A Proposal for a Selection Criterion in a Class of Dynamic Rational Expectations Models with Multiple Equilibria

The paper argues that multiple equilibria-whether non-stationary or stationary- are a generic property of dynamic rational expectations models. In light of this, this paper proposes a selection criterion for choosing between these multiple equilibria in an important class of dynamic rational expectations models. The criterion is based on the idea that agents can be assumed to coordinate their beliefs around the limit of a finite-horizon equilibrium. For three examples examined, all of which can have multiple stationary, i.e., non-explosive, rational expectations equilibria, there is, among the multiple equilibria of an infinite-horizon model, only one that is the limit of a finite-horizon model.