Rational Diverse Beliefs and Economic Volatility

Abstract:

This is a chapter in a new series of Handbook in Financial Economics entitled “ Dynamics and Evolution of Financial Markets.” It explores works which contribute to explaining market dynamics and volatility by employing models of rational but diverse beliefs. This excludes models of Behavioral Economics. The first part gives a complete exposition of Noisy Rational Expectations (i.e. REE) Asset Pricing Theory where diversity of beliefs arises from diverse private information. Examination of these models enable us to evaluate the assumptions made and the ability of the models to explain the data about market dynamics. The general conclusion is that, although Noisy REE asset pricing theories have many attractive features, they do not offer a satisfactory paradigm for market dynamics. Indeed, for most models increased amount of idiosyncratic private information leads to a reduction in volatility.

We next turn to models of diverse beliefs without private information. We explore work on Rational Belief Equilibria (i.e. RBE) starting with modeling beliefs. Agents do not know the structure of the dynamically complex economy but have data over a long time to enable an econometric computation of the empirical moments, on the basis of which they compute a commonly known empirical probability on sequences. A belief is then a model of deviations from the empirical probability. It is shown that to be rational a belief cannot be constant, creating an endogenous mechanism for market dynamics. For the development of an equilibrium theory, beliefs are treated as a vector of state variables and market belief is the distribution of individual beliefs. Market belief is an observable for which ample data is available and we review both data sources and applications which use such data. Much of the material reviewed covers the applications of this theory to the study of market dynamics. We trace the implication of one specific model to the explanation of a long list of stylized facts known about market volatility via simulations. It is seen that the results demonstrate that the model�s predictions are compatible with the available data. It is then argued that these results offer an unified and intuitive explanation for many diverse phenomena of market dynamics such as excess asset price volatility, the Equity Premium Puzzle, the predictability of asset returns and the cause of stochastic volatility.