$70B 5-Year Auction Yields 5.033%, Second-Biggest Tail on Record
The US Treasury's $70 billion 5-year note auction on September 23, 2026 priced at 5.033%, producing the second-largest tail on record and the weakest bid-to-cover since December 2018. The bond market is repricing sovereign risk in real time.

The US bond market just told Washington what it thinks of American sovereign debt at the offered price.
Key takeaways
- The US Treasury's $70 billion 5-year note auction on September 23, 2026 priced at 5.033%, per TreasuryDirect.
- The auction tailed by 3.1 basis points, one of the largest tails on record for this tenor, with the bid-to-cover ratio dropping to 2.212, as foreign and institutional buyers pulled back sharply.
- Primary Dealers absorbed 15.8% of the offering, meaning roughly $11 billion in supply landed on bank balance sheets the market didn't want, cascading pressure across the curve.
The US Treasury auctioned $70 billion in 5-year notes on September 23, 2026, pricing at 5.033%, up from 4.393% at the prior auction on August 26, according to TreasuryDirect. That 64-basis-point surge auction-over-auction on a single $70 billion tranche translates to roughly $448 million in additional annual interest cost on this offering alone ($70B × 0.64%). Across a $40 trillion-plus debt stack, the compounding math is orders of magnitude worse.
The official auction results are available at TreasuryDirect. The 7-year note auction follows on September 24.
The Internals Were the Real Story
A yield of 5% grabs headlines. The demand metrics underneath it are what matter.
The auction cleared at 5.033%, tailing by 3.1 basis points, meaning buyers required a steeper discount than pre-auction trading implied. Per Treasury Fiscal Data, that tail ranks among the largest on record for this tenor.
Bid-to-cover came in at 2.212, down from 2.371 prior. Indirect bidders, the category that captures foreign central banks and large institutional buyers, fell to 54.31% from 61.51%. Direct bidders jumped to 29.92%. That left Primary Dealers holding 15.8% of the auction.
Dealers are legally obligated to bid. When Indirects retreat and Directs cannot fill the gap, Dealers absorb the remainder. At 15.8% of $70 billion, that is approximately $11 billion of 5-year paper that landed on bank balance sheets because the market passed.
Those banks will need to offload it, and that selling pressure cascades into the broader curve.
The Feedback Loop Nobody in Washington Wants to Name
The borrowing-cost spiral is becoming self-reinforcing.
Every auction that clears with a large tail forces the next one to price at a higher yield. Higher yields expand the interest expense on the existing debt pile. A larger interest burden widens the deficit. A wider deficit requires more auctions.
And each new auction gets sold into a market that just demonstrated it will not absorb supply at the offered price without exacting a premium. TFTC has been tracking the structural mechanics of this dynamic, including the implicit yield curve control the Fed has been running in the background and what happens when it loses credibility.
The "safe haven" framing for US Treasuries depends on the assumption that buyers will always show up. Today's Indirect print puts that assumption under real pressure. When the reserve asset of the global dollar system fails its own auctions at offered prices, the implicit guarantee behind that system is exposed as conditional.
Foreign central banks and sovereign wealth funds are not panicking yet, but they are not stepping in to absorb supply the way they once did. The Hormuz crisis and Japan's own debt trap are both compressing the pool of buyers with the capacity and the incentive to hold long-duration US paper.
The falsifiable version of this thesis: if the September 24 7-year auction clears with a tail under 1 basis point, bid-to-cover above 2.4, and Indirects recovering above 60%, today reads as a liquidity air pocket. If the 7-year also fails, this becomes a pattern with two consecutive data points. A Fed emergency intervention or a sustained reversal in 10-year yields back below 4.7% on genuine demand (not an equity flight-to-quality bid) would also challenge the structural read.
What Comes Next
The 7-year auction on September 24 is the immediate test. Watch the tail and the Indirect allocation. Two consecutive weak auctions would make the "structural demand destruction" thesis considerably harder to dismiss.
The longer-horizon question is whether the Fed steps back in. If auctions keep failing, the policy options narrow fast: accept the higher rates and absorb the fiscal damage, or restart asset purchases and explicitly monetize the deficit. Neither path is painless.
One forces spending cuts that Congress has shown no appetite for. The other is yield curve control with the mask off, and it debases every dollar-denominated savings vehicle in the process. Bitcoin has no auction to fail and no coupon that needs to clear at a price the market will accept.
Update, September 24, 2026
The 7-year auction is in. The falsifiable test this article set up yesterday has its answer, and it is not the benign one.
The Treasury sold nearly $44 billion in 7-year notes on September 24, clearing at a high yield of 5.085%, the highest since April 1993.
The auction priced above the when-issued market yield of 5.078%, a signal of lower demand, and the bid-to-cover came in at 2.42, down from 2.50 at the prior auction.
