Empirical Applications Of Serial Persistence Portfolio Based On SV Model And Diversification

Constructing an appropriate portfolio to reduce risk is an important objective of the investment portfolio theory,but when investors select a portfolio,financial time series are often accompanied by a persistent feature of the portfolio return volatility.Correlative studies have found the features of financial time series would affect the earnings’ risk of portfolio in the future,then how to reduce the negative impact of the persistent volatility of portfolio’s return volatility,which can reduce portfolio risk in some extent is an issue faced by many investors.Therefore,it can reduce the impact emerged by financial volatility’s persistence on the future volatility of portfolio returns by constructing time series persistent based on stochastic volatility model.And the effectiveness of the method is tested through comparing the level of diversification of the portfolio’s risk.This research on portfolio selection has very important theoretical value and practical significance.In this paper,we use the method of random sampling to extract10 stocks randomly from the Shanghai and Shenzhen 300 Index stocks,and the period of the data is from 2011 to 2014.We construct the MV portfolio and series persistent portfolio based on stochastic volatility model,and after comparing the difference of basic statistical characteristics,parameters of Bayesian estimation,and the level of risk diversification between MV and SV portfolio we get the representative conclusions.We test the result through the method based on a sliding time window for the 10 stock portfolio,and make further efforts to test the empirical results by using the method of repeated sampling to construct 60 groups portfolio of”10 stock","15 stocks" and"20 stocks".We get some conclusions in the study of this paper.First,the returns’ distribution of the financial time series in China’s stock market has obvious fat tail,the volatility of earnings exhibit significant volatility clustering and persistence.Second,the volatility of the two kinds of portfolio is significantly less than that of the ten single stocks,higher than the overall average of single stock on earnings,which suggests that the two portfolios have good ability of risk control and receive a higher income than single stocks.On the whole,this kind of method to build a combination of series persistent based on the SV model has certain advantages.Third,we estimate the parameters of the SV model based on the MCMC method,and the persistent parameter index of SV portfolio is minimal of all the single stocks and the two portfolios,which is verified by the stability test.Fourth,SV portfolio is better than MV portfolio in the diversification of portfolios’ risk.In summary,the portfolio strategy of series persistent portfolio based on stochastic volatility model has good advantages in many aspects,and the conclusion of this paper is robust and stable.