【作者】 胡芝春；

【导师】 甘嵘静；

【作者基本信息】 西南财经大学 ， 技术经济及管理， 2007， 硕士

【摘要】 随着经济的发展,企业所处的内外部环境中不确定性因素越来越多,而且企业所能得到的关于市场和竞争对手的信息多为不完全的。这些变化使得传统的DCF评估方法的缺陷愈加明显。于是企业在进行投资决策时,应该考虑的因素也就越多,对投资项目的分析也就需要适应新投资环境的科学的决策方法。本文正是基于这种需要,在前人研究的基础上,运用不确定条件下投资理论的基础方法,引入了一个在原实物期权模型中加入不完全信息因素的模型。运用这种方法,可以把不完全信息量化到模型中去,对企业的战略投资行为具有更有现实意义的指导作用。实物期权概念的提出虽然仅仅三十年的时间,但它在提出后马上就成为了理论界的研究热点,足以显示出它对现实投资行为的解释能力的强大。国内外的学者不断地对实物期权理论模型进行完善,并运用到不同的领域,解释和指导不同的经济行为。在这个过程中,逐渐形成了对不确定条件下投资研究的一般方法,如随机过程、维纳过程、依藤过程、依藤引理、或有期权理论等。通过这些一般方法并结合不同的经济环境,就可以得出针对不同经济环境的最优的投资行为。于是本文首先总结和回顾了前人研究的主要内容,介绍了不确定条件下的基础研究方法。然后根据这些基本方法,本文逐步引出不完全信息下企业投资的实物期权模型。首先,考虑单个企业的投资行为,这也就是相当于垄断条件下的企业不采取战略行为的投资决策。单个企业垄断条件下的投资是所有模型的基础,而且这个模型跟前人的研究没有差别。在这个基础上,本文引入了抢先投资决策,也就是企业面临的不再是完全垄断的环境了,而是存在一个竞争对手。企业的抢先投资在战略上具有非常大的优势,本文假设可以完全占有某个细分市场。所以两个企业都想抢先投资,但又要考虑投资的经济性,即在最优的决策点进行抢先投资。同时假设两个企业所得到的关于竞争对手的信息又是不完全的。本文假设每个企业都完全清楚自己的成本和收益,但只知道其他企业的成本和收益服从某个分布,如韦伯分布和帕累托分布。把这些新因素加入到原单个企业投资决策的模型中,就得到了企业有战略行为的投资决策模型。本文发现同单个企业的无战略行为的投资行为相比,有战略行为的投资决策的投资临界点要低一些,也就是说,在竞争的条件下,企业会在未来收益更低的点上进行投资,因为现在除考虑自身的情况外,还要受到其他企业的行为的影响,为了获得战略优势,只能降低未来收益的预期。在两企业模型的基础上,通过适当的扩展就可以得到多个企业的实物期权模型。在多个企业的模型中,企业需要观察的变量比两个企业的情况要多,要找出其他多个企业的投资临界值的最低点,成本的最低点,通过这些估计来指导本企业的投资行为。根据贝叶斯纳什均衡处理不完全信息的方法,假设其他企业的成本服从帕累托分布,引入到两企业的模型中,经过一系列的数学分析方法,得到了企业投资决策的最优解,即投资临界点。这个投资临界点是介于无战略行为的投资临界点和马歇尔投资临界点之间的,而且无战略行为和马歇尔投资临界点只是本模型的两个特例,所以本模型更具有一般性。通过改变模型中参数的数值,可以得到各参数对最优解的影响。随着成本上限的逐渐增加,投资临界点会逐渐向I U→∞时的投资临界点所形成的直线靠近;对于有限的投资上限,投资临界点都会在I U与马歇尔临界点相交。调整帕累托分布的参数a可以发现,随着a的增加,投资临界点会逐渐减小;随着a的增加有战略行为的临界值逐渐从马歇尔临界值的1.5倍向马歇尔临界值靠近,项目的期权价值也逐渐由马歇尔临界点价值的0.42倍而趋于零。企业对竞争对手的成本分布的估计对本企业的最优决策点有比较大的影响。如果使资本的收益率δ、泊松过程的概率λ和项目收益的波动率σ增加时,项目的投资临界点都会增加。通过参数变化后,临界值的变化方向可知道,项目投资期内的不确定性越大,项目投资的临界值越高。对于本模型的应用性,本文通过两个案例来展示。案例一为新药品研发问题,药品研发面临非常大的不确定性,传统的NPV分析容易出现问题。根据本文的方法,得出了投放市场期权的价值及投资临界点。但同时也发现了本模型在处理复合期权的问题上存在不足,如能跟二项模型相结合,就可以更好的解决这类问题。案例二为现在比较热门的3G牌照问题,本案例并不是纯粹的案例性质,有点实证研究的特点。由于四家电信运营商都在香港上市,所以在资产替代上更加便利。根据四家公司过去5年的股市和财务表现,比较容易的估计出了3G项目的投资收益、成本及各参数的数值。根据这些估计的数值,结合蒙特卡罗方法,得到了NPV规则和本文模型的解。通过对解的分析再次验证了NPV规则的缺陷性,及实物期权方法和本模型的实际应用性。通过本模型的解,比较直观的反映了为什么四家企业会进行正在实施的投资决策,3G牌照背后的巨大的战略和经济价值是各运营商及早投资3G以获得3G牌照的推动力。本文的创新点有以下几个,一是在完全信息下实物期权模型的基础上加入了不完全信息的因素;二是对Lambrecht and Perraudin(2003)的模型进行了扩展,加入了泊松过程、放宽了一些假设条件及讨论了收益和成本服从多个分布的情况;三是对模型进行了实证检验和分析。对于本模型还有很多可扩充的地方,比如放宽一些假设条件的限制,同其他的一些实物期权模型的结合,考虑更多的战略行为和具体某个企业的战略规划、行业地位等等因素。通过这样可以使模型具有更广泛的应用性,可以解释和指导更多的投资行为。当然对本模型的实证研究也是一个很好的扩展点,因为对实物期权模型的实证研究本来就是比较困难的一件事。由于企业的投资决策更多的是在不完全信息下进行的,所以继续本文的研究思路对现实的企业投资行为将具有非常大的现实意义。更多还原

