On the seventeenth of October last year, four days after Prime Minister Sébastien Lecornu submitted a budget for 2026 to a National Assembly disinclined to receive it, S&P Global Ratings did something the agency had reserved, until then, for the calendar's prescribed dates. It downgraded France from AA minus to A plus, two months ahead of the scheduled review and on the strength of a single, troubling sentence: "the political and budgetary structures may lack the capacity to deliver deeper consolidation."¹ The market response was not a panic. It was something quieter and, in the long view, more consequential. The spread between the ten-year Obligation Assignable du Trésor and its German Bund counterpart, which had drifted in the high seventies through the summer, settled in the eighties and remained there. By the twenty-third of March this year, with Lecornu's government still standing only because it had bartered away the pension reform that justified its existence, the spread reached eighty-three basis points.² It is worth dwelling, for a moment, on what this very small number means.

A spread is not a verdict. It is a price, and prices in sovereign bond markets are unusually patient interpreters of political reality. Italian ten-year paper, for the first time in nearly two decades, now trades only one hundred and thirty to one hundred and fifty basis points wide of Bunds,³ a development that would have been unthinkable in the spring of 2012, when the same spread briefly exceeded five hundred. France, the second economy of the euro area and historically the bloc's senior fiscal partner, has watched its own premium climb from twenty basis points in 2017 to three times that figure today, even as Italy's has fallen by half. The two curves have not yet crossed. They are, however, on trajectories that converge.

The Banque de France was founded in 1800 by Napoleon, who understood, with the clarity that often attaches to those who require credit, that a state which cannot borrow cheaply cannot make war. For most of its history the institution has been the supplicant of governments, and for nearly two decades after the introduction of the euro it ceased to matter what supplicants the French Treasury had chosen, because European integration had achieved what generations of French governors had attempted in vain: a permanent compression of French borrowing costs to the level of the Bundesbank's. That compression is now, by stages, being unwound. The unwinding is not a crisis. It is something more like an audit.

The accompanying figure traces the two trajectories. They tell a story that the rating agencies and the IMF, in their separate registers, have been telling for some time. France's gross public debt stood at one hundred and sixteen and a half per cent of gross domestic product at the end of 2024;⁴ the European Commission expects the figure to reach one hundred and twenty per cent by 2027;⁵ and the IMF, looking further out, projects a path towards nearly one hundred and thirty per cent by 2030.⁶ The deficit, which the Bayrou government had hoped to bring to four point six per cent of GDP this year, is currently forecast by Brussels to land at four point nine,⁵ and the structural improvement implied by either trajectory has been described by the IMF as requiring "a frontloaded structural fiscal effort of 1.1 per cent of GDP in 2026, followed by an average of about 0.9 per cent of GDP per year over the medium term."⁴ The numbers are not catastrophic. They are merely incompatible with the political arithmetic of the National Assembly, which has now defeated, modified, or quietly buried every consolidation measure presented to it since the snap election of June 2024.

Since that election, the Agence France Trésor has issued four hundred and fifty seven billion euros of debt at tenors longer than one year,³ and has done so without difficulty. This is the curious feature of the present situation, and it is the feature most likely to mislead. France has no funding problem. It has, instead, a slowly compounding pricing problem. Every basis point of additional spread, applied to a stock of sovereign debt approaching three trillion euros, costs the French taxpayer roughly three hundred million euros per year in additional interest at full rollover. The eighty-three basis points of late March, compared with the twenty of 2017, imply an eventual incremental burden, when the existing stock has fully turned over, of approximately nineteen billion euros annually. That is more than the entire annual budget of the Ministry of Justice. It is, to put the matter plainly, the cost of legislative deadlock, paid by future taxpayers in instalments, and recorded in real time on the screens of bond traders in Frankfurt and London.

The historical parallel that recommends itself is not, as is sometimes suggested, the eurozone crisis of 2010 to 2012. That episode was a crisis of redenomination risk, and France, as a core sovereign with its own central bank standing behind the euro, was never plausibly subject to it. The more instructive parallel is older and, for that reason, more uncomfortable: it is France in 1981 and 1982, when the Mitterrand government, having pursued an expansionary programme financed by deficits, found the franc under sustained pressure within the European Monetary System and was compelled, in March 1983, to abandon expansion in favour of what came to be known as the "tournant de la rigueur." The tournant was not a moral conversion. It was the exhaustion of options. The French government discovered, as governments before and since have discovered, that there is no monetary or fiscal escape from a market that has decided to charge a premium until the underlying fiscal position improves. The lesson of 1983, learned at considerable political cost by the Socialists, was that markets do not require panic in order to enforce discipline. They require only persistence.

