Investing9 Oct 202627 min read

Compound interest is eating France, and the euro has bricked up the exit

France owes 119% of GDP and the interest bill is snowballing. The maths, why Frankfurt won’t help and what it means for your savings and your home.

A red convertible on a Riviera quay, buried under a tower of paper bills. A woman in a white dress and a porter in white walk past.
Not a photo. The pile is real, though: €3,595.5 billion of it.

Same formula, wrong side#

Compound interest is the best thing that can happen to your portfolio. Put away €300 a month for 30 years at 7% a year and you end up with about €351,000. You paid in €108,000. The 7% is an assumption, and historical returns are not promises.

Now flip the sign. A state that borrows every year and never pays anything down gets the same formula pointed at its own budget: interest on interest, paid with new debt. France is the live demonstration. Everybody who shares a currency with it is sitting in the splash zone.

What set me off was a Telegraph column, republished by Merkur on 8 October 2026: interest is strangling France, and the ECB can only step in once the crisis has fully escalated. I checked the numbers. A few overshoot. The accurate ones are worse, because nobody can wave them away.

I think the euro crisis of 2010 will look like the tutorial level next to this. And the bill ends up where it always does: with people who saved, and people who own something that can’t run away.

The numbers, before anyone shrugs “it’s just France”#

At the end of June 2026 France owed €3,595.5 billion, 119.0% of GDP. A record, says INSEE. The pile grew by €59.6 billion in three months. That’s about €650 million a day, Sundays included.

In 2007 France stood at 65.5% of GDP and Germany at 63.7% (Eurostat). Twins. In 2025 it was 115.6% against 63.5%. Same currency, same central bank, same crises. One of the two kept the receipts.

The deficit was 5.4% of GDP in 2023, 5.8% in 2024 and 5.1% in 2025 (INSEE). The “improvement” came mostly from tax rises, says the Cour des comptes, France’s court of auditors. For 2026 the budget law promised 5.0%. The government now expects 5.4%. The target for 2027 is 5.0% again, which France’s own fiscal council told parliament is the least to expect.

The column talks about a 7% deficit next year and debt on its way to 140%. No official paper says that. The darkest official scenario is the stress case of the Cour des comptes: 134.6% of GDP in 2031. It doesn’t need rounding up.

Guess first

YearInterest, € bnWho says so
202029.7Eurostat, outcome
202566.6Eurostat, outcome
202677 to 79Cour des comptes, budget papers
202791Budget plan
2029above 100Cour des comptes

Outcomes from Eurostat. The 79 is my subtraction: the fiscal council named 91 for 2027 and an increase of 12. That increase alone eats almost a third of the roughly €40 billion of effort planned for 2027.

Emmanuel Moulin, governor of the Banque de France, said on France Inter on 7 October 2026 that the 2027 bill, which he puts at €92 billion, is nearly the whole yield of income tax. He called it a “risque d’étranglement”, a risk of strangulation: every euro saved gets eaten by the higher interest charge (transcript). That’s the central bank talking, not a blogger with a grudge.

One thing the column skips: today France’s interest burden is still lighter than Italy’s. In 2025 it was 2.2% of GDP and 4.3% of government revenue, against 3.9% and 8.0% (Eurostat). The strangling is a story about 2027 to 2030.

But Italy took in 0.8% of GDP more than it spent in 2025, interest aside. France was 2.9% short before it had paid a cent of interest. Italy is paying for its past. France is still running up the tab.

The market has noticed.

Country10-year yield, %Gap to Germany, points
Germany3.470
Spain4.090.61
Greece4.400.93
Italy4.571.09
France4.851.38
United States5.251.78
United Kingdom5.441.97

Market levels of 9 October 2026, 11:15 GMT, from the aggregator worldgovernmentbonds.com.

France pays more than Italy and more than Greece. Not only on a bad Friday: on Eurostat’s monthly averages for August 2026 it was at 4.00%, Italy at 3.99%, Greece at 3.87%.

