French debt: the price of time

A government can lower its interest bill while committing itself to ask investors for money more often. The benefit shows up in the first few budgets. The trade-off sits in the repayment calendar, where a future administration may have to deal with very different market conditions.

On 7 October, Reuters reported that Roland Lescure was considering a marginal shortening of issuance. The finance ministry clarified that issuance plans remained unchanged and responsive to demand. A wholesale move into short-term borrowing has therefore not been announced. ^[1]

That clarification leaves a worthwhile economic question. When borrowing for longer costs more, how much should a government pay to secure several years of predictable financing? How much immediate relief justifies bringing part of the debt back to market sooner? The choice affects taxpayers, the investors holding the bonds and the room for manoeuvre of the next government.

The clearest way to understand it is to follow the same sum over the same period. Our illustration finances €10 billion for ten years. All rates are hypothetical. It compares a ten-year bond with five successive two-year bonds. Both strategies initially raise €10 billion and ultimately repay the same principal. What changes is the schedule of decisions in between.

Several different decisions fit under “shorter”

Agence France Trésor uses two main families of instruments. BTFs are bills maturing in less than a year, used in particular to manage the timing of receipts, spending and redemptions. OATs finance the medium and long term, with maturities ranging from two to fifty years. A three-year bond already belongs to that second family. ^[7]

Replacing thirty-year issuance with ten-year bonds, favouring three years over ten, and increasing bills with a maturity of a few months all bring repayments forward at very different speeds. The first choice still locks in a distant horizon. The last returns the financing need to the Treasury’s near-term calendar.

In its programme published on 29 September, AFT projects €340 billion of medium- and long-term issuance, net of buybacks, in 2027, and a €2.2 billion increase in BTFs outstanding. The detailed issuance programme is due in December. The two figures measure an annual issuance programme and a change in the stock, respectively. ^[2]

A three-month bill can mature and be replaced several times without increasing the year-end stock. Conversely, moving an issue from thirty to ten years shortens its maturity without adding any BTFs. The published totals therefore leave the future distribution of bonds across maturities open.

The financing task also extends beyond covering the deficit. The State’s projected funding requirement is €339.7 billion in 2027, with the increase partly reflecting €19.4 billion more medium- and long-term redemptions than in 2026. Much of the job is to replace securities reaching the end of their lives. The €340 billion figure is net of early buybacks, rather than scheduled redemptions. ^[2]

That distinction matters when reading a political announcement. The programme determines how much must be placed under the stated convention. Maturity determines when those securities will return for refinancing. The primary deficit, interest payments and other financial transactions determine the path of debt outstanding. Adjusting maturities affects that path through cost and risk; decisions on revenue and spending remain essential.

France already borrows across several horizons

AFT’s September bulletin provides a useful view of the composition. From January through August 2026, nominal medium- and long-term issuance was spread across several segments. The ten-year bucket accounted for €90.2 billion, compared with €12.2 billion at thirty to fifty years. Inflation-linked securities were reported separately. Our chart retains the nominal-security perimeter and the eight-month reporting period. ^[4]

This distribution describes choices already made. Establishing a recent change would require comparisons over matching periods, adjusted for major individual transactions and investor demand. A full year of 2025 and eight months of 2026 cannot on their own support a clean trend comparison. A large syndicated transaction can concentrate a substantial part of thirty-year issuance in February.

The indicative 2026 programme already provided for benchmark issues at three years, five or six years and ten years, as well as longer maturities depending on market conditions. Flexibility is built into this framework. The agency maintains a range of lines instead of trying to find a single buyer for an entirely new security whenever financing is needed. ^[11]

The stock tells another story. On 30 September, marketable State debt stood at €2,896.2 billion, including €217.7 billion of BTFs, or around 7.5%. Its average remaining maturity was eight years and 158 days. These are central State securities, a narrower perimeter than debt of the whole general-government sector. ^[3]

The stock embodies many years of decisions. Changing a small portion of issuance affects it at the margin; repeating the decision gradually changes the portfolio. The current month’s transaction, the year’s programme and inherited obligations must therefore be considered together. An average alone conceals their different time horizons.

An eight-year average can conceal much earlier repayments

Average maturity measures the time remaining until principal repayment, weighted by the amounts outstanding. It describes the portfolio’s centre of gravity. It does not answer the practical question of how much cash must be found over the next two years.

