Rethinking an old empirical puzzle: econometric evidence on the forward discount anomaly

摘要: Using both semiparametric and parametric estimation methods, this paper corroborates earlier findings of fractionally integrated behaviour in the forward premium. Two new explanations are also proposed to help reconcile conflicting empirical evidence on time series properties Traditional regression approaches used test rate unbiasedness hypothesis then evaluated, including levels, returns (Fama's, 1984, regression), error-correction format. Interesting statistical and/or interpretive implications found all three cases. For example, predictions appropriate nonstandard limit theory consistent with many standard results reported from Fama's regression, commonly occurring, yet puzzling negative correlations between spot It is suggested that principal failure unbiasedness, may be due instead difference persistence these two series. Copyright © 2001 John Wiley & Sons, Ltd.