Vector Autoregressive Modelling of the Interaction among Macroeconomic Stability Indicators in Nigeria (1981-2016)

Abstract

The dynamic behaviour of macroeconomic stability indicators particularly their; evolution, interaction and interdependence, obviously cause shocks among themselves. This study is a multivariate time-series modelling and investigation of the interaction and pattern of causality among exchange rates, inflation rate, interest rates, and implicit price deflator in Nigeria using unrestricted Variance Autoregression (VAR). Quarterly data on the variables spanning the period from 1981 to 2016 were sourced from CBN Statistical bulletin and used for the study. The study used both descriptive and analytical design. The result of the inverse root of AR characteristic polynomial indicated that the VAR model was stable. The Trace Statistics and Max Eigen result showed no co-integrating relationship. The Schwarz Information Criterion showed a lag length of 2. The VAR estimates indicated that the exchange rate was significantly affected by its first lag and second lag, while inflation rates was significantly affected by its first lag. The Wald statistics showed that both lags of each variable were jointly significant in affecting itself. The impulse response showed that all variables were instantaneously affected by own shocks, however, it ruled out the response in exchange rate to contemporaneous shocks in inflation rate, interest rate and implicit price deflator. The variance decomposition further showed that at least 80% of the impulse response were from own shocks. It was consequently recommended that government should regulates interest rates and exchange rates.