【Abstract】 With the development of the economy, more uncertainties were contained in the internal and external environment which enterprises faced. And the information about market and competitive rivals that enterprices could get was incomplete. These new changes had made the defects of traditional valuation method, DCF, become increasingly clear. So enterprices should consider more factors in investment decision, and need new methods to evaluate investment. This article was based on this need. And based on the previous research and the basical methods of investment under uncertainty, it introduced incomplete information into the real option models. Used this way, incomplete information was quantified to the model, which could give the strategical investment behaviors more practical guidance.Athough real option was proposed only 30 years ago, it immediately became a hot theoretical research, which showed its powerful ability to explain the real investment behavior. Scholars at home and abroad constantly improved real option theoretical models, applied them to different fields, and explained and guided different economic behaviors. In this process, the general methods of investmet under uncertainty were formed, such as stochastic process, Winener process (Brownion motion), Ito process, Ito lemma and contigent claims theory. Through these methods together wity different economic environment, we could have the optimal investment behavior. So at first, this article summarized the previous research and introduced the basical methods of investment under uncertainty.Based on these foundations, the paper graduatlly got the real option model of enterprice investment under incomplete information. Firstly, this article considered the investment behavior of singal firm, which equaled the condation that the firm invested under monopoly. Based on this model, this article introduced preemption strategy. In this condation, the environment was not monopoly, and there was a competitive rival. Firm benefited from preemption in strategy, and it coulde get the entire market. So two firms both wanted to preempt, and also wanted to invest at the optimal investment trigger. But the information that one firm collected about the rival was incomplete. This paper assumed that each firm knowed its own revenue and cost, but only learned the distribution of the other firm’s revenue and cost. Introduced these new factors to the sigal firm’s investment model, the investment decision modle with strategical behavior coulde be presented. Form the comprision of the two models, we found the strategical trigger was lower than the non-strategical trigger. This meaned that firm would invest at lower revenue under competitive environment, because firm should consider the behaviors of competitive rivals, in order to get the strategical edge, it invested at lower revenue expectation.Through appropriate expansion of the two firm’s model, the multi-frims real option model coulde be presented. In this model, firm need to consider more variables than two firm’s model. Firm needs to conjecture the lowest investment trigger and cost of other firms. In light of this conjecturing, firm dicised when to invest. Through the way for incomplete information in Bayesian Nash equilibrium, this paper assumed other firm’s cost is Pareto distribution. Then resolved the model, we coulde get the solution of the model, which was the optimal investment trigger. This trigger was between the non-strategical trigger and Marshall trigger, and non-strategical trigger and Marshall trigger only were the special cases. So the model was more general.Changed the parameters of the model could be influence in the optimal solution. As the increase of cost’s upper boundary, investment trgger woulde close to the straight line of I U→∞. For the finite cost’s upper boundary, investment trigger crossed the Marshall trigger at I U. As the increase of the parameter in Pareto distribution, a , investment trigger woulde be smaller, and the strategical investment trigger woulde close to Marshall trigger. The conjecturing of rival’s cost distribution had the obvious influence. As the increase of capital revenue rateδ, Poission process probabilityλand volatility of revenueσ, the investment trigger woulde increase too. This meaned the investment trigger was bigger under more uncertainty.Through two cases, this paper illustrated the application of this model. Case one was the research and development of new drugs. We got the investment tigger and the value of launch options with this model. At the same time, the defects of this model were showed on compound option problems. If this model would integrate with binomial models, it couled have more application. Case two was about the 3G licenses, which was a hot point. An empirical research was done on this question and this model. Because the four telecom companies were all listed on the Stock Exchange of Hongkong, the subsititute assets were easily gotten. This paper estimated the revenue, cost and kinds of parameters on the market and stock’s situation of the four companies in last five years. Integrated with Monte Carlo method, we resovled the solution of NPV rule and this model. The solutions showed the defects of NPV rule again, and the application of real option and this model. Through the solution, we coulde explain why telecom companies invested in 3G, and the value of 3G licences.The model can be expanded for many places, such as relaxing the restrictions on assumptions, together with other real option models, consider more strategical behaviors and the specific firms. Through these works, the model can be applicated in more fields and more conditions. Continue the research will be more helpful in real enterprice investment.更多还原