基于分形V/S技术的沪深股市长记忆性研究  

Long memory testing for Shanghai and Shenzhen stock markets based on the V/S analysis

作　　者：顾荣宝[1] 陈霁霞[1] 

机构地区：[1]南京财经大学金融学院,江苏南京210046

出　　处：《安徽大学学报（自然科学版）》2008年第3期18-21,共4页Journal of Anhui University(Natural Science Edition)

基　　金：江苏省高校自然科学基金资助项目(05KJB110033)

摘　　要：将Cajueiro和Tabak提出估计Hurst指数的新方法即V/S分析引入我国沪深股票市场非线性特征的研究.比较研究表明V/S分析不易受短期相关性的影响,较R/S分析更为稳健和有效.通过对沪深两市综合指数的日收益率和周收益率序列的V/S分析,得到沪深两市的Hurst指数均小于0.5,这表明我国股票市场呈现出反持久性特征而不是长记忆性.We investigated Shanghai and Shenzhen stock markets by using the V/S analysis, a new method introduced by Cajueiro and Tabak in calculating Hurst exponent. It showed that the V/S analysis was a powerful and stable method. We applied the V/S analysis method to analyzed the stock return for the first time in China and obtained that the Hurst exponents of both Shanghai and Shenzhen stock markets were less than 0. 5. It shows that Chinese stock markets hasn＇t long - range dependence.

关 键 词：V/S分析 HURST指数 长记忆性 

分 类 号：O212[理学—概率论与数理统计]