A1291
Title: Exact single- and multi-currency tests of forward rate unbiasedness using Fama regressions
Authors: Hsuan Fu - Laval University (Canada)
Richard Luger - Laval University (Canada) [presenting]
Abstract: The aim is to develop exact simulation-based tests of the forward rate unbiasedness hypothesis (FRUH) at both the single- and multi-currency levels using the classical Fama regression. The procedures deliver finite-sample valid inference and are invariant to conditional heteroskedasticity, feedback from errors to future regressors, unequal sample lengths across currencies, and stretches of missing or structurally uninformative data. Within this framework, an exact global test is proposed together with single-step and step-down multiplicity adjustments that ensure strong familywise error rate control when identifying currencies or parameters that deviate from the hypothesis. The methods are illustrated through a simulation study and an empirical application to an unbalanced panel of developed and emerging market currencies. Multiplicity-adjusted p-values reveal that rejections of the FRUH occur exclusively among emerging market currencies.
