Long Memory in Foreign-Exchange Rates

Using the Geweke-Porter-Hudak test, the author finds evidence of long memory in exchange-rate data. This implies t hat the empirical evidence of unit roots in exchange rates may not be robust to long-memory alternatives. Fractionally integrated autoregressive moving average models are estimated by both the time-domain exact maximum likelihood method and the frequency-domain approximate maximum likelihood method. Impulse-response functions and forecasts based.on these estimated fractionally integrated autoregressive moving average models are evaluated to gain insight into the long-memory characteristics of exchange rates. Some tentative explanations of the long memory found in the exchange rate s are discussed.