Forecasting Value Added Tax Revenue in Ghana

Abstract

Governments need accurate tax revenue forecast figures for good economic planning but there seems to be no consensus on which method is the most suitable to deliver reliable results leading to differences in the choice of technique from one country to another. This study therefore forecasts Ghana’s Value Added Tax (VAT) Revenue by comparing two methods, ARIMA with Intervention and Holt linear trend methods to establish the one with more precise predictive powers for VAT Revenue. Monthly VAT revenue data from the year 2002 to 2019 is used in the analysis. The findings show that ARIMA with Intervention method outperformed the Holt linear trend model in terms of accuracy and precision. A comparison of predicted results from the ARIMA with intervention model from 2017 to 2019 with Ghana Revenue Authority’s VAT revenue targets based on their in-house forecasting model for the same period reveals that the ARIMA with intervention approach performs better than the in-house forecasting model of the VAT authority. In this case, the study recommends the ARIMA with intervention method to the tax authority for consideration in its forecasting.