# France's deficit has run for fifty years. Why is the market nervous now?

Source: Realpha Blog (blog.getrealpha.com)
Original article and charts: https://blog.getrealpha.com/en/blog/oddlots-2026-10-16-the-french-political-crisis-that-s-turning-into-a-/

> Notes from the 16 October 2026 Bloomberg Odd Lots episode: former Macron adviser Shahin Vallée unpacks the rise in French bond spreads — what got repriced is who can pass a budget after the 2027 elections. Educational notes on how fiscal and political risk bind each other; no stock views and no buy or sell advice, readers should form their own judgement.

Published: 2026-10-16
Locale: en
Tags: France, European debt, fiscal policy, political risk, ECB


![An assembly chamber at dusk, empty tiers of seats receding into shadow, a single desk lamp on an open budget document](/covers/oddlots-2026-10-16-the-french-political-crisis-that-s-turning-into-a--cover.png)

> The mirror of the past is not far off; it lies in the age of the house of Xia.
>
> —— *Book of Songs*, "Greater Odes: Dang" (pre-Qin; translation mine)

On the 16 October 2026 episode of Bloomberg Odd Lots, Joe Weisenthal and Tracy Alloway hosted Shahin Vallée, senior fellow at the German Council on Foreign Relations, who advised Emmanuel Macron at the economy ministry in 2014–2015 and worked on his 2016 campaign. Vallée's read: the latest move in French bond spreads has nothing to do with new fiscal news, since the budget just announced came in close to expectations and arguably a touch better. What got repriced is whether anyone will be able to pass a budget after the 2027 elections. Tracy fact-checked her own intro on air: France has not run a budget surplus in more than fifty years. The conclusion rests on one condition — every ECB rescue tool is tied to the country first delivering a credible fiscal adjustment, so while the political deadlock holds, the rescue button cannot be pressed.

## Fifty years of deficits, so I expected the bad news to come out of the numbers

My prior going in was plain: a debt crisis is arithmetic. Watch debt to GDP, watch the deficit, watch how much of tax revenue goes to interest, and when one of them crosses a line the market responds. I had a rough sense of where France's numbers sit, so my reflex on this story was to go pull the latest fiscal figures and see what deteriorated.

Tracy asked the same question on the show, and she asked it more honestly than I had framed it. Fiscal fundamentals often look the same for years, she said; she has never in her lifetime seen France run a surplus, so what she wanted to know was what the actual catalyst is. Vallée's answer knocked my prior out: the budget just announced was in line with expectations, maybe slightly better. The numbers did not get worse. Something else moved.

![A two-panel line chart: in the top panel the deficit line stays below zero for fifty years, rising and falling but never changing height, while in the bottom panel the government bond spread runs flat for years and then suddenly kicks upward at the far right.](/figures/deficit-flat-spread-jumps-en.svg)

I make the same mistake with companies. I check leverage and interest coverage, see the numbers inside my tolerance, relax — and then get hit by an event where the numbers held steady and the people who decide the numbers changed. This episode made that habit visible to me.

## What got repriced is whether anyone can pass a budget after 2027

Vallée split the last few weeks into two layers, both political.

The first is that the odds of a centrist reaching the runoff got marked down. The long-standing assumption was Marine Le Pen against some centrist, identity unclear, with Édouard Philippe the best-placed of them. Since the summer those centrists have been slipping in the first round, and — the more worrying part — now losing the second round against Le Pen across the board. So the likeliest runoff has become Le Pen against Jean-Luc Mélenchon on the far left.

![Two panels compare the runoff matchup: on the left Le Pen faces a centrist block of roughly equal size, while on the right the centrist has shrunk to a sliver and the opponent has been replaced by a larger far-left block.](/figures/runoff-pairing-before-after-en.svg)

The second layer is the one I had been missing. A month after the presidential vote comes a general election. The new president will likely dissolve the current parliament and call a fresh one in order to govern. And Le Pen could win the presidency while struggling to win a majority in that general election, especially against a well-mobilised far left. Vallée says this is the first time people have seriously seen the risk that France after 2027 is in a worse position to decide fiscal policy than it is today. Today is already poor: since the 2024 snap election Macron has had no majority, and every budget since has gone to the wire. Markets had been giving France the benefit of the doubt, assuming that by 2027 there would be a president and a parliament able to pass a credible adjustment. That assumption is loosening.

