Where the Wild Things Are: Measuring Systemic Risk through Investor Sentiment

In this paper, I develop a systemic risk measure derived from investor sentiment that has predictive power over future economic activity and market returns. Unlike existing measures, it is not focused on flagging investors' heightened awareness of risk at the end of a boom episode but rather on capturing shifts in their trading behavior at the beginning of the episode. The method allows investors and regulators to observe industries in which risks could be building and provides regulators some lead time in deploying their macroprudential tools.