MULTIVARIATE TIME SERIES MODELLING OF SELECT MACROECONOMIC INDICATORS IN NIGERIA

MULTIVARIATE TIME SERIES MODELLING OF SELECT MACROECONOMIC INDICATORS IN NIGERIA

ABSTRACT

This work, in trying to ascertain some of the major macroeconomic factors that would drive the nation’s economic growth, examined the causal relationship among all share index (ASI), nominal exchange rate, foreign direct investment, oil exports and GDP using quarterly data spanning from 1985Q1 to 2014Q4. In a bid to determine if long-run relationships existed among the variables, the Johansen cointegration approach was used after subjecting the series to the Augmented Dickey Fuller (ADF) test. The variables of interest were found to be stationary after first differencing and there was evidence of cointegration. Furthermore, the impulse response and Error Variance Decomposition analyses are used to examine the dynamic relations between stock indices and various macroeconomic variables. Lagrange Multiplier tests are run to ensure that the residuals from the chosen lag length are serially uncorrelated.