On the asymmetric response of the exchange rate to shocks in the crude oil market

This paper aims to investigate the nonlinear dynamics in the effects of oil price shocks on the exchange rate for a sample from the Group of Twenty (G20) over the period 1994:1-2019:1.

Using monthly time series data covering the period1994:1-2019:1, the author first use the non-parametric triples test of Randles et al. (1980) to ascertain the existence of asymmetric properties in the sample of exchange rates. Then the author used the nonlinear ARDL cointegration approach developed by Shin et al. (2014) to examine the reaction of these exchange rates to the oil price shocks.

This study has identified significant evidence that the exchange rate is asymmetrically distributed, with the effect that high appreciation of the exchange rate is followed by slower depreciation. The NARDL results support such asymmetry even more strongly because in the test the exchange rate is shown to react differently in the long term to positive and negative shocks in oil prices. Another major finding was that the speed of adjustment differed over the sample, as the cumulative dynamic multipliers effect highlighted.

This change in direction and the employment of non-linear technique can be to obtain better insight into the model specification, which the author believes, will not only enhance the findings in the literature but also enhance forecasting and decision-making.

A practical implication of this change is the possibility that policymakers and participants concerned with exchange rate stability should intervene in the market to alleviate the unfavourable impact of oil price shocks on the exchange rate.

Addressing this nonlinear dynamic in the effects of oil price shocks on the exchange rate have at least the following two important reasons: asymmetry and regime change are types of nonlinearities that affect the market dynamics, especially, over marked sample period with such financial crises as the global financial crises of 2007, thereby violating the linear models. Adopting an asymmetric cointegration technique permits to incorporate cointegrated positive and negative components of the considered series.