Last week a federal court crossed into uncharted territory, for the first time calculating the proper reasonable and non-discriminatory (RAND) royalty on a standard-essential patent (SEP). [PDF here, via Ars] The dispute arose in litigation between Motorola and Microsoft over implementations of Motorola-owned patents incorporated into WiFi and video standards, which Motorola had agreed to license on RAND terms during the standard setting process. Never before had a federal court been compelled to establish by order the reasonable rate for an SEP; until now, parties had always been able to arrive at a privately negotiated rate.
The 207-page order is ably dissected by AU Professor Jorge Contreras on the Patently-O blog and by Matt Rizzolo at the Essential Patents blog (here and here) as well. Both note that the court takes “broad implementation” of a standard as a primary objective of standard setting. Stated otherwise, because a “reasonable” royalty cannot be one that dissuades adoption of a standard, the RAND royalty setting is no longer merely an exercise in private wealth maximization by the patent holder. This leads to the issue of “royalty stacking” – the phenomenon where hundreds of competing license claims on a standard could ultimately swamp and exceed the value of implementing the standard in the first place. Because of this risk, the court notes, “RAND negotiation would not be conducted in a vacuum. The parties would instead consider other SEP holders and the royalty rate that each of these patent holders might seek…”
It is reasonable to assume that royalty stacking will influence licensing negotiations. One problem with this reasoning is that it implicitly assumes that all relevant patents are on the table. However, undeclared yet essential patents may read on standards as well. This is not a new observation; standards ambush is a documented phenomenon. For example, in 2006 AT&T ambushed the MPEG-4 standard with previously undisclosed patents, and similar attempts occurred during the 1990s with respect to standards for both the GIF and JPEG file formats. In fact, several non-practicing entities (NPEs) have ambushed the WiFi standard at issue in this case. These license obligations will stack too.
Thus, when both RAND-encumbered and non-encumbered patents read on patents, the value of the former must be pro-rated in light of other obligations — and to give any meaning to the court’s objective of promoting standard adoption, “other obligations” must extend beyond RAND commitments, to include royalties due to patent-holders who ambush the standard.
Of course, there is no statutory basis for limiting the royalty of patent holders without RAND obligations who claim standard essential patents. To use a bankruptcy analogy, SEP patents are rather like the claims of unsecured creditors, whereas non-encumbered patents are fully secured, and are not subject to a cram-down, pro-rata distribution. The patentee who ambushes the standard, on the other hand, has no RAND commitment and is not bound to accept a rate that would “promote adoption.” He may ask any royalty, and this royalty is likely to be quite high due to the sunk costs and lock-in costs of having adopted the standard. Should SEP patent-holders negotiate over a reasonable RAND royalty after such an ambush has occurred, last week’s decision suggests they may reasonably ask only for whatever scraps remain.
By contrast, what happens if a patent holder ambushes a standard after the RAND rates have been set? It is still entirely likely that total royalty obligations will stack to an extent that nevertheless deters adoption.
These outcomes all seem unfortunate. If standard-setting participation means one’s patents take a backseat to those who sat the process out, it is likely to deter RAND-based standard-setting participation in the first place. Conversely, the incentive to ambush is – at best – unchanged.
This is not to say the RAND decision was necessarily wrong, although the ‘administrability’ of this court’s rule remains to be seen. But ultimately, one of two evils must persist: either obligations will continue to stack in a way that can deter standards participation, or RAND constraints will create uncertainty about a patent’s value to a degree that will likely deter others from making RAND commitment in the future.
This is further demonstration of the fact that patents don’t enable. What standards implementers need is a guarantee of “quiet title” to their investment in the standard, and litigation involving one patentee cannot provide that certainty. These problems are illustrative of the continuing difficulties of resolving the current patent morass: what appears to be a facially reasonable decision may result in second-order problems for which no solution is readily apparent.