Some thoughts on rational expectations models, and alternate formulations

We present, in this paper, an alternative, optimization based formulation for "forward looking" models in economics, more commonly known as rational expectations models. For one basic scalar model, we study both finite and infinite horizon formulations under two different information patterns, and in each case we obtain explicit expressions for the unique solution without making any a priori assumptions on its structure. We then compare these results with other possible solutions obtainable using the traditional approach to rational expectations. The approach introduced here can handle higher order models as well as nonlinear ones, and also those where there is an additional exogeneous input controlled by a different set of agents. Some of these possible extensions are briefly discussed in the latter part of the paper; others are left as challenging but highly promising problems for future research.