CMStatistics 2020: Start Registration
View Submission - CFE
A1091
Title: Multiple testing of the forward rate unbiasedness hypothesis across currencies Authors:  Hsuan Fu - Laval University (Canada)
Richard Luger - Laval University (Canada) [presenting]
Abstract: Distribution-free procedures are developed to test the forward rate unbiasedness hypothesis (FRUH) across multiple currencies, jointly. These tests directly assess whether forward exchange rates provide unbiased predictions of future spot exchange rates, in levels, as predicted by the FRUH under rational expectations and risk neutrality. The approach proceeds with test statistics for individual currency FRUH assessments. It then uses Monte Carlo resampling techniques to combine the marginal $p$-values in a way that controls the joint significance level. Our framework allows for missing data and the presence of multivariate GARCH-type effects in the spot and forward rates. The usefulness of the new procedures is illustrated with a simulation study and with an assessment of the FRUH across 13 currencies that exist over differing time periods. We find support for the joint FRUH.