An application of Hamilton Model of Switching with Autoregressive Dynamics on Exchange Rate Movement in Nigeria

Abstract

This study looks at the relationships existing between the exchange rate and two important determinants such as crude oil price and foreign external reserves in Nigeria between 2006 and 2022. The exchange rate movement has been a major economic growth driver in many countries. In Nigeria, various governments through the Central Bank of Nigeria, in a bid to sustain a stable exchange rate regime came up with various policies. Our objective is to develop a powerful predictive model using the Time-invariant Hamilton model of switching with autoregressive dynamic techniques to generate probabilities of transiting from one state to another. Specifically, a two-state Time-invariant Markov-Switching model was estimated under the assumption that the errors are serially correlated with order four. In our specification, the economy is in regimes 1 and 2 whenever the exchange rate drops and rises respectively and the transition between the two regimes is modelled as the solution of the first-order Markov process. If it is in the “rise” state, this signifies depreciation in Naira but if in the “drop” state, then the currency is said to appreciate against the USDollar. The regime-specific, common and transition parameters were estimated by maximizing iteratively the normal mixture log-likelihood function. The analysis of the constant transition between regimes suggests that the exchange rate is more likely to transit from a “drop” state to a “rise” state. The likelihood that the exchange rate remains in either a “rise” or “drop” regime of origin is high. Generally, the exchange rate is expected to spend more time in the “rise” state than in the “drop” state.