Uncovering bias in uncovered interest parity tests

Purpose This paper aims to re-examine the uncovered interest parity (UIP) hypothesis, which posits efficiency in forward foreign exchange and rational expectations. Testing these assumptions involves estimating parameters in a k-step-ahead forecasting model, where forecast errors are expected to be serially correlated up to lags k 1. When errors are correlated beyond these lags, OLS is no longer consistent unless the regressors are exogenous. Design/methodology/approach The authors extend the FGLS procedure developed in Perron and González-Coya (2022) to a setting in which lagged dependent variables are included as regressors. The authors thus provide a consistent and efficient framework to estimate the parameters of a general k-step-ahead linear forecasting equation. Following the work of Perron and Olivari (2023), the authors introduce an instrumental variable (IV)-based approach for this problem that requires pre-determined but not necessarily exogenous IVs for consistency. Findings The authors apply the authors’ FGLS procedures to the analysis of the two main specifications to test the UIP. Contrary to most empirical results available in the literature, in particular those based on some OLS regression or GMM, the authors’ robust and efficient procedure cannot reject the null hypothesis that the UIP holds. Originality/value Overall, this study’s results can be viewed as overturning the so-called forward discount anomaly. The methods proposed can also be applied to a wide variety of contexts.