![A left-to-right branching diagram: the presidential election leads to a parliamentary election one month later, which splits into a thin branch for winning a majority and a thicker branch for falling short, and the thicker one leads to even weaker decision-making.](/figures/two-elections-branch-en.svg)

Loosening with it is the "Meloni script." The expectation was that a Le Pen government would resemble Italy's — centre-right to far-right, broadly pro-European, broadly compliant with EU fiscal rules. Vallée says that script is being de-priced, and he thinks that is right, because France sits in a different seat: France is a net contributor to the EU budget, while Meloni arrived when Italy was the net recipient of €300 billion of new money from the recovery and resilience facility. Someone receiving money and someone writing the cheque negotiate differently. The EU budget is being negotiated now, and Le Pen has room to pick that fight in a way Meloni did not.

Her fiscal posture is the more interesting part. On 6 October she put forward a fiscal rule she wants written into the constitution by referendum: bring the deficit back to the debt-stabilising level, which means roughly halving it almost immediately; until then, a minimum adjustment of 0.5% of GDP a year; and keep going until debt to GDP returns to 60%. The governance is tight — no budget violating the rule can pass, with the Constitutional Court blocking it; an emergency exception for war, recession or an energy crisis needs three-fifths of parliament, lasts two years, and after that requires a referendum. Vallée's verdict is that this sets the bar higher than the German debt brake. Yet she has said nothing substantial about pension reform. His word for it is cognitive dissonance: a claim bold enough to strain credibility, with no policies behind it. That tension cannot last, and at some point she has to disappoint someone — tell markets she cannot deliver what she promised, or tell her electorate that keeping the rule means the pension age goes back to 65 or higher.

![At the centre is a circle for the EU budget; an arrow on the left sends money from France into the circle, while an arrow on the right sends 300 billion out of the circle to Italy, so the two arrows point opposite ways.](/figures/payer-versus-payee-en.svg)

The left side is more theatrical. Mélenchon seized on the fact that roughly 20% of French debt is owed to the French central bank, because the ECB intervened during both the euro crisis and COVID. Debt we owe ourselves, the argument goes, can be cancelled. First he said he would throw the Banque de France debt "into the fire." Then the phrasing changed to "into the freezer" — not cancellation, but asking the ECB to roll it in perpetuity. Vallée reads the provocation as deliberate and thinks Mélenchon did not anticipate that it would raise market worries about sustainability rather than calm them. Then the new governor, Emmanuel Moulin, told the Financial Times he worried France would be strangled by its debt without fiscal action. Mélenchon took that as betrayal and threatened the governor with a treason trial in a tweet. Vallée places him next to Greece's Yanis Varoufakis: Varoufakis had real political support, won the referendum he called, and then could not convert that win into anything. Mélenchon may collect political credit for standing up to markets; delivery is a separate question.

One number on the collapse of the centre stuck with me. The Socialist Party held its primary that weekend and 120,000 people voted; at the last presidential cycle it was more than a million. Vallée calls Macron France's Renzi — the last force holding the centre together, failing to hold it, and leaving both the centre-left and the centre-right in disarray on the way out.

![On the left a solid wall of fiscal rules is stacked three steps high, while on the right, on the same baseline, there is only an empty dashed box, leaving an obvious gap in height between them.](/figures/fiscal-wall-pension-gap-en.svg)

The bank channel is where I was most wrong. Tracy raised the sovereign-bank doom loop — bonds sell off, banks take losses, lending dries up, and the sovereign that is already stretched has to step in. Vallée's answer: French banks hold relatively little government debt compared with Italian or Spanish peers, could step up to stabilise the market and still sit inside their 2011–2012 peak, and would have €300–400 billion of room if they went as far as Italian banks do. He also corrected my picture of 2011. What squeezed French banks that summer, as the crisis moved from the periphery into the core, was dollars — US money market mutual funds stopped funding European and particularly French banks. That was a dollar run, a different animal from the textbook doom loop. It is absent today, and spreads on French bank debt have recently been tighter than on the sovereign.

## The ticket to a rescue is the thing they cannot do

The part that made me stop was why the ECB cannot run the 2012 play again. Tracy's half-joke was that all they have to do is say "whatever it takes."

Vallée's answer is that the euro crisis ended with a new architecture nobody quite announced. First a backstop that could lend to member states — eventually the ESM, a European version of the IMF. Then banking union, a single supervisor and in principle a single resolution authority, to break the loop between banks and sovereigns. And finally the unspoken acceptance that the central bank could be called on as lender of last resort. He agrees the architecture is more robust. But the political conditions came attached: you lend against an adjustment programme. OMT — the ECB buying a country's bonds outright, announced by Draghi — was designed to activate only if the country had an ESM programme. The ECB has since invented the transmission protection instrument, TPI, never used, which can be activated without a programme but only if the country is deemed compliant with the European fiscal rules. Compliance requires the political capacity to commit to a credible medium-term path.

