Without searching through my old posts, I’m confident that I’ve already made this point many times in passing, but I just want to restate this point up front — highlighted and underscored. Any economic model must satisfy the following rational-expectations condition:
If the agents in the model expect the equilibrium outcome of the model (or, if there are multiple equilibrium outcomes, they all expect the same one of those equilibrium outcomes), that expected equilibrium outcome will be realized.
When the agents in an economic model all expect the equilibrium outcome of the model, the agents may be said to have rational expectations, and those rational expectations are self-fulfilling. Any economic model that lacks this contingent RATEX property is incoherent. But unless an economic model provides a theory of expectation formation whereby the agents all form correct expectations of the equilibrium outcome of the model, RATEX is a merely contingent, not an essential, property of the model.
Although an actual expectation-formation theory of rational expectations has never, to my knowledge, been derived from plausible assumptions, the RATEX assumption is disingenuously insisted upon as a property of rational decision-making implied by neoclassical theory. Such tyrannizing methodological intimidation is groundless and entails the reductio ad absurdum of the Milgrom and Stokey No-Trade Theorem.
Uneasy Money 
David,
Can I have rational expectations that are unfulfilled or irrational expectations that are fulfilled?
It seems that a model founded on rational expectations tries to incorporate the unmeasurable since,
1. Unfulfilled expectations are unlikely to show up in any market indicator
2. Fulfilled expectations are likely to be justified as rational post hoc
LikeLike
Your expectations can be rational and unfulfilled if new information arrives between their formation and the result. Are the irrational expectations that are fulfilled held by just one person or by everybody? One person’s rational expectations are not really rational in the sense that those expectations are not equilibrium expectations. Irrational expectations held by everyone would not be truly rational even though they are equilibrium, because it would just happen by luck not because everybody really understands how the world works. At least that’s how I understand what rational expectations means, but the concept is so problematic I’m not sure that there is any clear way of applying it to the real world.
LikeLike