特质波动的估计与结构突变检验——基于中国股市的实证研究

Estimation and Structural Change Tests of Idiosyncratic Volatilities:Empirical Research Based on Chinese Stock Market

基于中国股市1994年7月至2013年12月数据,选取直接分解法和间接分解法得到24个市场层面特质波动序列。相关度和结构突变分析认为,24个特质波动序列可分为"高度相关类"、"相关类"和"不相关类"三组,前2组包含的22个"历史"特质波动序列的结构突变点位置较为一致,而第三组对应的2个"预期"特质波动序列则较为特殊。典型特质波动序列表现出阶段性趋势,且基本反映了同期股市制度变革和宏观经济运行状况。

Direct and indirect decomposition methods are used to get 24 kinds of idiosyncratic volatility time series( IVTSs) with data of Chinese stock market from July 1994 to December 2013. Correlation analysis and structural change test are made and the outcomes indicate that there are 3 types IVTSs: highly correlated group,correlated group and uncorrelated group. The former two groups include 22 kinds of historical IVTSs,whose structural change points are in common,while the latter,uncorrelated group,has 2 kinds of expected IVTSs by contrast. Typical IVTS exhibits periodical trends,and the movement of IVTS above reflects the institutional reform of stock market and conditions of macroeconomic development in the corresponding period.