The Daily Return Model Of New Shanghai Composite Index And Shenzhen Component Index Based On G-ARMA-GARCH Model

During the changing of prices in stock markets,the volatility is a very important indicator,and many scholars from home and abroad keep studying the hot issue.There are fluctuating and asymmetric characteristics,time-varying characteristics of variance and clustering characteristics in stock markets at different development levels,and the negative returns in the series tend to bring greater volatility than the positive returns,thus the leverage effect exists in the price series.Compared with the developed countries,Shanghai and Shenzhen stock markets of China started relatively late,and the development is not mature enough,the regulatory measures in these two stock markets are not perfect,so we have a huge gap between stock markets of our country and mature stock markets.Considering the current situation of Shanghai and Shenzhen stock markets,we study the asymmetric and fluctuating characteristics of prices in Shanghai and Shenzhen stock markets.And this helps us broaden our understanding of volatility of Shanghai and Shenzhen stock markets.In Shanghai and Shenzhen stock markets of China,the Shanghai Composite Index and Shenzhen Component Index are two major indices.We start with the reform of Shanghai Stock Exchange,end with revising the preparation plan of Shenzhen Stock Exchange,and we use the descriptive method and empirical analysis based on the G-ARMA-GARCH family models to describe the fluctuation of the daily return of these two indices from January 4,2006 to May 19,2015,then compare these two indices.Finally,we predict the daily closing prices of these two indices for May 20,2015 to December 31,2015 with the best-fit models.During the descriptive analysis,we discuss the normality,stability,heteroscedasticity,autocorrelation and partial autocorrelation of these two indices.We conclude from the analysis that the daily return of New Shanghai Composite Index and Shenzhen Composite Index are stable series,which indicates that the initial fitting of conditional mean can be carried out by the ARMA model.At the same time,there are left partial features and spike thick tail characteristics in the daily return of these two indices,indicating that heteroscedasticity maybe exists in the series,and we fit conditional variance with the GARCH family models.In the process of empirical analysis with the G-ARMA-GARCH family models,we add the gradient factor to the ARMA-GARCH family models to reflect the effects of data and night gap,considering the possible relationship between today's opening price and historical closing prices.The results show that EGARCH model in GARCH family models is more suitable than TGARCH model to fit the conditional variance of New Shanghai Composite Index and Shenzhen Composite Index,and the fitting effect is better when the residuals of the series obey with the GED distribution.In addition,we find that the leverage effect exists in daily return of these two indices,and the leverage effect of Shenzhen Composite Index is stronger than that of New Shanghai Composite Index At the same time,the volatility of conditional variance of Shenzhen Composite Index is also relatively larger than that of New Shanghai Composite Index,which indicates that the risk level of Shenzhen Component Index is higher.By comparing the accuracy of ARMA-GARCH family models and G-ARMA-GARCH family models in predicting closing prices,we can conclude that G-ARMA-GARCH family models with the gradient factor are more accurate and more stable.According to descriptive analysis and empirical analysis,we can find that there are still many problems in Shanghai and Shenzhen stock markets of China.In the future,we need to discuss these problems through deeper research to promote further development of Shanghai and Shenzhen stock markets of China...