Consider two fictional €100 billion portfolios. One repays €20 billion in years four, six, eight, ten and twelve. The other repays €50 billion in year two and €50 billion in year fourteen. Both have an eight-year average maturity. Yet the second must refinance half its principal in two years, while the first has more time. Neither portfolio represents France.

The actual calendar has its own concentrations. The September bulletin’s snapshot, dated 31 August 2026, shows more debt maturing in 2029 than in any of 2027, 2028 or 2030. The amounts in the chart include inflation-linked securities using the indexation coefficients at the observation date. They locate maturities within the reported stock; they are not final forecasts of redemption cash. ^[4]

Buybacks can bring part of the refinancing forward. New issues can add to existing lines. Inflation changes the principal of indexed securities. An observed 2029 maturity is therefore an obligation to manage whose amount and financing will evolve. Comparing it with the budget’s 2027 funding requirement would require reconciling these transactions and accounting conventions.

This becomes especially important when the State chooses new two- or three-year borrowing. Those issues fall into years that already have obligations, spending commitments and uncertainties. Preserving market access over time requires managing concentrations as well as averages. The French Treasury explicitly described maturity smoothing as an objective in its January 2022 study of issuance strategy. ^[8]

The saving in the first two years

Return to the illustrative €10 billion. Assume issuance at par: €10 billion of cash for €10 billion of principal, annual coupons and no indexation. A ten-year bond paying 5% costs €500 million a year in interest. A two-year bond paying 3% costs €300 million.

The second choice saves €400 million over the first two years. That saving is real within the model. At the end of year two, however, the €10 billion must be repaid. To finance the same underlying requirement, the State borrows another €10 billion at the terms then available. It repeats the exercise in years four, six and eight. Final principal repayment falls at the same ten-year horizon as the long bond.

Gross issuance consequently adds up to €50 billion under the shorter strategy, including the initial issue, against €10 billion under the longer one. Outstanding principal remains €10 billion until the final repayment. Intermediate issues replace maturing principal. Their frequency measures dependence on market access, rather than €50 billion of additional spending.

Interest is assumed to be paid from the budget, without being capitalised into fresh debt. Fees, cash buffers and any market response to the strategy are excluded. These simplifications isolate the maturity decision. They also prevent the result being treated as a forecast of France’s interest bill.

Three paths show the stakes. If subsequent two-year bonds are issued at 2%, cumulative interest under the short strategy reaches €2.2 billion over ten years. At 3%, it reaches €3 billion. If every refinancing after year two costs 6%, total interest rises to €5.4 billion, against €5 billion for the long bond. The initially cheaper choice becomes the more expensive one over the full period.

The mechanism is gradual. In the 6% path, the early €400 million saving is eaten away by an additional €100 million of interest each year from year three. It disappears by the end of year six. The final four years leave the short strategy €400 million behind. No individual bond breaches its terms: the cost comes from the sequence of new contracts.

The rate that erases the head start

Under these assumptions, refinancing at a constant 5.5% from year three onwards produces exactly the same nominal interest bill as borrowing for ten years at 5% from the start. The first two years at 3% allow the short strategy to bear a higher rate later before its early advantage disappears.

The threshold can be read another way. Every additional percentage point on the €10 billion costs €100 million per year while that rate applies. Spread across the remaining eight years, the €400 million head start can absorb half a percentage point above the long bond’s 5% coupon. The calculation concerns the whole sequence, rather than the first refinancing in isolation.

For changing refinancing rates, the comparison requires the actual interest paid over each two-year period. A single costly renewal does not automatically make the ten-year total worse. It may be followed by cheaper funding. A low first coupon is equally insufficient to establish that the full strategy saves money.

Discounting changes the comparison slightly because later payments receive a smaller weight. With a hypothetical 4% annual discount rate, the constant future refinancing rate that equalises the present value of interest is about 5.606%. Both strategies repay the same principal at the same final date. Intermediate principal repayments in the short strategy are matched one for one by new borrowing, so their net cash flows cancel in this particular model.

The nominal and discounted thresholds answer different questions. One adds up payments; the other gives them a chosen time value. The table below the figure reports the nominal break-even rate and the thresholds at several discount rates. The 4% assumption is neither a forecast nor an official AFT decision rate.

Another exposure remains: investors must be willing to finance each new maturity. Even the 6% scenario assumes that all principal can be placed. Refinancing risk includes a temporary inability to raise the required amount, which calls for a liquidity response. A coupon calculation alone leaves that risk unmeasured.

