Application Of "Mean-CVaR" Portfolio Optimization Model In China's A-share Market

There is a famous saying in the West: Don't put eggs in the same basket.The meaning of this sentence in finance is to achieve the purpose of diversifying risks through a portfolio of diversified investment packages.How to choose the type of component assets and the weight of the component assets in the investment portfolio,so that the investment portfolio will bring investors the largest rate of return or face the smallest risks is a very important issue.In this paper,by testing the normality of the average return time series of sample stocks,we can know that most of the stock return time series in China's securities market have peaks and thick tails,which do not meet the assumption of normal distribution.Therefore,many domestic Research under this assumption is not valid.In the risk measurement method,the CVaR method makes up for the shortcomings that Va R does not meet sub-additivity,inability to make full use of tail information,and difficulty in predicting certain small probability extreme losses.Portfolio optimization is more effective.In order to further demonstrate the application of the mean-CVaR model in China's Ashare market,ten financial indicators were selected to comprehensively measure all the stocks in the SSE A-share stock pool,and the principal component analysis method was used to reduce the multi-index data to extract dimensions into several A few comprehensive indicators to reflect most of the original data,construct a new principal component equation,and then score all the stocks of the Shanghai A-shares,sort them based on the score results,and select five top-ranked stocks from different industries for the average-Empirical analysis of CVaR portfolio optimization model.In the process of empirical analysis,firstly retrieve the data of the five stocks before the restoration of the stock price,and obtain the time series of the yield according to the average yield formula.The Matlab programming software is used to solve the optimal portfolio composition of a certain yield for a given confidence level.The weight ratio of the asset and the corresponding CVaR value.At the same time,the following conclusions are drawn: First,at the selected confidence level,as the expected return on the investment portfolio gradually increases,the investment weight ratio of each component asset in the optimal investment portfolio will also change significantly,and the investor 's funds Will be more inclined to stocks with higher average returns and smaller variances.When the CVaR value of the investment portfolio chosen by the investor is higher,the demand for expected returns is naturally higher,and the greater the willingness to bear;Under the selected confidence level,as the expected return of the investment portfolio gradually increases,the value of CVaR will slowly increase,but the value of Va R does not necessarily increase with the expected return.Large,and the value of CVaR is always higher than the value of Va R,which is in line with the definition of CVaR as the expected loss of the portfolio over Va R;Third,under the selected expected rate of return,with the change of the confidence level,the mean-CVaR investment There will also be significant differences in the effective frontier of the combination model.The effective frontier moves to the right as the confidence level increases,and the effective feasible set also moves to the right.The portfolio income range available to investors is gradually increasing.Little while,CVaR value also increases as the confidence level increases,investors' funds are more concentrated in a small number of higher-yielding assets.