Non-Fundamental Volatility in Financial Markets

Abstract

I use coarse Bayesian updating to explain three stylized facts: trading anomalies around major stock market milestones, excess price volatility and trade volume, and heavy-tailed prices. When expectations are coarse and traders are heterogeneous, traders make heterogeneous mistakes. This disagreement generates substantial trade volume and I show that the ensuing price discovery process converges in distribution to an empirically relevant class of Levy processes. Notably, I obtain this result without any exogenous shocks or processes; the only exogenous random variable is a heterogeneity parameter drawn at time zero. I then establish a perfectly non-revealing equilibrium: because traders are different, their aggregated trades generate volatile prices, yet this volatility impedes traders from learning their differences.

Keywords: Non-fundamental volatility; Behavioral finance; Coarse Bayesian updating; Price barriers

JEL Classification: G14, G41, D84, D91