我国证券市场结构突变与波动性实证研究

An Empirical Study on Structural Breaks and Volatility in Chinese Stock Markets

重大事件的发生会使证券市场波动率发生结构突变。通过修正ICSS算法检验,可以发现我国证券市场波动存在明显的结构变化。将结构突变因素加入波动率模型进行比较后,可以发现含结构突变的波动率模型能更准确地刻画波动率特征。再分别按照结构突变发生的时间点分割样本区间与构造虚拟变量两种方法,对含结构突变的波动率模型进行比较研究的结果表明:构造虚拟变量方法对我国证券市场的波动率建模能取得较好的统计效果。

Important events can cause structural breaks of volatility in the security market. Using a modified ICSS algorithm to detect the structural breaks of Chinese stock markets, the author has found that there exist significant structural breaks in Chinese stock markets. Moreover, a comparison of volatility models after considering the factor of structural breaks reveals that the models can describe the feature of volatility more rigorously. Finally, using segmenting samples zone and constructing dummy variables respectively in terms of structural breaks, the author conducted a research on the GARCH models. The empirical research shows that the GARCH models with dummy variables can bring about better statistical findings in Chinese stock markets.