France pays 23.5 bp more to borrow than Italy, with a far better rating. The OAT-Bund spread hit a 14-year high of 152 bp. The arithmetic behind a eurozone rescue, and why this one cannot be talked down the way 2012 was.
On the second of October, French ten-year paper yielded 4.989 % against Germany's 3.63 %. A spread of 152 basis points. The widest since 2011, the year the eurozone nearly came apart. That number has since come in to 140.9. The level matters less than what the level is made of.
In 2012, the spread on southern European debt was a panic. Capital was running, no one knew who held what, and Mario Draghi could say the words "whatever it takes" and mean that the price was wrong. The spread fell on the sentence alone. It fell because the fear was not backed by the arithmetic.
This one is backed by the arithmetic.
Start with the inversion, because everything else follows from it. As of midday on the fifth of October, France borrowed at 4.863 % for ten years. Italy borrowed at 4.628 %. France pays 23.5 basis points more to borrow than a country it outranks by three to five notches on every major rating. That inversion is without precedent in the euro's lifetime.
That has not happened before in the euro's lifetime. In 2011 it was Italy that needed rescuing; today the roles have reversed. In September 2022 the Italian spread over Germany touched 251 basis points. It has since fallen to 117.3. France went the other way: 55 basis points at the start of this year, 100 by 19 September, 120 on the thirtieth, 152 on the second of October.
The market is not confused about the ratings. It is saying something a rating cannot: the trajectory matters more than the level.
A rating describes where a sovereign has been. The deficit describes where it is going. On the trajectory, the two countries have swapped places. Italy ran a general government deficit of 3.1 % of GDP in 2025, against a debt load of 137 %. France ran 5.1 % against 119 %. Italy is closing; France is opening. The rating agencies are looking in the rear-view mirror. The bond market is looking through the windscreen.
Here is the part that gets less attention than the headline spread, and deserves more.
France's average interest rate on its outstanding stock is nowhere near 4.86 %. It is about 1.65 %, because France borrowed most of this money when rates were below zero and coupons ran below 1.5 %. The interest bill for 2025 came to €66.6 billion on a stock of €3.60 trillion, which is where that 1.65 % comes from. That is Eurostat's figure rather than an estimate.
So France pays 1.65 % on the old money and would pay 4.86 % on new money. The gap is 3.2 percentage points. Every bond that matures and rolls over moves a slice of the stock from the first number toward the second. Nothing has to go wrong for the bill to rise. The rise is already scheduled.
That is why the debt-service line behaves the way it does. €36.2 billion in 2020. €59.3 billion in 2026 on the central-state basis, according to the Senate's own budget report. €74 billion by 2027. Projections put the whole-of-government figure near €100 billion by 2030.
In 2027, France is scheduled to spend more on interest than on its schools, and more than on its defence. The interest line becomes the largest single item in the budget, and it is the one line no parliament can vote down.
The identity is unforgiving and worth writing out. A debt ratio stabilises when the primary balance (revenue minus spending, excluding interest) equals:
primary balance = d × (r − g) / (1 + g)
where d is debt over GDP, r the average rate paid on the stock, and g nominal growth.
Take France's measured numbers. d is 119.3 % of GDP. Nominal growth g is 4.3 %: 0.9 % real, 3.4 % inflation, both from official sources. The average rate is 1.65 %. That combination permits a primary deficit of about 2.9 % of GDP.
France's actual primary deficit is about 2.9 % of GDP.
France is currently in balance on this identity. Barely, but in balance. That is the honest reading, and it is more surprising than the alarm in the press.
Now let the average rate drift. Every rollover pushes it up. At an average rate of 3 %, the permitted primary deficit falls to 1.4 %. At 4 %, to 0.3 %. At 4.3 % — where r equals g — the permitted deficit is zero.
France's primary deficit is 2.9 %. So the required correction goes:
average rate 1.65 % → correction needed 0.0 % of GDP
average rate 3.0 % → correction needed 1.5 % of GDP
average rate 4.0 % → correction needed 2.6 % of GDP
average rate 4.86 % → correction needed 3.2 % of GDP
The last line costs roughly €96 billion a year. That is the arithmetic of a full rollover at today's rate; it is an identity rather than a forecast. The required consolidation roughly triples without anyone making a single additional mistake, purely because old cheap debt is replaced by new expensive debt.
This is the thing to hold onto. France does not have to do anything wrong to get into trouble. It only has to keep refinancing.
The eurozone has a tool for exactly this, and it was built after the last crisis. The Transmission Protection Instrument, agreed in July 2022, allows the ECB to buy unlimited quantities of a member state's bonds when financing conditions deteriorate in a way "not warranted by country-specific fundamentals."
Two words carry the whole weight: not warranted. That is the door Draghi walked through in 2012. It is also the door that has become harder to open.
The ECB's own eligibility text lists four criteria. The first:
"compliance with the EU fiscal framework: not being subject to an excessive deficit procedure."
