What Happens to Governments When Refinancing National Debt Gets Much More Expensive?

Photo: Farcaster at en.wikipedia, Wikimedia Commons, Public domain

01 A refinancing wall, not a borrowing boom

The defining fiscal fact of the second half of this decade is not how much governments are borrowing — it is how much of what they already borrowed comes due. Roughly a third of outstanding U.S. marketable debt matures within twelve months, and trillions more in notes issued at 1-to-2-percent coupons during 2020 and 2021 reach their maturity dates over the next few years. Each maturing security must be refinanced at prevailing yields — yields that sit far above the coupons being retired.

This is the refinancing wall: the cost of standing still, before a single new dollar of program spending is approved. Interest outlays have already more than doubled in five years — from $345 billion in fiscal 2020 to $882 billion in fiscal 2024 — and the line is still climbing because the average rate on the debt continues to reprice upward as cheap securities roll into expensive ones.

Analysis — not prediction. N43 and Hermes AI grounds every scenario in the documented record and verified reporting as of September 21, 2026; where evidence is incomplete we say so.

02 Why rollover is the whole game

Interest cost is a stock times a rate: the debt outstanding multiplied by the average interest rate paid on it. Politicians argue about the deficit — the flow — but the interest bill is set by the stock, and the stock reprices only as fast as the debt matures. The average rate on marketable debt rose from about 1.6 percent in 2021 to roughly 3.4 percent by 2025, and it keeps climbing because the market yield at which new debt is issued — near 4.8 percent on the ten-year Treasury in September 2026 — remains well above the blended average being retired.

The Treasury has also shortened its own stack. After 2023 it leaned heavily on bills to fund deficits, which means an even larger share of the debt reprices almost continuously. A government financed at the short end of the curve gets no protection from having locked in low rates years ago; its interest bill moves with the market within months, not decades.

03 The compounding arithmetic of interest

The arithmetic compounds because interest is itself borrowed. When net interest approaches $1 trillion a year, the Treasury issues debt to pay it, adding to the stock, which adds to next year's interest. At roughly $40 trillion of debt, each single percentage point on the average rate is about $400 billion a year — larger than the annual budgets of most federal departments, and it buys nothing.

And the average rate is still below the marginal rate. Even if yields never rise another basis point from here, the interest bill keeps growing for years, purely from the mechanical replacement of retired low-coupon debt with market-rate debt. That is what makes the refinancing wall different from a rate shock: the shock has already happened, and the bill is simply arriving in installments.

04 Term premium: the price of duration

How much refinancing costs depends not only on the policy rate but on the term premium — the extra yield investors demand for holding long-dated bonds instead of rolling short ones. Through the 2010s the term premium was negative: inflation was dormant, the Federal Reserve owned trillions of Treasuries, and price-insensitive buyers absorbed duration at almost any price. Refinancing long was cheap, and governments locked in low rates for decades.

That regime has reversed. The Fed is shrinking its portfolio, foreign official demand has grown more selective, fiscal supply is flooding the market at record pace, and inflation — still volatile in 2026 — has reminded investors that long bonds can lose real value. Term premium measures compiled by the New York Fed turned positive in 2024 and 2025 and have stayed there. The practical consequence: a government that wants to borrow for ten or thirty years now pays materially more than the expected path of short rates alone would imply — or borrows short and gives up the rate certainty that used to be nearly free.

05 The CBO trajectory toward $1 trillion and beyond

The Congressional Budget Office's baseline projections carried net interest from $882 billion in fiscal 2024 past the trillion-dollar mark in the middle of this decade, toward roughly $1.4 trillion by fiscal 2029 — and toward a share of the economy last seen at the early-1990s interest-rate peak. By 2024, net interest had already passed national defense spending, and on current baselines it approaches the scale of the largest retirement programs within a decade.

These projections assume no recession, no new war, no financial crisis — the boring scenario. Stress scenarios are worse. The point of the baseline is not that it is a forecast but that the wall is now embedded in the stock of debt: most of the projected interest spending is the mechanical arithmetic of refinancing what has already been borrowed, and it would take radical policy change simply to hold the line flat.

06 The emerging-market version of the squeeze

The United States refinances in the currency it prints, with the deepest bond market on earth. Emerging markets have the same wall and none of the privilege. Sovereigns that issued Eurobonds in 2020-21 at 5-to-7 percent have been rolling maturities in a market where stressed credits face double-digit yields; the episodes of Egypt, Kenya and Pakistan through 2024-2026 all involved refinancing foreign-currency debt at rates that made the debt ratio worse by arithmetic alone.

The International Monetary Fund has warned for several reporting cycles that elevated sovereign refinancing needs extend through the decade for lower-income countries. A hard-currency rollover at a yield above the growth rate is a slow-motion default math: the debt ratio rises even if the government balances its primary budget. Where the United States gets a compounding interest bill, an emerging market without reserve-currency status gets a rollover crisis — the same wall, hit at speed.

07 What would bend the wall

Four things historically bend a refinancing wall. Primary surpluses shrink the stock that must roll — politically rare, and none is on the current U.S. baseline. Faster nominal growth erodes the debt-to-GDP ratio — but growth sufficient to outpace 4-to-5-percent borrowing costs is a demanding assumption. Lower rates require the disinflation and the term-premium compression of the 1990s, and 2026 has so far delivered the opposite. And financial repression — capping yields, directing captive buyers into government debt — works, at the documented cost of distorted markets, as the post-war decades showed.

What to watch instead is narrower and more honest: the bill share of Treasury issuance, the term premium, and the tails at long-maturity auctions. They will say whether markets are still volunteering to hold the refinancing at a price the government can pay — or whether the wall, currently arriving in installments, starts arriving in lumps.