上海股市收益率和波动性长记忆特征实证研究

An Empirical Study on the Long Memory Character in the Returns and Volatility of Shanghai Stock Market

本文运用标准对数周期图法以及tapered对数周期图法对上海证券市场综合指数以及一些分类指数进行了长记忆检验,并进而建立ARFIMA-FIGARCH模型来刻画股市的长记忆特征。研究结果表明,上海股市指数收益率序列的长记忆特征不显著,但其波动性过程却具有显著的长记忆特征。

We test for long memory in the daily returns and volatility series of Shanghai Stock Market. The measures of long- term persistence employed are the classical log- periodogram regression estimator and the tapered pefiodogram based estimator. Further analysis is conducted via ARFIMA- FIGARCH model. There is no significant evidence of long - range dependence in the returns while signitficant long memory is conclusively demonstrated in the volatility measures.