Modelling the Inflation Rate in Sudan by a Seasonal ARIMA Model

Sudanese Monthly Inflation series is modelled by Seasonal Autoregressive Integrated Moving Average methodology. The realization analyzed spans from 2005 to 2015. The time plot shows a generally positive trend. An inspection of the series reveals a yearly seasonal pattern. Augmented Dickey Fuller test suggests that this original series is not stationary. A seasonal (i.e. twelve-monthly) differencing proves not to be enough to render the series stationary. A further non-seasonal differencing renders the series stationary. The autocorrelation structure of this resultant time series suggests some SARIMA models including those of orders: (1,1,0)x(1,1,1) 12 , (1,1,1)x(1,1,1) 12 and (0,1,1)x(1,1,1) 12 . Diagnostic checking procedures applied suggest the comparative adequacy of the SARIMA(1,1,0)x(1,1,1) 12 model. Forecasting and simulation of the series may therefore be based on it.