A Credibility Mean-Minimax Risk-Skewness Multi-Stage Portfolio Selection Model with Short-Selling Limitations

In this paper, the multi-stage portfolio selection problem with short-selling constraints is studied. Firstly, triangular fuzzy numbers is used to measure the expected returns of risky assets. Meanwhile, the practical constraints are introduced to construct a Credible Mean-Minimax Risk-Skewness multi-stage portfolio selection model (CMSVS). These practical constraints include short-selling total constraints, single-stage minimum return constraints, threshold constraints, and budget constraints. Secondly, an improved multi-objective dragonfly algorithm is designed to solve the proposed model. Thirdly, the empirical analysis is conducted by using real stock data from the Shanghai Stock Exchange. And the optimal solutions obtained by the improved multi-objective dragonfly algorithm are compared with those derived from the traditional multi-objective dragonfly algorithm. The effectiveness and predictive power of the proposed optimization model and algorithm are validated through these results.