Persistence is precisely what the bond market has now adopted as its posture towards France. The chart above shows no spike, no crisis, no moment of capitulation; it shows a steady and undramatic widening, punctuated by political events, that has resisted every attempt by the French Treasury to talk it down. The Banque de France's December projections, which expected GDP growth of eight tenths of a per cent in 2026 and one and a half per cent in 2027,⁷ assume a fiscal trajectory that has not yet been legislated. The European Commission's forecast for a four point nine per cent deficit this year⁵ assumes a degree of expenditure restraint that the National Assembly has yet to authorise. The market, looking at this gap between assumption and achievement, has assigned it a price, and the price is reset weekly.

The Optimist's Case

It would be an error, however, to read the spread as a unanimous indictment. There is a respectable case, advanced by some on the optimistic side of the rating agencies' ledger, that the OAT remains unusually well positioned. France retains an exceptionally deep domestic savings pool, an investor base that includes large insurance companies with long-dated euro liabilities, and an issuance programme run by the AFT with a duration profile that has been gradually extended to lock in lower funding costs. Lombard Odier, in October of last year, observed that the eurozone as a whole was stabilising even as France remained "the weak point,"⁸ a formulation that contains within it both a warning and a reassurance. The bloc's collective creditworthiness has not been called into question. It is the relative position of France within the bloc that has shifted, and the absolute level of French borrowing costs remains, by historical standards, low.

There is also the institutional consideration, which one underestimates at one's peril. The European Central Bank, having created in July 2022 the Transmission Protection Instrument, possesses an explicit mandate to intervene in sovereign bond markets where it judges the transmission of monetary policy to be impaired by "unwarranted, disorderly market dynamics." Whether the present widening of the OAT-Bund spread would qualify is a question the Governing Council has wisely declined to answer in advance. What is not in doubt is the existence of the instrument itself, and the discipline it imposes on speculators considering whether to test it. The TPI has not been used. Its mere availability has, it is reasonable to argue, contributed to the orderliness of the present widening.

What, then, is to be watched? Three variables, in my view, will determine whether the spread converges with Italy's, stabilises in the eighties, or returns toward the historical norm of twenty to forty basis points. The first is whether the Lecornu government, or its successor, can pass an annual budget that the rating agencies, the IMF, and the Commission accept as a credible consolidation path; the answer, on present evidence, is uncertain at best. The second is the trajectory of French growth, which the Banque de France has revised down repeatedly and which now depends, more than is comfortable, on a recovery in private investment that has been deferred year after year. The third is the conduct of the European Central Bank itself, which has held its key rates unchanged at its March meeting⁹ and which, in any tightening scenario, would amplify the cost of every basis point of French spread by raising the underlying Bund yield against which it is measured.

The Reichsbank Parallel, Properly Drawn

The Reichsbank, in the years between 1924 and 1929, was widely admired as the best-managed central bank in Europe. Hjalmar Schacht's stabilisation of the mark, his judicious use of foreign loans, and the apparent solidity of the Dawes Plan led contemporaries to believe that Germany had returned to financial respectability. The market, throughout this period, charged Germany a steadily declining premium over British and American debt. What it did not record, because it was not asked to, was the fragility of the political settlement on which the financial respectability rested. When that settlement collapsed in 1930, the premium reasserted itself with terrifying speed.

I do not draw the parallel to suggest that France in 2026 resembles Germany in 1929. The institutional framework is incomparably stronger, the political fragmentation incomparably less severe, and the external environment incomparably more supportive. I draw the parallel because the OAT-Bund spread is the only honest record, in real time, of how a market is pricing the gap between a nation's stated fiscal intentions and its observed political capacity to deliver them. That gap, for France, has been widening for two years. It has been widening slowly. It is being recorded patiently. And it is, in the end, the kind of widening that no central bank instrument, however well designed, can reverse without the cooperation of the institution that issues the bonds.

That institution, at the time of writing, sits in the Palais Bourbon, divided into three roughly equal blocs that have demonstrated, with admirable consistency, that they can prevent any consolidation from passing while declining to accept responsibility for the consequences. The bond market, watching this, is doing what bond markets do. It is naming the price.