Characteristics Test And Prediction Of China’s Security Market Based On The R/S Analysis Method

As we all know, an efficient security market can push the limited capital resource to the most efficient industries and enterprises, regulate supply and demand, realize the rational distribution of resources, promote economic development. Although China’s security market has been gradually forming a more perfect market system, It was founded in a very short time,the gap between China’s security market and the efficient security market is still quite large, and there are a lot of theoretical blank.More importantly, the fluctuations of price in Stock market are often rendered in a messy situation, make it difficult for ordinary investors to grasp the law of it,and usually suffered great losses in the investment activities. Therefore, to forecast capital prices in stock market have important theoretical and practical significance. But according to the efficient market theory(Efficient Markets Hypothesis/EMH) in the West, in the most efficient market, price fluctuations are random, no one can predict the stock price and get long-term profit by using the historical information. Therefore, before forecasting the stock market prices we need to answer an important question-is the stock market efficient? If we cannot determine the effectiveness of the security market, an empirical prediction is a castle in the air. EMH took up a very important position in the basic framework of modern mainstream financial theory since it had been advanced, however in the1970of the20th century, EMH was challenged form theory and empirical area, since it unable to solve many problems in the present world, an new research area-behavior financial theory emerge because of demand. Analyzing the characteristics of the security market with the fractal theory(Fractal Market Hypothesis/FMH) of non-linear complex system is becoming the frontier and hotspot of financial scientific research. Relative to EMH, FMH put greater emphasis on the influence of liquidity and term structure to the investors, it is a kind of model that correspond to practical situation and can accurately describe the behavior of the investors and the price fluctuation of the securities.This paper first elaborated the EMH and the challenges it faced, descanted the FMH, demonstrated the applicability of FMH in complex financial system. Then tested the basic statistical characteristics, non-linear features and fractal structure of the return series of shanghai and shenzhen security markets, found that China’s security market is not consistent with the efficient market hypothesis, it’s more realistic to use the fractal market theory to depict it. Then, based on the division of development course of the stock market, used the R/S analysis calculated the Hurst Indexs in different stages, the result indicates that China’s security market is far from efficient markets, that we can use existing information to forecast the stock price for the future. Finally, by comparing various pros and cons of stock price forecasting model we choose the Vector Autoregressive Model (VAR model) to do the fitting and prediction of stock price in China. Empirical results show that:The return series of stock in China do not follow a normal distribution, which is with sharp peak and heavy tail, and non-linear correlation, shows an obvious fractal characteristics, and therefore do not meet the assumption of efficient markets theory, non-linear methods should be adopted to research it; Use R/S analysis to calculates Hurst Index of Shanghai and Shenzhen stock market returns series, we found the Hurst Index in the three phases are greater than0.5, indicating that the return series is not a random walk process, the market is not a efficient market but with long-term memory effect. Therefore, we can make forecasts using existing information; The degrees of market efficiency of the two markets in Shanghai and Shenzhen demonstrate a trend from weak to strong and then return to weak, but in different stages the degrees of market efficiency are different, and the degrees of the changes of market efficiency is not the same too; By using VAR model to fit and forecast the price of China’s security market, we found that VAR model has strong applicability for economic forecasts, when using the model to forecast the next period price of the market index, the relative errors is only0.058%, shows that it has a high precision sexual when make forecasting, it indeed can help us to make investment decision and get profit.