AN AUTOREGRESSIVE INTEGRATED MOVING AVERAGE INTERVENTION MODEL OF 2016 BRAZILIAN REAL AND NIGERIAN NAIRA EXCHANGE RATES

This work is an attempt to fit an autoregressive integrated moving average (ARIMA) intervention model to daily Brazilian real (BRL) / Nigerian naira (NGN) due to the 2016 Nigerian economic recession. A time plot of the exchange rates shows the time of intervention as 22nd June 2016. The pre-intervention rates are non-seasonal at level but seasonal at first difference. The correlogram of the first differences shows a white noise structure, which means that all post-intervention forecasts are each equal to the last rate of the original series. The transfer function of the intervention model was determined and superimposition of the graphs of the intervention model and the post-intervention rates on one another shows a close agreement between them. This indicates model adequacy for the data. The model may be used by planners for planning.