming: 3 – 6%
- Tablets: Mobile communications: not applicable; Wi-Fi: 4 – 10%; Streaming: 4 – 8%
- Laptops/PCs: mobile communications: not applicable; Wi-Fi: 3 – 8%; Streaming: 4 – 8%

138 This results in total ARBs ranging from 13–20% for mobile phones, 8–18% for tablets, and 7–16% for laptops/PCs. As explained above, this wide range is based on the fact that the figures provided by the parties were used.

139 The chamber thinks that a rate of 18% appears appropriate for mobile phones (mobile communications 8%, Wi-Fi 5%, and streaming 5%). As a result, this means that the total ARB for mobile phones is thus calculated at 18% of $170 and hence amounts to US$30.60. For the other product groups, further factual submissions in proceedings are required before a determination can be made.

cc. Regarding the application discounts and surcharges

140 In the review of the top-down approach, discounts and surcharges are also taken into account. The chamber is aware that discounts and surcharges do not structurally fit the nature of the top-down approach, as this approach is based solely on the value of the standard for the usability of the device. The fact that discounts and surcharges are nevertheless taken into account is due to the fact that the FRAND calculation is dominated by the use of comparable license agreements. Therefore, the elements that must be considered in an analysis of comparable licenses must also be applied when implementing the top-down approach.

141 In summary, the top-down approach is product-specific, while the discounts and surcharges are technology-user-specific.

dd. Discounts

142 Discounts on the determined per-unit price may be justified, particularly when the technology user must take a license for very large quantities. In the ZTE/Samsung decision, the chamber assumed a 30% discount. Since Samsung is either the largest or one of the largest manufacturers of mobile phones worldwide, this represents the maximum discount. It can be assumed that the requirements for a discount should not be set too high and are generally met if the number of units licensed per year is in the seven-digit range. Even in such cases, a 10% discount is likely to be generally permissible.

143 Surcharges are justified if the user’s conduct has unduly delayed the conclusion of a settlement agreement. This is because, in that case, the patent holder has incurred costs that were not factored into the conventional calculation. Determining when the threshold for permissible delay has been exceeded requires a discretionary assessment.

144 In the ZTE/Samsung decision, this threshold was already exceeded simply because the defendant in that case had filed an exclusion request with the standardization body ETSI. Consequently, a 10% surcharge (over US$60 million) was justified in that case. For the same conduct, a higher percentage surcharge would also have been possible with a lower license sum, since these surcharges essentially serve to cover the additional costs incurred by the patent holder. However, these actual additional costs do not cap the surcharge either.

145 It need not be determined whether merely filing a request with an incompetent court to determine a rate justifies a penalty surcharge. Although there are good reasons to support this view, this question did not arise in the ZTE/Samsung case because both parties proceeded in the same manner in this regard.

b. Determination of the share/value of the patent holder’s portfolio

146 Once the ARB has been calculated by determining the per-unit price and the percentage share, the value to be applied for a one-percent share in the standard is also established. For 5G mobile communications, based on the calculation method from ZTE/Samsung, one percent in the 5G mobile communications standard has a value of US$0.136 (13.6 U.S. cents). This value must now generally be applied to the size of the patent holder’s portfolio.

147 The central question is how many patents (and patent applications) constitute the standard and which share the patent holder holds.

aa. Determination of the relevant reference value

148 The total number of patents and patent applications that constitute the standard must be determined. The onus of presentation and burden of proof lie with the plaintiff, and the principles of a graduated onus of presentation and burden of proof apply. This means that a substantiated statement of facts requires the defendant, in turn, to make a substantiated submission.

149 In a large number of proceedings, the parties agree to use the numerical data provided by a specific supplier. The discussion usually centers on the question which items should be taken into account. In individual cases, there is, among other things, a dispute over whether declared or actual standard essentiality should be the basis for determination; however, this issue presently remains open because it ultimately depends on the specific submissions made in the proceedings.

150 The most important rules in this regard are:

- The same method of determination must be applied to both the total portfolio and the plaintiff’s portfolio.
- Reductions in the plaintiff’s portfolio must always be applied to all patent holders in the same situation, such that any change to the plaintiff’s portfolio must always be accompanied by a recalculation of the total portfolio.

151 With regard to individual items, the following applies:

- Patents are to be taken into account only for the period during which they are in force. For the future, the status as of the date of the last oral hearing may generally be used as a basis. In this respect, it is to be assumed that the expiration and grant of patents balance each other out. Any contrary approach must be substantiated by the parties.
- In addition to granted patents, patent applications must also be included. Where appropriate, a discount of 20 to 30% is justified (i.e., a granted patent counts as 1; a patent application counts as 0.7 or 0.8). This is particularly important for rapidly growing companies.
- Whether the assessment is based on patents declared to be standard-essential or on those that are actually standard-essential (however that is evaluated) is a matter to be determined on a case-by-case basis and requires a substantiated submission by the parties.

