We investigate the behavior of the long-run U.S./U.K. real exchange rate from 1885 to 1995. Our long-run real exchange rate series is derived from an unobserved components model which divides the real exchange rate into permanent and transitory components. The transitory component is modeled as having variances which switch, according to a Markov-switching process, among low, medium and high variance states. The underlying assumptions of our time-series model are based on an economic theory in which the permanent component represents real influences, while the transitory component represents primarily short-run movements due to nominal exchange rate fluctuations. Because the model is difficult to estimate by standard methods, we describe how the method of Gibbs sampling can handle this model. We find that our long-run real exchange rate series moves similarly to other measures proposed in the literature based on economic models.
