Learning Rational Expectations

This paper explores the possibilities for global stability of rational expectations under boundedly rational learning. The authors focus on a specific asymmetric example of L. Blume and D. Easley's (1982) environment, based on R. Radner (1979). Their main results point out an intriguing relationship between stability in learning and classical conditions for uniqueness and stability of Walrasian equilibrium. In particular, the authors find that gross substitutability, suitably extended to incorporate the informational role of prices, is sufficient to ensure global stability of revealing rational expectations equilibrium under learning. Their analysis is significantly aided by a technique developed here, which allows convenient graphical treatment of the issues surrounding rational expectations. Copyright 1990 by The London School of Economics and Political Science

Econometrics · Economics · Equilibrium point · Focus (optics) · Machine learning · Mathematical economics · Microeconomics · Point (geometry) · Rational expectations · Stability (learning theory) · Uniqueness · Advanced Topology and Set Theory · Computer Science · Economic theories and models · Economic Theory and Institutions · Mathematics