So it loops back to the same point, and Joe put it most bluntly: you can have a backstop in exchange for fiscal consolidation, but if you could do the consolidation we would not be having this conversation. Tracy capped it with a bilingual joke — a catch-vendeux, catch-22 with the 22 in French. The knot survives the laugh.

![Four boxes form a closed loop: central bank backing leads to doing the fiscal consolidation first, which needs a political commitment, which runs into political fragmentation and returns to the start, while an arrow trying to cut across is crossed out.](/figures/rescue-catch-loop-en.svg)

Vallée adds a layer I had not considered: this is not France's problem alone. Greece, Portugal and Spain were pushed into harsh adjustments by a mix of market and European political force, and it is hard to explain to them why France would be different. For a German chancellor, letting the ECB backstop France for free is the surest way to grow support for the AfD — a party founded to oppose the euro and the Greek rescues, now campaigning mostly on immigration. That signature is not one this chancellor can give.

Where I land: I have rewritten the story from "France's numbers are bad" to "France's decision-making power is fragmenting, and the rescue conditions only accept intact decision-making power." Vallée's own framing is that the building blocks for solving this — a cross-party fiscal consensus plus the ECB's tools — are within reach, and he cannot see a catalyst that assembles them before the elections. His first block is Macron asking every party to nominate one or two experts and having them agree a consensual debt rule for the next decade. He thinks the German debt brake was poorly designed, but what Germany built with it was a national consensus on fiscal policy, and that is the useful half.

The two sacred cows deserve a mention. Vallée disagreed with Macron about abolishing the wealth tax, symbolically and economically, and thinks the "president of the rich" label that followed cost Macron the pension reform he had in his 2017 programme and never passed. He expects consolidation to require some form of wealth tax coming back, and pension reform alongside it. Even Mélenchon has not fully embraced the more aggressive version proposed by the economist Gabriel Zucman. Tracy's closing observation is the one I will keep: if AI does more of the work, taxing labour income gets harder, and the conversation about taxing capital gets louder.

If your first instinct on a headline like this is to pull the numbers, the reading from this episode transfers. You have probably hit the version of it where earnings hold up and the stock falls anyway, or good news lands and nothing moves. My old explanation was "the market is irrational," and I stopped thinking there. Now I sort what I watch into two piles. One pile is levels — debt ratios, revenue, margins — which move slowly and get seen in advance. The other is decision rights — who can pass a budget, who controls capital allocation, how the board is composed — which sit still and then jump. In France the levels have not moved in fifty years; what jumped is the probability distribution over decision rights. That is why checking the latest figures turns up nothing on this question.

![Two kinds of curve share one panel: the level is a smooth line rising slowly, while control is a step line that runs flat for a long time and then drops a step all at once.](/figures/level-versus-control-en.svg)

Plenty stays uncertain for me. I do not know which two people actually reach the 2027 runoff. I do not know whether market stress will push fiscal policy into the centre of the campaign the way Vallée suspects — he was struck that with spreads at stressed levels the policy debate has stayed timid on this. And I do not know what a compromise between a far-left parliament and a far-right president would look like. Vallée says he struggles to see how it emerges without a more adult conversation about fiscal policy, one France has gone without for a long time.

## Worth looking up

- Bloomberg Odd Lots, 16 October 2026, with Joe Weisenthal and Tracy Alloway; guest Shahin Vallée, senior fellow at the German Council on Foreign Relations and former Macron adviser
- The ECB's two instruments: OMT (announced by Draghi in 2012, activation tied to a rescue programme) and TPI (created in 2022, never used, conditional on compliance with EU fiscal rules)
- Germany's constitutional debt brake, as a comparison for the fiscal rule Le Pen presented on 6 October
- Banque de France governor Emmanuel Moulin's Financial Times comments about France being strangled by its debt
- Gabriel Zucman's wealth tax proposal and its version in the Californian referendum
- The 2015 Greek referendum and the Varoufakis episode, the mirror this conversation keeps returning to

## One thing to take away

Whether something goes wrong often has little to do with the number itself, and a lot to do with who holds the power to change it and whose consent they need. France's deficit has run for fifty years; the market is nervous now because the odds of someone being able to pass a budget got marked down.

Here is something I have tried that needs no portfolio. Pick a thing you are currently worried about — moving house, the project stuck at work, the conversation in a relationship you keep postponing — and write two lines on paper. Line one: who holds the decision. Line two: whose agreement that person needs in order to act. Then look at line two. If the answer is "two people who never speak to each other," you know why it has not moved, and you know which side to push instead of staring at the surface metrics. I have done this a few times, and twice the surprise was finding the decision was mine all along, while I had been filing it under someone else's name.

![A node holding the decision is linked by one line to each of two nodes that must nod, and the line between those two is broken and crossed out, so the decision never gets through.](/figures/who-must-nod-en.svg)