Several forces shape the slope of the curve

Why might ten-year borrowing cost more? The lender agrees to receive coupons fixed today for a longer period. If rates rise, the security becomes less attractive to a new buyer. Inflation and future resale conditions add uncertainty.

A long yield incorporates expectations of future short rates and compensation for holding duration, usually called the term premium. BIS research shows that estimates of this premium depend on the model and can reflect shifts in supply and demand. Subtracting a two-year yield from a ten-year yield therefore does not directly reveal its size. ^[10]

Investors in a particular sovereign also consider credit quality, security liquidity and conditions specific to each maturity. Treating the entire French curve slope as a political-risk premium, or as a pure insurance premium against higher rates, would conflate different forces.

AFT’s daily series reports a 4.90% TEC10 on 8 October, compared with 4.83% on 7 October. BTFs auctioned on 5 October carried published rates ranging from 2.786% at thirteen weeks to 3.262% at fifty-two weeks. These observations illustrate different financing conditions. They concern different dates, maturities and conventions; a simple subtraction would not provide a defensible term-premium estimate. ^{[12] [6]}

The strongest argument for some shortening deserves a fair hearing. When investors demand unusually large compensation for committing funds far into the future, buying that protection mechanically can be costly. The OECD discusses this trade-off in its March 2026 report, drawing mainly on developments in 2025. It also stresses liquidity and predictable issuance. ^[9]

A debt office may therefore favour an intermediate maturity over an unattractively priced very long issue while continuing to issue long bonds. The value of that flexibility depends on how much is shifted, the existing maturity profile and the buyers available.

A 2% coupon can still mean borrowing at 5.40%

Reading an auction requires one further distinction. The interest rate in an OAT’s name is its coupon. The State can reopen that line much later at a price well below its redemption value.

The 1 October 2026 auction provides an example. The 2% OAT due 25 May 2048 had a weighted average price of 57.23% of face value and a weighted average yield of 5.40%. An initial €2.063 billion of face value was allocated on that line, before additional non-competitive allocations. Investors were still funding this long maturity at the published terms. ^[5]

AFT’s final-results table, checked on 8 October, brings this line to €2.795 billion of face value after €732 million of additional non-competitive allocations. Total face value for the 1 October OAT auction rises from €11.999 billion to €15.606 billion; the 5 October BTF auction rises from €6.700 billion to €8.642 billion. These volumes include the additional allocations; the figure retains the initial auction’s prices and yields. ^[13]

For €100 of principal, the quoted price is €57.23 excluding accrued interest. The bond pays an annual €2 coupon and repays €100 at maturity. Its yield reflects the full sequence of cash flows, including the gap between purchase price and final repayment. Actual settlement also includes accrued interest.

Reopening an older low-coupon line can thus reduce annual coupon payments per euro of face value while raising less initial cash for that same face value. Any apparent saving must be measured against the principal promised at the end. Price and the full payment schedule make two financing options comparable.

Issue discounts also appear in AFT’s projected 2027 cash resources. They are a reminder that an issuance programme expressed in face value need not equal the cash received. The lowest stated coupon and the lowest funding cost are different criteria. ^[2]

Who bears the risk while the clock runs?

A long fixed-rate bond protects its issuer’s payment schedule. Its market value can move substantially for the investor holding it. A shorter bond usually reduces that holder’s price exposure, while returning the borrower to a refinancing decision sooner.

Our next test uses two hypothetical securities issued at par: two years at 3% and ten years at 5%. Their market yields rise immediately by one percentage point, with no passage of time and no default. The first bond then trades at about 98.11 per 100 of face value; the second at 92.64. Their contractual annual coupons remain €3 and €5 respectively. Every input in this test is illustrative.

The long-bond investor therefore absorbs more market-value variation in this shock. Selling before maturity can realise the loss; its accounting treatment depends on how the asset is held. The State continues to owe the same coupon on the existing bond. Higher yields affect new financing and debt that must be renewed.

For an insurer with distant liabilities, holding long bonds can nevertheless reduce risk across the whole balance sheet. The insurer is also trying to match asset receipts to payments promised in the future. A security’s standalone price risk and its usefulness within a portfolio can therefore diverge. Research on term premia highlights the importance of investors’ different horizons. ^[10]

Indexed debt adds another clock. Its maturity may be distant while inflation affects principal and coupons much sooner. Legal maturity alone is an incomplete guide to how quickly every shock reaches the budget. Our rate comparison deliberately holds nominal fixed-rate securities constant. ^[7][14]

Investor demand helps set the price of certainty

Governments finance themselves through investors with different constraints. Money-market funds, banks, insurers and asset managers do not all want the same maturities. Issuance that ignores this demand may require a price concession to clear.

