A reconsideration of the failure of uncovered interest parity for the U.S. dollar

Abstract

We reexamine the time-series evidence on uncovered interest rate parity for the U.S. dollar versus major currencies at short-, medium- and long-horizons. The evidence that interest rate differentials predict foreign exchange returns is not stable over time and disappears altogether when interest rates are near the zero-lower bound. However, we find that year-on-year inflation rate differentials predict excess returns – when the U.S. y.o.y. inflation rate is relatively high, subsequent returns on U.S. deposits tend to be high. We interpret this as consistent with the hypothesis that markets underreact initially to predictable changes in future monetary policy. The predictive power of y.o.y. inflation begins in the mid-1980s when central banks began to target inflation consistently and continues in the post-ZLB period when interest rates lose their primacy as a policy instrument. We attempt to address some econometric problems that might bias the conventional Fama (1984) test.