Are real interest rates cointegrated? Further evidence based on paneleconometric methods

We examine the validity of real interest parity as a long run condition for the G7 countries. If real interest parity holds, differences of real interest rates are stationary. This is investigated by the means of conventional and panel unit root tests, where heterogeneity and contemporaneous correlation across the panel members is taken into account. Performing ADF- and KPSS-style panel tests on ex post rates, the evidence suggests a mixture of stationary and nonstationary series. However strong linkages between individual real interest rates can be found in the European economies.