Based On The Long Memory Properties Of Stock Market Volatility Empirical Research

This paper is an empirical study on the volatility of stock market in China—based on the long-term memory theory. Stock market is full of uncertainty. The rapid flow of information and capital leads to frequent price changes in stock market, which in turn results in market fluctuation. Fluctuation of stock price is a normal characteristic of stock market. Moderate stock price fluctuation can promote the activity and enhance the liquidity of the market. Nevertheless, dramatic and frequent fluctuation can distort the market price mechanism. Being an emerging market, China's stock market is characterized by high risk of market fluctuation. Therefore, the research on demographic characteristics of stock market fluctuation has significance both theoretically and practically in finding an effective way to counter the excess fluctuation.This paper is long-term memory oriented, analyzes the demographic characteristics of the rational border of stock market fluctuations, and on this basis, analyzes efficiency and effective level of the dynamic process of stock market in China. This paper recalled the development history of the efficient market hypothesis and the classic theory of capital market, then elaborated the long-term memory theory on the basis of fractal market hypothesis, and introduced the R/S analysis for empirical research on long-term memory. In the framework of the EMH, using the methods of co-integration and the poor standard test the selected samples and the simulation of stock market combination, including whether or not the establishment of the border of variance and the present value model. EMH hypothesis believes that if a market is efficient, the market price changes are random, the more effective the market, the stronger random changes in prices. Based on the history of asset prices, the future price forecast is impossible. The existence of the long-term memory in financial markets is a major challenge to efficient market hypothesis. If there is a long-term memory exists in the financial markets, asset price changes follow some kind of law, the time series model that descript short-term memory in the past will no longer apply, long-term memory is negative to the EMH.With R/S method, empirical research was made on return time series of the Shanghai stock exchange comprehensive index and Shenzhen stock exchange component index. The empirical results show that: Hurst exponents are all larger than 0.5, which indicates that there's long-term memory in time series. This paper thinks that people's nonlinear response to information, the lack of hedging tools and the imperfect information disclosure are the main reasons that stock return series take on long-term memory characteristics.Based on empirical research, the paper achieves the conclusion that long-term memory and excessive fluctuation exist in the stock market of our nation, then refuses market efficiency. The author believes that the methods to reduce financial market volatility and promote market stability and efficiency should mainly include: to strengthen information disclosure, to strengthen education on market participants, to foster listed companies and strengthen government regulation.