Innovative Statistical Modelling of Exchange Rates and Agricultural Sector Performance in Nigeria: The Markov-Switching Vector Autoregressive Modelling

Nigeria has increasingly been interconnected in the global economy. Consequent upon this, understanding the dynamic interdependence between exchange rates fluctuations and the real sector performance in Nigeria is necessary for researchers, economic managers and policymakers. This study on innovative statistical modelling employed the Markov Switching Vector Autoregression (MS-VAR) to explore the dynamic relationship between exchange rates and agricultural performance, capture the nonlinear interactions, regime shift, duration of stay in a regime, and the probability of transition from a regime. Quarterly data on exchange rates and agricultural sector gross domestic product spanning from 2000 to 2024, where sourced from the statistical bulletin of the Central Bank of Nigeria. The result of the statistical analysis showed that exchange rate and agricultural performance where stationary at the first difference, there was no cointegration, the result identified 2 regimes, and lag selection criteria showed 3 lags while the model selection based on information criteria selected MSI(2)-VAR(3). The result showed varying significant but negative effects of exchange rate on agricultural performance in regime 1 and 2. The result also showed 55.8% probability of staying in regime 1 for 7 months, and an 88.2% probability of staying in regime 2 for 17 months. Both exchange rates and the agricultural sector were self-explained, however the forecast variance indicated that agricultural performance was more exogenous. The study recommended policies that will reduce exchange rates as it showed inverse effects on agricultural performance. Also, regime possibility test should be part of pre-diagnostic tests.