Of France’s 4.85%, 3.47 points are the German base that everybody pays. France’s own surcharge is the 1.38 on top. On 25 February 2026 it was 0.55.

How the snowball rolls#

France’s existing debt costs 2.3% on average in 2026, after 2.0% in 2025. That’s the cheap money of the zero-rate years: in 2021 the state issued its bonds at an average yield of zero, notes the Cour des comptes.

New money is another story. Ten years cost 4.85% on 9 October. The debt agency Agence France Trésor has to place €310 billion of medium and long-term bonds in 2026 and €340 billion in 2027, a record, mostly because bonds from the pandemic and the energy crisis fall due.

The state’s debt runs for 8 years and 142 days on average. So every year a slice of the 2% pile gets swapped for paper at 3.55%, the average of this year’s issues up to September, or more. It’s a variable-rate mortgage in slow motion, and the reset letters have started arriving.

At today’s yield the bill doubles within five years, from €79 billion to €158 billion. No new policy needed. Just time.

Whether debt runs away comes down to two numbers: the interest rate on the pile and the growth of the economy including inflation. In 2022 and 2023 inflation pushed that growth to 5.8% and 6.8% while the debt cost 1.8%. France could overspend by 3% of GDP before interest and the ratio still fell. In 2025 that free lunch ended: growth (1.9%) dropped below the interest rate (2.0%). To hold the ratio France would have needed a small surplus before interest. It ran a deficit of 2.9%.

Now you. The starting numbers: debt of 119.0% of GDP (INSEE, end of June 2026), a deficit before interest of 2.9% (2025, Cour des comptes), average interest of 2.3% (2026, fiscal council) and nominal growth of 2.0%, roughly what the fiscal council’s president gave for 2026.

Two things to try. Push the interest slider to 4.9, where the market is. Then set the deficit before interest to zero and watch the debt keep climbing anyway.

At 4.9% interest and 2% growth, holding the ratio takes a surplus before interest of about 3.4% of GDP. From a deficit of 2.9%, that’s a swing of 6.3 points: roughly €190 billion a year at the 2026 GDP of about €3,034 billion. The average rate gets there bond by bond, not overnight. Once it shows in the budget, it’s locked in for years.

Why the euro makes it worse#

France shares its central bank with 20 other countries, and that bank has a job description: inflation.

Euro-area inflation was 3.8% in September 2026. The ECB raised rates in June and again in September, to 2.50%, and expects inflation to stay well above target for an extended period.

It’s also getting rid of bonds. At the peak in January 2023 the euro central banks held €838.9 billion of French public debt, about 28% of the total. At the end of September 2026 it was €597.2 billion (ECB data, added up across the two programmes). That’s €242 billion that private buyers have to take on top of all the new bonds.

Then there’s the rescue tool, the Transmission Protection Instrument of July 2022. It has never bought a single bond. The column says France fails all its conditions. On paper it doesn’t. France has been in an EU deficit procedure since July 2024, but Brussels put it on hold on 3 June 2026 and hasn’t declared that France failed to act, which is what the wording asks for. And the ECB calls its list of criteria only an input. Brussels grades on a curve.

The real obstacle is the first sentence of the tool. It’s for market stress that is “not warranted by country-specific fundamentals”. A spread that widens because the deficit is 5% and parliament has no majority is warranted. By the ECB’s own definition.

France’s own central bank governor says so. Moulin, in the same interview: don’t go looking for solutions in Frankfurt. The ECB’s mandate is inflation, not the French budget.

And there’s Karlsruhe. In 2020 Germany’s Constitutional Court listed what kept the ECB’s bond buying on the right side of the ban on financing governments: a limited volume, purchases spread across countries by a fixed key, buying tied to the inflation target. A rescue of France would be unlimited by design, for one country, with inflation above target. That’s my reading of the two texts. No court has ruled on the tool.

“But Britain can print”#

It can. Look at the invoice.

