The Treasury's buyback failure signals a fiscal-credibility break at 5% bond yields

When the Treasury wanted to keep the ten-year note below 5%, it deployed the largest buyback under the expanded program. On September 9th Scott Bessent, the treasury secretary, announced a buyback of up to $6bn of 10- to 20-year bonds, three times the size of a normal operation, to be executed the next day. The move followed a selloff that had pushed the 30-year yield to levels not seen since 2007 and the ten-year to its highest since late 2023. The market understood the operation as a test: how hard would the fiscal authority fight for 5%, a level the administration had openly named as one it does not want crossed?

The answer came within a day, and it was not reassuring. Dealers offered the Treasury $10.5bn of bonds. It accepted $5.2bn—barely half the offers, and, in a rare turn, less than its own $6bn ceiling. Dealers' bids sat above the price the Treasury was willing to pay, so the department, armed with its enlarged tool, accepted only the bids it judged competitive and rejected the rest. The result was the opposite of the plan: yields rose, and the next day the ten-year touched 4.99%, its highest since 2023 and a whisker from the very line the operation had been meant to defend.

The detail that matters is that this was not an under-subscribed sale. The market was not short of the bonds; it had plenty to sell. The Treasury rejected roughly half of a $10.5bn pile of genuine supply on price, and still lost the argument. That combination—a selective buyer, an overflowing market, a yield that rose anyway—is the unit of evidence from which a fiscal-credibility story is built.

A tool of the wrong size

A buyback is not a monetary tool, and it cannot do the work of one. Its stated purpose is liquidity support: buying older, less-liquid securities so dealers have somewhere to park inventory. Its capacity is trivial next to the problem it now faces. A $6bn operation set against $32 trillion of marketable debt held by the public is a rounding error. A fair sceptic could object that with such an instrument, no amount of buying could have held yields down, so its "failure" proves only that the tool is too small, not that the fiscal path has broken. That objection deserves weight, and it is precisely why the buyback alone proves little.

The trouble is what surrounds it. Record issuance has pushed the national debt to $40 trillion, more than double its level at the start of the decade; a $25bn auction of 30-year bonds this year cleared at its richest yield since 2021, and interest on the debt now exceeds $1 trillion a year, the second-largest line in the federal budget. Foreign buyers, who hold nearly a third of Treasuries, have been buying less, throwing more supply onto price-sensitive domestic investors. The result is legible in the term premium—the extra compensation investors demand for holding long-dated bonds rather than rolling a series of short ones—which has climbed to its highest level in more than a dozen years. That premium, as the Kansas City Fed's work shows, is the channel through which heavy supply transmits to long yields, independent of what the central bank does with short rates.

This is the subtlety at the heart of the 5% move. A hot inflation reading (August core CPI rose 0.3% month on month, above forecasts) and oil above $100 have pushed markets to price roughly a 90% chance that the Fed raises rates on September 15th. On its face that is a tightening—the opposite of a central bank captive to the deficit. Yet the ten-year has climbed even so, because the long end is driven less by expectations of near-term policy than by supply and term premium. That is fiscal strain at its boundary: not a treasury printing money to fund spending, but a market that prices long debt as if the fiscal path, and its inflationary consequences, are the binding constraint. The Fed cannot offset the fiscal strain by tightening if tightening is itself a measure of how far the strain has gone.

The test is not yet over

The competing case deserves to be stated at full strength. Inflation is genuinely hot—tariffs, an oil shock from the war with Iran, an AI-infrastructure buildout all push the same way—so a rising ten-year is partly a correct repricing of policy rather than pure fiscal doubt. And a credible, independent Fed that hikes and compresses the term premium would falsify the fiscal-dominance reading cleanly. The test is therefore live, not settled. If the ten-year durably retreats below roughly 4.8%, or the term premium compresses even as the central bank tightens, then this week was a repairable liquidity-and-inflation squeeze, and the crossing was a number rather than a verdict. Nothing in the evidence yet confirms that escape.

What leans the other way is that the Treasury has turned a liquidity facility into an informal price target without the capacity to enforce it. Bessent expanded buybacks, named 5% as a boundary, then—offered far more supply than he would take, and losing the trade anyway—advertised the limits of his own tools. That advertisement is what moves expectations, and it compounds. If buybacks keep failing to contain the selloff, Bank of America's Mark Cabana suggests the next steps could be cuts in long-end auction sizes or a rethink of the 20-year bond—a retreat from the supply side that would itself confess the squeeze. Each escalation confirms that the fiscal authority is, in effect, negotiating with its creditors for the price of its own debt, and negotiating from a position it may not hold.

That is a condition worth watching closely, though not for the reason headline-chasers assume. The question is not whether the ten-year prints 5%; it has already come within a basis point of it. It is whether, once there, the term premium stays high—which would tell creditors (and the reader) that the fiscal path has been judged not self-correcting—or whether a hawkish but independent Fed can compress that premium, which would show monetary policy still has a floor under the long end. The buyback's failure points in one direction only because it now sits inside a larger fact: the market has begun to price the fiscal path as the constraint that binds. That price may yet be wrong. But the pressure being applied to the central bank that could prove it wrong—by an administration openly demanding lower rates—is itself the pressure the market is being paid to carry.