Research On Fuzzy Multi-period Portfolio Selection Model Based On A Theory Of Possibility

How to reasonably allocate limited wealth among different financial assets,so as to effectively avoid risks to maximize investment returns,has always been the ultimate goal pursued by experts and scholars in the field of modern financial investment.Since Markowitz's classical mean-variance model,more and more experts and scholars have improved and expanded it to solve the problems of portfolio selection in random environments,and achieved rich scientific achievements.However,there are many uncertain factors in the real financial market that are difficult to describe with probability distribution.Therefore,the portfolio selection models under the framework of probability theory cannot accurately describe the uncertain factors affecting investment decisions.After Zadeh proposed the fuzzy set theory and the possibility theory,some scholars used the possibility distribution instead of the probability distribution to study the portfolio selection problem under the fuzzy environment.However,due to the late start,most of the current research on portfolio selection in fuzzy environments is still limited to a single period of investment behavior.Based on the fuzzy set theory and possibility theory proposed by Zadeh,this paper analyzes the optimal solution for multi-period portfolio selection problem in fuzzy environment.First of all,the fuzzy set theory and the possibility theory were used to deal with various uncertainties in financial markets.And considering the transaction cost of the portfolio and the leptokurtosis and fat-tail phenomenon of the return rate in the real financial market,the mean-semivariance-skewness-kurtosis model is constructed.Then using the historical data in the Chinese stock exchange market for empirical analysis.Secondly,the dynamic feedback adjustment strategy was used to take the deviation between the actual return and the expected return as one of the influence factors of the investment strategy.Then,using the fuzzy entropy based on possibility theory to describe the degree of decentralization of the portfolio,the mean-semivariance-fuzzy entropy model is constructed,and the fuzzy programming method is used to transform the model into a single-objective programming model.Finally,this paper compares the mean-semivariance-skewness-kurtosis model and the mean-semivariance-fuzzy entropy model based on dynamic feedback adjustment strategy with the classical mean-variance model,and proves the advantages of the two models.