152 With regard to different versions of the same standard (e.g., 3G, 4G, and 5G or Wi-Fi 6 and Wi-Fi 7), the following principles apply:

- It is assumed that, in case of doubt, licensing of the highest standard always encompasses all patents of an earlier standard as well.
- If a standard other than the highest one is licensed, this is to be understood – in case of doubt – to mean that the devices covered by the licensed standard are those for which that standard is the highest standard. In other words: If 4G is licensed, all devices that have 4G as their highest standard are covered. 5G-capable devices are not licensed at all, unless expressly agreed otherwise in the agreement. The 4G patents in the 5G device are therefore not licensed.
- Based on these principles, patents that are relevant to multiple standards are to be considered only in the highest standard in each case.

bb. Correction of the result found

153 One argument against using the top-down approach might be that a very aggressive filing behavior can be advantageous because the focus is solely on the sheer number of patents. In fact, the chamber also recognizes that the quality of patent portfolios varies. In this regard, from a procedural standpoint, it is the responsibility of technology users to submit substantiated arguments to demonstrate that the size of the plaintiff’s patent portfolio is attributable to specific filing behavior. To date, this has not occurred in any case, although there are numerous lines of argument that could be pursued:

- For publicly traded companies, the number of patent applications could be compared to the published research and development figures.
- The number of contributions to standardization meetings could be measured.
- The number of citations of the developments by the patent holder in other works could be identified.
- Etc.

154 In the wake of the ZTE/Samsung decision, further findings have emerged that justify a rebuttable presumption that companies from the People’s Republic of China possess artificially inflated patent portfolios. This is based on the fact that, within the framework of the People’s Republic of China’s state-directed economy, patent applications have been financially subsidized and thereby favored. As a result, a particularly large number of patents were filed. Although this government program is probably no longer in effect, its effects persist in the current portfolio of intellectual property rights and place market participants from countries whose governments do not require or promote such measures at a disadvantage. In the chamber’s view, this interference with the free market economy justifies a rebuttable presumption that a 15% discount should be applied to the patent portfolios of Chinese companies. However, there is no basis for a further adjustment through a corresponding increase in the values of other companies’ portfolios.

c. Use cases of the top-down approach

155 The top-down approach has several areas of application. It is important to emphasize once again the starting point, namely that the result obtained does not represent the actual burden of license payments. Rather, the result obtained is an amount determined on the basis of well-founded assumptions but is, in essence, a fictitious amount that provides insight into what fee for the use of a specific standard would be readily permissible.

156 The chamber assumes that the current reality of licensing is such that no medium-sized technology user pays, in total, the amount recognized as the permissible aggregate royalty burden for the respective standard. In the current licensing landscape, the barriers to enforcement for patent holders are quite high, and the public perception of cases in which high licensing claims have been asserted does not correspond to their actual economic significance.

157 If, in exceptional cases, a defendant were to have a total burden that exceeded the top-down result, the respective user would be required to make a submission demonstrating the total amount paid for the use of a specific standard.

aa. Review of the result derived from comparable license agreements

158 Typically, the top-down approach serves to verify whether a result derived from comparable agreements is reasonable overall. The rationale behind this is that, when performing the calculation using the top-down approach – at least as specified by this chamber – most assumptions have already been made in favor of the technology users. In particular, the use of a fixed per-unit price at a relatively low level favors manufacturers of products in the mid- and higher-price categories. To the extent that particularly manufacturers of the lowest-priced products criticize a value that is high in relation to the sales price, no interests worthy of protection appear to be affected. This is because, generally, these companies in particular have a very low degree of licensing, and the low price is primarily due to – and made possible by – a refusal to obtain the necessary licenses.

159 To the extent that the result derived from comparable agreements does not exceed the result determined by the top-down approach, the amount is presumed to be reasonable. If the result derived from comparable license agreements is higher, a distinction must be made between whether the patent holder has a 1 to 2 percent share in the standard or more, or whether the patent holder is a “small” patent holder. This is because holders of smaller patent portfolios (“outsiders”) must be assessed differently than holders of larger patent portfolios, since with larger patent portfolios it can be assumed that the innovation clout and the number of patent applications roughly correspond to each other. For holders of smaller portfolios, whose innovation clout is not adequately reflected by the number of patent applications, a more detailed analysis is required.

