Treasury Is Offered Far More Than It Buys Back

On August 19 the Treasury said it would at least double the maximum size of its liquidity-support buybacks in the 10-to-20-year and 20-to-30-year sectors, from $2 billion to at least $4 billion per operation, from September 9 through the November refunding. The stated reason was the “significant volume of high-quality offers Treasury routinely receives in longer-dated buyback operations.” We wrote on August 31 that the Treasury had started managing the long end by buying it. The operations data answer a narrower question that matters more: of what dealers offer, how much does Treasury actually take?

• Offered
• Accepted

Dealers offer several times what Treasury buys in every month of the program. September's total covers operations through the 17th.

Snapshot: 2026-09-20 · buybacks-ls-offered source data · buybacks-ls-accepted source data

Data as of Sep 17, 2026 · Snapshot updated Sep 20, 2026 · Source: U.S. Treasury, Fiscal Data: Treasury Securities Buybacks (Operations)

Show the numbers (last 24)

Date	Offered	Accepted
2026-09-01	22.3	8.1
2026-08-01	81.6	11.1
2026-07-01	100.2	10.7
2026-06-01	70.6	7.3
2026-05-01	79.6	8.8
2026-04-01	94.8	6.4
2026-03-01	104.5	6.9
2026-02-01	149.5	12.3
2026-01-01	94.6	9.7
2025-12-01	66.2	7.8
2025-11-01	64.3	7.1
2025-10-01	107.9	11.0
2025-09-01	73.6	6.8
2025-08-01	63.0	8.6
2025-07-01	59.7	9.8
2025-06-01	36.2	5.1
2025-05-01	73.8	8.7
2025-04-01	37.4	6.9
2025-03-01	35.8	7.4
2025-02-01	60.5	8.5
2025-01-01	36.2	9.2
2024-12-01	12.3	3.0
2024-11-01	44.3	8.2
2024-10-01	24.0	9.0

Not much. Across 133 liquidity-support operations between May 29, 2024 and September 17, 2026, counterparties offered $1,725.5 billion of par and Treasury accepted $222.7 billion, or 12.9 percent. Against the maximums Treasury announced for those same operations, a combined $311.0 billion, it used 71.6 percent of its own ceiling. The acceptance rate has fallen as the program matured: 21 percent of what was offered in 2024, 14 percent in 2025, and 9 percent in 2026 through August 18, as offer volumes grew faster than the caps.

The five operations since the August announcement are a small sample, and five operations are not a trend. But they point the same way. Treasury was offered $40.9 billion, accepted $11.1 billion, and had announced maximums totalling $18.5 billion. That is 27 percent of what was offered, a higher share than earlier in the year, yet only 60 percent of the maximums, a lower share than the program average. Only one of the five, the September 10 operation, fell in a sector where the cap had actually been raised; the others were shorter nominal sectors and one TIPS operation at their existing limits. The first operation under the doubled limits, on September 10 in the 10-to-20-year bucket, received $10.5 billion of offers against a $6 billion maximum and took $5.2 billion. A week later, in 7-to-10-year paper with a $4 billion cap, it took $2.4 billion of the $9.7 billion offered.

Each point is one operation. Treasury regularly leaves part of its own ceiling unused; the first operation under a doubled cap, on September 10, used 86 percent of it.

Snapshot: 2026-09-20 · buybacks-ls-fill source data

Data as of Sep 17, 2026 · Snapshot updated Sep 20, 2026 · Source: U.S. Treasury, Fiscal Data: Treasury Securities Buybacks (accepted ÷ maximum par to be redeemed)

Show the numbers (last 24)

Date	Buybacks, liquidity support: accepted as a share of the maximum
2026-09-17	60
2026-09-15	100
2026-09-10	87
2026-08-25	30
2026-08-20	47
2026-08-18	100
2026-08-11	100
2026-08-06	100
2026-07-28	100
2026-07-23	100
2026-07-22	54
2026-07-16	100
2026-07-09	57
2026-07-01	100
2026-06-25	100
2026-06-24	100
2026-06-16	14
2026-06-09	100
2026-06-03	100
2026-05-28	18
2026-05-21	8
2026-05-19	42
2026-05-13	98
2026-05-07	100

The reason is in the program’s own rules rather than in the market. Treasury “may buy back less than the maximum amount announced” and may decide “not to buy back any securities during an operation”; offers are evaluated “based on their proximity to prevailing market prices at the close of the operation, as well as measures of relative value,” and the department describes itself as a price-sensitive buyer. The binding constraint on how much gets bought is price, not size. Raising the cap raises what Treasury can absorb when offers are cheap enough. It does not commit Treasury to absorb anything.

That is the correction to our August reading. A buyer with a ceiling and full discretion is a backstop, not a standing bid, and the scale is modest against the stock it sits beneath: the largest single purchase to date was $5.2 billion, against a $6 billion maximum, in a market with $5.5 trillion of Treasury bonds outstanding at the end of August. What the August 19 decision did change is the message. Treasury chose to say, in a press release, that long-end offers are consistently strong and that it wants to support that sector. Whether the November refunding turns a temporary increase into a permanent one will tell us more than any single operation.