Market data as of 22 September 2026. Debt and deficit figures are Eurostat's, issuance the French Treasury's, lending rates the ECB's — all official. A few figures come from published Amundi research and a commercial spread tracker; those are flagged where they appear, because they are not official statistics and it matters which is which. Yields move in real time.
At the close on 22 September 2026, France paid 102.0 basis points more than Germany to borrow for ten years — a basis point being a hundredth of a percentage point, so 4.47% against 3.45%.
That deserves a minute of your attention for one reason. It is the highest in a year: the twelve-month range was 59.0 to 102.0, averaging 73.8. We are at the top of it, not somewhere comfortable in the middle.
One table, and only the right-hand column matters
Eurostat publishes every member state's debt on the same basis each quarter. Debt is the total owed; the deficit below is one year's overspend. Q1 2026:
| Country | Debt / GDP | Change in one year |
|---|---|---|
| Greece | 143.5% | −9.4 pp |
| Italy | 138.9% | +1.7 pp |
| France | 117.6% | +4.0 pp |
| Belgium | 109.1% | +3.1 pp |
| Spain | 101.6% | −1.7 pp |
| Euro area | 88.9% | +1.7 pp |
| Germany | 64.4% | +2.4 pp |
Ignore the middle column, the one everybody quotes. Greece took 9.4 points off its ratio in a year, Spain 1.7 off, France added 4.0 — more than twice Italy's. France is not Europe's most indebted country; among the big economies it is the one pedalling hardest in the wrong direction, while the countries that were the emergency a decade ago quietly stopped being one. (EU-wide it is not the fastest riser — Eurostat records larger quarterly increases in Finland and Bulgaria — but neither of those is 30% of the euro area's high-grade collateral. See below.)
The deficit is blunter: −5.1% of GDP in 2025, against Germany's −2.7%, Italy's −3.1%, Spain's −2.4% and the euro area's −2.9%. France's deficit is the worst of the big four. Worse than Italy's. Which reverses roughly everything twenty years of headlines taught you about who has the fiscal problem in Europe — and it is not an opinion, it is the number France files itself.
Three things that cut the other way
Leaving these out would be the failure we spend our days catching in company accounts. Most of the debt reprices slowly: Amundi puts the average maturity of France's medium- and long-term debt at 8.5 to 9 years, so for that portion today's yield hits only what gets refinanced. That is why a country can look uncomfortable for years without anything breaking — good news and bad news in one sentence. Not all of it behaves that way: France also issues short-term bills and inflation-linked paper, €23.3 billion of the latter in 2025, and those move sooner. The interest bill is not extraordinary: Amundi's research puts French interest payments, against GDP, below those of the US, the UK and Italy (not an official comparison — one house's maths). The absolute number is huge because the pile is €3.54 trillion; the ratio is not the outlier the headline implies. And this is fixable: the proof is the top row of the table. Nine points in twelve months.
Why this is not only France's problem
France does not drop by the bond market occasionally. Its Treasury plans €310bn of net medium- and long-term issuance in 2026, and in 2025 it placed €347.7bn gross (the €310bn is net of buybacks, so gross is the better guide to how much paper must find a buyer). When the news turns bad, France cannot politely step back and try again next month.
Two figures from Amundi rather than official statistics: about 56% of French government debt is held by non-resident investors, and French paper is more than 30% of the euro area's highest-quality government debt — a pool defined as AAA down to A, worth saying out loud, since France sits inside that range rather than at its top. That is the contagion story, and it is arithmetic rather than prophecy. Amundi describes French paper as a major euro-area benchmark and as paper frequently used as high-quality collateral; the inference we draw from the 30% share is our own. If close to a third of that pool reprices, it is hard to see how the repricing stays inside one country's bond market — you cannot have a fire in a third of the building and call it a problem with one room. That is not a prediction that other countries get dragged down; it is an observation that France is too large a share of the collateral base for a French repricing to be a French-only event.
About 2011
Today's 102.0 is less than half the 225.1 of 17 November 2011, the highest in records going back to 2005. On price, then, the comparison fails: this is a premium, not a rupture. On the debt ratio it holds uncomfortably well, because France owes far more now. We are not saying the market is wrong — that would be a view on an instrument we do not hold and cannot value better than the people trading it. We are saying the two numbers moved in opposite directions, and writers quote whichever one fits the conclusion they arrived with.
The binary
Either the deficit path bends — Amundi reads the government as intending measures taking the deficit to 4.9% next year, though we found no official source stating it — France stops adding four points in a year the way it did to March, and the gap drifts back toward 74. Greece and Spain have shown countries can do this.
Or it does not, the ratio compounds — one institutional projection has French debt above 130% of GDP by 2030 without measures — and at some point France stops being priced as core Europe and starts being priced as something else. That reclassification is the event worth caring about, and per the section above, it would not be French-only.
Both at once, no. And notice what decides it: not earnings, not a central bank, but whether a parliament passes a budget. No financial model has a column for that, which is exactly why the spread has drifted up rather than gapped.
The only part you can act on
If you run a business, that spread is not your problem; the rate your bank quotes you is. They are connected, indirectly but really, because government yields are the benchmark bank funding is priced off. The ECB's composite rate for new loans to non-financial corporations in the euro area was 3.79% in June 2026, up 15 basis points on the month:
| New company loans, floating or fixed up to 3 months (June 2026) | Rate | On the month |
|---|---|---|
| Up to €250,000 — small borrowers | 3.91% | +13 bps |
| Over €1 million — large borrowers | 3.54% | +26 bps |
A small company pays 37 basis points more than a large one for the same euro. The ECB publishes the levels, not the reason; reading it as the price of being harder to assess is ours. And it moves — that month the large-borrower rate rose twice as fast, narrowing the gap.
Which is the practical point, and it has nothing to do with Paris. The slice of your borrowing cost that sovereign politics sets is small and none of your business. The slice your own numbers set is bigger and entirely yours. A company arriving at a renewal with a rolling 13-week cash forecast, margin by line and clean management accounts negotiates from a different chair than one arriving with last year's filed accounts — and unlike the French budget, that is solvable this quarter.
The three things we would watch
- The spread against 74, not against 225. The twelve-month average is the honest yardstick; the 2011 peak gets quoted because it is frightening.
- The next Eurostat release — the change, not the level. Those 4.0 points are one reading, not an annual rate. If the increase slows, the binary resolves the good way with nobody announcing anything.
- Whether a budget passes, and what deficit it assumes. The 4.9% is Amundi's reading of an intention, not a verifiable target. The only political item of the three, and it decides the other two.
In one sentence
"French debt added four points of GDP in a year while Greece shed nine, its deficit is the worst of the big four, and it is selling hundreds of billions into a market charging the most it has in twelve months — and still less than half what it charged in 2011."