This paper examines the forward discount anomaly, i.e. the fact that the fo
rward exchange rate is a biased predictor of the future spot rate. We run a
series of rolling regressions which we use to predict the value of the fut
ure spot rate based upon this bias. We show that the average return from an
investment strategy based on the bias in forward exchange rates is in many
cases insignificantly different from zero. In other cases, however, the re
turn is significantly positive. Hence the in-sample bias does not necessari
ly lead to a money-making strategy for all currencies. (C) 1998 Elsevier Sc
ience S.A. All rights reserved.