The European Central Bank published a Working Paper setting out a Bayesian mixed-frequency vector autoregression model for fiscal monitoring that aims to nowcast the annual government deficit-to-GDP ratio in real time and explain the drivers of its revisions. The approach combines monthly cash fiscal data with quarterly accrual-based fiscal statistics and other high-frequency macroeconomic and financial indicators to produce monthly density nowcasts, rather than the biannual point forecasts typically published by official institutions. Using 208 real-time monthly data vintages for Italy (with an evaluation sample covering 2010–2023), the paper reports that the model’s median nowcasts are similarly accurate and, in some months, more accurate than the European Commission’s forecasts, while also providing uncertainty bands and “news” decompositions that attribute month-to-month nowcast revisions to incoming releases. Cash revenue and expenditure releases are highlighted as important contributors to nowcast updates, and aggregated results indicate that revenue and expenditure dynamics generally drive movements in the deficit ratio more than changes in GDP and the deflator; performance issues in 2022 are linked to ex-post deficit revisions associated with the “Superbonus 110%”. The paper also illustrates scenario analysis, finding that a recession triggered by a monetary policy tightening generates a larger and more persistent rise in the deficit ratio than a “typical” recession of the same GDP magnitude, largely due to higher interest-related expenditure and the absence of stabilising rate cuts in the monetary-tightening scenario.