On unit roots and real exchange rates: Empirical evidence and Monte Carlo analysis

This paper presents some empirical evidence that real exchange rate series contain a unit root in their time series representation, which cancels out on first differencing, using the augmented Dickey-Fuller test. The power of this test is also tested using Monte Carlo methods and it is found to be quite powerful against a range of stationary local alternatives. The findings imply the absence of any tendency of the nominal exchange rate to converge on purchasing power parity, even in the long run.