Portfolio Model Based On Risk Entropy And High Order Moment

Markwitz's portfolio theory removes finance from the previous state of purely descriptive research and empirical operation,and lays the foundation for the development of modern portfolio theory.A necessary condition of the mean-variance model is that the return on risk assets follows a normal distribution.Unfortunately,this condition is not of practical significance.Many empirical studies at home and abroad have shown that the return on risk assets does not obey the normal distribution.Based on the above problems,the Markowitz portfolio model is improved in accordance with the characteristics of the Chinese financial market.In view of this,investors should consider the impact of high-order moment risk when making portfolio optimization decisions,otherwise suboptimal decisions will be resulted.At present,many scholars at home and abroad have studied the optimization problems of portfolios based on high-order moments.These studies mainly focus on using the higher-order moments of portfolio returns as objective functions or constraints.On the other hand,many scholars have studied entropy as a measure of risk.The reason is that the rate of return on assets does not have to obey normal distribution when entropy measures risk.Based on previous research results,this article expands the research on portfolio optimization problems based on high-order moments and information entropy,which makes portfolio theory more complete and practical.Aiming at the complexity and variability of financial markets,in order to measure financial risks more accurately,this article combines risk entropy and high-order moments to measure asset risk.In actual investment,considering that investors are more concerned about whether the actual return of the asset has reached the expected return,the negative half entropy and the lower half variance are used to measure that part of the risk,therefore,constructing negative half-entropy lower half-variance approximate skewness portfolio model and negative half-entropy lower half-variance theoretical skewness portfolio model.On the other hand,Yager entropy,generalized entropy,proportional entropy,and negative half entropy measure the risk that the actual return of the asset is lower than expected returns,and then extract the principal component entropy by principal component analysis.So the principal component entropy lower half-variance skewness portfolio model is established by the principal component entropy.By empirical analysis of eighteen stocks selected from CSI300 by the coefficient of variation factor model,it is found that,compared with the traditional Markowitz model and the semi-variance model,the combined assets of the new model have greater excess returns and lower risks.It can be seen that the empirical results also show the rationality of the new model and the feasibility of the algorithm.