160 For a larger patent portfolio, the price determined using the top-down approach – possibly with a small surcharge – usually represents the upper limit of the admissible price. A different assessment may be warranted only if there is a large number of freely negotiated comparable license agreements.

bb. Impact of a portfolio split on the top-down approach

161 Once a patent has been part of a larger patent portfolio, it does not acquire the privileged status of an “outsider patent” because the larger portfolio is split up. Rather, it remains tied to the portfolio price.

162 However, it must be acknowledged that holders of larger patent portfolios may have an interest in selling parts of their portfolio, for example, to raise funds for enforcing their intellectual property rights. In such a situation, the acquirer has a legitimate interest in receiving compensation for the proportionate additional effort involved. However, the price determined using the top-down approach must not be arbitrarily increased. A reasonable approach might be a multiplier of 3 to 5 times the value corresponding to the result of the top-down approach. The exact amount is dependent on the size of the portfolio and the size of the acquirer. Ultimately, the goal is to appropriately account for the additional effort and profit expectations of the new patent holder.

cc. Initial determination of a license fee using the top-down approach

163 If there is no established licensing practice, an initial fee may be set using the top-down approach. This was the situation underlying the ZTE/Samsung decision.

In detail:

164 In the ZTE/Samsung decision, the plaintiff made the submission that there were no suitable comparable license agreements. The defendant in that case referred to a license agreement concluded between the parties in 2021 and sought to have the rates specified therein applied to the future as well. In doing so, the defendant allocated the amount agreed upon in 2021 particularly to the very high unit quantities that Samsung sold in the 2010s (the “past release” issue).

165 In the aforementioned decision, the chamber did not consider this agreement to be a suitable comparable license agreement and concluded that other license agreements submitted in the proceedings were also unsuitable for indicating the value of the plaintiff’s portfolio.

166 Consequently, the reasonable license rate could be determined using the top-down approach. In this respect, a special situation also existed because ZTE entered the mobile communications patent licensing market relatively late. Some other companies already have license agreements with users that have been renewed multiple times. This may mean that ZTE can possibly actually charge a higher license rate than would have been the case had it begun actual licensing earlier.

167 Following the ZTE/Samsung decision, there were proceedings before the chamber in which parties that had entered into numerous license agreements made the submission that none of the concluded license agreements were comparable. The apparent reason for this is that the value of the license agreements actually concluded is lower than the amount that would result from the top-down approach. In the chamber’s view, such an approach is impermissible: if comparable license agreements exist, then the value is determined based on those agreements.

dd. Price adjustment exceeding the 15% threshold

168 According to the chamber’s case law, the patent holder has the option to increase the average value determined by comparable license agreements by 15%. This can be done in a bilateral process by agreeing on a higher price in new agreements. In this way, the demanded license fee can be increased by approximately 30 to 50% over a longer period of 10 years (assuming the portfolio maintains the same scope through new applications), which roughly corresponds to the rate of inflation.

169 This limitation becomes problematic when the starting value is low and does not fully exploit the potential of the patent portfolio. Since growth is based solely on a percentage increase from a (low) starting value, it may not be possible to realize the actual licensing potential.

170 Therefore, in the chamber’s view, it is permissible to completely renew a licensing program. This must be done in such a way that a patent holder discloses the current rates in a sufficiently transparent manner and, for a transitional period of at least six months, offers all interested parties the opportunity to obtain a license at the previous rates (taking into account the 15% increase, if applicable). After this period expires, the price for the patent portfolio is no longer limited by the previous license agreements; instead, the newly established (and openly communicated) rates take effect.

7. Specific situations

171 The following section briefly presents and analyzes the ASUS I, ASUS II, RENAULT, and ZTE/Samsung decisions. Subsequently, the principles are applied to obvious scenarios.

a. ASUS I

172 In the ASUS I decision, the chamber’s principles set forth in the preliminary rulings in cases 7 O 64/25 and 7 O 2750/25 of July 14, 2025, were specified for the first time. With regard to the FRAND calculation, this constituted an obiter dictum because the defendant had not even cleared the hurdle of an external willingness to obtain a license. The case was suitable for in-depth analysis because the plaintiff was a member of the SISVEL WiFi 6 Pool, whose rates were publicly known. In fact, however, the parties negotiated only a bilateral license.

In detail, the following applies:

173 The defendant primarily manufactured laptops and PCs. The plaintiff was a member of the SISVEL WiFi 6 Pool. To clarify the calculation method, the chamber first conducted an analysis of the pool rate. The pool was priced such that there was – and still is – a standard price of 0.60 euros and a reduced rate of 0.50 euros. The reduced rate applied from the time the license was obtained.