Indirect bidders, the category that captures foreign central banks and large institutional buyers, took just 57.2% of the offering, down from roughly 61% at the prior auction. That is a second consecutive session of Indirects pulling back across consecutive coupon maturities. The "isolated air pocket" exit is closed. This is now a pattern.
The damage was not contained to the auction itself. The benchmark 10-year Treasury yield surged to 5.139%, its highest level since July 2007, and the 30-year pushed to a post-2004 peak of 5.438%. After the 7-year results crossed, the 10-year yield sat at around 5.146%. Those are generational thresholds being crossed in back-to-back sessions. The breakdown is also coordinated across sovereign markets: Japan's 10-year JGB yield rose 8 basis points to 3.055%, its highest since August 1996, while UK Gilts and German Bunds hit fresh multi-year highs. Tokyo cannot step in to absorb Washington's supply when its own bond market is in freefall. The article's thesis about Japan's debt trap compressing the global buyer pool is now showing up in real-time price action.
The structural context behind every number above remains unchanged. UOB flagged that the initial Wednesday selloff "was driven by rebounding oil prices, stronger-than-expected US PMI data, and weak demand" at the 5-year auction. Those same forces carried straight into Thursday's session. The Institute of International Finance reported that global debt hit a record $365 trillion in the first half of 2026, with advanced economies spending more than $3.5 trillion on interest payments over the prior 12 months. Every government selling bonds simultaneously means every auction competes for a shrinking pool of willing buyers.
Two consecutive auctions. Two consecutive tails. Two consecutive Indirect bid retreats. The original article's falsifiable threshold has been crossed on every metric that matters. The policy options the Fed faces have not changed, but the clock on making that choice is shorter. Accept the higher rates and absorb the fiscal damage, or restart asset purchases and explicitly monetize the deficit. The latter is yield curve control with the mask off, debasing every dollar-denominated savings vehicle in the process. Bitcoin has no auction to fail and no coupon that needs to clear at a price the market will accept.
Update, October 3, 2026
The selloff has not paused. The 10-year yield closed at 5.28% on October 2 , and that was not the week's high. The benchmark 10-year yield climbed above 5.34% earlier this week, its highest level since 2002, amid expectations of further Fed tightening, the lack of a resolution to the Middle East conflict, concerns over the US fiscal and debt outlook, and resilient economic data. The Federal Reserve's own H.15 release shows the 10-year printing 5.29% on September 30 before partially pulling back, with the intraday spike to 5.34% on October 1 confirmed by multiple market sources. The 30-year US bond rate also reached a 24-year high.
The move carries a clear supply story underneath it. Heavy government and corporate bond issuance has become a bigger driver this year, according to Macquarie's Thierry Wizman.
AI-related borrowing is adding to bond supply, while stronger growth and persistent inflation are keeping pressure on yields. The curve is also steepening: the 10-year minus 2-year spread closed at 45 basis points on October 2 , up from 32 basis points just a week earlier on September 28, meaning the long end is repricing faster than the short end. That is the market demanding a term premium for duration, not a recession trade.
From the auction failures on September 23 and 24, the 10-year has added roughly 17 basis points. Over the past month the yield has edged up by 51 basis points and is 116 basis points higher than a year ago. The arithmetic of that on a $40 trillion-plus debt stack requires no elaboration. Every rollover now prices into a market that has moved from questioning offered yields to outright repricing generational risk.
Sources
- TreasuryDirect, Auction Announcements, Data & Results
- Treasury Fiscal Data, Record-Setting Auction Data
- Treasury Fiscal Data, Treasury Securities Auctions Data
- US Treasury Tentative Auction Schedule
- Treasury Fiscal Data, Debt to the Penny
- First reported by ZeroHedge (September 23, 2026), citing official Treasury press release figures for tail, bid-to-cover, Indirects, Directs, and Dealer takedown; confirm all internal metrics against the official TreasuryDirect results PDF before final publish.
Frequently Asked Questions
What does a "tail" mean in a Treasury auction, and why does a 3.1 basis-point tail matter?
The "tail" is the gap between the yield at which Treasuries were trading just before an auction (the When Issued rate) and the yield at which the auction actually cleared. A positive tail means buyers demanded a higher yield, and therefore a lower price, than the pre-auction market implied. At 3.1 basis points on a $70 billion offering, today's tail ranks among the largest on record for 5-year notes, per Treasury Fiscal Data. It signals that demand was materially weaker than expected, even with yields already surging heading into the sale.
Could the Fed intervene to support the bond market if auctions keep deteriorating?
The Fed could restart asset purchases (quantitative easing) to absorb Treasury supply and cap yields. That option is always on the table.
But deploying it with inflation still above target would be politically and economically explosive, effectively choosing to monetize the deficit and debase the currency to avoid the discipline the market is trying to impose. The TFTC read: either the market forces fiscal adjustment the political system refuses to make voluntarily, or the Fed prints to bridge the gap. Both outcomes strengthen the case for a fixed-supply, non-sovereign asset. Full historical auction data is available at Treasury Fiscal Data.


