Financial Crises, Risk Premia, and the Term Structure of Risky Assets∗

The literature on rare disasters shows that low probability events can explain high, time-varying risk premia. I find that large spikes in risk premia occur around financial crises but not around other disasters such as wars. A model with financial intermediaries generates endogenous financial crises that quantitatively match those in the data, while also replicating high equity risk premia and volatility. Compared to a standard disasters framework, the model makes additional empirical predictions which I confirm in the data. First, the equity of the intermediary sector strongly forecasts stock returns. Second, financial crises are temporary, which implies that the term structure of risky assets is downward sloping during financial crises when risk premia are concentrated in the near term. The model explains the level and slope of the term structure of risky assets including equities, corporate bonds, and VIX, both unconditionally and in a crisis. I then use the term structure of risky assets to infer the daily probability and persistence of a financial crisis in real time, providing a useful tool to analyze policy responses in a crisis. ∗Kellogg School of Management, Department of Finance. I would especially like to thank my committee, Arvind Krishnamurthy, Andrea Eisfeldt, Dimitris Papanikolaou, and Ravi Jagannathan. I would also like to thank Jason Chen, Anna Cieslak, Rob Dam, Jules van Binsbergen, Jonathan Parker, Victor Todorov, Torben Andersen, Nicola Fusari, Arik Bendor, Annette Vissing-Jorgensen, and seminar participants at Kellogg. I thank Barclays and Moody’s Analytics for providing data on credit spreads, returns, and default probabilities. All errors are mine. The most recent version of this paper can be found at www.kellogg.northwestern.edu/faculty/muir/Muir_JMP.pdf .