The Research Of Multi-period Portfolio Efficiency Evaluation In Stochastic DEA

In recent years,the issue of portfolio evaluation has become a hot research topic.The current financial market has become more and more complex and has a lot of information.Therefore,how to build a portfolio reasonably and how investors can achieve the least risk of investors' income maximization in the financial market is particularly important.This has led more and more researchers to research the issue of portfolio evaluation.Based on Markowitz's mean-variance portfolio theory,this paper constructs a multi-stage stochastic DEA portfolio efficiency evaluation model.The DEA portfolio efficiency evaluation of the existing literature uses risk as input and income as output,but the actual investment process input and output are wealth values.This paper starts from the actual investment process of portfolio problem to construct a multi-stage stochastic DEA model of portfolio problem.Aiming at the problem that stochastic DEA is difficult to solve,this paper transforms the stochastic DEA model into its deterministic equivalence model according to the chances constraint method proposed by Charnes and Cooper.Thirdly,based on the randomness of the fund's rate of return,the Bayesian method are used to estimate the statistical parameters of the fund data.Since the analytical solution of nonlinear programming is generally difficult to solve,this paper uses the rarely used interior point solver(KNI-TRO)to solve the deterministic equivalence model of stochastic DEA.We found that this solver can solve the optimization problem without any convergence problem.The efficiency of 23 hybrid securities investment funds in China's financial market was analyzed using the proposed stochastic DEA model.The results show that the stochastic DEA model based on Bayesian inference method is feasible.However,in general,the multi-stage stochastic DEA model is more in line with the actual situation of the investment process.