Testing for Nonlinear Granger Causality Between Bitcoin Market and Crude Oil Market

This paper investigates the causality between bitcoin market and WTI crude oil market through multi-scale analysis and causality testing. The complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) method is employed to decompose the two price series at different time-scales. In causality testing, a nonlinear Granger causality test is formulated to investigate the relationship among each pair of matched components. And we also divide the information components of different series into high-frequency components, low-frequency components and long-term trend according to the Fine-to-coarse reconstruction. In the end, a set of hypothetical scenarios are created and a statistical test for causality is performed.