Identifying Shocks to Systematic Risk in Times of Crisis

Abstract

We characterize how risk evolves during a crisis. Using high-frequency data, we find that the first two principal components (PCs) of the covariance matrix of global asset returns experience large, sudden, and temporary spikes coinciding with well-known crises – Covid-19 pandemic, Global Financial Crisis, and Brexit. Despite the origin of these crises being very different, the risk dynamics share remarkably common features: PC1 shocks come solely from asset volatility, while PC2 shocks come from changing loadings/composition, effectively making it a “crisis” factor. Using the exogenous nature of Covid-19, we provide novel identification of risk dynamics by linking these changes to news about the virus and epidemiological model forecast errors over time and across countries. We conclude with investment implications, where shocks to systematic risk sharply reduce diversification benefits and ex ante attempts to hedge it are futile, which may be a defining characteristic of a crisis – that it is unavoidable.

Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.