Evidence of time-varying conditional discrete jump dynamics in sub-Saharan African foreign exchange markets

Abstract

An Autoregressive Jump Intensity-GJRGARCH model is used to examine the time-varying conditional discrete jump dynamics in the foreign exchange markets of Ghana, Kenya, Nigeria and South Africa. The findings suggest that conditional discrete jump is time-varying, and time-varying conditional discrete jump is sensitive to past shocks for all the four countries’ foreign exchange markets. Time-varying conditional discrete jump sensitivity is persistent in all the four markets, and all four markets exhibit asymmetric time-varying conditional discrete jump volatility. We also find that all the foreign exchange markets exhibit asymmetry in volatility, the so-called leverage effects. The findings shed some light on the volatility in these markets which are very relevant for hedging, portfolio allocation, pricing of currency derivatives and forecasting.