Budget Digital Twin e a inovação na previsão das receitas públicas

This article presents an analytical model for municipal public revenue forecasting, aimed at enhancing public managers’ decision-making through the production of structured information capable of making the budget cycle more responsive and efficient. The proposed approach is grounded in the use of historical revenue series and probabilistic simulations, employing the Monte Carlo method to identify the statistically most likely revenue trajectory for a given fiscal year, as well as to define revenue realization bands. This framework is conceived as the first stage in the development of a Budget Digital Twin (BDT). The model is applied to six municipalities with different population sizes, analyzing the relationship between forecasted revenues, realized revenues, and projections for the subsequent fiscal year, compared with the actual outcomes observed. The results indicate that improving the revenue forecasting stage significantly contributes to reducing systematic deviations between budget planning and execution, providing more realistic inputs for fiscal decision-making. The study concludes that data-driven and probabilistic analytical approaches strengthen fiscal governance and expand the preventive role and concurrent oversight capacity of Audit Institutions, while also establishing methodological foundations for future extensions of the model to other phases of the budget cycle.