The Study On Regulating And Controlling Of China's Stock Market Based On Chaos

The obvious character in China stock market is that share prices fluctuated frequently and unusually . The status regards directly to that the stock market is whether or not and certain extent supervised and regulated and controlled by the government. Stock market system is a complicated phenomenon and it is significant for theory and practice to grasp this phenomenon for regulators to control the stock market effectively.In this thesis Chaos and Fractals are surveyed and the relations between Chaos and Fractal are clarified , among which three nonlinear dynamics numeric experimentation ways are introduced. These include two keys problem in phase space reconstruction theory, that is embedding dimension and optimal delay time, and numeric recognition of chaos, that is Lyapunov exponent, as well as numeric recognition of fractal ,that is correlative dimension of chaotic attractors. Then R/S analysis , phase space reconstruction of the system, chaos analysis and fractals analysis are done through Matlab program , based on original data of HuShen stock markets compositive index from year 1991 to year 2002 . And the author draws a conclusion based on original data that China's stock market obeys low-dimension fractals and ebb-chaos in terms of the experimentation result: Hurst exponents are between 0 and 1, memory cycles are obvious, Lyapunov exponents are more than zero and chaotic attractors correlative dimensions are between 2 and 3 in HuShen stock markets In this thesis the concept Information noises is put forward . Stock market information about policy and company of the last ten years is packed up and classified for regulators make decisions in terms of power the factor influences the stock market index. As a result, interest rate is the lest sensitive factor , the speech of relative power man is the most sensitive factor, and behavior of the leader shares or problem shares of the market are important as well. Also ways which are used to assist regulators make decisions and evaluate what have been always made are put forward, that are single apex of time series probability cut and chaotic degree . The two ways are demonstrated through index data from which information noises are cut out .The author put up some policy advice include fluctuate exchange tax rate and enhance rules and scale-up the stock market for stock market regulating. Accordingly,policy analysis which core is information noises based on chaos and company analysis which core is core competence based on AHP are put up for share investing.