The proponents of floating exchange rates before 1973 did not promise that exchange rates would necessarily be stable under such a system, but only that they would be as stable as the underlying macroeconomic fundamentals.' Nevertheless, the widespread feeling is that exchange rates have turned out to be more volatile than necessary. Many practitioners believe that exchange rates are driven by psychological factors and other irrelevant market dynamics, rather than by economic fundamentals. Support seems to have grown in the 1980s for "target-zone" proposals, or some other sort of government action to stabilize exchange rates.2
The Meaning of "Excessive Variability"Economists have understood for some time that under conditions of high international capital mobility, currency values will move sharply and unexpectedly in response to new information. Even so, actual movements of exchange rates have been puzzling in two major respects. First, the proportion of exchange-rate changes that we are able to predict seems to be not just low, but zero. According to rational expectations theory we should be able to use our models to predict that proportion of exchangerate changes that is correctly predicted by exchange market participants. Yet neither models based on economic fundamentals, nor simple timeseries models, nor the forecasts of market participants as reflected in the forward discount or in survey data, seem able to predict better than the lagged spot rate. Also, the proportion of exchange-rate movements that