استفاده از روش فازی برای ارزیابی ثبت اختراع تحت احتمال شکایت های قانونی

The value of a patent, including such unmeasurable things like the chance of litigation, is calculated using a methodology which combines real option theory and fuzzy numbers. The vagueness about the patent holder’s future profits, the validity and scope of the patent, the litigation costs, the court’s decision under imperfect enforcement of property rights are specified introducing fuzzy numbers. This method is embedded in a real option computation, where the value of a patent includes the option value of litigation. We study how the value of a patent is affected by the timing and incidence of litigation. The main results are compared with the empirical findings of previous results.

In the recent debate on patent reform the urgency to explicitly recognize the policy dimension of legal uncertainty is a central issue. Virtually, all property rights contain some elements of uncertainty and both litigation and settlement decisions occur in an environment characterized by imprecise information (Bebchuk, 1984, Farrell and Shapiro, 2008 and Shavell, 1989). Fuzzy boundaries of the patent property right seem to be a main cause of the recent explosion of patent litigation (Bessen and Meurer, 2005, Bessen and Meurer, 2007 and Bessen and Meurer, 2008). Recently, the European Commission has started studying the feasibility of alternative schemes against patent litigation risks, including insurance, in view of the dramatic explosion of patent litigation (CJA Ltd., 2006). As far as patent policy is concerned, it is emphasized that the effects of uncertainty should be incorporated in regulation and enforcement rules. This issue requires an accurate quantitative determination of the patent value. In this paper we propose a new comprehensive model, focusing on the different sources of uncertainty. Our main result is a valuation formula for patents which can be used for practical applications. A patent is usually defined as a right to make exclusive use of an innovation at a predetermined cost for a predetermined period of time, i.e. the life of the patent. As such it is viewed as a real option (see Dixit & Pindyck, 1994). The patent holder may commercialize some products or licence her technology or use it for further developments. The interpretation of patents as real options presupposes an enforceable property right. Yet, an increased number of patents have registered a high frequency of disputes and litigation involving patent holders and alleged infringers, so that the risk that a patent will be declared invalid is substantial. There is a wide variation across patents in their exposure to risk: as Lanjouw and Schankerman, 2001 and Allison and Lemley, 1998 have shown through detailed empirical evidence, for high-value patents and specific types of patentees the litigation risk can be quite high, in some cases almost offsetting what would otherwise be the R&D incentive provided by patent ownership. Bessen and Meurer (2008) in a most comprehensive empirical research have found that technology that rely heavily on software are vexed by huge patent litigation costs, so that “the patent system has turned from a source of net subsidy to R&D to a net tax” (2008, p. 1). Moreover, “roughly half of all litigated patents are found to be invalid, including some of great commercial significance” (Lemley & Shapiro, 2005, p. 76). Thus, because of uncertainty in the enforcement of property rights, it has been stated that “a patent does not confer upon its owner the right to exclude but rather a right to try to exclude by asserting the patent in court” ( Lemley & Shapiro, 2005, p. 75). Accordingly, the clarification of the norms about intellectual property right has been indicated as the main challenge for lawyers and politicians in the next decades ( Landes & Posner, 2003). Because of imperfect enforcement of property rights, most patents represent highly uncertain or probabilistic property rights. Lemley and Shapiro (2005) use the term probabilistic patents. Modeling patents as probabilistic rights requires to rethink how to reform the patent granting process and the patent litigation procedures. In this paper we translate the vague notion of probabilistic patents into a mathematical model, where the valuation of patents can be performed by a combination of real options and a fuzzy methodology. In order to capture the notion of vagueness about the validity and scope of patents under a regime of imperfect enforcement of property rights, we introduce a promising concept of uncertainty, alternative to probability theory, through the theory of fuzzy sets. In this way, we are able to capture the vague and imprecise ideas the patent holder possesses about her future profits, the validity of the patent, the litigation costs and the court’s decision. Moreover, we embed such methodology within a real option approach, where the value of a patent includes the option value of litigation. There are various papers applying the theory of real options to the valuation of patents although very few of them introduce the patent enforcement process explicitly. Pakes (1986) first estimated the distribution of the returns earned from holding patents as options which are renewed at alternative ages and require renewal fees. Bloom and Van Reenen (2002) builds on Pakes (1986) and derives empirical predictions on the relationship between patents and market uncertainty. Schwartz (2004) implements a simulation model to value patents as complex options, taking into account uncertainty in the cost-to-completion of the project and the possibility of abandoning the project. Takalo and Kanniainen, 2000 and Lambrecht, 2000 investigate the patenting decision under technological and market uncertainty with competing firms. None of the above-mentioned papers introduce the risk of litigation. To the best of our knowledge, the only analyses of the option value of litigation are Marco, 2005 and Baecker, 2007. However, Marco (2005) is mainly focused on the timing and incidence of patent litigation and is concerned with the empirical estimates of patent litigation. Baecker (2007) develops at length both the theory and the numerical implementation of some jump-diffusion models, where the risk of litigation is exogenously given and negatively affects the value of the patent in the form of discontinuities or jumps in the value process. He also addresses some issues of endogenous patent risk through a model where the patent holder possesses full knowledge about the probability distribution of the litigation risk. Our paper is the first that combines a real option to litigate with a fuzzy valuation. The need for a fuzzy valuation comes from the common observation that patent claims are sometimes vaguely defined, the outcomes of a trial are difficult to forecast, legal costs are not easily predictable, it may be years before litigation is concluded, there may be divergence in parties’ expectations about the court decision and future cash flows from commercialization are imprecise. Although the existing literature has identified the main determinants of litigation, it has not investigated how the value of a patent is affected by the timing and incidence of litigation under an appropriate framework of uncertainty. Section 2 presents the model of a patent under imperfect enforcement of property rights, where the relevant parameters are fuzzy. The model is solved analytically for infinitely lived patents and in Section 3 the main results are compared with the empirical findings of previous studies. Finally, in Section 4 some concluding remarks are presented.

The notion of probabilistic patents requires a deep understanding of patent risk and an appropriate way of modeling the different sources of uncertainty, regarding both the commercial significance of the invention being patented and the validity and scope of the legal right being granted. Recent debates on the appropriate methods for calculating infringement damages have engaged the patent community: actually, patent litigation is one piece of the patent reform puzzle. In this paper we provide a valuation formula for patents within a framework that combines a real option to litigate with a fuzzy methodology and give a specific characterization of legal uncertainty. Our model is consistent with the view that the strength of patent protection and the presence of more codified and defined boundaries are crucial determinants of the patent value, the probability of licensing and patent validity. A comprehensive model and the resulting valuation formula can have significant implications for patent policies. To the extent that the effects of uncertainty and enforcement are neglected in patent policy, incentives for innovation may become inconsistent, issues of patent invalidity arbitrary and the economic cost of patent litigation will be exacerbated. The contribution of uncertainty to total patent value and its effect on the litigation decisions need to be taken into account to set a good patent system and make it an effective tool for providing positive incentives. In this paper we have pointed out a few relevant differences with respect to traditional real option modeling of patents. These are due to imprecise information about the driving parameters, that is, the probability of successful litigation, the gains from successful infringement suits or patent challenges, the litigation costs. A realistic estimate of these parameter values, which our model makes possible, is necessary to ensure that patent policy will be effective.