EXPLICIT SOLUTION OF THE MEAN-VARIANCE OPTIMAL INVESTMENT MODEL FOR DEFINED-CONTRIBUTION PENSION UNDER NON-EXTENSIVE STATISTICAL MECHANICS

With the increasingly serious population aging, people pay more attention to the operation and management of pension. The optimization of pension investment has attracted the research interest of many scholars. Firstly, an asset price model is established by using the non-extensive statistical theory, which can well describe the high-peak and fat-tail characteristics of asset returns. Then, under the mean-variance criterion, the optimal investment model of a defined-contribution pension is constructed. Moreover, the explicit solution to the optimal investment strategy of definedcontribution pension is obtained by using dynamic programming, Legendre transformation, and duality theory. This conclusion not only broadens the application of non-extensive statistics in the financial field, but also provides a new theory for the investment of pension funds.