Tail dependence structures between economic policy uncertainty and foreign exchange markets: Nonparametric quantiles methods

Abstract

This study examines the extreme dependence and nonlinear causality between economic policy uncertainty (EPU) and major real foreign exchange markets (FER) in Australia, Canada, China, the E.U., Japan, Mexico, the U.K., and the U.S. For a deepen analysis, we also explore the financial uncertainty (FU)-FER nexus. To do this, we used both the Quantile-on-Quantile (QQ) approach and the nonparametric causality-in-quantiles tests. Using the QQ method, the results show negative average and extreme dependence between EPU and FERs. Moreover, the structure of dependence between the considered variables is found to be asymmetric across the quantiles. By applying the nonparametric causality-in-quantile tests, we found a weak evidence of causality-in-mean (at middle quantiles) and a strong evidence of causality-in-variance (for almost all quantiles) from both local and U.S. financial and EPU to FERs. Finally, the linkages between EPU and FERs intensified during our analysis of the 2008–2009 global financial crisis (GFC). These results have important implications for currency traders and monetary policy.