Study Of Industrial Characters Of Chinese Stock Market Based On Nonlinear Theory

Traditional theories about stock market are mostly based on EMH( Efficient Market Hypothesis) and consider the stock price submit to a normal school. But many domestic and overseas empirical researches show that traditional theories don't accord with the realities well. More and more researches on the stock price daily return indicates its distribution is not a normal school but an obvious leptokurtosis progress. So using non-linear dynamical theory to study the stock market as a complex system has great significance.However the present domestic researches are mostly about testing the market's validity and analysing fractal characters of the whole market. There are few non-linear researches on the industrial characters. Industrial character analysis and industrial risks are always important factors considered by investors when making investment decisions. In a mature stock market, list companies among different industies have obvious differences on their stock price risks and returns. Previous studies on the stock industrial characters and industrial risks are mostly abased on traditional linear theory. The variance or beta in the CAPM model are the primary exponents to estimate the stock investment risks in those theories.This paper use sample datas from the Shanghai stock market to make empirical testing on the Chinese stock market. Through analysis on the Shanghai synthetical index and five kinds of industrial indexes we get Hurst exponent ,long memory property and so on.1) The results approve Chines stock market to be an invalid market and show that stock price volatility is an antipersistence time serial( 0