Empirical Study Of Long Memory In Chinese Stock Market

Stock market is often the main subject which is researched by many people. As the core of stock market, the stock price's movement has attracted large quantities of scholars to investigate. Along with the fast development of Chinese economic Chinese capital market is growing with a high speed. It will become one of the most important market in the world. But As a emerging market ,Chinese capital market has its own property. The foreign theories are often based on the developed market, So it's very necessary to do research work in Chinese market. As related to fractional theory, long memory now become popular more and more . Propose the use of the model developed from such theories in Chinese stock market and get the result has no-doubt the great significance in much aspect. Such as ,supervising the market, predicating the price's trend, guiding invest. In this article we propose the use of a version of the tests of Robinson for testing the fractional differencing parameter in financial time series data to test the long memory of Chinese market .This method is much efficiency in test finite sample financial time series data with non-Gaussian disturbances. We draw some conclusion in follow by analysis the result: 1) The indices are showing strong property of long memory. The logarithm return rates of stock indices also has long memory 2) The ShangHai stock market and ShenZheng stock market have a same fractional structure. 3) During the analysis, we find that I(0) process with white noise disturbance has the smallest d value ,this is a very strange result.