The French Treasury describes the logic in terms of investors’ preferred horizons. Its 2022 study treats a diverse investor base and liquid benchmark lines as elements of financing security. This is useful evidence about the issuer’s approach, rather than a numerical evaluation of the 2026 decision. ^[8]

The OECD also identifies international changes in demand, including those associated with pension portfolios in some countries. Their French impact would require data by maturity and investor group. Non-residents’ aggregate holdings, for example, describe geography while leaving behaviour in a stressed market unresolved. ^[9]

Auction results provide more immediate evidence. A bid-to-cover ratio compares bids received with the amount allocated. It should be read alongside price, the size offered and prevailing conditions. Abundant demand at a high yield does not establish that the same financing would have been equally easy at a much lower yield.

Reducing a long issue can therefore reflect a decision on price and placement rather than a loss of market access. A persistent sequence of price concessions and shortened maturities would, however, deserve monitoring. Identifying a change in financing conditions requires observing both repetition and its effect on future redemptions.

Savings matter most when the budget is already stretched

The budgetary debt charge is projected at €72.9 billion in 2027, against €62.6 billion in the updated 2026 estimate. Those forecasts reflect much more than any potential change in maturity choice. Their scale nevertheless explains the political appeal of immediate savings. ^[2]

In the €10 billion model, saving €200 million a year initially is small against that total. Repeating the operation increases the initial gain, but also the principal that must be placed at nearer dates. The market’s ability to absorb issuance can change with the volume. Multiplying our result by the entire €340 billion programme would misleadingly assume that every maturity is freely interchangeable at unchanged prices.

The fiscal risk becomes most relevant when difficulties arrive together. In an adverse scenario, refinancing rates rise as receipts disappoint or new spending becomes necessary. Extra interest falls due just as room for manoeuvre contracts. The value of long debt then lies in the period for which this part of the bill remains known.

The favourable scenario also matters: disinflation and stronger fiscal prospects can allow cheaper refinancing. Shorter borrowing benefits from that improvement sooner. Maturity choice therefore determines how much of future market conditions the budget accepts. A diversified debt strategy spreads that exposure instead of concentrating it on one date.

Issuance will show how maturity choices evolve

Any shift will become visible in the actual distribution of issues, amounts falling due over the next few years, buybacks and placement terms. The detailed programme expected in December will make it possible to compare intentions with amounts. The documents examined so far contain neither a quantified shift across segments nor an official saving attributable to a new maturity decision. ^{[2] [11]}

For taxpayers, the useful question is precise: how much principal returns to market earlier in exchange for the saving sought, and what resources can absorb an unfavourable refinancing? Average maturity remains informative when accompanied by a principal repayment calendar and a sufficiently demanding rate test.

Shorter borrowing can be a rational choice. Its quality depends on the savings that survive later refinancings and the ability to get through a difficult maturity. The first year’s coupon shows what the State gains now. The calendar shows when the next decision must be made.

Method, data and limitations

Research is cut off at 8 October 2026. Stock and issuance data come from AFT and retain their own reference dates. The September bulletin reports issuance and stocks through August, while the separate stock summary is dated 30 September. Figures for 2027 are projections. The original Wall Street Journal interview was not consulted in full; the news framing relies on Reuters’ updated account including the ministry’s clarification.

The models use constant principal, issuance at par and annual coupons. Principal refinancings are separated from interest. No probabilities are assigned to interest-rate paths, and loss of market access is not priced. The valuation test assumes an immediate parallel yield shock. The equal-average-maturity portfolios are fictional. The expandable tables below each figure provide the data, assumptions and calculations. Annual interest is principal multiplied by the coupon rate; its present value is that payment divided by (1 + discount rate) raised to the number of years elapsed.

The public documents examined do not provide the price assumptions, placement constraints, future deficits and liquidity choices needed to optimise France’s whole debt portfolio. The results explain and test mechanisms. They do not identify a single optimal maturity for the State.

For the public-finance context, read France’s 2027 budget and bond-market risk. Our investigation into European banks and sovereign debt follows the consequences for holders. The guide to reading European sovereign debt explains the indicators to compare.

Sources and documents