Expectation Formation Following Large Unexpected Shocks

Abstract

By matching a large database of individual forecaster data with the universe of sizable natural disasters across 54 countries, we identify a set of new stylized facts: (i) forecasters are persistently heterogeneous in how often they issue or revise a forecast; (ii) information rigidity declines significantly following large, unexpected natural disaster shocks; (iii) the response of forecast disagreement displays interesting patterns: attentive forecasters tend to move away from the previous consensus following a disaster while the opposite is true for inattentive forecasters. We develop a learning model that captures the two channels through which natural disaster shocks affect expectation formation: attention effect { the visibly large shocks induce immediate and synchronized updating of information for inattentive agents, and uncertainty effect { the occurrence of those shocks generates increased uncertainty among attentive agents.

Keywords: Expectation Formation, Heterogeneous Agents, Information Rigidity, Learning, Natural Disasters, Uncertainty Shock

JEL Classification: C32, C53, D83, D84, E17, E37