A Note on Exchange-Rate Expectations and Nominal Interest Differentials: A Test of the Fisher Hypothesis