The Empirical Analysis Of SMSE And GEM

To forecast the financial market’s risk accurately has been the subject that the financial theoretical and the business circles are very concerned about. Because part accurate forecast on it will help us derive substantial revenue from the financial market, so the forecast to it seems apparently very important. This article will focus on the empirical analysis of high-frequency time-series data of the China’s SMSE and the GEM stock index, based on the time-series studies of high and low frequency at home and abroad in recent years, studying with1minute,5minutes,15minutes,30minutes,60minutes and day data of the SMSE and GEM stock index,from the view of logarithmic yield and its volatility, by establishing the HAR-WRV-GARCH-VaR model which has more practical significance, through further improving the model on the basis of financial time-series analysis and we draw some conclusions in the end. This article is mainly studying from the following aspects:First of all, we use the traditional statistical methods to make preliminary statistics on the high-frequency time-series of SMSE and GEM stock index, finding that China’s stock market high-frequency time-series are different from the low-frequency time-series in some characteristics, the former has the characteristics of high skewness, high kurtosis, negative first order autocorrelation, calendar effect and so on, this shows that the original research methods well performed in the field of low-frequency time-series can’t be directly applied to the high-frequency time series. Next, we establish ARIMA (1,1,1) model for the SMSE and the GEM stock index and solve, in addition, we analysis the effect of the model and find that the model is ineffective. Moreover, we make the further improvement about the model gradually by introducing the concept of long memory and then propose HAR-WRV-GARCH model. Following by this, we have an empirical study on the GEM stock market volatility and establish HAR-WRV-GARCH(1,1) model,then solve it. Finally, we establish the HAR-WRV-GARCH-VaR model and obtain the computing formula of GEM daily VaR by choosing GEM as an example on the basis of the summary of the previous model, making sure that the proposed model has more significance in the practical application.