In autumn 2022 the Bank of England announced up to £65 billion of emergency bond purchases. It bought £19.3 billion in 13 trading days and had sold everything again by January 2023, while raising rates (Bank of England). Inflation peaked at 11.1% that October. A central bank can stop a fire sale. It can’t fund a state for free.

The years of printing before that are on the bill too. Between October 2022 and August 2026 the Treasury paid £110.7 billion to cover the Bank’s losses on the bonds it had bought (added up from the ONS public finance tables). Debt interest ran at about £110 billion in 2025-26, says the budget watchdog OBR. That’s 1.76 times the defence budget.

And the yields? Scroll up. Britain and the United States own printing presses, and both pay more than France.

So the press isn’t a way out. It’s a different way of sending you the bill: inflation instead of default. The saver pays either way. The euro only removes the option of doing it quietly. What’s left is cutting, or a crisis.

Why 2010 would look like the tutorial level#

Size.

France is 18.8% of the euro area’s GDP (Eurostat, 2025). Greece, Ireland, Portugal and Cyprus together were 6.1% in 2010. Add Spain, which got a bank programme, and you reach 17.3%. Every country that needed rescuing back then, combined, was a bit less than one France.

The rescue fund, the ESM, has €433 billion of lending capacity left. France sells €340 billion of bonds in 2027 alone. So the fund covers about 1.3 years of French borrowing, or 12% of French debt. All five rescues from 2010 to 2018 together paid out about €482 billion.

And who backs the fund? France, with 19.9%, the second-largest shareholder.

Then the banks. Foreign banks had claims of $600 billion on France’s government and central bank at the end of March 2026, says the BIS. On Greece’s in March 2010, when that crisis broke: $92.5 billion. Six and a half times, not adjusted for inflation.

Now the fact that speaks against panic. France’s gap to Germany is 1.38 points. On the monthly averages of 2011 and 2012, Italy peaked at 5.19, Spain at 5.55, Portugal at 12.03 and Greece at 27.39. France’s own worst month back then was 1.54. And bond auctions go through: demand in 2026 has been 2.5 times supply.

That’s my point. This is the smoke. In 2010 it came from a few small rooms at the edge of the building. This time it comes from the second-biggest room in the house, and the extinguisher was sized for the small ones.

Who’s next if it spreads? Not the usual suspects. Greece ran a surplus before interest of 4.9% of GDP in 2025, Portugal one of 2.6%. The holes are elsewhere: Belgium with debt of 107.9% and a deficit of 5.2%, Austria with a deficit of 4.2%, and Italy with 137.1% of debt to refinance.

And nobody in Berlin gets to feel smug. The federal government’s own plan has interest spending going from €30.3 billion in 2026 to €80.7 billion in 2030, which is 12.7% of the core budget. Net borrowing in 2027: €204 billion (KfW Research). In September the federal court of auditors told the government to get off the road into the debt trap.

Germany’s debt is 63.5% of GDP. Its problem is the direction, not the level. But it’s the largest guarantor of the rescue fund, with 26.5%.

So yes, I expect a French crisis to pull the others down with it: through the banks, through the guarantees and through an ECB forced to choose between fighting inflation and saving a member. I don’t have a date. Bond markets can stay calm for years, and then not.

The exit nobody votes for#

A state has three honest ways to stop a snowball: grow, tax more, spend less.

Growth: the government cut its 2026 forecast from 1.0% to 0.5%. Next.

Taxes: France collected 45.3% of GDP in taxes and social contributions in 2024, second in the EU behind Denmark (Eurostat). The euro-area average is 40.8%. Even the Cour des comptes writes that the repair can’t rest mainly on new tax rises.

That leaves spending: 57.2% of GDP in 2025, second in the EU behind Finland and 7.4 points above the euro-area average. That’s about €225 billion a year.

Pensions cost €422 billion in 2025: 14.1% of GDP and almost a quarter of all public spending, says the pension council COR. The average living standard of French retirees was 100.2% of that of the whole population in 2023. That compares household income after taxes and transfers, adjusted for household size, with everybody, children and the unemployed included. It doesn’t compare pensioners with workers, which is how the column tells it.

