中国股票市场的非对称效应研究

摘要

为检测重大风险事件对我国股票市场影响的非对称效应,运用贝叶斯MCMC推断技术对我国上证综指和深证成指进行了实证研究.研究结果显示:基于t-分布的门限随机波动模型比基于正态分布的随机波动模型能更加合理地刻画经济事件、政治事件和自然灾难对我国股票市场收益和波动影响的非对称特性.并且发现,经济事件、政治事件和自然灾难对我国股票市场收益和波动的影响均具有显著的非对称性.相对而言,政治事件对股票市场收益的影响存在正向杠杆效应,而经济事件和自然灾难却存在反向杠杆效应；经济事件和政治事件对股票市场波动的冲击存在正向杠杆效应,而自然灾难则存在反向杠杆效应.此外,除在牛市环境下利好经济事件和利空经济事件对股票市场收益和波动的影响具有反向杠杆效应之外,其它风险事件在熊市和牛市环境下对收益和波动的影响均具有正向杠杆效应.%For investigating the asymmetry of impact of major risk events on ChTo investigate the asymmetry of the impacts of major risk events on Chinese stock market, this paper studies the SSE composite index and SZSE component index using the threshold stochastic volatility (THSV) model with student distribution approach developed based on Bayesian MCMC technique. The empirical results show that the THSV model with student distribution are more comprehensive than with normal distribution for describing the asymmetry of the impacts of economic events, political events and natural disasters on the returns and volatilities in Chinese stock market. There are prominent asymmetries of impact of economic events, politic events and natural disasters on the returns and volatilities. Correspondingly, there are positive leverage effects in political events and reversed leverage effects in both economic and natural disaster events on returns. And there are positive leverage effects in both economic and political events and reversed leverage effects in natural disaster events on volatilities. In addition, other risk events all have a positive leverage effect on returns and volatilities in both bull and bear markets, except the reversed leverage effect of good and bad economic events on both returns and volatilities in bull markets.