Abstract
The paper adopts a choice-theoretic, information-oriented approach to the issue of stationarity of real interest rates. It is shown that a constant real rate of interest, even for short run and within the context of a simple two-market framework, requires overly demanding assumptions which are unlikely to be satisfied if efficient market hypothesis is explicitly considered. Such a model which indirectly supports the short-run variability of real interest rates in response to random information signals is tested empirically by utilizing multiple time series models for the 1959–87 observation period. The empirical results suggest a favourable interpretation of the model.