Study on the optimization of portfolio based on entropy theory and mean-variance model

This paper integrates the entropy theory into Markowitz portfolio model to make a better performance in simulation for the relation between investment return and risk. With historical data of stocks in different industrial sectors, empirical analysis is conducted for the portfolio optimization. The study is superior to the standard approach, performing well with fairly few examples, particularly when momentum factor is employed and adjusted. The portfolio risk is measured under the constraints of return-oriented and systematic risk. The model provides a natural probabilistic interpretation for daily return which usually changes from positive to negative, and it indicates that the entropy can be used as a complement to the mean-variance portfolio model.