Nonlinear causality between crude oil price and exchange rate: A comparative study of China and India - A Reassessment

De Vita and Trachanas's (hereafter DV-T, 2016) paper published in (Energy Economics, Volume 56, May 2016, pages, 150-160) criticizes Bal and Rath's paper (Energy Economics, Volume 51, September 2015, pages, 149-156) (hereafter, BR, 2015) by undertaking a 'pure replication' and a 'reanalysis' using (BR, 2015) data set. The aim of this paper is to reassess (BR, 2015) by providing comments and additional evidence. We revisit (BR, 2015) with the aim of applying additional unit root, cointegration and nonlinear causality tests. The results derived from these supplementary tests clearly reveal that the oil price series is non-stationary at level. The bivariate noisy Mackey-Glass model proposed by Kyrtsou and Terraza (2003) reveals bi-directional non-linear causality exists between real oil price and exchange rate in case of China, whereas for India, only unidirectional nonlinear causality running from oil price to exchange rate.