توسعه محصول چند مرحله ای با اکتشاف، افزایش ارزش، گزینه های نوآوری و انحصاری

We provide a real options framework for the analysis of product development that incorporates research and exploration actions, product attribute value-enhancing actions with uncertain outcome, as well as preemption and innovation options. We derive two-stage analytic formulas and propose a general multi-period solution using a numerical lattice approach. Our analysis reveals that exploration actions are more important when the project is out or at-the-money (near zero NPV) and less important for high project values. In a multi-stage setting, exploration actions are important even for in-the-money projects, when follow-on actions exist that can enhance the expected value of the project. With path-dependency, early actions are more valuable since they enhance the impact or reduce the cost of subsequent actions. Preemptive controls affecting rare event (jump) frequency and innovations that introduce positive jumps are more valuable for firms with higher frequency of competitive threats involving low volatility.

We develop a real options model to study costly interacting managerial control actions. These actions involve pure research or exploration actions, investments that are expected to enhance value or reduce the cost of a project with an uncertain outcome, preemptive investments that reduce the damage of rare events (jumps) and innovation investments that introduce (or increase) the frequency of value improvements (positive jumps). In our model, the information revelation of exploration actions and the volatility of value-enhancing actions interact with exogenous demand-driven uncertainty (e.g., capturing changing consumer preferences). The latter is modeled using a Brownian motion or a jump-diffusion process. In the more general jump-diffusion model, the firm can also make preemptive investments that help control the frequency and size of negative jumps and innovations that increase the likelihood of positive jumps. Our model allows for optimal timing of staged product introductions (with earlier products providing information about future products) and for abandonment options resulting in partial recovery of invested capital. Pure research or exploration actions include investments in early product versions (pilot projects), experimentation using new processes and marketing research. These actions help resolve uncertainty about the true project value or cost, enabling management to capitalize on new information before large or irreversible investment is undertaken. For example, Samsung conducted marketing research concerning what consumers considered most important attributes of a flat-screen TV that resulted in a more focused development that achieved a higher market penetration (Moon, 2006). Childs et al., 2001 and Bernardo and Chowdhry, 2002 use a filtering approach to study information acquisition in a real options model with noisy assets. Pindyck (1993) examines sequential multi-stage investments involving technical uncertainty that decreases as the project approaches completion. Pindyck assumes continuous reduction of technical uncertainty while we allow for different levels of technical uncertainty resolution between stages. We also allow for interacting actions and derive analytic formulas for the two-stage problem. Childs and Triantis (1999) consider accelerated versus sequential strategies and learning spillovers between projects. They assume that actions affect the Brownian volatility while we allow for value enhancement maintaining a separate exogenous demand driven uncertainty; furthermore, we also consider path-dependency. Direct value-enhancing actions include R&D efforts to improve the attributes or quality of a product, enhancement of customer perceptions through advertising, or efforts to reduce cost through adoption of new technologies in production. For example, Google invests in new technology in promoting online display advertising to enhance its revenues (Hof, 2009). As in Huchzermeier and Loch (2001), these actions aim at enhancing project value, but have an uncertain outcome. We assume that investment decisions are taken at discrete points in time and their outcome is realized immediately. Impulse-type actions with uncertain outcome were introduced in the real options literature by Martzoukos (2000). Childs and Triantis, 1999 and Berk et al., 2004 consider projects that require completion of development stages before the commercialization of the product. In our setting, the firm may decide to develop the product immediately, delay development exploring further development opportunities or introduce early product versions. The expected impact, volatility and costs of investments and the cash flows of early product versions depend on the sequencing of decisions (path-dependency). For example, the firm may expect a higher impact of R&D, if prior marketing research has been implemented. New information following the results of an experimentation process may also reduce next-stage costs. Grenadier and Weiss (1997) provide a model for the adoption of technological innovations where firms adopting innovations early are better able to benefit from future innovations. In the presence of severe competition causing negative jumps in value, the firm may engage in preemptive or innovation investments. Preemptive controls exercised at an optimal time allow management to reduce the frequency and size of competitive threats (negative jumps). Brown and Petersen (2010) discuss how new entrants use high levels of R&D, posing a danger on established firms regarding their profitability and market share. Theoretical arguments in Lambrecht and Perraudin (2003) demonstrate that, in a competitive environment with preemptive investments, equity returns will exhibit jump discontinuities and skewness. Empirical evidence on the presence of jumps in equity prices has been documented in the literature (e.g., Ball and Torous, 1985, Bates, 1991 and Nimalendran, 1994). Our model extends the standard jump-diffusion models by incorporating heterogeneity in the frequency and size of jumps between periods, allowing these characteristics to be controlled by the firm through preemptive and innovation investments. Nimalendran (1994) provides empirical evidence of jump heterogeneity that is affected by corporate events and managerial actions. Camara (2009) models the stock price following a jump-diffusion process where upward and downward jumps can have different means and standard deviations, showing that it can better capture volatility smiles and skews in option markets. Pennings and Lint (1997), drawing on experiences from their involvement in Philips’ R&D program, also propose discrete and instantaneous updating of information, based on the arrival of new information about competitors and the impact of R&D. Our framework extends theirs in several dimensions, including the incorporation of multiple classes of jumps, jump heterogeneity between decision points and controls that affect the frequency and impact of these jumps.1 Our framework also allows for innovation investments, which increase value by increasing the frequency and size of positive jumps. They involve, for example, investments in human capital or technological infrastructure that intend to create future growth potential. We derive analytic solutions for a two-stage problem that involves multiple value-enhancing actions. Our analytic solutions nest several known results as special cases, including those of Geske, 1979 and Longstaff, 1990 (see also Chung and Johnson (2011) for the multi-stage extendible option). We further incorporate path-dependency and optimal timing of managerial exploration and value-enhancing actions. We extend our model to a multi-stage framework using a numerical lattice approach and provide applications with multiple actions, path-dependency and controls on jump diffusion parameters. Consistent with results in Bernardo and Chowdhry, 2002 and Huchzermeier and Loch, 2001, we show that managerial exploration actions may be more valuable for projects that are marginal or break-even. However, we show that in the case of interacting actions, exploration actions may be important even in deep in-the-money projects, when follow-on value-enhancing actions are involved. Furthermore, we show that multiple and interchanging decision regions (as a function of project value) between delay, early development, exploration and value-enhancing actions are possible. Path-dependency has a substantial impact on these regions. In the jump-diffusion case, we use prior empirical study estimates of jump-diffusion parameters and analyze two classes of firms: firms with low frequency of jumps and high impact/volatility and firms with high frequency and low impact/volatility. We find that preemption actions are more important for at-the-money to in-the-money range options and for the high frequency group. Innovation investments that bring about positive size and preemption options that open-up future opportunities are the most valuable and may be exercised in the out-of-the-money range as well. The rest of the paper is organized as follows: Section 2 describes the model and assumptions. Section 3 derives analytic formulas for the two-stage problems and discusses the decision regions and the impact of path-dependency. Section 4 provides numerical results for the general multi-stage applications involving path-dependencies in new product development, involving preemption and innovation options. The last section provides the conclusion. An Appendix A includes the numerical lattice-based model for the diffusion model. Appendix B includes the numerical lattice-based solution for the jump-diffusion. Appendix C presents analytic solutions for the jump-diffusion and Appendix D includes an investigation of the numerical accuracy of the lattice-based model.

We have considered investment options with four types of managerial controls embedded in a real options model: explorative research (e.g., experimentation or learning), value-enhancing (attribute or quality improvement) actions, preemption and innovation options. Our framework is suitable for analyzing product development and determining the optimal sequencing of decisions. Our framework can account for path-dependencies in the mean impact, volatility and the cost of actions. We derive analytic solutions for two-stage sequential options and develop a numerical lattice-based approach analyzing the more general multi-stage problem. Our model allows for early development, abandonment and early versions of the product that provide own cash value and help resolve technical uncertainty. Exogenous market uncertainty is modeled using diffusion (or jump-diffusion) processes. Our analytic solutions for the two-stage problem account for exploration investments, value-enhancing managerial controls, as well as early development, abandonment and early versions of the product. Our analytic solution nests as special cases the compound option of Geske (1979) and the extendible option of Longstaff (1990). We use this model and study the probability of development and the incremental benefits of investing in managerial control actions as a function of the volatility of actions. We confirm that the probability of development decreases in the volatility of the actions for in-the-money projects and increases for out-of-the-money projects. The incremental benefits of an action increase in volatility, but this effect decreases in the presence of subsequent actions. We investigate an interchanging range of optimal decision regions. In general, exploration actions are worthwhile when the NPV of the project is close to zero. We show that exploration actions may be important, even for in-the-money projects, when subsequent actions may enhance further the expected value of the project. The optimal sequencing of actions is important in the presence of path-dependency. Preemptive investments to minimize competitive damage are particularly important when the frequency of negative jumps is high and the volatility of their impact is low. Preemptive investments are typically optimal in the at-the-money to in-the-money range, unless they eliminate severe negative impact, create innovations with enhanced beneficial volatility or create future growth options.We have considered investment options with four types of managerial controls embedded in a real options model: explorative research (e.g., experimentation or learning), value-enhancing (attribute or quality improvement) actions, preemption and innovation options. Our framework is suitable for analyzing product development and determining the optimal sequencing of decisions. Our framework can account for path-dependencies in the mean impact, volatility and the cost of actions. We derive analytic solutions for two-stage sequential options and develop a numerical lattice-based approach analyzing the more general multi-stage problem. Our model allows for early development, abandonment and early versions of the product that provide own cash value and help resolve technical uncertainty. Exogenous market uncertainty is modeled using diffusion (or jump-diffusion) processes. Our analytic solutions for the two-stage problem account for exploration investments, value-enhancing managerial controls, as well as early development, abandonment and early versions of the product. Our analytic solution nests as special cases the compound option of Geske (1979) and the extendible option of Longstaff (1990). We use this model and study the probability of development and the incremental benefits of investing in managerial control actions as a function of the volatility of actions. We confirm that the probability of development decreases in the volatility of the actions for in-the-money projects and increases for out-of-the-money projects. The incremental benefits of an action increase in volatility, but this effect decreases in the presence of subsequent actions. We investigate an interchanging range of optimal decision regions. In general, exploration actions are worthwhile when the NPV of the project is close to zero. We show that exploration actions may be important, even for in-the-money projects, when subsequent actions may enhance further the expected value of the project. The optimal sequencing of actions is important in the presence of path-dependency. Preemptive investments to minimize competitive damage are particularly important when the frequency of negative jumps is high and the volatility of their impact is low. Preemptive investments are typically optimal in the at-the-money to in-the-money range, unless they eliminate severe negative impact, create innovations with enhanced beneficial volatility or create future growth options.