Research On Multi-period Portfolio Selection Model Based On Uncertainty Theory

The mean-variance model proposed by Markowitz has made the investment portfolio enter a quantitative era.Scholars in the financial sector have used quantitative analysis to appropriately allocate wealth to investors in order to maximize investment returns and minimize investment risks.A large number of literature studies show that portfolio models are basically based on probability theory.This requires that the return of the portfolio is a random variable,but the premise of using probability theory as the research background is that the variables have sufficient sample data..However,in fact,in financial markets that are constantly changing and complicated,usually we cannot obtain enough sample data.At this time,probability theory cannot be used to deal with such problems,and it is not certain that the theory just makes up for this deficiency.It relies on the experience and reliability of experts to estimate missing data.Based on the above reasons,this article assumes the return of the investment portfolio as an uncertain variable,and builds a relevant portfolio selection model with the background of uncertainty theory.The main research work and innovations of this paper are summarized as follows:(1)First derived the clear form of the lower half-variance of the uncertainty theory,and then use the uncertain half-variance to measure the risk in the investment portfolio.At the same time,considering the complex constraints such as transaction costs,entropy constraints,and cardinality,a model based on The theoretical mean-semi-variance multi-stage portfolio model was determined.Finally,the Firefly Algorithm(FA)was improved and used to solve the model in empirical analysis.(2)Based on the proposed model,this paper continuously improves and innovates.The utility function in the cumulative outlook theory is used in a multi-stage portfolio model.The expected utility value of the final return of the portfolio is used as the objective function.The semi-variance proposed above is used to measure risk,and a multi-stage utility function-semi-variance portfolio model is constructed.At the same time,the condition of bankruptcy constraint is considered in the model.The artificial bee colony algorithm(ABC)was improved.Finally,in the empirical analysis,the improved ABC is used to solve the model.