The Empirical Study Of The Characteristics Of Several Kinds Of Stock Index Based On EGARCH-M Model

With the rapid development of Chinese economy and the capitalization enhancementof stock financing, share price also becomes national economic development barometer.Because of increasing enhancement of the financial strength, the stock market impact hasbeen not only limited in the financial field, but also widely existed in the economy, politicsand all aspects of life. To some extent, a country’s development status and size can directlyor indirectly reflect the country’s economic strength and economic development status.Traditionally, the research of stock market mainly includes the security investmentanalysis and time series prediction method. Thus, the two methods have some kind oflimitations. At the same time, the prevalence of fat-tail characteristic in the stock market isthe one big difficult problem. Time series prediction method is difficult to determine thepredictive model, and we cannot access if this method can accurately predict the stockindex. Most of the time series prediction models are still based on the traditional linearhypothesis. However, a large amount of data is accompanied by noise. ARCH model andGARCH model is often used as the main model of the random disturbance to modelinganalysis, more fully the information extraction model residuals, more suitable for fittingthe financial sequence fat tails.In this thesis, we use five parts to study the SSE Composite Index, Shenzhencomposite index, the phase change characteristics of a stock bull market, influence ofpolicy effect on the stock market fluctuation in China and the dynamic relationships amongglobal three big stock markets. The obtained conclusions are drawn as follows:(1)Shanghai and Shenzhen composite index has the following features: a ARCH item is notsignificant, but GARCH is significant; bull market phase of the ARCH and GARCHcoefficients are statistically significant. Hence, the impact on the back period of theconditional variance of the effect is lasting.(2) China Securities Regulatory Commission’swarning and the GEM listed has significant influence on Chinese stock market, and this isaffected by the stock price formation mechanism and effect of the conditional variance offluctuations in the stock market.(3) the fixed cost of Dow Jones index long termfluctuation is highest. American stock market fluctuation is accompanied by relatively highfluctuations of the cost and the volatility clustering effect is largest for Shanghai StockExchange Composite Index, and information impacts of the three big stock markets haveall leverage effect.