节点文献
基于行为金融的资产选择模型研究
Study on Portfolio Selection Models Based on Behavioral Finance
【作者】 彭飞;
【导师】 史本山;
【作者基本信息】 西南交通大学 , 管理科学与工程, 2005, 博士
【摘要】 现代金融(Modern Finance)承袭主流经济学(特别是新古典经济学)的分析方法与数理分析技术,经过近40年的努力,建立起了一套从公理假设到应用理论都很完善的理论体系。许多经典的现代金融理论和模型均是建立在理性人假设和Fama提出的“有效市场假说”这两个理论基础之上,其研究范式也局限于“完全理性”的分析框架。现代金融理论的理性人假设使其得以忽略人的具体决策行为而方便构建理论体系,但也因此使得决策过程成为了黑箱。 然而,现实生活中人们很难做到完全理性。在以现代金融理论的模型和方法对金融市场进行实证检验中,金融学家们遇到了许多不能被现代金融理论解释的“异象”(Anomalies),同时对经济的预测也出现了较大误差。在此种情况下,一些经济学家为使理论更好地符合现实,将决策过程中的主观因素和心理因素纳入了研究视角,于是20世纪80年代中期逐渐兴起了行为金融(Behavioral Finance)的研究热潮。经过近20年的发展,基于“有限理性”和“有限套利”(有称“非有效市场”)这两个与现代金融理论恰好相反的理论基础,行为金融逐步形成了如下的基本观点:投资者不是理性人,而是普通的正常人;投资者也不是同质的;投资者是损失厌恶型的,而不是风险厌恶型;投资者的风险态度并不是一致的;市场并不是有效的。 作为现代金融中的核心理论之一,现代资产选择理论的研究基础同样建立于现代金融理论基础之上,也不可避免地带有现代金融理论的局限性:即对金融市场上投资者的决策过程进行了理想化假设,在基于理性预期(Rational Expectation)、风险回避(Risk Aversion)、效用最大化等假设基础之上,投资者的决策过程被抽象成为一个追求主观期望效用(SEU)最大化的过程。行为金融的研究表明现代资产选择理论的假设基础并不符合人们的实际决策过程中的心理和行为,理论模型在现实中缺乏可操作性。 改变现代资产选择理论的假设基础,在符合人们决策心理和行为的基础上,研究适合的资产选择模型以满足实际投资应用的需要则成为了一个既有理论意
【Abstract】 After 40’year endeavor, modern finance (MF) has been established a suit of integrity theory system( from axiom theories to application theories), which inherit the mainstream economic (especialy Neoclassic Economic) approach and mathematical analysis technology. Many classical theories and models of MF are based on the rational agent suppose and Efficient Market Hypothesis (EMH) put forward by Fama, which research pattern also localized on the analysis frame of complete rationality. Rational agent suppose makes modern finance theories overlook man’s concrete decision-making behavior, then decision-making course is turned into a black-box.However, people are quite difficult to live up to complete rationality in the real life. Finance specialists have been encountered many anomalies which cannot be explained by modern finance theories, when they applied the models and approachs of modern finance theories to make empirical tests on finance market. At the same time, finance specialists found there were rather errors when they were applied to forecast economy. Under such situation, in order to make theories more accord with realism, some economist introduced individual’ subjective factors and psychology factors into finance study fields. Then, the upsurge of study Behavoral Finance (BF) rised gradually in the later 1980s. After near 20 years development, based on bounded rationality and limits to arbitrage, which are in opposition to modern finance theory, behavioral finance gradually formed its’ basic viewpoints: investors are common agent rather are rational agent; investors are not homogenous; investors are loss aversion rather risk aversion; investors’ risk attitude are not all consistent; and the market is also not efficient.Among modern finance theories, Modern portofio theory (MPT) is one of the most important theories, which is also based on the basises of modern finance theories. So MPT inevitably have its’ localization: investors need only to maxmize their subjective expected utility in their decision-making course, on the assumption that investors’ decision-making accord with ideal perfect finance market, namely investors possess such characters as rational expectation, risk aversion and utility maximum. Researches in behavioral finance indicate that hypothesis basises ofmodern portfolio theories do not accord with people’s real decision-making psychology and behavior, and it is short of maneuverability.Based on according with people’ decision-making psychology and behavior, it becomes an important task to study suited portfolio model to meet the needs of real invest application through change the hypothesis basises of MPT. Shefrin and Statman has made some valuable explore along such derection. We first suppose that investors can make good forecast for future and they violate Expected Utility Theory on the risk choose and risk judgement. Then based on those exited research fruits of BF and different risk measurement, and through combining BF with MPT, we present some behavioral portfolio models which can give investors better guidance for their real invest decision-making. The main research contents are as follows:1. The main theories accomplishment and basic viewpoints of MF are summarized. The challenge to MF from BF and basic viewpoints of BF are discussed briefly. The development history of MPT are also summrized ,and sum up the latest thoughts on MPT. We definitely bring forward that the development of finance should be combined BF research with MF research.2. The development historal of behavioral finance are summarized detailedly and the study system of BF are sum up generally.3. Aimed at such characters as losses aversion and reference-point setup which investors possess, setting out from the notion of variance risk, we put forward the mean-value variance model, the mean-subjective value variance model and mean-double target value variance model. We also discuss the utility function form of mean-value variance model and prove that the two fund separation theorem holds in the market when all investors adopt the same invest behavior.4. Based on deviation risk measurement, we bring forward a new risk meansureindex, namely value deviation , and put forward mean-absolute value deviation model and mean-maximize value deviation model, we also introduce such notion as return trade-off factor into the upper two model to improve them, which were put forward in decision-making science field, finally, on the assumpion that there are the same reference point for all the security, we present some portfolio models in which can be applied SP/A theory and vulue function.5. Introducing return balance factor, we substitute value deviation for return
【Key words】 behavioral finance; behavioral portfolio; risk measurement; value function; mental accounting; SP/A theory.;