Empirical Research On The Nonlinear Structure Of The China Stock Market

Traditional capital market theories are studied in the frame of neoclassical economics. As the core of neoclassical economics, Efficient Market Hypothesis (EMH) has always being the foundation of capital market theories. Many classical capital market theories are developed on the basis of EMH or on some bases closely related with it. Viewing from this, traditional capital market theories are studied by means of analysis on which is based linear model, but the practical capital market can be viewed as a complicated nonlinear dynamic system, the linear analysis couldn't be used to precisely describe the practical capital market. Many empirical analysis, home and abroad, show the limitation of EMH. So it's an inevitable selection to study capital market in nonlinear ways instead. The paper mainly consists of five portions and the main content is as follows: The first portion, as problem bringing forward, expounds the origin and meaning of this paper. Reviews the Domestic and international relevant research result, simply expounds the structure and innovations. The second portion particularly expounds the compare of the EMH and FMH, and shows the origin and theory basis of the R/S analysis. The third portion expounds the relevant concepts about content that the R/S analysis empirically researches the stock market in china. The fourth portion is the empirical analysis, The paper by the R/S analysis empirically researches the stock market in china, and gives the empirical analysis of the weekly logarithm returns of Shanghai General Index and Shenzhen Component Index and conduct comparative analysis of three portions before and after May 1994 and December 2001. The fifth portion is the conclusion. According to the analysis result we can receive the following conclusion:(1) Both Shanghai and Shenzhen Stock Market, from the whole extent, logarithm returns have the obvious nonlinear character.(2)From the whole extent, Shanghai Stock Market has much obvious nonlinear character than Shenzhen Stock Market and Shanghai Stock Market have average circle length about 280 days and Shenzhen Stock Market average circle length is 350 days (3) By using the BDS test of both weekly logarithm return series shows that the randomicity of weekly logarithm returns of Shanghai and Shenzhen Stock Market is obvious improved since 2001. (4) From the establishment of our stock market to 2001, the analysis shows that the market has the obvious low-dimension chaos process, and the Hurst index of the weekly logarithm returns also reveals that it has the anti-continuity character.