Can The Markov Switching Model with Time Varying Transition Probabilities Forecast Exchange Rates?

We use Lee's (1991) Markov switching model with time varying transition probabilities (the TVTP Markov model) to analyze the behavior of the U.S. dollar/ British Pound exchange rate. We employ the magnitude of the deviation of the exchange rate from a monetary equilibrium value as the economic fundamental with which the transition probabilities vary. The empirical results we obtained from this paper are the following: first, the deviation of the exchange rate from a monetary equilibrium value can affect the transition probabilities of the exchange rate from one state to another state. When the exchange rate is overvalued (undervalued) relative to the monetary equilibrium value, the probability of the exchange rate to appreciating (depreciating) will be low (high). Second, the TVTP Markov model can identify both the appreciation state and the depreciation state of the exchange rate better than the Markov model with fixed transition probabilities (the FTP Markav model). Third, the forecasts of the TVTP Markov model are superior at predicting the direction of change of the exchange rate to those of the FTP Markov model.