When the European Central Bank raised rates in 2022–23, the five largest euro-area national central banks lost about €56 billion in 2024 alone, and most stopped paying their treasuries. This paper shows that the euro area already runs a hidden fiscal union of a limited kind. The Eurosystem's income rules divide the largest cost of quantitative easing — the interest on the reserves that financed it — among national treasuries by capital key. Three things stay national: the yields each central bank locked in on its own government's bonds, the credit risk on those bonds, and the rules by which it absorbs losses.The result is a paradox. Using the audited accounts of the Bundesbank, Banque de France, Banca d'Italia, Banco de España and De Nederlandsche Bank, we estimate that each percentage point on the deposit rate costs the system about €21 billion a year at 2025 balance sheets, and we divide that cost among treasuries. Per unit of GDP it falls most on Italy and Spain, because the capital key weights population. Yet France and Germany reported the largest losses, because their central banks hold bonds yielding about half a percent, while Italy's two-percent bonds return part of its risk premium through the pool. And the German and Dutch treasuries, which bear the least, are projected to receive nothing from their central banks until the late 2030s, while the Italian treasury, which bears the most, was paid every year and the Spanish treasury waited only two. Two of the three rankings bear no relation to debt/GDP; the third resembles it only because of the population weight in the capital key. We argue that no indicator of the EU fiscal framework measures any of them, and pre-register a test of that claim. The framework also clarifies part of a live dispute: the 2026 French proposal to convert the Banque de France's holdings of government bonds into perpetual bonds. Within the consolidated French public sector, the coupon decides only whether the treasury or the central bank records the cost. The consolidated saving, about €1.6–4.8 billion a year depending on the spread at which maturing bonds would otherwise be refinanced, is equally available by reinvesting them, and it would be lost if the proposal raised yields on market-held French debt by more than about 5½–16½ basis points. At a common coupon, the proposed euro-area-wide version would raise what the German and French treasuries pay their central banks and lower what the Italian treasury pays.
No takes yet. Share an insight, caveat, or question.
Thomas Filsecker (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: