A New Tail-Based Correlation Measure and Its Application in Global Equity Markets
提出了一种基于预期亏损的尾部相关性度量方法，模拟显示其比现有方法更准确，并能捕捉相关性变化，实证发现国际股市在低收益时相关性显著增加。
Abstract Many studies have found that large negative returns tend to occur together. This article develops a correlation measure focusing on tails using the expected shortfall (ES), referred to as the ES-implied correlation. The new correlation measure provides closer estimates to the true correlation than does the existing value at risk-implied correlation measure in the simulations. In addition, the new correlation measure performs as well as the linear correlation measure when correlation is constant but captures changes in correlations when correlations are not constant. A series of test statistics are developed to measure and test correlation asymmetries. The empirical analysis shows that correlations among international equity markets increase significantly when returns are low.