Modelling volatility persistence and asymmetry of Naira-Dollar exchange rate

This paper modelled the volatility persistence and asymmetry of naira-dollar exchange rate in interbank and Bureau de Change (BDC) using monthly data between January 2004 and November 2017. The study employed Generalized Autoregressive Conditional Heteroscedasticity [GARCH (1,1)], Thresh- old GARCH [TGARCH (1,1)] and Exponential GARCH [EGARCH (1,1)]. The results of Bai-Perron (2003) structural break identified two significant breaks in each market. Interestingly, the breaks, particularly for the interbank exchange rate(December, 2014 and January, 2015), seem to have coincided with the period of depreciation in the country's exchange rate which could be linked to the precipitous movement in the international crude oil prices. The findings showed that persistence is generally explosive in the BDC market as compared to interbank market where the persistence was high but not explosive especially under asymmetric models. Based on our model selection criteria, the symmetric GARCH model, appears to be better than the asymmetric ones in dealing with exchange rate volatility in the interbank market while asymmetric GARCH, especially TGARCH, seems to be better in the case of BDC market. By implication, it is important that the monetary authority considers the developments and reactions of the markets to news most especially the BDC, when designing appropriate exchange rate policies for the country.