节点文献
投资者有限理性与证券价格行为研究
【作者】 张宗强;
【导师】 伍??;
【作者基本信息】 青岛大学 , 系统理论, 2007, 博士
【摘要】 现代金融学自Markowitz开始,通过在金融分析中引入“均值-方差”分析方法,先后出现了CAPM、APT、BS模型等支柱性成果,形成了一个完备的EMH假说框架。但随着研究的进展,市场上出现了众多EMH无法解释的市场异象,在理论界也开始了对EMH框架的反思,其中最具冲击力的是行为金融理论。行为金融理论系统地研究了投资者有限理性假设下的偏好、信念和行为偏差,指出普通投资者在投资过程中系统违背了理性人假设,并且不使用贝叶斯法则进行推理,这样,在投资者效用刻画上,行为金融理论提出了展望理论以区别于EMH理性人假设的SEU期望理论。在相应微观假设基础上,学者们研究了证券价格行为,给出一些理论模型,如充分理性假设下CAPM资产定价模型、非完全信息下的噪声交易价格冲击模型、有限理性假设下的证券价格反应过度和反应不足模型以及BAPM、还有物理金融学视角的协同市场假说。在此基础上,论文针对有限理性投资者只寻求满意解而非最优解的经典假设,建立基于模糊数学的投资者有限理性行为模型,在充分吸收了有关投资者偏好、信念和行为偏差的基础上,重点刻画了有限理性投资者的效用函数,认为其效用受其总财富水平、当期赢利情况、历史赢利水平、参考收益等多因素的综合影响。通过该模型我们认为:投资者心理效用变动,如贪婪、恐惧等,通过“参考收益”因素影响投资者决策,是投资者行为偏差发生的重要原因;有限理性投资者的满意程度随市场行情的变化而变化,并导致了投资者风险证券投资仓位的相应调整;在一个完整的市场波动周期后,有限理性投资者必然亏损;风险证券的价格变动与投资者平均仓位的变动过程呈正反馈关系,仓位变化推动了价格变化,价格变化又反过来推动了仓位调整。对模型的相关假设及结论,论文还通过实证检验与仿真研究进行了验证。对净买量与股票价格的实证研究表明,股价当日波动幅度与该股票的当日净买量显著正相关,存在逻辑关系。对模型的MATLAB仿真研究表明,存在平均持仓量与证券价格正反馈关系的市场中,大部分的投资者必然亏损,从而验证了“一赚二平七亏”的股谚。文章最后对投资者的常用投资策略进行了实证检验,其中主要是动量交易和反转交易策略。实证结果表明,我国的证券市场上,长期来看,动量交易并不能带来明显获利,但显然存在长期的价格反转现象。
【Abstract】 Since from Markowitz, modern finance formed a comprehensive framework of EMHby using of the "mean-variance" researching ideas and quantitative analysis method,and some important models such as CAPM, APT and BS model were established. Butwith the progress of the research, many market anomalies which can’t be explained bythe EMH theory were found. Then, researchers began to reflect on the EMH theory,and proposed the most influential theory, behavioral finance.Behavioral finance theory did the research on investors’ preference bias, belief biasand behavioral bias under the bounded rationality assumptions, pointing out the factthat the ordinary investors in the investment process systematically contraries to therationality assumption and will not use Bayesian rules for reasoning. Then, inbehavioral finance theory, the prospect theory was proposed to distinguish the SEUtheory in EMH theory for characterizing the utilities of investors.Based on the corresponding micro-assumptions about investors, scholars did lots ofresearches on securities’ price behaviors, and then, gave some theoretical models suchas CAPM under the fully rationality assumption, noise trading price shocks modelunder the incomplete information, over- and under-reaction model of stock priceunder the bounded rationality assumptions, BAPM, and cohesive market hypothesisfrom the physical financial perspective.By all above, under the important assumptions that bounded rational investor seeksonly satisfactory solution instead of the optimal solution, we established the investorbounded rationality behavioral model based on the fuzzy math theory, depicting theutility function of the bounded rational investors by analysis their preference, beliefand behavioral bias. And then, we got the result that their utilities were affected bytheir total wealth level, current profit, historical profit margins, and reference returns.By establishing this model, we believe that the changes of investors’ psychologicaleffects such as greed and fear will be the important reasons for investors’ behavioralbias through the "reference return" factor affecting investors’ decisions; their degreesof satisfaction will be changed by the changing of the market, and following withpositions changing of the risk securities investment; after a complete cycle of marketvolatility, bounded rational investors will suffer their inevitable losses; and there willbe a positive feedback process between the pricing changes of risk securities and thechanges of the investors’ average positions, which will affect each other.In this paper, relative assumptions and conclusions of the model were tested byempirical testing and simulation research. The empirical research on net buyingvolume and the prices of stocks shows to us that in the logic, there was a significantlypositive correlation existed between the volatility of stock’s price and the net buyingvolume. And MATLAB model of the simulation shows that in the market where theaverage positions have a positive feedback with the securities prices, most of theinvestors will get their inevitable losses and the words, "Minorities earning, majoritylosses" have its mean. Finally, we did the empirical test on investors’ common investment strategies, whichmainly focused on momentum trading strategy and contrary investment strategy. Then,we got the result that long-term momentum trading strategy can not get obvious profitin China’s securities market, while the long term contrary investment strategy can.
【Key words】 behavioral finance; bounded rationality; fuzzy math; investor utility; securities prices; investment strategy;