French men leave the labour market at 61.9 on average and can then expect 22.5 years of retirement, the second-longest in the OECD (2024). The OECD averages are 64.7 and 18.6.

And the state itself? The “consolidation” budget for 2027 adds 8,025 full-time posts to the state payroll, says the think tank IFRAP, which read the annexes. The government shows minus 1,076 by leaving out the teacher-training reform and the military build-up. The savings plan hires.

Milei, with the full bill#

Javier Milei proved it can be done. He also got an invoice.

Argentina’s real primary spending fell 27.4% in 2024, says the congressional budget office OPC: from 18.1% of GDP in 2023 to 13.1% in 2025. Monthly inflation dropped from 25.5% in December 2023 to 1.7% in August 2026. I admire the nerve.

About a fifth of the 2024 cut was pensions rising more slowly than prices, with prices rising by over 200% a year. Inflation held the chainsaw. At 3.8% inflation you’d have to announce the cuts in plain numbers.

Poverty went from 41.7% to 52.9% in the first half of 2024, fell to 28.2% and was back at 32.3% in the first half of 2026. Yearly inflation is stuck around 33.5%. In 2025 Argentina needed outside money twice: a $20 billion IMF programme in April and a $20 billion credit line from the US Treasury in October.

His party still won the midterms of October 2025, with 40.7%. Argentines had lived through 211% inflation. People vote for the chainsaw after the currency has burned, not before.

And inside the euro there’s no devaluation to soften the landing. Greece’s economy shrank by 27% between 2008 and 2013 (Eurostat). Latvia, which kept its peg to the euro, lost 25% from peak to trough. Ireland lost about 9% between 2008 and 2010.

Who cut, and who didn’t#

It’s not impossible in Europe. Between 2009 and 2025 Denmark took public spending from 56.3% to 48.1% of GDP, Portugal from 50.3% to 42.7%, Greece from 54.8% to 48.3% and the Netherlands from 48.4% to 44.9% (Eurostat).

The two big ones didn’t. France: 58.0% to 57.2%. Germany: 48.3% to 50.5%.

The recent record, with sums:

Where and whenOn the tableWhat came out
France, 2018Fuel tax riseCancelled after the gilets jaunes protests. Concessions: €10.3 billion in 2019
France, 2025€43.8 billion savings planPrime minister voted out, 194 to 364
France, 2025Pension age 64 on schedulePushed back one birth year. Cost in 2027: €1.9 billion, government count
Britain, 2024Winter fuel payment cut, saving £1.5 billion a yearLargely reversed in 2025, cost about £1.25 billion
Britain, 2025Welfare cuts of £4.8 billion in 2029-30Main clause dropped in July 2025
Germany, 2025A pension packagePassed. Cost in 2030: €14.5 billion, by the bill’s own numbers
Germany, 2026Welfare reform, advertised by a CDU general secretary at €15 billion in savingsPassed. Saving in 2026 per the bill: €86 million

The Treasury’s own Budget table carries a line for the welfare policy changes of summer 2025: a cost of £6.9 billion in 2029-30.

Two facts for precision. France’s pension reform of 2023 did pass, against 1.28 million demonstrators on the biggest day by the interior ministry’s count. Even after the “suspension” of 2025, 64 is still reached, by people born in 1969 instead of 1968. And the governments of 2024 and 2025 fell in parliament, not on the street: 331 votes against Barnier, 364 against Bayrou.

That makes it worse. A street protest ends when it rains. A parliament that mirrors its voters doesn’t.

Everyone wants their slice#

Follow the incentives. Take Germany, because the count is public: at the federal election of 2025, 42.6% of eligible voters were 60 or older. Under 30: 13.0%.

In France 57% of people are net beneficiaries of redistribution, says the statistics office Insee. Careful what that measures: pensions count, and so do schools and hospitals. Among households led by someone over 65 it’s 90%. In Britain the figure is 53.3%, on a similar definition.

