Real Interest Rate Parity: Evidence from Industrialized Countries

Salah A. Nusair

This paper tests real interest parity (RIP) for a group of industrialized countries using quarterly data on long-term and short-term interest rates from 1957:1 to 2003:1. The paper looks at such issues as the lack of power in the standard unit root tests, spans of data, the base country, and the high volatility of exchange rates under the current float as possible reasons for the weak support for RIP. Overall, the standard ADF unit root test provides more supporting evidence in favor of RIP than the more powerful ADF-GLS test and the KPSS stationarity test, and the results do not seem to be sensitive to the choice of the base country. Lack of power seems to be an issue in short samples for the standard ADF test and the ADF-GLS test. The paper also investigates the behavior of real interest differentials (RIDs) across the Bretton Woods era and the current float. The results are consistent with the ‘neutrality proposition’, and indicate smaller RIDs post-Bretton Woods; thus, the claim that the current float caused RIDs to widen is not supported. Estimated speeds of convergence to RIP are between two to three quarters for both long-term and short-term RIDs and regardless of the base country.