Modeling nonlinear Granger causality between the oil price and U.S. dollar: A wavelet based approach

In this paper, we use a wavelet approach to study the linear and nonlinear Granger causality between the real oil price and the real effective U.S. Dollar exchange rate. Instead of analyzing the time series at their original level, as it is usually done, we first decompose the two macroeconomic variables at various scales of resolution using wavelet decomposition and then we study the relationships among the decomposed series on a scale by scale basis. A major finding of this paper is that the linear and nonlinear causal relationships between the real oil price and the real effective U.S. Dollar exchange rate vary over frequency bands as it depends on the time scales. Indeed, there is a strong bidirectional causal relationship between the real oil price and the real dollar exchange rate for large time horizons, i.e. corresponding to fundamentalist traders, especially fund managers and institutional investors. But, for the first frequency band which corresponds to a class of traders whom investment horizon is about 3-months and whom trading is principally speculative (noise traders), the causality runs only from the real oil prices to real effective U.S dollar exchange rate.

► A nonlinear Granger causality study is carried out between oil price and U.S. dollar. ► Causality is explored among time series of Oil and dollar decomposed by wavelets at various scales. ► The link is very complex: For high frequency bands, the oil causes the dollar, at medium and low frequency bands, a strong bidirectional causality is found.