The road less travelled: Institutional aspects of data and their influence on empirical estimates with an application to tests of forward rate unbiasedness

The difference between data used in empirical studies and that envisioned in the theory can influence empirical estimates, sometimes by enough that the direction of future research is altered. This paper illustrates how, in the context of tests of forward rate unbiasedness, a lack of articulation about the mechanics underlying foreign exchange market speculation and about the institutional aspects governing spot and forward rate contracts leads to sampling of the data in a way that is not consistent with the theory. We find that the sampling problems account for some, but not all of the bias in the coefficient on the forward premium.