Equilibrium Stock Return Dynamics Under Alternative Rules of Learning About Hidden States

Abstract

We examine the properties of equilibrium stock returns in an incomplete information economy in which agents need to learn the hidden state of the endowment process. Both Bayesian and suboptimal learning rules are considered, including near-rational learning, conservatism, representativeness, optimism, and pessimism. We demonstrate that Bayesian learning performs reasonably well in producing realistic variation in the conditional equity risk premium, return volatility, and Sharpe ratio. Alternative learning behaviors alter significantly both the level and time-variation of the conditional moments of returns. However, when we allow the agents to be conscious of their learning mistakes and to price assets accordingly, the properties of equilibrium stock returns under alternative learning are virtually indistinguishable from those under Bayesian learning.