Too good to be true: A theory

Abstract

We use a Gaussian mixture prior with two clusters to explain market fears. We show that a surprisingly positive signal can shake investors’ confidence in their understanding of the market, and in the process, potentially lower their expectation of an asset’s value.