And work? Germany has the fewest hours per worker in the OECD: 1,332 a year in 2025, against an average of 1,736. But that divides all hours by everybody with a job, 29.2% of German workers are part-time and the employment rate is 81.1%. A part-time country, not a lazy one.

Put it together. Everybody likes their share of government money: the pensioner his pension, the farmer his diesel, the civil servant his post, the tenant his rent cap. Everybody agrees that saving is necessary, and that the neighbour should start. Politicians can count votes, so the reform gets announced, softened, postponed and “suspended”. Spending other people’s money works fine until the other people want 4.85%.

I don’t expect anybody to change this by choice. The bond market does it for them. That’s how it went in Greece, in Portugal and in Britain in 2022. A country that can’t agree on €40 billion in a quiet year gets to find much more in a loud one.

What this has to do with you#

A state that runs out of room doesn’t close. It reaches for what it can get: your pension, your savings, your house.

Guess first

CountryRent, %Own with a loan, %Own outright, %Rent or owe, %
EU31.524.843.756.3
Germany52.824.023.376.8
France38.630.830.569.4
Netherlands31.257.910.989.1
Spain26.428.145.554.5
Italy22.917.359.940.2
Poland12.812.574.725.3
Switzerland58.037.84.295.8

Share of people, 2025, Switzerland 2024, from Eurostat. The last column is my sum.

Most Europeans own, and the largest single group in the EU has no mortgage at all. “Rents or still owes the bank” is a northern pattern. In the north your housing costs depend on a landlord or a bank. In the south and east you own the one asset that can’t leave the country.

The cushion is thin#

29.2% of people in the EU can’t pay an unexpected bill from their own money (Eurostat, 2025). The typical euro-area household has liquid assets worth 3.4 months of gross income, says the ECB’s household survey.

Thrift isn’t the problem everywhere. Germans save 18.9% of their income. It’s where the money sits. Of all financial assets of euro-area households, 30.9% are cash and bank deposits and 4.8% are listed shares (Eurostat, end of 2025). In Germany it’s 36.4% and 7.1%. French households keep 26.0% in life insurance.

The rest of the plan is the state. Cash benefits plus public services like health care and schools equal 44.4% of what EU households have at their disposal (my calculation from Eurostat, 2024, before the taxes that pay for it). That’s the flow a state without room cuts.

Where a French crisis meets the saver#

A life insurance policy is mostly a bond fund with a nicer brochure. French life insurance holds €2,171 billion (August 2026), a sum as big as 60% of the public debt. More than half of what French insurers and pension funds invest is bonds, and about 11% is French government bonds, says the Banque de France.

And there’s a law for the day everybody wants out at once. Since 2016, France’s financial stability council can temporarily limit what insurers pay out when savers cash in their policies: for three months, renewable, and for no more than six months in a row (article 49 of the Sapin II law). It’s a temporary limit on withdrawals, not a confiscation. The law gives no power to cut the value of a policy. But “your money is safe, you just can’t have it right now” is a sentence worth knowing before you need it.

Bank deposits up to €100,000 are protected by EU law, and in every euro-era crisis small deposits kept their face value. Not for lack of trying: the first Cyprus plan of March 2013 took 6.75% from every deposit under €100,000 (finance ministry announcement). Parliament voted it down.

And a guarantee is only as good as the fund behind it. France’s guarantee fund holds €7.456 billion for €1,472.6 billion of covered deposits: 0.506%, right at the lowest target EU law permits. Germany’s schemes hold 0.811%, Spain’s 0.877% (EBA, end of 2025). Fine for one bank. Not built for a system.

Mortgages and pay#

A new German mortgage cost 1.32% in December 2021 and 4.00% in August 2026, on the ECB’s figures. The Bundesbank expects about 20% of the outstanding volume to come up for renewal by 2027, from around 2.5% to more than 4%. It also expects repayments and higher incomes to cover that for most households.

