【作者】 杨涛；

【导师】 黄健柏；

【作者基本信息】 中南大学 ， 管理科学与工程， 2008， 博士

【摘要】 科斯(Coase)曾把20世纪初形成的主流经济学称为“黑板经济学”,由于假定委托人和代理人都是理性的经济人,没有充分考虑到现实中委托人和代理人的某些非理性行为和心理因素对模型的影响,因此很多“异常”的经济现象得不到合理的解释。行为契约理论的兴起为解决这一问题提供了借鉴思路。作者希望通过本文的研究,能够在理论上补充和夯实委托代理理论体系,并为行为契约理论的后续研究搭建一个理论平台。在实践上为降低上市公司管理者的行为风险、改善上市公司内部委托代理关系和制定有效的上市公司管理者报酬契约从而提高激励效率提供一些理论指导和有益思路。本文以企业委托代理理论、激励理论及行为经济理论为研究基础,以上市公司管理者为研究对象,并结合问卷调查、博弈论和计量经济分析等方法,将我国上市公司管理者普遍存在的“过度自信心理偏差”融入到激励契约理论中来,开发并检验了我国上市公司管理者过度自信度量问卷,探索了过度自信的上市公司管理者的行为风险,然后对过度自信的上市公司管理者的报酬业绩敏感度展开了相关研究,最后设计了相应的报酬契约机制以规避行为风险。本文主要的研究成果如下:(1)回顾了委托代理理论的发展历程,系统而全面地总结了过度自信的研究成果,指出行为契约理论是委托代理理论研究的前沿。(2)开发了我国上市公司管理者过度自信度量问卷,问卷包括能力过度自信分问卷和风险识控过度自信分问卷两个部分,对总问卷和分问卷均进行了信度和效度检验,检验结果表明:问卷信度较高,结构效度、内容效度、外部关联效度等良好。通过过度自信问卷调查对我国上市公司管理者过度自信进行了差异分析,统计结果表明:我国上市公司管理者过度自信在性别和企业性质上差异显著;在地域、年龄、教育背景、职称的差异不显著。(3)系统完整地对既往研究中过度自信的度量方法进行了归纳总结,指出传统方法的缺陷和不足,结合我国上市公司管理者过度自信的特点,提出我国上市公司管理者过度自信的度量方法。(4)首次运用数理推导证明:过度自信的上市公司管理者报酬业绩敏感度比理性的上市公司管理者报酬业绩敏感度高;过度自信的上市公司管理者报酬业绩敏感度与风险成反向关系。并利用了2002-2007年我国上市公司高管的相关数据进行了计量检验。在此基础上提出提高我国上市公司管理者报酬业绩敏感度的策略建议。(5)分析了过度自信的我国上市公司管理者行为风险行为,根据其主要的危害和重要程度提出了“聘用”与“决策”二维模型,并指出“聘用”维的行为风险是工作不尽职,“决策”维的行为风险主要是投资过度等行为。提出规避行为风险的报酬契约设计方法:规避“聘用”维行为风险的方法是考虑过度自信对报酬契约设计的影响,并探究了影响机制;规避“决策”维行为风险的方法是采用联合基数法,既能保证管理者付出最大努力工作,又能保证管理者作出合理的决策。更多还原

【Abstract】 Coase once called the mainstream economics formed at the beginning of twentieth century as blackboard economics. As the principle and agent are all rational and they have commonness in both economic behaviors and psychological ideas, the mainstream economics didn’t fully consider that the irrational behaviors done by the principal and agent and psychological factors can influence the model. As a result, much abnormal economic phenomena can not be explained reasonably.Theory of behavior contract springs up and offers ways to solve this problem. The author hopes that this paper can complement the theoretical series of principle and agent, and build a platform for the further study of behavior contract theory. Practically, this paper will supply some theoretical advices and beneficial ideas to lower the manager’s behavioral risks of the public company, to improve their inner relationship between the principal and agent and design effective incentive mechanism in order to enhance the efficiency of motivation.With the principle and agent theory, incentive theory, behavioral economic theory for study foundation, with public company managers for research objects, using questionnaire, game and econometrics, this paper introduces the overconfidence psychological warp of public company managers into the incentive theory and design the overconfidence questionnaire of public company managers and analyze their behavioral risks. According to the results, this paper studies how to enhance the overconfident manager’s reward-performance sensitivity and designs compensation contract mechanism to avoid behavioral risks finally.Main results of this paper are as follows:(1)This paper systemically and fully summarizes the development process of principle and agent theory and the research results about overconfidence, points out that the behavioral contract theory is the new direction of principle-agent theory.(2)This paper has designed questionnaire to measure the overconfidence of public company managers, which is made up of two parts including overconfidence about ability and overconfidence of risk control. Both total and sub questionnaire are carried on reliability and validity test. The results show that reliability is high and the structure validity, content validity and exterior connection validity are good.Based on these, this paper analyzes the overconfidence of public company managers in several aspects. It shows that overconfidence lever is remarkable different in sex and kind of enterprise, while not remarkable in area, age, education background and position title.(3) This paper systematically and completely sums up the measure method for overconfidence in the past and points out the bugs and shortage of traditional method. At the same time, considering the peculiarity of Chinese public company, this paper suggests the right way to test those managers’ overconfidence.(4) The first used Mathematical Analysis shows that: the sensitivity of reward-performance of public company managers is higher that rational ones; the sensitivity of reward-performance of public company managers is opposite to risk. This paper uses relevant dates from 2002 to2007 to test these assumptions. At last the strategies of raising the lever of sensitivity of reward-performance are put forward.(5) This paper analyzes the behavioral risk of Chinese overconfident public company managers, and brings forward two-dimension model of "employment" and "decision-making" according to the main damage and importance. Model shows that the behavioral risk of "employment" dimension is not line of duty, while the behavioral risk of "decision-making" dimension is over-investment. This paper puts forward how to design reward contract so as to avoid behavioral risks: the effective way to avoid behavioral risk of "employment" dimension is to consider the overconfidence influence on reward contract. The influence mechanism of overconfidence to compensation designing are analyzed. The effective way to avoid behavioral risk of "decision-making" dimension is to use HU, it can either make the managers do their best or ensure the managers make rational decision.更多还原