Seasonal Vector Error Correction Modeling Of Gross Domestic Product And Money Supply

ABSTRACT

This study compared the performance of a vector error correction model with the seasonal vector error correction model when there is presence of seasonal unit root and seasonal co-integrations in the series. Quarterly data on GDP and Money supply from the year 1986 to 2016 were collected for the purpose of this research. The series after being subjected to Augmented Dickey-Fuller test for unit root, Canova-Hansen test and HEGY test for seasonal unit root as well as Engel-Granger test and Johansen test for non-seasonal co-integration and EGHL test for seasonal co-integration. The results were identified to have the presence of non-seasonal unit root, seasonal unit at zero frequency and biannual cycle, co-integration and seasonal co-integration. Finally, both models were fitted and their results were compared to determine the best model. The findings of this study showed that seasonal vector error correction model appeared to be better than the vector error correction model.