中国股市行业指数及个股长程记忆研究

On Chinese Stock Market Industrial Indices and Long-term Memory of Single Stock

股票市场收益记忆特征对于非线性结构的确定以及市场有效性的研究具有重要意义。该文选取沪深股市的三只行业指数和隆平高科及大江股份的日、周收益率时间序列,通过R/S分析方法分别计算出其相应Hurst指数。结果表明收益率时间序列具有长程记忆特征,且存在一个大周期和一个小周期,也验证了收益率时间序列不符合布朗运动。这对风险管理及投资时机的选择有一定的指导作用。

Memory characteristics of the stock market gains are of great significance for determining the non-linear structure and studying the effectiveness of market.In this paper,we selected the yield-week time series and yield-day time series of three industrial indices and two agricultural stocks in Shanghai and Shenzhen stock market,and we calculated their Hurst exponent by R/S analysis methods.The results shows that yield time series have long-term memory characteristics,and there are a big cycle and a small cycle.We also verified the yield time series which are not consistent with the Brownian motion.It has a certain roles in the risk Management and choosing;Investment opportunity.