Change and Rationality in Macroeconomics and Finance Theory : A New Rational Expectations Hypothesis

We call attention to the class of models that serve as the foundation for the rational expectations hypothesis (REH). Models in this class rule out completely any structural change that cannot be fully anticipated with a probabilistic or other quantitative rule. REH models are abstractions of rational decision-making, but only in a hypothetical world in which participants can fully anticipate when and how they might revise their understanding of the process driving outcomes. We propose a new rational expectations hypothesis (NREH) as a way to represent rational decision-making in real-world markets. NREH builds on the insights of Muth (1961) and Lucas (1972, 2001) and imposes internal coherence between the economist's understanding of outcomes and that of the market. However, like Soros's (1987) conceptual framework, NREH models recognize that any quantitative understanding of the process driving outcomes is necessarily provisional, eventually becomes inadequate, and thus requires revision. Consequently, NREH does so in the context of models that are partly open to unanticipated structural change. NREH models accord participants' expectations an autonomous role in internally coherent models. They also incorporate REH's and behavioral economists' insights about the importance of fundamental and psychological considerations, without presuming that market participants are irrational