174 Since the parties had not submitted any relevant figures other than the defendant’s unit quantities, the chamber used publicly available figures to determine the reasonableness of the plaintiff’s offer. In doing so, the chamber made all assumptions in favor of the defendant.

175 In favor of the defendant, it was assumed that the SISVEL Pool WiFi 6 represents only 10% of the standard. In fact, this figure is likely to be significantly higher (somewhere between 15 and 20%). Although the defendant primarily manufactures laptops and PCs (i.e., products with a high per-unit price), a 50% share for laptops and PCs was assumed in favor of the defendant (a share of over 80% is likely to be realistic). The other devices were calculated based on the rate of the least expensive product group.

176 At that time, the chamber assumed a specific product price of

- mobile phones: US$150 – US$200
- laptops/PCs: US$500 – US$550

and an ARB for Wi-Fi 6 of

- mobile phones: 2 – 6%
- laptops/PCs: 2 – 7%

and thus an ARB unit price of

- mobile phones: US$3 – 12
- laptops/PCs: US$10 – $38.5

177 In favor of the defendant, the lowest values were used as the basis for the calculations in each case (lowest price and lowest ARB). Thus, 50% laptops/PCs and 50% mobile phones were assumed, resulting in an average price of US$0.065 per one percent in the standard. Based on the assumed 10%, a unit price of US$0.65 would be justified; thus, the chamber concluded that the pool rates published on the Internet (which are stated in euros) were not exceeded, taking into account the share of the two rates.

178 With regard to the relation between the pool rate and the bilateral rate, reference is made to the decision. It should also be noted that the chamber is of the opinion that the pool rate should, on principle, be more favorable than the combined bilateral rates of the three largest pool members, because otherwise there would be too great an incentive to circumvent pool licensing.

From today’s perspective, the following should be added:

179 In retrospect, the chamber would make some different assumptions today. For the product price of mobile phones, a price of US$170 would be set. With regard to the ARB, the chamber would no longer accept the defendant’s price but would instead adopt a low value that is, however, at least within the realm of economic reality. According to the chamber’s current view, these values would be at least 4% for mobile phones and 3% for laptops (in this regard, further submissions from the parties may be necessary). Overall, this would result in the following values:

- mobile phones: Us$170, 4%: US$6.80
- laptops/PCs: US$500, 3%: US$15

180 Based on the calculation method described above, this results in a calculated value of US$10.90 as the total ARB. If this is then allocated using a realistic figure of 80% of products sold in the Laptops/PCs category, the result is a value of US$13.36 (total ARB). This corresponds to a value of US$0.1336 per one percent in the standard. Taking into account the actual share in the standard of at least 15%, a reasonable unit price is US$2.00.

181 Against this backdrop, it can be assumed that even if a discount were applied for the Chinese patents in the portfolio, the pool does not fully realize its inherent licensing potential. Even for the least valuable product group, the value amounts to US$0.867 (15% of US$6.8, minus 15%).

b. ASUS II

182 The ASUS II decision – which concerned the HEVC standard – structurally differs from the ASUS I decision, as it is based on an analysis of over 90 comparable license agreements submitted by the plaintiff. However, the fact that the plaintiff admittedly introduced the agreements into the proceedings relatively late, and thus shortly before the oral hearing, had no impact on the calculation.

183 In that case, the defendant had paid the undisputed amount to the plaintiff. Based on the totality of the circumstances, the chamber assumed that the requirement of external willingness to take a license had been met. In this respect, it was decisive for the decision whether the plaintiff’s offer was FRAND. Substantially, it is the first decision for which the question whether the plaintiff’s offer was FRAND was decisive.

184 An analysis of the comparable license agreements showed that the plaintiff in that case established and implemented a licensing program that differentiates based on the Average Sales Price (ASP) and applies a uniform license rate to all products with a sales price of 90 euros. With regard to the criteria for analyzing the submitted license agreements, reference is made to the decision, which states:

- There is a wide variety of licensees operating in different technological fields.
- The license agreement with the lowest value covers a quantity of significantly fewer than 1,000 units.
- Companies are concerned that can be described as reputable and well-known in the market and which can be assumed to possess sufficient business acumen.
- Some of the license agreements were of significant economic importance and involved substantial payment.
o a license agreement with a 9-digit volume
o a license agreement with a high 8-digit volume
o several license agreements with a low 8-digit volume
o numerous license agreements with a 7-digit value
- No license agreements were submitted in which the licensed products essentially required mobile phone use.
- What all the license agreements had in common was that the unit prices actually paid were close to the rate demanded by the plaintiff as a starting point.
- The highest discount was less than 20%.
- The highest discount was given to a manufacturer with high volume.
- For manufacturers of very low-cost devices (low ASP), the low rates of the licensing program were applied. No further discounts were granted.

