Friday 7 May 2010

Being on FRANDly terms with friendly trolls.

The famous patent battle opposing NOKIA to German IP asset management firm IPCOM illustrates well the lack of practical understanding as to what fair, reasonable and non-discriminatory (or FRAND) licensing terms practically means in Europe.

The conflict started at the end of 2006 when IPCOM, a company founded by Bernhard Frohwitter and Christoph Schoeller, acquired the patent portfolio developed by BOSCH in the late 80s and 90s for its mobile and in-car telephone systems. This portfolio comprises 160 patent families and 35 of them are said to be essential to the GSM and UMTS communication standards. The acquisition of a license from IPCOM is thus a necessity for all firms manufacturing and selling products conforming to the new technological norms. Meanwhile IPCOM must accept to license these essential patents under FRAND conditions.

The negotiations between NOKIA and IPCOM turned sour very quickly, as IPCOM asked for 5% of NOKIA’s mobile phone sales in all countries covered by its portfolio and for each year of use. This represents a cost of 600 million euros per year for NOKIA, that is to say an astronomic total of 12 billion euros over 20 years. The Finnish company stopped negotiating with IPCOM and accused the German firm of not honoring the FRAND commitment that previous patent holder BOSCH made to the ETSI (European Telecommunications Standards Institute) as well as to all companies that participated to the establishement of the 2G/3G telecommunication standards. This accusation has been vehemently rejected by IPCOM. According to Frohwitter IPCOM has always been dealing with potential licensees on FRAND terms, the amicable settlement found with Siemens AG, Motorola Inc., Philips Electronics and Alcatel-Lucent amongst others being an irrefutable proof of IPCOM's willingness to cooperate on a fair and reasonable basis. There is a wide divergence between the two firms on the meaning of FRAND and proceedings of patent infringement/invalidation are now in progress. NOKIA took an advantage as the two most contentious patents have been recently invalidated in UK (see here). Regrettably for those (like me) who were hoping for some guidance, the meaning of a FRAND commitment has not been debated in court.

The best solution to find the meaning of FRAND is to delve deep into IT and legal resources. However the literature addressing this topic is already consequent and the opinion of specialists on this “hot” topic is diverging. It is difficult for a FRAND neophyte like to discern the truth from lies. Fortunately for beginners, Roger G. Brooks and Damien Gerardin recently published an excellent piece (that you can download here) breaking all myths commonly held on this tricky subject. For some, the fact that both authors are partners have been representing chip manufacturing company Qualcomm Inc. in many FRAND-related issues in the past might call their impartiality into question. However the arguments developed in this paper are unbiased, simple and quite convincing (to a certain extent).

In a nutshell Gerardin and Brooks refute many commonly accepted statements, notably that in order to satisfy a "fair and reasonable" commitment, a patent holder "must set his royalty rate based on a mathematical proportion of all patents essential to the practice of a standard or that a patent holder is not entitled to seek injunctive relief against a standard implementer should they fail to agree on licence terms. Both statements are not supported by economic or legal evidence and are in contradiction with the actual behaviour patent owners in a standard-setting context. Numerical proportionality in the calculation of a royalty rate is not a sign of fairness, since it tends to ignore the real value of a specific patent in favour the average value of all patents used to comply with a technological standard. The statement that a FRAND commitment should be construed as a waiver of injuctive relief from the patentee is according to both authors equally as misleading, because this assertion is contradicted by hard facts found in the history of the IPR policy of many standard-setting organizations (or SSOs) such as the ETSI, as both authors demonstrate.
All things considered, the plain truth about FRAND commitments is that they are the result of a voluntary contract between essential patent holders and SSOs to negociate in good faith about the terms and conditions of a licensing agreement. As Brooks and Gerardin observe, “fair and reasonable” are on their face flexible terms the specific content of which is substantially left to the negotiation between the parties. All economic interpretations of FRAND commitment proposed till now are thus restrictive limitations, whose validity cannot be verified against the IPR policy of SSOs.

The simple definition provided by Brooks and Gerardin is very compelling and hardly refutable. However in the very specific context of the IPCOM case, which does look like an ex-post hold up to me, it does not provide the helpful guidance one could have hoped. The balance between the two actors in the licensing negotiations is skewed from the beginning because IPCOM finds itself outside the telecom market. NOKIA cannot use its own patent portfolio as a bargaining chip to negotiate the terms and conditions of a license (and above all the price of such a license), like it certainly does with other market players. Indeed NOKIA's competitors de facto need those patents in order to operate on the market, whereas IPCOM simply does not. Given this initial distortion giving much leeway to IPCOM, it is not surprising to see NOKIA withdrawing from the negotiations. This case shows the importance of the context in which the negotiations are taking place and to my mind Brooks and Gerardin's bare definition is hardly useful here, because it fails at taking this context into account.

In my opinion the concept of “FRAND terms of agreement” has been included to the SSOs' IPR policies, so as to help negotiating parties to reach an agreement satisfying them both. I agree with Brooks and Gerardin that the meaning of such a concept cannot be too specific, but if you dry up its meaning, this concept will not fulfill its primary purpose, and this would be a failure as well.

No comments: