【作者】 蔡强；

【导师】 曾勇；

【作者基本信息】 电子科技大学 ， 管理科学与工程， 2009， 博士

【摘要】 在专利研发、专利购买和专利商业化等专利投资活动中，来自于内外部环境的诸多不确定性和竞争将影响企业的专利投资决策。前者如专利技术和专利价值的不确定性、有关专利技术未来市场前景的不完全信息等，而隐含于这些不确定性中的实物期权等待价值会使企业产生等待和推迟投资的动机和愿望。同时，专利的排他性、独占性特征使企业间为争夺专利而展开的竞争比通常意义下的企业竞争更趋激烈，为了不落后于对手，企业只得放弃部分甚至全部期权价值，将投资时机提前。这种专利投资中的期权博弈在各种不确定性的作用下将产生特征各异的过程和结果。本博士论文以期权博弈为工具，围绕专利投资决策中的若干前沿问题开展研究，内容包括：1、针对不完全信息专利竞赛，在只考虑先发优势时，外部随机发生的事件中蕴含着专利未来前景的不完全信息，这些信息和研发能否成功的技术不确定性会影响企业的专利研发投资决策。建立实物期权投资决策模型，分别得到单个企业和双寡头企业投资专利研发所需的临界信念，并进一步分析了两对称企业专利竞赛可能出现的均衡类型及产生条件。当先发优势和后发优势同时存在时，它们的相对大小将决定两对称企业在专利研发投资时机选择博弈中所产生的均衡类型并影响企业的研发投资决策。引入信息披露及其滞后效应使得追随者和领导者分别具有后发优势和先发优势。通过对各最优投资临界信念值的计算和分析，结论表明，一定参数条件下，占先博弈和消耗战均有可能发生。信息披露滞后时间越长、技术不确定性程度越低、随机信号到达速度越快、信号质量越高，领导者先发优势就越明显，越易产生占先博弈；反之，追随者后发优势就越明显，越有可能发生消耗战。2、针对有记忆专利竞赛，通过引入非齐次泊松过程并考虑企业R&D活动中的“学习效应”或“记忆性”，研究了两企业的专利竞赛问题。分别就单个企业的投资时机选择和两家企业在非合作博弈条件下所形成的有记忆专利竞赛进行了分析，结论表明：R&D活动的“记忆性”使得单个企业的投资临界值降低，且无论是连续区域企业的等待期权价值还是停止区域执行期权后的项目价值均比无记忆时有所增加；而在有记忆专利竞赛中，由于领导者价值始终大于追随者价值，占先的动机加强，投资临界值降低，最终使博弈在较低专利价值处产生双方均不愿接受的同时投资均衡。3、针对专利的购买性投资，考虑不完全信息和竞争，分析两对称企业竞买同一个专利可能出现的均衡类型及产生条件。结果表明，竞争的力量和专利的独占性、排他性特征不仅使得投资所需临界信念降低，同时，出现错误投资决策的概率也随之增大。4、针对专利商业化投资，考虑到有限的专利保护宽度使专利的商业价值同样会受到竞争的影响。针对各自拥有面向同一新兴市场但具有不同市场表现的专利技术的双寡头企业，构建投资时机选择期权博弈模型并分析企业的专利商业化投资决策特征。结果表明，专利技术的非对称程度将决定双寡头企业在专利商业化投资时机选择博弈中的均衡类型。5、针对专利投资中的社会福利问题，考虑到专利技术的研发投资策略不仅影响企业价值还会影响消费者剩余，两种影响通常会呈现出两种不同的效应，从而导致专利研发投资决策对社会福利的不同影响。针对旨在通过研发专利技术从而垄断新兴市场的两非对称企业，构建投资时机选择期权博弈模型和消费者效用模型得到非合作博弈(即专利竞赛)和合作博弈两种模式下的总投资净现值和消费者剩余，分析社会福利与影响其变化的诸多因素间的动态变化特征并提出相应的创新政策建议。结果表明，若以社会福利最大化为目标，则专利技术的成长性越好、无风险利率越小、专利产品质量越低时，政府政策越应有利于合作研发模式；反之，政策越应鼓励竞争。更多还原

【Abstract】 In the patent-related investment activities as patent R&D, patent purchase and patentcommercialization, competition and many uncertainties from internal and externalenvironment will affect the decisions of the enterprise’ patent investment. Theuncertaities in patent technology and patent value, incomplete information about themarket prospect of the patent can affect the real options value of the patent and can inrum cause the enterprise the motivation and intention of waiting and delaying ininvestment. At the same time, the exclusive and proprietary characteristics of patentmake enterprises scramble for patent and compete against each other more violently. Inorder to beat the rivals, an enterprise must give up some or all options value, andadvance the time of investment. With various uncertainties, this option game in patentinvestment will produce processes and results with different features. Based on optiongame theoretical models and methods, this dissertation studies patent investmentdecisions. The main contents are as follows.1. As to incomplete information patent race, when considering the first moveradvantage only, the uncertainties that will influence the enterprises’ patent R&Dinvestment decisions include: information uncertainties as incomplete information aboutthe future prospect of the patent combined with the external randomly happening events,and technical uncertainties as whether the R&D can succeed. By considering incompleteinformation, technical uncertainty and competition, this dissertation develops aninvestment decision-making model based on real options analysis, and obtains thecritical beliefs needed for patent R&D investments of single enterprise and duopolyenterprises respectively, and further analyzes the equilibrium types and relatedconditions that may appear in the patent race between two symmetric enterprises. Whenfirst and second mover advantages exist in the patent race simultaneously and theselevels would determine the equilibrium types arises in the game of investment timingfor patent R&D between two symmetric enterprises as well as the investment decisionin the process. Under incomplete information and technical uncertainty, it is essential tointroduce information spillovers and lag effect for the followers and leaders to occupy the second and first mover advantages respectively. Through calculation and analysis ofvarious optimal investment critical beliefs, it is shown that both a preemption game anda war of attrition can arise for specific parameters of the model. The longer the lag timefor information spillovers will be, the smaller the technical uncertainty will be; thefaster the random signal received will be, the higher the quality of the signal will be;and the leader’s first mover advantage will be more apparent, the possibility of thepreemption game will be higher. Contrarily, the follower’s second mover advantage willbe more apparent, and the possibility of the war of attrition will be higher.2. As to memorable patent race, by introducing the non-homogeneous Poissonprocess and considering the "leaning effect" or "memorable effect" in firms’ R&Dactivities, this dissertation studies patent race formed by two firms and a risky patentR&D project. The decision of a single firm on investment timing and the memorablepatent race in non-cooperative game are analyzed respectively. It is shown that the"memorable effect" results in a decrease in the firm’s investment critical value, andwhether the value of the option to wait in the continuous area or the value of the projectafter option exercising in the stopped area is more than that in the memoryless case.Furthermore, the fact that the value of the leader always exceeds that of the followerenhances the firms’ motives to invest in preemption, and results in decreases in theircritical values for investment, and finally leads to a simultaneous investmentequilibrium with relatively low patent value, which would not have liked to be acceptedby the two parties in the race.3. As to purchasing investment of patent, considering incomplete information andcompetition, and further analyzes the equilibrium types and generation conditions mayappear from the competition purchasing aimed at a same patent between two symmetricenterprises. The result shows that the power of competition and inherent characteristicof patent, the "winner takes all",not only leads the critical belief needed for purchasingpatent decreases,but also the follower’s second mover advantages usually caused in"information spillover effect" generated by the leader’s investment won’t be existenced;Simultaneity, the probability of a wrong decision increases with it.4. As to patent commercialization investment, limited patent breadth makes patentcommercial value influenced by competition as well. Aimed at patent of duopolyenterprises with same emerging market and different market performance, it is advised to establish an investment timing decision-making option game model and analyzedecision feature of enterprise patent commercialization investment. The results showthat the dissymmetry degree of patent will decide equilibrium type of game that duopolyenterprises in patent commercialization investment timing decision-making.5. As to social welfare in patent investment, considering the different impacts ofpatent investment strategies on enterprise values and consumer surplus result indeferring impacts of patent investment decisions on social welfare. By combining theinvestment option game model and the consumer surplus model, this dissertationobtains the total investment NPV and consumer surplus under both modes ofnon-cooperative game (patent race) and cooperative game for two asymmetricalenterprises. The dynamic characteristics of social welfare and the influential factors areanalyzed and relevant innovative policies are put forward. The result shows if the targetis the optimization of social welfare, the less the heterogeneity of R&D capabilities is,the better the patent growth prospect is, the smaller the risk free interest rate is, and thelower the quality of the patent product is, policies that are more conducive tocooperative R&D modes should social planners come up with. Contrarily, policiesencouraging competition should be made.更多还原