185 Of particular significance for the qualitative assessment of the licensing program were the facts that there were agreements with very high revenue and that, overall, only moderate discounts were granted.

186 The defendant’s main argument was the claim that the requested license rate was unreasonable because it was overpriced compared to the rates of the HEVC pools operated by Access Advance and the VIA Licensing Alliance. In this regard, it was indeed found that the price of the plaintiff’s portfolio – when broken down to the “price per individual patent” – was significantly higher than that of the two pools. In this context, the chamber conducted a brief analysis of the pool rates and concluded that neither pool is fully exploiting its licensing potential. However, the existence of a favorable pool landscape has no impact on the price that non-pool members can charge for licenses.

187 As a result, the chamber did not conduct a top-down analysis of the HEVC standard as a control review but rather based its decision on the fact that the plaintiff, as the owner of a small portfolio (an “outsider”), is explicitly not bound by the top-down approach.

188 If the case were to be decided again, the chamber would perform a control calculation using the top-down approach. In doing so, the chamber would arrive at the following figures:

- mobile phones: US$170, streaming: 3 – 6% (specifically 4%) = US $6.8
- tablets: US$150, streaming: 4 – 8% (specifically 6%) = US$9
- laptops/PCs: US$500, streaming: 4 – 8% (specifically 4%) = US$20

189 As a result, the plaintiff could – according to the top-down approach – charge $0.068 for mobile phones, $0.09 for tablets, and $0.20 for laptops/PCs based on its 1% share in the standard. Because the amount actually demanded exceeds these amounts, a qualified analysis of the license agreements submitted by the plaintiff for comparison would be necessary. For the reasons already stated, this analysis would be in favor of the plaintiff.

190 For the sake of clarity, it should also be noted that the specific decision primarily concerned the “laptops/PCs” product group. Within this group, the gap between the demanded amount and the result of the top-down approach was smallest. In this regard, it must also be taken into account that the agreement in question was a streaming-only license agreement. Particularly in the case of the plaintiff, which operates in various fields, it can be assumed that, for example, with regard to the “mobile phones” product group, different comparable agreements would be presented because the mobile communications standard would be the focus there. The same applies to tablets with regard to the Wi-Fi standard. In this respect, entirely different questions would arise regarding the more affordable product groups when performing the comparison calculation using the top-down approach.

c. RENAULT

191 In the RENAULT decision, the reasonableness of the plaintiff’s license offer is demonstrated by various license agreements of the plaintiff. The content of these agreements differed quite significantly from one another, and therefore a qualified analysis of the various agreements had to be conducted. The specific procedure for determining the average value and the argument pursued by the defendant in that case regarding the need for a geographic adjustment have already been explained.

192 Since the licensed portfolio in that case consisted of a mix of various standards, including non-standard-essential patents, a comparative calculation using the top-down approach was not possible.

d. ZTE/Samsung

193 In the ZTE/Samsung decision, the chamber applied the principles established in previous decisions. Since the parties are litigating before various courts worldwide, this decision has received particular attention. Ultimately, the chamber deemed several comparable license agreements irrelevant and therefore permitted a license calculation using the top-down approach.

194 An amount of US$798,642,530.40 was specified as the upper limit of the FRAND corridor. In addition, the chamber stated that it would consider a lump-sum payment of US$640 million for a period of 5 years (2024 to 2028) appropriate.

195 This decision reflects the submissions made by the parties at that time. Although the defendant raised objections regarding the quality of the plaintiff’s portfolio, it failed to provide specific figures; the defendant must accept this as a disadvantage. Therefore, no discount was warranted at the time of the oral hearing on April 30, 2026.

196 If the same situation were to be decided again, the chamber would apply a 15% discount to the plaintiff’s portfolio – regardless of the defendant’s specific submissions – in order to address the issue of the expansive filing practices of companies from the People’s Republic of China. However, this would not fundamentally alter the decision. Even if a 15% discount were applied to the maximum framework amount of US$798,642,530.40 (US$678,846,150.80), the plaintiff’s offer would still fall within the reduced framework. Taking the discount into account, however, the proposed rate for the same period would have been set at US$550 million instead of US$640 million.

e. AVANCI 5G pool licenses

197 The aforementioned principles apply to AVANCI’s 5G pool regarding the licensing of mobile communications technology for automobiles. According to the information on AVANCI’s website, the standard price is US$32 per vehicle for lifetime use of that specific vehicle. In addition, under certain circumstances, it is possible to obtain an “early licensee price” of US$29.

