Bayesian Expectations and Strategic Complementarity: Implications for Macroeconomic Stability

Abstract

This paper examines the heterogeneous market in which economic agents of different information-processing abilities interact. In the theoretical framework, the market is composed of three different types of agents, “sophisticated” agents with rational expectations, “naive” agents with adaptive expectations, and Bayesian agents endowed with learning abilities. The behavior of these agents in the context of an important economic problem of nominal price adjustment after a fully anticipated one-time negative monetary shock is examined. If sophisticated agents with their perfect foresight find it profitable to imitate the biased behavior of naive agents, then the interaction of agents exhibits strategic complementarity. Thus the naive agents will have a disproportionately large effect on sluggish price adjustment towards equilibrium. However, the introduction of Bayesian agents with learning abilities into the market will have a compensatory effect by mitigating the price rigidity. Since Bayesian learning is allowed in heterogeneous market, Bayesian agents that first start as naive will undergo a learning process to become sophisticated after a certain period. In conclusion, the proportion of naive agents decreases in favor of sophisticated agents as depicted in the simulation model. As a result, the price adaptation towards equilibrium is accellerated.