ETS - ARIMA Intervention Modelling of Bangladesh Taka/Nigerian Naira Exchange Rates

Abstract

In real-world scenarios, numerous external events disrupt many time series, causing fluctuations in the series' mean level. When modeling such series using the traditional ARIMA model, this can result in distortions in the model's parameter estimations, the structure of the fitted model, and future value projections. Any unusual values in the series that might have arisen as a result of the special event could be adjusted using the Box-Tiao intervention modeling technique. This study investigates time series intervention modelling based on ETS and ARIMA models aimed at studying the response of the comparative value of the Bangladesh Taka to the Nigerian Naira due to the 2016 economic recession. The dataset for this study is the daily exchange rate of Bangladesh Taka to Nigerian Naira from January to December 2016. The BDT/NGN2016 exchange rates have been considered, with a step intervention being the introduction of the economic recession in June 2016. Results revealed an initial impact of 0.5217. The intervention caused a 68.49% depreciation in the value of the Naira exchanged with the Bangladesh Taka in the exchange rate market, with a decay rate of 0.6. The intervention effect was persistent, with a long-run effect of 1.2862. Hence, the intervention had a gradual start and a permanent effect.