Mean-Var Multi-Stage Portfolio Model Based on Simulated Annealing Algorithm

Simulated annealing algorithm is derived from the process of crystal cooling. If the solid is heated and then cooled, regular crystals with high density and low energy can be formed, which corresponds to the global optimal solution in the algorithm. This algorithm has the advantages of simple calculation process, strong robustness and suitable for solving complex nonlinear optimization problems. Based on a series of practical constraints, this paper constructs a finite short selling mean var multi-stage portfolio optimization model, and uses this model to obtain the optimal investment strategy of assets in each stage under different paths. Through empirical research, this paper finds that the prediction results of the model are in line with the actual investment behavior of investors, which shows that the model proposed in this paper is reasonable.