Getting Paranoid When Disasters Strike: A Learning-Based Explanation for Asset Pricing Puzzles
62 Pages Posted: 16 Mar 2012
Date Written: November 15, 2011
This article studies the pricing implications of learning about arrivals of economic disasters and the subsequent recoveries. We model a disaster as a separate phase, and transitions between the disaster and the normal phase introduce structual changes to the consumption process which triggers jumps in the equity market, a concern for a representative agent with recursive utillity. Learning creates a positive covariance between the expected arrivals of future events (either disaster or recovery) and their actual realizations. In particular, the agent becomes more pessimistic (optimistic) the moment a disaster (recovery) occurs. The positive covariance generates the extra learning-induced compensation for jump risks which proves helpful to understanding various asset return puzzles.
Suggested Citation: Suggested Citation