The Research Of Distributed Characteristics On Returns Serial In Stock Market

After more than ten years development of Chinese stock market, it has obtained compelling achievement. With the efforts of all participants, the market is becoming more and more mature and perfect and the researches are gradually close to the root. Although stocks' returns are the main concerns for almost all of the academical researches about stock market, returns' random processes are even more important for correct investment decision making, because they provide the basic information about capital risk. This paper analyzed the characteristics of the ditribution of stock market returns serial from different aspects.Firstly, the development of Chinese stock market was reviewed and normal test was applied to Shanghai and Shenzhen stock market respectively. The result showed that Chinese stock market's returns serial does not submitted to normal distribution. For this reason, stable distribution was introduced to research Chinese stock market and parameters was estimated to discover the stable characteristics of stock market returns serial. Secondly, Multifractal Model of Asset Returns was introduced to research Shenzhen stock market and scaling function and multifractal spectrum was used to demonstrate the multifractality of stock market returns serial. Finally, Pearson Type IV Distribution and a new test method of Rao Score Test was introduced. Monte Carlo simulation was used to compare the Rao Score Test with test, and Rao Score Test was chosed to identify the leptokurtic of stock market returns serial.The result of research showed that the stock market returns serial has these characteristics: leptokurtic, fat tails, asymmetry and multifractality. The researches on these characteristics are the base of analysis of capital asset pricing, option pricing and prevention of financial risk, etc.