198 According to the pool’s own statement, more than 90 patent holders have joined the pool, and coverage probably exceeds 90% of mobile communications standards. In particular, the pool includes patent holders whose patent portfolios are generally considered to be the strongest (Qualcomm, Nokia, Interdigital, Ericsson, etc.). In other words: Given a patent portfolio of over 50,000 patents, the pool is as close to a comprehensive license as it can be.

Comparable license agreements

199 The fact that the pool license has been adopted by a large number of car manufacturers already speaks to the reasonableness of the demanded rate. The chamber does not ignore the fact that, in some cases, this occurred only under the pressure of injunctions. As far as can be seen, however, the underlying causes of these disputes were not due to the amount of the demand, but rather the specific contractual structure of car manufacturers and suppliers.

200 The demanded rate also holds up to a review using the top-down approach. In fact, there are significant indications that even the AVANCI Pool is not fully realizing the portfolio’s licensing potential.

Top-down approach

201 A first approximation to the demanded US$32 leads to a comparison with the price of a 5G license for a mobile phone. The total ARB for 5G (backward-compatible) was set at US$13.60 in the ZTE/Samsung decision, and based solely on this magnitude, a 2.5-fold increase seems reasonable for an automobile. This is because a vehicle has an average useful life that is approximately three times as long as that of a mobile phone. It can be assumed that a mobile phone has an average useful life of 3 to 5 years and a passenger car has an average useful life of 10 to 15 years.

202 Based on this, an amount of US$36.72 (3 × US$13.60 × 0.9 (share in standard)) would be FRAND.

203 A significantly higher value might possibly result if the calculation were based on the value of the components required to implement 5G technology in a vehicle. However, this would require a further statement of fact by the parties.

204 In addition, beyond the criterion of a longer useful life, consideration should also be given to the potential cost savings (efficiency gains) made possible by the licensed technology. This is because maintenance represents a recurring revenue stream for car manufacturers, and mobile communications access to vehicles that have already been sold allows car manufacturers to perform some maintenance tasks remotely (or have it performed remotely) and to prepare for upcoming maintenance in some cases. It requires careful consideration whether there is a difference in this respect from the mobile phone business, where the costs of operating system updates are not passed on to device users.

205 From the perspective of a possibility to commercialize a product, it could also be discussed whether the real-time transmission of user data represents additional value for car manufacturers. This, too, can constitute a value to be used as the basis for the license calculation. In this regard, a follow-up consideration would be whether, on principle, a licensing model might be possible that combines a one-time payment for certain uses and a time-dependent element for other uses. For should this be possible, elements containing a time component could be satisfied by increasing the one-time payment. However, this would have to be decided on the basis of actual situations supported by substantiated factual submissions. The fact that such a licensing model could, on principle, be possible illustrates that the AVANCI Pool’s current licensing practice is particularly simple and advantageous for licensees – not least because it is designed to be inflation-neutral.

f. “Streaming as a service”

206 The chamber defines “streaming as a service” as offers by companies that stream content to consumers and whose content benefits from advanced and efficient encoding and decoding technology. To use these services, devices from other providers are generally required which provide the necessary hardware components.

207 The chamber has already dealt with numerous proceedings involving complaints against manufacturers of streaming-capable products, such as mobile phones, tablets, laptops, and PCs. With regard to these proceedings, there are clear rules, comparable license agreements, and ultimately a fundamental understanding on the market that licensing is required.

208 Recently, there has also been an increase in complaints against streaming content providers. Typical and well-known streaming providers include Netflix, PrimeVideo, Disney+, Apple TV, DAZN, Sky, Magenta, YouTube, etc. In these complaints, a FRAND defense is often not raised. The formal justification for this is that there is said to be a difference between encoding and decoding with regard to standard relevance. While this argument is not convincing, since an objection cannot be raised against the will of the party entitled to raise it, there have been no decisions to date on the issues associated with the licensing of streaming services.

209 Hereinafter, the chamber has performed a sample calculation based on publicly available figures; while this calculation is not binding for the chamber, it should be understood as a guideline indicating which facts would need to be submitted in any potential proceedings.

Exhaustion under patent law

210 In the chamber’s view, there are no reasons why these providers cannot be held liable simply because the devices used to play the content already have a license. This is because providing streaming capability and offering a streaming service are different types of use, such that the principles of exhaustion under patent law do not apply.

Licensing practice

211 To date, there has been little experience regarding the specific licensing of “streaming as a service.” As far as can be determined, individual providers who also manufacture end devices in large quantities have been able to also include “streaming as a service” in the licensing package as part of license negotiations covering multiple standards. Since these are multi-standard licenses, these license agreements are generally not suitable for single-standard licenses.

