Exchange rates and sovereign risk : a nonlinear approach based on local Gaussian correlations

We empirically assess the interlinkages between sovereign risk, measured in terms of CDS spreads, and exchange rates for a sample of emerging markets. Our period of analysis includes periods of severe stress, such as the Global Financial Crisis, the COVID-19 pandemic and the Ukrainian War. Using the most recent developments in local Gaussian partial correlation analysis and the associated nonlinear Granger causality tests, we are able to uncover linkages between assets across different segments of their joint distributions. Disentangling the effect of global factors, we show that the information on sovereign risk of other emerging economies is more relevant for the sovereign risk-exchange rate relationship than the state of developed markets risk for all countries in our sample and for all segments of the assets distribution. The same considerations apply for the movements of the US dollar relative to other currencies, where knowledge on movements of emerging currencies is of particular interest. Nonlinear Granger causality tests show bi-directional causality for most countries, confirming the importance of multiple transmission channels. Taken together, our results are of interest for international investors and policymakers, showing all interlinkages between sovereign risk and exchange rates across their entire distribution.