Theory of rational expectation assumes or argues that:
a) Prediction errors are always equal to zero.
b) Errors in prediction (or expectations) of some variable could be on average different from zero only if there is some relevant economic theory that explains developments of this variable.
c) Publicly available recommendations of analysts could be very good guideline for investments on the capital markets (in the case when these markets are efficient).
d) Errors in prediction (or expectations) cannot be correlated with any other information available at the time of construction of the prediction.
e) For any two different future moments the errors in prediction of some variable in those periods positively correlated (if the shock appears in one moment there is positive probability that it appears also in the second).