212 Recently, the Access Advance “Video Distribution Patent Pool” (covering the HEVC standard, VVC standard, VP9 specification, and AV1 specification) and the “Avanci Video” pool (covering AV1, HEVC, VVC, MPEG-DASH, and VP9) were established. The “Avanci Video Pool” sets its standard license rate at either 1.6% to 2.0% of revenue or US$0.12 to US$0.15 per user per month. The Access Advance “Video Distribution Patent Pool” is less expensive, with rates going up to 0.5% of revenue. Both pools still have relatively limited coverage of licensors and have only a few licensees that are also of less economic importance.

213 Overall, it can therefore be noted that there is no established licensing practice. Consequently, a reasonable licensing rate could be calculated using the top-down approach.

Top-down approach

214 In order to perform a comparative calculation using a top-down approach, it is necessary to determine a per-unit price and to determine the reasonable Aggregate Royalty Burden. Determining the per-unit price presents numerous challenges because the various offerings are structured in completely different ways. In some cases, users must pay monthly fees; in others, the offers are financed through advertising and the analysis of user data. In some cases, the pricing is a hybrid model. Particularly for smaller providers, the price consists overwhelmingly of the costs for streaming rights, especially for sports events. When determining the ARB, empirical values from the licensing of standards – where there is a sufficiently long history of licensing experience – could be transferred.

Per-unit price:

215 The following discussion focuses on prices in the Federal Republic of Germany. The chamber is aware that the price structure is high and that significantly lower prices apply in the vast majority of countries.

216 In the chamber’s view, the streaming providers Netflix and Disney+ are best suited for an initial analysis. This is because these are services that are fully paid for by the fees charged to users. In contrast, Prime Video, Apple TV, and Magenta TV are so closely linked to other offers by their respective providers that determining the value of the actual streaming service would be even more difficult. In the case of YouTube, advertising – and possibly data collection as well – plays such a significant role that an assessment appears difficult.

217 Netflix currently offers three different plans:

- standard subscription with ads for 4.99 euros/month
- standard for 13.99 euros/month
- premium (4K and HDR) for 19.99 euros/month

218 Disney+ also offers three plans, with the regular prices (excluding special offers) listed as follows:

- standard subscription with ads for 6.99 euros/month
- standard subscription for 10.99 euros/month
- premium (4K and HDR) for 15.99 euros/month

219 In total, a standard annual subscription costs 167.88 euros on Netflix and 131.88 euros on Disney+. The premium annual subscription costs 239.88 euros and 191.88 euros, respectively. Excluding VAT, this amounts to:

- Netflix: 141.08 euros and 201.58 euros
- Disney+: 110.82 euros and 161.24 euros

220 It remains to be seen whether these prices actually cover costs, or whether pricing must also take into account that this is a new market in which market shares are currently being allocated and will continue to be allocated in the future.

221 Besides, there are the prices set by the chamber for devices of medium type and quality:

- mobile phone: US$170
- tablet: US$150 – US$ 200

222 Although these prices represent the retail price excluding value-added tax (ASP), it is important to take into account that devices are typically sold through retailers, while streaming services are sold directly to end users.

223 If we factor in a retailer margin of approximately 10 to 20% for devices, the manufacturer ultimately retains US$136 – US$153 for a mobile phone and US$120 – US$135 for a low-cost tablet. This amount must be sufficient to cover manufacturing and distribution costs, the necessary licenses, and profit. In this respect, there is a difference compared to streaming services, which deliver directly to end customers. In addition to the costs of film rights, there are distribution costs and costs for storage space, although there are significantly higher economies of scale. The main costs arise from setting up the infrastructure; once operations are running, there is little difference whether 10,000 or 1 million users utilize the service. More detailed information on the structures and calculation bases will have to be provided by the parties – where appropriate, with reference to expert opinions – in relevant proceedings.

224 The most significant difference between end devices and streaming services is probably that the transmitted content (movies, series, documentaries, etc.) is particularly valuable. The chamber takes this into account by applying a 2/3 discount. In other words, most user fees are charged for the content. The remaining portion is:

- Netflix: standard: 47.03 euros and premium: 67.19 euros
- Disney+: standard: 36.94 euros and premium: 53.75 euros

225 In this regard, it should also be taken into account that the difference between the standard and premium prices is primarily due to the difference in video quality (i.e., the video standard). Specifically, the difference is:

- Netflix: 20.16 euros per year / 1.68 euros per month
- Disney+: 16.81 euros per year / 1.40 euros per month

Determination of a reasonable ARB

226 In the ZTE/Samsung decision, the chamber set an ARB of 8% for the use of the 5G mobile communications standard (backward compatible). As explained above, the chamber assumes that additional licensing costs must be factored in for Wi-Fi and streaming. Therefore, as explained above, the total ARB probably ranges from 13% to 20%. Even assuming a low average of 16% (instead of the realistic 18%), this would correspond to a total amount of US$27.20.

