Statistical Intervention Analysis of Nigerian Monthly Inflation
Ette Harrison Etuk
Department of Mathematics
Rivers State University, Port Harcourt
Email: etuk.ette@ust.edu.ng
ABSTRACT
Since 2016, Nigeria has been in economic recession. Coincidentally, its inflation rates have been rising too with time. The goal of government has been to maintain the inflation in a single digit range. From 2013 up to January 2016, this goal has been achieved according to the records of the Central Bank of Nigeria after which it has swung back to two digits until now. The index is rising on a monthly basis. It is known that recession could worsen inflation. This paper is an attempt to propose an intervention model to explain this relationship between the two variables: inflation and recession. Clearly the point of intervention is February 2016, the change in the mean level being observed to be statistically significant. The pre-intervention series is observed to be non-stationary and thereby differenced to make it stationary. The correlogram of this resultant series has a spike in the partial autocorrelation function at lag 3 and no spike at all in the autocorrelation function, suggesting an AR(3) model, which fitted is observed to be adequate, the residuals following a Gaussian distribution. The difference of the post-intervention observations and the AR(3) forecasts are analyzed for the intervention model. There is very close agreement between the observations and the intervention model forecasts. It is recommended that the Nigerian economy be upped with appropriate economic policies in order to ameliorate the inflation situation.
Keywords: ARIMA modelling, economic recession, intervention analysis, Nigerian Inflation
