- Twenty-five years after the Treasury shocked markets by ending 30-year sales, strategists are debating whether the 20-year is next. Deutsche Bank’s George Saravelos reported mumblings of a coming suspension after client meetings in New York, while Citigroup is advising clients to position for the 20-year to outperform the 10-year on the prospect of a reduction or cancellation.
- The anomaly is that the 20-year has mostly yielded more than both the 10-year and the 30-year since its 2020 reintroduction. A shorter maturity costing more than a longer one in the same credit indicates an instrument without a natural buyer base rather than any credit signal, and the Treasury pays that premium for no benefit.
- Even Steven Mnuchin, who revived the 20-year as Treasury secretary in 2020, told Bloomberg by 2024 that it should be cancelled because it is costly to the taxpayer.
- Attention centres on the November 4 quarterly refunding statement, with Citigroup anticipating a $3 billion reduction per auction for both 20- and 30-year bonds, funded by additional bill issuance. In August the Treasury changed its guidance language from evaluating potential future increases in coupon sales to evaluating changes, which opened the door to cuts.
What Happened?
Padhraic Garvey of ING said that if the Treasury wanted to act, the sensible move would be to leave everything else and cut the 20-year, since nobody likes it, though his baseline view is that it stays. Citigroup’s head of US rates strategy Jason Williams suggested Bessent may cite poor pension fund Treasury demand, possibly caused by increased long-end corporate issuance. Vishal Khanduja of Morgan Stanley Investment Management said there is definitely a competition for capital happening, pointing to AI-related issuance by hyperscalers and the structural shift away from defined-benefit pension plans that once provided reliable long-dated demand. BNP Paribas strategists led by Guneet Dhingra warned that cancellation could be perceived as panic and signal exhaustion of the Treasury’s toolkit, encouraging bond vigilantes, with unintended consequences including lower market liquidity. The Treasury did not immediately respond to a request for comment. A dealer questionnaire is due Friday, borrowing projections on November 2 and the refunding statement with a Treasury Borrowing Advisory Committee statement on November 4.
Why It Matters?
The competition for capital point is the one with the widest implications and it is stated plainly here. Hyperscalers issuing large volumes of long-dated bonds to fund AI infrastructure are competing with the US government for the same pension and asset manager demand, which means corporate AI capital spending has become a factor in sovereign borrowing costs. That is an unusual thing to be able to say and almost nobody is modelling it. It also explains a pattern visible across three separate borrowers: hyperscalers have moved to 46% short-dated issuance from 30% in the first half, France is selling less 30-year debt because dealers report weak demand, and the US Treasury is now considering the same retreat from the long end. All three are responding to the same shortage of buyers for duration. The structural explanation matters because it determines whether this is fixable. Defined-benefit pension plans once bought long-dated government debt to match liabilities, and their decline removed a buyer whose demand was price-insensitive. Nothing has replaced it, which is why the long end struggles to absorb supply even at the highest auction yields since 2000. Cancelling the 20-year addresses the symptom efficiently, since it is an orphan security serving neither the pension bid at 30 years nor the benchmark liquidity of the 10-year. The BNP warning is the honest counterweight and the 2001 precedent supports it. The 30-year was cancelled, officials were criticised, and the bond was eventually reinstated. Each discretionary intervention, including the surprise doubling of buybacks two weeks after the last refunding statement, trades a modest interest saving against the Treasury’s reputation for being regular and predictable, and that reputation is itself worth basis points. Investors have already warned about shifting debt management practices.
What Next?
Friday’s primary dealer questionnaire is the first signal, and a question on the 20-year would indicate the Treasury is laying groundwork rather than planning a surprise. Borrowing projections arrive November 2 and the refunding statement with the advisory committee’s views on November 4. Watch what replaces any reduction, since more bills, a SOFR-linked floating rate note or larger two- to five-year auctions all shorten the government’s duration profile and increase how often the debt stock reprices. The spread between 20- and 30-year yields is the live market measure of whether this is being priced. Consumer price data on October 14 and the Federal Reserve meeting later this month set the backdrop against which all of it is decided.
Affected Tickers and Coins: ZB, ZN, TLT
Source: Bloomberg
