A Portfolio Model With Entropy Constraints And Its Empirical Analysis

The securities market has become an indispensable part of social and economic development.It is effectively configuring social resources,and plays an important role in broadening the channels of capital flow.However,the stock market returns and risk go hand in hand,so how to control risks in a reasonable range is very important for investor.This thesis mainly studies the portfolio investment from the perspective of dispersed investment risk.The main work is as follows:(1)On the traditional portfolio model,sometimes there will be too much money in the securities that are historically well behaved,thus leading to a larger investment risk,this thesis is based on the traditional mean absolute deviation model,we control the allocation of funds by increasing the Yager entropy constraint,to reach the goal of controlling risk.At the same time,according to the investor's demand for the liquidity of the securities,liquidity index represented by turnover rate,a mean absolute deviation model with entropy and liquidity constraints is established,and the validity of the demonstration.(2)Membership degree calculation method of possibility as fuzzy number,possible non duality that puzzles investors,the possibility is eliminated by credibility instead of possibility.Due to the different preferences of investors,introduction of skewness constraint,the probability of increasing positive returns,so that the allocation of funds can be controlled in the range of investors' needs.Finally,a reasonable investment strategy is obtained by dispersing risk through Yager entropy.