Bayesian Pricing of News

Abstract

This paper characterizes the equilibrium stock price reaction to arbitrarily distributed signals when the prior distribution of the payoff is normal and the utility is exponential. This stock price reaction is shown to be proportional to the Fisher score of the news calculated under a risk-neutral probability measure. As an application of our analysis, we (i) characterize the stock price reaction to news whose arrival is content-dependent, (ii) develop a model of "agenda-setting" disclosures, and (iii) construct an equilibrium in a voluntary disclosure model with multidimensional information and risk averse investors.