“Higher incomes” is carrying a lot of weight there. Real pay per employee, 2019 to 2025: Germany plus 1.2%, France minus 1.3%, Italy minus 3.6% (my calculation from Eurostat data). Six years, nothing. Poland managed plus 17.3%, so it isn’t a law of nature. Germany lost 511,000 factory jobs in the same period, 83% of the EU’s whole net loss.

That’s the standard plan in the old core: a salary that’s supposed to rise, a pension from a state that borrows to pay it, and savings in deposits and bond-heavy insurance policies. Three bets on one horse.

Is your house yours?#

My hunch was that owning in Germany is more a right of use than property. The yearly tax doesn’t prove it. Recurring taxes on land and buildings are 0.37% of GDP in Germany, against 1.95% in France and 2.80% in Britain (OECD, 2024). Germany is near the bottom.

Germany charges at the door. The transfer tax is up to 6.5%, and 15 of 16 states have raised it. On a €400,000 home that’s €26,000 before you’ve bought a single tin of paint. Then come rent caps if you let, and heating rules that parliament wrote in 2023 and tore up again in July 2026.

Elsewhere the meter runs every year. France’s property tax rose 37.3% in ten years, according to the owners’ association UNPI.

The stronger argument is history: what states did to savers and owners once they were out of options.

Guess first

  • Germany, 1948. 100 Reichsmark of savings became 6.50 Deutsche Mark. Owners of houses, land and businesses then got their bill: 50% of the assessed 1948 value, payable in instalments over 30 years (DIW).
  • Italy, 1992. A decree published on 11 July took 0.6% of every bank balance as it had stood two days earlier, plus a one-off 0.2% to 0.3% of the value of property.
  • Cyprus, 2013. At the largest bank, 47.5% of deposits above €100,000 were turned into shares.
  • Greece, 2010 to 2018. Twelve rounds of pension cuts. Shares down 91%. Home prices down 42%. Private holders of government bonds lost 53.5% of face value. And in 2011 a new property levy arrived, collected through the electricity bill.

No two cases are alike, but the order is. Claims on the state and the banks go first: bonds, pensions, big deposits. Property keeps its substance and then gets the invoice, because a house can’t leave the country and the state knows the address.

What I do with that#

No product tips. Principles.

  1. Own productive assets, across countries and currencies. Shares in businesses all over the world instead of one state’s promises. The setup is in investing in Europe. Greek shares lost 91%, so “across countries” is the important half.
  2. Keep a cash buffer. It’s what stops you from selling at the worst moment.
  3. No debt that depends on next year’s interest rate. France is showing what refinancing looks like when the rate has doubled.
  4. Earn more. Income you control beats a promise somebody else has to keep. Start at Build.
  5. Know your own numbers. What you own, what you owe, at which rate, until when. And how much of your retirement is a promise from a state with this spreadsheet. That maths, for 33 countries, is in the pension article and the pension check.

Why I don’t count on the state is in Own your future. The crash plan is in the everything bubble.

And here the loop closes. The formula that’s eating France doesn’t care whose side it’s on. A debtor who never pays down gets buried by it. A saver who keeps paying in gets carried by it. You can’t fix the French budget. You can pick your side of the equation.

What if I’m wrong?#

Then parliament passes a real budget, growth comes back, inflation falls and the ECB stops shrinking its bond pile. French savers keep funding the state: they save 17.4% of their income and put a net €43.5 billion into life insurance in the first eight months of 2026. The long maturity of the debt buys years.

In that world France muddles along at 120% and I’ve been careful for nothing. I can live with that. It’s the plan I’d want anyway.

The other mistake costs more: trusting that the people who built this spreadsheet will fix it before it reaches you.

This article is education, not investment, tax or legal advice. It contains my opinion, scenarios with the sums shown and no product tips.

Sources#

Secondary sources, where no primary document was at hand:

Education, not advice. I don’t know your situation, and past returns promise nothing. Check my numbers, then make your own call. You’re a grown-up.