Dynamic link between oil prices and exchange rates: A non-linear approach

Abstract

We explore the non-linear relationship between crude oil prices and exchange rates of major currencies from quantitative and structural perspectives, by utilizing bivariate normal mixture model. The correlation coefficients between oil prices and exchange rates demonstrate that their dependences typically start from 2004 then dynamically change over time. Then we investigate whether business cycle and oil price shocks as two possible exogenous factors affect the dependence structure of oil-FX linkage. We find significant structural heterogeneity during economic expansion while little evidence of heterogeneity in recession. This finding provides alternative interpretations for the enhanced dependence between oil prices and exchange rates and generates implications of financialization in commodity market. In terms of oil price shocks (Kilian, 2009), the normal mixture model captures significant heterogeneity, implying that unstable oil-FX dependence structure is frequently associated with oil aggregate demand shocks and oil-specific demand shocks. With an emphasis on the dynamic weights of two underlying states, we interestingly find that structural heterogeneities coincide with a variety of geopolitical and economic events. The application of normal mixture not only provide us knowledge of non-linear relationship between oil prices and exchange rates but also guidance for investors in risk management and portfolio diversification complementary to the traditional portfolio theory based on normal distribution.