Talk:Rational expectations

From Wikipedia, the free encyclopedia

[edit] Is it plagirism?

This page was basically copied and pasted from its external link.

http://www.colombialink.com/01_INDEX/index_finanzas_eng/rational_expectations.html

Foraday 16:52, 26 March 2007 (UTC)

[edit] 'Equilibrium' can mean steady state, or can mean mutual consistency of decisions

I am removing (for the second time) a sequence of assertions by user jmccauley, who claims that rational expectations involves a misunderstanding of the concept of equilibrium. The text I removed is shown below:

The basis for the rational expectations model is based on a mathematical inconsistency (in Muth, 1961): a time varying price is derived from the assumption of market equilibrium. Equilibrium means time translationalinvariance, simple averages cannot change with time. Muth's (and Lucas's and Sargent's) violation of the definition of equilibrium is common in the economics and finance literature. For the correct definition of equilibrium (dynamic or statistical) see texts on ordinary diffeential equations, dynamical systems, statistical physics, or stochastic processes. Lucas tried to maintain the essence of neo-classical ideology by attempting to define away nonstationary processes, but this resulted only in self-organized confusion, not understanding of markets. Finance markets (and presumably all other markets) are nonstationary, are far from equilibrium, and all attempts by economists to understand and explain markets have failed.

References:

1. J.L. McCauley, K.E. Bassler, and G.H. Gunaratne, Martingales, Detranding Data, and the Efficient Market Hypothesis, Physica A37, 202, 2008.

2. K.E. Bassler, J. L. McCauley, & G.H. Gunaratne, Nonstationary Increments, Scaling Distributions, and Variable Diffusion Processes in Financial Markets, PNAS 104, 17297, 23 Oct. 2007.

This text asserts that 'equilibrium' refers to a system which is not changing over time. Many years ago, this was indeed the usual meaning of equilibrium in economics. But today, most economists use the word 'equilibrium' in a different sense, essentially the same way it is used in game theory: mutual consistency between agents' decisions. Economists do not claim to use the word in the same way statistical physicists use it.

The whole point of the rational expectations revolution was that some earlier adaptive expectations models implied behavior that was far from optimal. Imposing rational expectations means that agents know how the economy behaves, and make their decisions optimally taking this knowledge into account. In other words, the expectations on which agents base their decisions are consistent with the actual behavior of the economy. That is equilibrium in (roughly speaking) the sense of Nash. It is NOT equilibrium in the sense of statistical physics: there is no assumption that the economy is behaving in a constant way over time. An economic situation that is unchanging over time is nowadays usually called a 'steady state', which is NOT a synonym for 'equilibrium' in the Nash sense of the word. --Rinconsoleao (talk) 14:42, 26 December 2007 (UTC)

[edit] Errors in the article on rational expectations

This article succeeds in getting across the notion that RE is controversial in economics, and communicates well some of its basic ideas. But there are a few errors that will mislead those encountering the concept for the first time via this entry.

First, it is not correct to say that RE lacks any microfoundation. The dominant strain of modelling at the moment adopts RE at the micro level, and then aggregates to derive the implications for the behaviour of macroeconomic variables.

Second, the criticism that economists do not deal scientifically with falsifications of RE is not fair as put. This may be true of some, but is certainly not a feature of mainstream RE economics. Modern quantitative macroeconomics, for example, is concerned entirely with figuring out the best response to observing that its models are false : figuring out the most likely causes of model failure. If there is a problem with economic modelling of this kind, it is that all economic models are by practical necessity false. The quest is for the most useful false model. Actually the use of rational expectations in macroeconomic models has been helpful because it generates very sharp predictions for the behaviour of macroeconomic models, and can thus be more clearly falsified. In mainstream macro debates, those who adopt non-RE models face the challenge that there is always a non-RE assumption that you can find to plug the gap between any model and the data: because there are infinitely many such non RE assumptions. Now it must certainly be true that RE does not hold in reality. But without sharp evidence from behavioural science, macroeconomists can get lost in what Chris Sims called a 'wildnerness' of unfalsifiable alternatives.

Third, the reference to the Lucas Critique is not well made here. The Lucas Critique would apply to non-RE models as much as RE models.

Fourth, the article does not explain that there is a very active and serious field of research in macroeconomics and finance exploring the implications of non RE models. Is this relevant to an article about RE? I'd argue it is, first, because this research field has inherited some of the spirit of early RE models: it tends to adopt as a standard the desire to make minimal departures from RE; and retains the assumption that agents are rational.

Fifth, the reference to credibility in the discussion of adaptive expectations is not helpful. Actually it was rational expectations that gave birth to the first coherent treatments of the issue of credibility - for which kydland and prescott got half of their nobel prize.

At some point this article deserves a re-write. —Preceding unsigned comment added by Eekonermissed (talk • contribs) 17:09, 12 April 2008 (UTC)