The Wolf Street analysis of the Bessent Treasury buyback programme cuts straight to the mechanism the bond market cheerleaders prefer not to discuss: this is not money creation, it is debt swapped for more expensive debt, dressed up as a policy signal.
The Treasury Department announced it would ‘at least double’ its buyback auctions of 10-year notes, 20-year bonds, and 30-year bonds, beginning 9 September and running through 4 November. Maximum auction sizes move from $2 billion to $4 billion per session. Seven auctions are scheduled across the two maturity sectors in that window, lifting total buybacks from $14 billion to $28 billion.
Why the Bessent Treasury Buyback Programme Is Not QE
The popular conflation of buybacks with quantitative easing deserves a direct rebuttal. QE involves money creation: a central bank purchases securities with newly created reserves. The Treasury cannot create money. Every dollar it spends on buybacks must first be borrowed, because 100% of tax receipts are already committed to existing obligations. What the Treasury is doing is issuing new debt, typically short-duration T-bills at close to 4.0%, to retire older long-duration bonds carrying coupons as low as 1.875%. The debt stock shifts in maturity and cost, but the money supply does not move.
The scale reinforces the point. There are roughly $4.4 trillion of 10-year notes outstanding and $5.5 trillion of 20-year and 30-year bonds outstanding, a combined total approaching $10 trillion. The additional $14 billion in buybacks over the September-to-November window represents approximately 0.14% of those securities. That is not a market-clearing force. It is a signal dressed as a programme.
The Arithmetic of Buying Cheap Bonds Dear
The buyback conducted today illustrates the interest-expense trap neatly. The Treasury purchased $175 million worth of a 30-year bond issued in February 2021, carrying a coupon of 1.875%, at 52.375 cents on the dollar. To fund that purchase, it borrowed at approximately 4.0%, today’s T-bill rate. In dollar terms: about $7 million in annual interest on the new borrowing to retire about $6.3 million in annual interest on the old. The face-value debt declines modestly because of the discount, but the interest burden in dollar terms edges upward. It is a transaction that costs more to service, not less.
Wolf Richter, writing for Wolf Street, frames this as the defining irony of the programme: the Treasury is swapping its cheapest liabilities for its most expensive ones, and describing the exercise as liquidity support.
The precedent is instructive. Treasury Secretary Yellen initiated this approach in April 2024, after the 10-year yield briefly touched 5% in October 2023. According to Bit, the standing buyback programme has been running since May 2024 on a regular published schedule, and by 2025 it amounted to roughly 0.6% of total Treasuries outstanding in a given year. Bessent is now scaling up a tool his predecessor installed, not inventing one.
The longer history is worth setting alongside that continuity. U.S. Department of Treasury Fiscal Data records show that from 2003 to 2013 there were no buyback operations at all. The Treasury had run a buyback programme in 2000 through 2002, then suspended it entirely for a decade before reviving it on a negligible scale (amounts of roughly $25 million a year) from 2014 onward, purely for operational maintenance. The current programme, whatever its merits, represents a more sustained political commitment to the tool than the post-2014 skeleton regime suggested.
The near-term market effect has landed as intended. The 10-year Treasury yield dipped by about 5 basis points on the announcement, and the 30-year yield fell by around 8 basis points. The same pattern followed Bessent’s earlier move: the joint US-Japan intervention at the start of August produced a brief rally in bond prices and a corresponding dip in yields, before reversing. The 30-year yield subsequently rose to a new two-decade high, and 30-year bonds were sold at the highest yield since 2001 at the following Treasury auction.
The consensus read today treats the doubled auction sizes as a substantive liquidity backstop. The more uncomfortable read is that $28 billion across seven auctions, against a $10 trillion outstanding stock, cannot materially alter supply-demand dynamics, and that the Treasury’s own interest arithmetic works against the stated rationale. Whether the bond market’s brief enthusiasm for the buy signal survives contact with the next long-duration auction is the question the cheerleaders are not asking.