France is subject to an excessive deficit procedure. Its deficit was 5.1 % in 2025 and is forecast at 5.4 % for 2026 against a ceiling of 3 %. The third criterion asks for a sustainable debt trajectory. The Commission's own forecasts have French debt at 119.3 % of GDP this year and 121.7 % next, the highest since 1978 by INSEE's count.
So the criteria are not met. But the ECB wrote its own escape hatch: it calls those criteria an input to the Governing Council's decision, "dynamically adjusted to the unfolding risks." It is a political call wearing a legal jacket, and everyone involved knows it.
Which leaves the question that actually matters. Not can the ECB act, for it always can, but what would it demand in return. Every eurozone rescue has carried conditionality. OMT in 2012 required an ESM programme with terms attached. If the price of a French backstop is a programme, then the vote in April 2027 becomes a referendum on accepting external supervision of the French budget. That is a question no French politician wants to be on the wrong side of, which is precisely why the market is pricing it.
A serious person can argue the alarm is overdone, and the case is better than the headlines suggest.
France's debt is overwhelmingly held domestically, by insurers and banks with long-dated euro liabilities and no plausible same-currency exit. The investor base is captive in a way Greece's never was. The AFT, France's debt agency, has deliberately extended its duration profile to lock in the cheap coupons of the last decade. The eurozone as a whole is stable; it is France's position within the bloc that has moved, not the bloc's creditworthiness. And the arithmetic above says France sits near balance today: the required correction is zero, which is far from catastrophic. If nominal growth holds and the average rate drifts only slowly, the picture stabilises.
There is also a technical point in the agencies' favour. Analysts at Dorval Asset Management calculate that France's rating alone justifies a spread of about 50 basis points over Germany, and 63 once negative rating momentum is included. The current 140.9 is well above that — meaning today's spread embeds a bet that France loses another notch or two. If it does not, the spread has room to fall.
That is a real case, and it rests on two assumptions: that French politics produces a credible consolidation, and that growth does not disappoint. Both are testable, and neither is settled.
Macquarie's Thierry Wizman put it bluntly in a note early in October: the widening spread should be read as a "guilty verdict" on the direction of French presidential politics. His reasoning is that neither populist contender is a fiscal hawk. Marine Le Pen, leading the polls, promises tax cuts and a retirement age back to 60. Jean-Luc Mélenchon wants the central bank to cancel the state's debt outright. A runoff between them is expected. Wizman puts the odds of an RN presidency at close to 50 % and an outright default at low.
That is the shape of the risk. Not default. In a country that borrows in its own central bank's currency, default is a political choice rather than an accident. The risk is a permanent repricing: a sovereign that borrows more expensively than its rating implies for a decade, because the market has stopped expecting the political system to solve anything.
And here the word bailout earns its place. Not a cheque written to France — nobody is writing one. A bailout in the eurozone's post-2010 grammar is a conditional purchase programme: the ECB buys the bonds, and in exchange the government accepts supervision it would never have chosen. That is what Greece signed, and what Ireland and Portugal signed. The market is not pricing a French default. It is pricing the probability that France ends up needing one of those programmes, and the political price of accepting it.
French state interest costs were €36.2 billion in 2020. They are €59.3 billion this year on the central-state measure and heading for €74 billion in 2027, and €100 billion by 2030 on the wider measure. Those numbers do not care who wins an election. They arrive on schedule.
Draghi's sentence worked because the spread was a fear, and fear can be talked down. This one is a price for a risk the market can name: a deficit of 5.4 %, a debt ratio at a post-war record, three downgrades in twelve months, and two political blocs that both campaign on spending more.
A bailout of France would be the largest sovereign rescue in European history, on a debt stock four times Greece's. And it would arrive not as a crisis but as a negotiation — because that is the only form a eurozone rescue takes.
The ECB can still act. It has the tool, the balance sheet, and the legal ambiguity it wrote for itself. The price of acting, though, has changed. In 2012 the question was whether the ECB would defend the euro. In 2026, the question is whether France will accept the terms of its own rescue — and that is a question for French voters, not the Governing Council.
The spread is not a prediction. It is the market writing down what it believes about a political system. And the line has been drifting in one direction for ten months.
Sources: CNBC government bond quotes, October 2026, 13:00 CEST (France 4.863 %, Germany 3.4542 %, Italy 4.6277 %, Spain 4.1035 %); Eurostat gov_10dd_edpt1 (French general government interest 2025: €66.6bn at 2.2 % of GDP; deficit 2025: −5.1 %; debt 2025: 115.6 %); French Senate budget report on the 2026 finance bill (central-state debt charge €59.3bn); Cour des comptes, June 2026 (2026 interest €77.4bn on the wider measure); INSEE (debt at 119.3 % of GDP, highest since 1978); ECB press release, July 2022 (TPI eligibility criteria, quoted verbatim); Macquarie via Fortune, October 2026 (CDS at 81bp, default-risk reading); Dorval Asset Management, September 2026 (fair-value spread 50–63bp).