Oxford Economists have been dismayed at the failure of extant models to explain exchange-rate movements. These models
encounter two basic problems. First, many models concentrate on modelling short-run movements in exchange rates.
In the short run, exchange rates are determined largely by speculative capital flows, depending upon expectations. It is
very difficult to claim that these expectations are based upon a wide set off undamentals. Since the ‘fundamentals’
evolve gradually, short-run movements in exchange rates are largely noise. Second, the real fundamentals are generally
ignored by assuming purchasing-power parity, that the mean and variance ofthe real exchange rate are invariant over
time, and that the real exchange rate converges relatively rapidly to the unchanging mean. That is, it is assumed that the
real exchange rate is stationary. Very little attempt has been made to explain what economic forces determine the
mean. Since the period of floating, it is apparent that the real exchange rates ofthe major countries are not stationary in
the sense defined.
The short-run models ofm onetary dynamics with rational expectations have given way to the newer representative
agent intertemporal optimization models. The newer models have generally not been subject to empirical verification.
In the few cases where they have been, they are inconsistent with the evidence. The reasons are that these models make
highly restrictive assumptions and their crucial variables are not objectively measurable.