Realized GARCH-HAR-RV Based On Jumping Behavior Model Of Chinese Stock Market Volatilit

In recent decades,with the outbreak of economic crisis in various countries,the volatility of financial market returns has become the most concerned issue for financial investment traders.Therefore,the study of stock market volatility in financial markets is extremely important.In the real stock market,there are many reasons for the stock market volatility,macroeconomic fluctuations,the announcement of government policies,etc.will all lead to a significant change in stock prices in a short period of time.The existence of jumping behavior has a direct impact on the estimation of the volatility of asset returns and the accuracy of forecasting.Therefore,jump behavior cannot be ignored in the study of stock market volatility.This paper analyzes the volatility of the stock market based on the jump behavior of China's securities market combined with the realized volatility.The main content of the paper includes several aspects:First,a new jump VJt is constructed based on volume,and Jt,CJt,SJVt,SJV-t,SJV+t,MedJt,and RJt are constructed.There are a total of eight jumps,which describe the Chinese stock market from Jumping behavior multiple angles and levels.Secondly,based on the HAR-RV model proposed by Corsi under the heterogeneous market hypothesis,the influence of long-term traders,medium-term traders,and short-term traders on the volatility of the stock market was discussed from the perspective of heterogeneity.In view of the impact of jump on the fluctuation can not be ignored.Eight jumps and principal component jumps were added to the HAR-RV model in order to construct HAR-RV-J,HAR-RV-CJ,HAR-RV-SJVt,HAR-RV-SJV+t,and HAR-RV-SJV-t,HAR-RV-MedJt,HAR-RV-VJt,HAR-RV-RJt,HAR-RV-Primary Component Jump.The empirical results show that jumping is inevitable,and there is heterogeneity between jumping and fluctuating.Through the comparison of models,it is found that the goodness of the HAR-RV-PCG jump model is better than that of other HAR-RV jump models,indicating that the principal component jump reflects more effective information in the stock market and can better characterize the heterogeneity of the stock market.Jumping sex.Next,the principal component jump is introduced into the realized measure equation of the realized GARCH model.Based on this,the realized GARCH-HAR-RV principal component jump model is established.Through empirical research,the model describes the main jump and has been Achieve volatility correlation and impact on profitability,and can successfully predict volatility.Finally,using the out-of-sample forecasting to achieve the volatility,the empirical results show that the high prediction accuracy indicates that the GARCH-HAR-RV principal component jump model has achieved a good fit.The Markov prediction mechanism is also used to successfully predict the state of the main component jumps and fluctuations,and to analyze the correlation between the jump state and the fluctuation state.