FRAND determination in TCL v. Ericsson and Unwired Planet v. Huawei: Same Same But Different?

May 13, 2018

Peter Georg Picht focuses on the treatment of the two core approaches in FRAND calculation, namely the so-called “top-down” and “comparable licenses”. As to the relative importance of these two calculation approaches, Peter Georg Picht mentions that comparable licenses seem more relevant regarding the facts at issue. If good comparable licenses are at hand they should loom large and produce the default results. Comparables and, in particular, the way in which they were concluded come into play because they can document patterns of customary market behavior. Regarding the top-down valuation methodology, he argues that it creates more problems than solutions to treat publicly announced royalty rates as a form of binding “pledge”, and that extensive “mid-point guessing” by judges/ parties can be problematic. Peter Georg Picht notes that it is quite questionable whether top-down calculations should loom larger than the collective “market intelligence” embodied in comparable licenses, especially when the licenses were negotiated by experienced players and absent impending litigation or similar pressure factors.


According to the author, comparable licenses can arguably inform FRAND determination in three – closely interconnected – ways


  1. Their conditions provide data points showing how the market values a particular patent portfolio. Since the “fair” and “reasonable” conditions for licensing a portfolio are not preordained by the gods but depend on the economic potential a licensee can reap by using the licensed technologies and since the licensee’s willingness to offer the patent holder cash and non-cash benefits expressed in the licensing conditions is directly related to this potential, comparable licenses are valid indicators for the “FR” prong of FRAND

  2. When combined with information on the market position and other characteristics of the licensees, comparable licenses become the single most important reference for assessing the “ND” prong of FRAND

  3. Comparable licenses can also have an informative value with regard to the “procedural” aspect of FRAND. A key goal of the FRAND mechanism, the peaceable and effective conclusion of license contracts providing a reliable framework for standard-based market activity, cannot be achieved without parties acting in a proactive and cooperative manner.


    Concerning the top-down valuation methodology, Georg Picht argues that uncertainties in the determination of an appropriate aggregate royalty rate, of the total and the patentee owned SEP portfolio, and of the portfolio patents’ value can add up and make top-down results quite fragile. In particular:


  1. The first question goes to the legal basis for a binding effect of the royalty announcement. Does the announcement constitute a third-party beneficial contract as it is the case (in the view of Judge Selna) with ETSI’s FRAND commitment? Or is the binding effect rather estoppel-based?
  2. Both a contractual and an estoppel interpretation of the “pledge” would probably suggest asking what the addressees of the announcement, i.e. (potential) implementers and licensees, could have reasonably understood it to mean. For instance, are ex-ante royalty announcements intended to be binding “forever” or for a limited period of time? Of the construing effort necessary to answer such intricate questions one finds rather little in the TCL decision. Instead, it is more the judge’s than the addressees’ viewpoint that seems to matter in interpreting the announcement.
  3. Would it be legitimate to give past announcements an (unconditionally) binding effect although TCL v Ericsson’s “pledge approach” seems not to have formed part of (US) FRAND license law so far and patentees could therefore not foresee the ramifications of their doing? If this seemed too far-reaching, should pre-TCL v Ericsson announcements have no binding effect at all or should the effect be attenuated by reading into the “pledge” conditions or subsequent modifications reasonably to be expected by its addressees
  4. An assumption that each patentee is bound by its individual announcement and that the patentee has some leeway in framing this announcement could result in differing aggregate royalty rates for different patentees. Evidently, such an outcome contradicts the concept of a coherent, top-down determined royalty framework for all SEP holders in the respective market, endangers the goal to prevent royalty-stacking (since patentees may race for higher aggregate royalty “pledges”), and tends to disincentivize reasonable, FRAND-oriented aggregate royalty announcements.
  5. Adjusting, instead, the individual announcements into a blended, uniform aggregate royalty rate does largely away with the concept of a contract- or estoppel-based “pledge” because it removes (most of) a patentee’s freedom to decide upon the content of its offer/promise. Furthermore, it is misleading to consider individual announcements as “pledges” which add some substantial form of legal commitment to the obligations springing already from the ETSI FRAND declaration.
  6. The shortcomings in party-submitted top-down models get even more severe as they proceed from SEP-counting to valuing the patented technology. A counting-only approach that neglects the argument that fair royalties should depend on a patent’s market value weakens the reliability of top-down royalty calculations.
Author(s) / Organization: 
Publication Type: