The Rational Beliefs Theory is founded on the observation that in most situations the same data
may be rationally explained (in terms of a statistical model) in many different ways. This is obvious
in a world where there is a limited number of observations. The novel idea of the Rational Beliefs
Theory is to use a technical device (namely the concept of a non-stationary but stable measure) to
formulate rational differences of opinion also in a world with «unlimited» data on the past performance
of the economy. In this article we provide a non-technical introduction to the Theory of Rational
Beliefs as well as a brief overview of the papers that have been written using this theory.
This study examines what role the concept of endogenous uncertainty can have in explaining a
phenomenon of international financial markets, the forward discount bias. The forward discount bias
puzzle is unexplained by models assuming economic agents have full knowledge of the structure of
the economy. This puzzle is explained here by constructing an overlapping generations model of a
Rational Belief Equilibrium in a non-stationary environment where agents do not know the true
probability distribution underlying asset price movements. The agents’ beliefs are rational in the
sense that their predictions of long term averages of economic variables match historical data. The
equilibrium of the model is solved numerically using the nonlinear equation solver TENSOLVE,
which is well-suited to models containing multiple financial assets. The results confirm the underlying
hypothesis that propagation of the beliefs of economic agents can play an important role in understanding
international financial markets
Is it theoretically legitimate for an investor to attempt to outperform the market index? Or is the
investor simply deluding himself? The purpose of this paper is to apply the new Theory of Rational
Beliefs (RB) to demonstrate the following result: Even when all agents have symmetric information,
there exist three canonical strategies in an RB environment that make it theoretically legitimate to
try to outperform the market. More specifically, while the three strategies that we will identify are
not directly implied by RB theory which is a positive theory, the strategies are logically compatible
with RB theory when interpreted in a normative manner from the standpoint of an individual agent
attempting to exploit the existence of structural change in the environment. This is what we mean
by the claim that the proposed strategies are «theoretically legitimate». Moreover, we propose that
these three strategies are canonical in the sense that all theoretically legitimate strategies for outperforming
the market (including all successful hedge fund strategies that we know of) are combinations
of these three basic strategies.
This paper presents simple models with competitive markets, fully flexible prices and fully observable
macro-economic variables. Agents hold heterogeneous beliefs because the exogenous
shocks are stable but non-stationary and the true law of motion is unknown to agents. We show that
the diversity of beliefs enables monetary policy to have real effects and the fluctuations in beliefs
cause excess volatilities in real economy. By log-linearization we derive a version of the aggregate
supply curve which relates aggregate employment to the discrepancies between the market forecast
of inflation rate and the stationary forecast of it. The computational results demonstrate that monetary
policy rules can dramatically reduce consumption volatility and there is a tradeoff between stability
of consumption and inflation. The results are thus comparable with Kurz - Jin - Motolese
(2003).
This paper studies how communication amongst agents influences the equilibrium of a financial
economy. We set up a standard overlapping generations (OLG) model with assets, while allowing
for heterogeneous beliefs. The paper explicitly describes how communication generates correlation
of beliefs, and show that communication can be embedded in the models of rational beliefs that do
not model communication explicitly a priori. We confine our attention to a Markovian economy, in
which the beliefs of the agents are all Markovian. Simulation results are provided to examine the effects
of communication, while classifying the beliefs in accord to the reactions to communication.
By means of an example, it is demonstrated how rational beliefs may be incorporated into a
Bayesian learning model. The empirical distribution is learned over time and the subjective rational
belief is modified accordingly.
Fascicolo digitale
|
