A comparative study on GARCH effect and chaos in China´s stock markets

The paper comparatively studies ARCH effect and chaos characteristic of China´s stock markets based on independent and identical distribution test, GARCH model, correlation dimension and Lyapunov exponent methods. The results indicate that the GARCH model captures the part information of the stock markets and the chaos methods can measure the nonlinear characteristics. With the analysis of correlation dimension and the largest Lyapunov exponent, the studies show that the weekly stock prices in China´s stock markets are more complex than the daily stock prices and that the complexity of Shenzhen stock market is higher than that of Shanghai stock market.