Time-Varying Rare Disaster Risk and Stock Returns

Abstract

This study provides empirical support for recent theoretical models that allow for time-varying rare disaster risk. Using a unique database of 447 international political crises during the period 1918–2006, we create a crisis index that shows substantial variation over time. We show that changes in this crisis index, our proxy for changes in perceived disaster probability, have a statistically significant and economically large impact on both the mean and volatility of world stock market returns. World markets react more strongly when crises are more severe and when there is more at stake due to the involvement of major powers. Using predictive regressions, there is no significant relation between crisis risk and future market returns, however crisis risk is positively correlated with the earnings-price ratio and the dividend yield. Cross-sectional tests provide further support for the hypothesis that crisis risk is priced: US industries with higher crisis risk sensitivity generate a higher excess return. Our crisis measures also correlate negatively with U.S. consumer confidence and GDP-growth forecasts, both proxies for expected U.S. consumption growth.