Patent Settlements as a Barrier to Entry

Abstract

Empirical studies have found that most patent litigations are resolved through settlements rather than trials. Moreover, license fees negotiated in settlement agreements have become very large. We formulate a model of sequential entry with an incumbent patent holder, multiple potential entrants and asymmetric information about the validity of the contested patent (patent strength) between firms that are already in the market and future potential entrants. Within this framework we show that patent settlements between the incumbent and the first entrant can be mutually beneficial even when the cost of trial is zero and the settlement agreement takes the form of a simple fixed license fee. For intermediate patent strengths, settlements are a tool for further entry deterrence. The two parties agree on a high settlement amount which sends a credible signal to 'outsiders' that the incumbent's patent is not weak and therefore further entry will not be profitable. This provides a novel explanation for the role of settlements and to the recent observation of high license fees negotiated in settlement agreements. Our analysis demonstrates than even non-reverse settlements that entail only a fixed fee can be anticompetitive because they are used to block further entry.

Patent Licensing with Bertrand Competitors

“…For example, an interesting ground of research may regard ad valorem licensing as an entry deterrence tool. While the entry deterrence impact of per‐unit licensing and fixed fee licensing has been considered in literature (Duchene and Serfes, ; Duchene et al ., ), the role of ad valorem licensing is yet unexplored. Another question concerns the role of ad valorem licensing in a purely collusive set‐up.…”

Licensing standard‐essential patents with costly enforcement

“…study the impact of "pay-for-delay" deals on the litigation strategy of a patent holder. They show that these deals generate a payoff externality between entrants and allow the patent holder to adopt a divide-and-conquer strategy.10 See alsoDuchêne and Serfes (2012), who extends Choi (1998)'s framework by introducing asymmetric information and the possibility of settlement.11 SeeSpier (2007) for a comprehensive review of this literature.12 This means that the downstream firms can be active if, and only if, they use the patented technology. There is no backstop technology that they could use.13 The model can be easily be extended to ≥ n 2 downstream firms by assuming that the first n − 1 downstream firms make the same licensing decision (i.e., to accept or reject the license offer).14 In line with the literature on probabilistic patents, we assume that the patent is not ironclad, that is, θ < 1.…”

“…10 10 See also Duchêne and Serfes (2012), who extends Choi (1998)'s framework by introducing asymmetric information and the possibility of settlement.…”

Patent Licensing and Entry Deterrence: The Role of Low Royalties

“…However, there is no entry threat in their model. Duchêne and Serfes () show that high fixed fees in patent settlement agreements can deter entry by sending a credible signal to outsiders that the patent is not weak.…”