Working Paper: NBER ID: w22143
Authors: William N. Goetzmann; Dasol Kim; Robert J. Shiller
Abstract: Historical data suggest that the base rate for a severe, single-day stock market crash is relatively low. Surveys of individual and institutional investors, conducted regularly over a 26-year period in the United States, show that they assess the probability to be much higher. We examine factors influencing investor responses and test the role of media influence, finding evidence consistent with an availability bias. Adverse market events made salient by financial press are associated with higher subjective crash probabilities. Exogenous shocks related to earthquakes are also associated with higher probabilities. Finally, subjective crash probabilities are negatively associated with mutual fund flows.
Keywords: crash beliefs; investor surveys; media influence; availability bias
JEL Codes: E03; G00; G02; G11; G23
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| negative media sentiment (G41) | increased crash probability assessments (C29) |
| adverse market events highlighted by financial press (G14) | higher subjective crash probabilities among investors (G41) |
| recent rare events (e.g., earthquakes) (G14) | elevated crash probability estimates (C13) |
| media sentiment (D79) | investor beliefs about crash probabilities (G17) |
| negative media sentiment (G41) | reduced investment in equity funds (G23) |
| past market returns (G17) | media sentiment (D79) |
| media coverage temporal ordering (C41) | investor responses (G11) |