Long-term Memory And Trend Analysis Of Chinese And American Stock Markets

The main goal of this paper is to explore the long-term memory of the return series of stock markets in China and the United States.We also present a prediction of the market trends of the two countries,and differences of the stock markets.Through a variety of quantitative methods such as modified R/S analysis,GPH method,Local Whittle and Exact Local Whittle method,this paper conducts the estimation of long-term memory parameters and spurious long-term memory test based on the original rate of return series and the break-free series of four representative stock indices in the stock markets of China and the United States respectively.The test results show that Chinese stock markets have steady long-term process.There are no evidence indicating the stock markets of the United States have long-term memory.At the same time,this paper established the model of time-varying long-term memory parameters,and it maintains the accuracy of more than 80% for the prediction of the trend of the Chinese stock markets.However,when forecasting the trend of the United States stock markets,the forecast results are relatively poor.In addition,this paper also takes SSECI as an example,the prediction ability of long-term memory modelARFIMA and traditional short-term memory model ARMA for SSECI weekly closing price series is compared and analyzed.The results show that compared with the traditional short-term memory model,the long-term memory model ARFIMA model has better prediction ability.Finally,this paper probes into the reasons for the different performance of the Chinese and American stock markets in long-term memory and the overall operation trend,and puts forward some corresponding policy suggestions to promote the healthy and steady development of the Chinese stock market.