应用变点模型来研究沪深股股市波动性突变行为

股票价格的每日变化呈现随机漫步的特征，其一定程度的波动反映了股票变化的内在规律，但未能预期的股价的突然变化对投资者及社会经济却产生着巨大的影响，如股价突然大幅下跌会令投资者损失惨重。笔者提出一种方差变点模型(波动性突变点模型)，并用该模型分析从1992年到2002年上证和深证综合指数的方差变点，应用二分分段法结合西沃兹信息标准(Chen和Gupta1997)，找出股指收益率序列中的所有方差(波动性)突变点，对这些变点的经济意义进行解释。

Testing and Locating Variance Change Points with Application to the Volatility of Chinese Stock Market

Stock prices are different on a daily basis with small ups, small downs, or abrupt changes. While small fluctuations are typical according to economic theory and understandable to most investors, abrupt changes are unexpected and can be either harmful (if dramatic downward change occurs) to a society's economy (and investors) or beneficial (if vast upward change occurs) to a society's economy (and investors). We introduce the variance change point model (volatility change point model), and use it to analyze two stocks: Shanghai Stock Composite Indices and Shenzhen Stock Composite Indices from 1992 to 2002. The authors use a binary procedure combined with Schwarz Information Criterion (SIC) as in Chen and Gupta (1997) to search all the possible variance (volatility) change points exist in the sequence. Changes in both indices are successfully identified and their economic explanations are given.