The rational expectations hypothesis presupposes — basically for reasons of consistency — that agents have complete knowledge of all of the relevant probability distribution functions. When trying to incorporate learning in these models — trying to take the heat of some of the criticism launched against it up to date — it is always a […]