Risk Aversion Features Based On Behavior Portfolio Theory

Based on the fact that return distribution is featured by heavy tail and skewness,portfolio selection problem in Behavior Portfolio Theory(BPT)is difficult to solve,while Mean-Variance(MV)model can be expressed as a quadratic programming problem which can be solved with mature and practical algorithms to different scenarios.By merging BPT and MVT,this thesis aims to relate the risk measures of BPT and MVT.Empirically applying the of return data set of stocks in Shanghai and Shenzhen Stock Market to BPT model to calculate the ”implied” risk aversion coefficient ? in MVT based on the quasi-VAR constraint in BPT,we find the values of ? are unstable.On account of variance as the classical risk measure,we substitute ? with risk preference factor ?,which is defined as the ratio of the variance of BPT optimal portfolio to the minimum variance of MV effective frontier.By the regression analysis of risk measure parameters,we set up the relation between risk constraint parameters in BPT based on Safety First Criterion and the risk measure in MVT,so as to determine the risk measure ? of investor and get the optimal portfolio through MV model.It provides a new solution for portfolio selection.