Is the covariance of international stock market returns regime dependent?
应用SWARCH模型研究五大股票市场日度数据，发现市场波动率状态切换具有同步性，且国际协方差矩阵不稳定，高波动率导致平均相关系数上升，时间和状态因素对相关性有显著解释力。
The application of a SWARCH model to stock market returns allows one to endogenously determine the regime dependence of the stock market volatility. Comparison of the results from a sample of daily data from five major stock markets shows that the majority of the markets switch regimes simultaneously. This fact is used to investigate the relation between market volatility and the behaviour of the variance—;covariance matrix. It is found that the international variance—;covariance matrix is not stable and that changes in the matrix are dependent on the volatility regime. A high level of variance causes an increase in the average correlation coefficient. The co-movement of the markets is further described by a steady increase in the covariance over the whole sample period. It can be shown that both the time component and the regime dependence of the average correlation have separate and significant explanatory power.