Vat Revenue Forecasting In Ghana

ABSTRACT

It is prudent for a government to budget approximately for its expenditure based on expected or an accurately forecasted revenue. This will reduce the recurring budget deficit of the economy and bring the fiscal profile of an economy as Ghana to an acceptable level. There seem to be no clear guidelines from literature as to which econometric method provides the most accurate fiscal variables forecasts, but rather each forecasting case is considered and evaluated separately for individual country case. This study after a thorough scrutiny of applicable econometric methods of VAT Revenue forecasting, settled on comparing ARIMA with Intervention Analysis method and Holt linear Trend method to ascertain which model gives a better prediction of VAT Revenue in Ghana in terms of precision and accuracy. Monthly VAT revenue data which spanned the period 2002 to 2017 was used in the analysis. The study further explored the effect of Real GDP on VAT Revenue and the correlation that exists between the two economic variables to know if Real GDP can be an appropriate proxy VAT base in Ghana using quarterly data from 2006 to 2017 for both variables. ARIMA with Intervention Analysis method of forecasting outperformed the Holt trend model in forecast accuracy and precision, thus was used to forecast monthly VAT revenues for the next 24 months. (i.e. 2018 and 2019). Also, a positive effect and a high correlation between VAT Revenue and Real GDP was realised, suggesting that Real GDP can be used as a proxy VAT base in Ghana. The study recommends, therefore, that ARIMA with intervention analysis model should be compared with the in-house model used at Ghana Revenue Authority for forecast accuracy and prediction and if it outperforms the in-house model, it should be adopted by the Ghana Revenue Authority in forecasting monthly VAT revenues for the purpose of drawing half year and full year government budgets so that the ministry of finance budgets approximately expenditure based on expected or forecasted revenue. Also, because of the high correlation between Real GDP and VAT Revenue, real GDP can be used as proxy VAT base in forecasting VAT revenues.