Non-linear Structure Of Stock Market In China-Empirical Research And Forecasting

The validity of the financial market is the foundation of financial market. It is the analysis of fluctuation characteristic and basic structure of the financial market. The basic structures, feedback mechanism to historical information, fluctuation characteristic, inherent mechanism of the financial market are the most important question to the financial market research.The Efficient Market Theory has been occupying the leading position in this field all the time, the Efficient Market Theory (Fama, 1970) believe that when market price is relevant to information (historical information that reveal, disclose information and inside information) promptly, the capital market is effective. These theories are at least set up on the basic supposing. First, the change of the price is totally decided by new information, today's price and yesterday's price are separate. Second, the price is normal distribution. Third, investor is completely reasonable. Such linear form has occupied financial economics as long as half a century.However, the Efficient Market Theory can't explain numerous phenomena appearing in the real financial market. Thus, the capital market theory based on the Efficient Market Theory now receives query and challenge in varying degrees.Since 1990s, non-linear theory and methods are applied to financial market research extensively, such as nonlinear kinetics, Chaos theory, fractal theory, etc. With the applying of the non-linear theory in researching the structure of financial market, the stock price forecasting model develops towards non-linear direction too.The development of the fractal theory establishes the Fractal Market Hypothesis (FMH) which conforms to the real market better. FMH has broken the pessimistic viewpoint of the Effective Market Hypothesis and provides a basis of forecasting the stock market. This article does empirical research on fractal structure and memory character of the stock market in China by high frequency data of the Shanghai Stock Exchange Composite Index. Next we use a Nerve Net model to forecast the Shanghai Stock Exchange Composite Index. The results of forecasting sustain the fractal structure and memory character of the Shanghai Stock Exchange Composite Index which we conclude.The structure of article is as follows: Chapter one is the introduction. We introduce the background and meaning of the research. Chapter two, we systematically explain the fractal theory, including basic content, characteristic and meaning of the fractal theory. Chapter three, we do a lot of empirical research on fractal structure and memory character of the stock market in China. Chapter four, we study the forecasting models of the Chinese stock market. Finally we summarize the result of study and some problems needing further studying.