Exchange rates and prices in purchasing power parity framework: Are bilateral real exchange rates stationary?

This study empirically examines the relationship between nominal exchange rate and prices as postulated by the Purchasing Power Parity (PPP) hypothesis. We have also estimated a panel error correction model to test for Granger causality in the presence of cointegration among the variables. The result of unit root test suggests that strict PPP does not hold. However, the results from panel cointegration test suggest that the nominal exchange rate and relative prices do move together in the long-run, which apparently provides some support to weak PPP. Furthermore, the result from fully modified ordinary least square estimation indicates that the failure of strict PPP is mainly driven by the rejection of the homogeneity restrictions imposed on the real exchange rates, since most of the coefficients are found to be statistically different from unity.