ECONOMETRIC MODELING OF THE DYNAMICS OF VAT BUDGET REFUNDS

Abstract

This study examines the dynamics of budgetary value added tax (VAT) refunds in Ukraine and their relationship with key macroeconomic and fiscal indicators over the period 2018–2025. The purpose of the study is to identify the main factors influencing VAT refund volumes and to develop econometric models capable of explaining their behavior under different economic conditions. The research is based on official monthly and quarterly statistical data on VAT revenues, VAT on imports, VAT on domestically produced goods and services, and VAT refunds, adjusted for inflation to ensure comparability. The methodological framework includes correlation analysis, trend and seasonal decomposition of time series, and multifactor regression modeling. Econometric estimation was carried out using the LINEST function in Excel, t- and F-tests, and the Durbin–Watson test for autocorrelation. The results reveal pronounced structural breaks in VAT refund dynamics. During the pre-war period of 2018–2021, refund volumes were relatively stable with noticeable seasonal peaks at the end of fiscal periods. In 2022, the outbreak of the full-scale war caused a sharp collapse in VAT refunds, followed by an uneven recovery in 2023–2025. Trend-seasonal models incorporating dummy variables that capture wartime shocks and recovery phases explain more then 80% of the variation in VAT refunds. Correlation analysis indicates weak relationships between VAT refunds and total tax revenues or domestic VAT, while a moderate positive relationship is observed with VAT on imports. Traditional macroeconomic indicators such as GDP, exports, and imports do not show statistically significant direct effects in VAT refund models. At the same time, regression models that include the share of exports in GDP demonstrate high explanatory power, emphasizing the structural role of export orientation. Further improvement is achieved by adding the industrial production index, increasing the coefficient of determination. The findings confirm that VAT refunds are shaped by structural, seasonal, and extraordinary factors in addition to general macroeconomic conditions and provide a basis for scenario analysis and improved fiscal planning.