Nonlinearities in Emerging Currency Markets Under Uncertainty

Abstract This study provides a comprehensive assessment of the asymmetric and nonlinear impacts that major sources of economic and financial uncertainty exert on exchange rates in 23 emerging and frontier market currencies. Through a flexible modeling approach, we uncover critical threshold effects and complex transmission mechanisms underlying the uncertainty-exchange rate nexus. Our results indicate that traditional linear models substantially understate these dynamics. We make three key contributions. First, we model nonlinearities across three distinct data transformations (level, logarithmic, and return series of exchange rates), offering a nuanced understanding of exchange rate behavior at different scales – capturing long-term trends, short-term fluctuations, and volatility dynamics. Second, we introduce a novel data-driven calibration of B-spline generalized additive models (GAMs), which optimizes the number and placement of knots based on information criteria and cross-validation. This approach yields a parsimonious yet highly adaptive fit that quantifies the extent of nonlinearity with unprecedented granularity. Third, using an extensive daily dataset from 2011 to 2023 that spans multiple global crises, we deliver generalizable insights into the pervasive asymmetries and nonlinearities governing emerging-market forex movements. The evidence highlights that uncertainty stemming from equity markets, commodities, and interest rates has markedly nonlinear effects on currencies – effects that intensify beyond certain uncertainty thresholds. Our findings guide policymakers in fortifying resilience against disruptive capital flows and exchange rate pressures, and they advance the literature on uncertainty and exchange rates by revealing hidden nonlinear patterns that linear models would overlook.