227 Streaming service providers do not incur licensing costs for standard-essential patents beyond those associated with the various streaming standards; thus, a service provider has a different cost structure than a device manufacturer. In this regard, the chamber assumes that the amount attributable to all standards in the case of a mobile phone is the correct benchmark for determining the share for all streaming standards. This means that, at a minimum, the figure of 16% has to be applied for a standard subscription. Whether an increase in this amount is necessary and reasonable is subject of discussion between the parties. Specifically, this implies:

- Netflix: standard: 47.03 euros = 7.52 euros per year = 0.63 euros per month
- Disney+: standard: 36.94 euros = 5.91 euros per year = 0.49 euros per month

228 In the chamber’s view, a different standard applies to premium subscriptions. Since the price difference is essentially defined by streaming quality, a share of 25% of the difference must be paid as a license fee. Any higher license fee for the corresponding content is already covered by the flat-rate deduction of 2/3. This results in the following figures:

- Netflix: Basis: 0.63 euros plus 25% of 1.68 euros: 1.05 euros per month
- Disney+: Basis: 0.49 euros plus 25% of 1.40 euros: 0.84 euros per month

229 Consequently, these values could then serve as a starting point for allocation among the individual patent holders based on the size of their portfolios. In making this allocation, the principles set forth in this preliminary decision – including the details regarding surcharges and discounts – would need to be taken into account.

Summary:
230 Many questions remain unanswered in the field of streaming. The above remarks are not binding to the chamber but represent an application of general principles to issues related to streaming. A specific statement of fact is required to assess the reasonableness of the demanded license payments.

IV. Concluding remarks

231 These guidelines are intended to provide the parties in FRAND/SEP proceedings with the necessary assistance to review whether the plaintiff’s offer is FRAND. From a practical standpoint, they also clarify which facts must be submitted in proceedings. Due to the complexity of the possible procedural scenarios, the chamber has refrained from including a checklist in these guidelines.

232 However, the general principle applies that a detailed presentation of facts at an early stage helps conduct the proceedings more effectively. In particular, the option of a preliminary “FRAND FIRST” hearing can serve to resolve unclear issues quickly.

233 These guidelines are based on a consensus reached by all members of the chamber and were finalized on August 13, 2026. The final version is saved as a PDF and is available for download. The guidelines will be sent ex officio in all pending FRAND proceedings.

Dr. Schön
Presiding Judge at the Regional Court

Dr. Schweyer
Judge at the Regional Court

Töszér
Judge at the Regional Court

Date
11/12/2026

PDF

FRAND Guidelines of the 7th civil chamber of the Regional Court of Munich I

Preamble:

I. Introduction

1. SEP/FRAND complaints as the “tip of the iceberg”

2. Super-FRAND, successful hold-out, and valuation

II. The legal framework for the FRAND review

1. The (well-known) assessment framework

2. Conventional interpretation

3. Distinction between external and internal willingness to take a license

4. The two aspects of FRAND

5. Need for a “SAFE HARBOUR” concept?

III. Examination of the plaintiff’s offer

1. Calculation approach

2. Criteria for determining a reasonable license fee

3. What are appropriate license agreements for a comparison?

4. Analysis of the original agreement

5. FRAND determination within a framework

6. Top-down approach

7. Specific situations

IV. Concluding remarks

Organisation

S'abonner à la newsletter

FRAND Guidelines of the 7th civil chamber of the Regional Court of Munich I

Preamble:

I. Introduction

1. SEP/FRAND complaints as the “tip of the iceberg”

2. Super-FRAND, successful hold-out, and valuation

II. The legal framework for the FRAND review

1. The (well-known) assessment framework

2. Conventional interpretation

3. Distinction between external and internal willingness to take a license

4. The two aspects of FRAND

5. Need for a “SAFE HARBOUR” concept?

III. Examination of the plaintiff’s offer

1. Calculation approach

2. Criteria for determining a reasonable license fee

3. What are appropriate license agreements for a comparison?

4. Analysis of the original agreement

5. FRAND determination within a framework

6. Top-down approach

7. Specific situations

IV. Concluding remarks

Related News

Organisation

S'abonner à la newsletter