For Treasurys, A Month To Remember. And Forget.

One way or another, September will be remembered as a pivotal month for US government bonds.

September 2026 was the month a US Treasury secretary determined to arrest rising yields was humbled: “Bet against me if you want,” Scott Bessent declared, on September 9.

Bessent was talking about the yen, which he’s also trying to micromanage, but the bond market took his remark as a dare. Fast forward three weeks and the yields Bessent’s trying to cap were busy hitting new post-2007, and in the case of the US long bond, post-2004, highs.

20-years-and-out US Treasurys headed into Q3’s final trading sessions nursing an 8% quarterly decline, the worst since Q4 2024, when yields rose sharply into year-end following Donald Trump’s reelection.

10-year yields, which last week pushed through 5.20% for the first time since 2007, are up more than 40bps this month, twice the average increase for Septembers going back a decade.

The figure above shows you the extent to which 10-year US Treasurys — the benchmark of all benchmarks — underperformed even in the context of a rough seasonal.

If past Fed hiking cycles are precedent, 10-year yields may have further to climb. Perhaps a lot further. And on most accounts, that’ll bode ill for risk assets eventually, with ramifications for an economy that arguably depends on stock prices staying buoyant.

“The quickest way to end the US boom is a surge in bond yields,” BofA’s Michael Hartnett said. “The Trump administration knows macro and stocks are ‘too big to fail.'”

The figure above, from Hartnett’s note, gives you some context for what this decade’s macro regime shift did to duration. The greatest bull market of all time — the four-decade bond bull — melted in the presence of scorching-hot nominal growth.

What happens from here’s anyone’s guess (crude’s obviously the wild card), but one thing’s for sure: Bond vol has to calm down.

“Outside of the optical calm in equities vol and spot index, fixed income / macro vol has absolutely woken up, evidenc[ing] massive ‘fear of the unknown,'” Nomura’s Charlie McElligott said.

The figure above uses the simplest, most widely-cited measure of UST vol just to keep things simple. Harley Bassman’s MOVE just experienced one of its largest two-day jumps of the decade.

“VaR shock risk is always high when volatility of the collateral of the financial system jumps 35% in two days,” Hartnett went on, referring to the MOVE, adding that “policymaker panic to cap yields has begun.”

Considering that panic, it might be time to “nibble” at bonds, he suggested. Should US yields fall 100bps over the next 12 months, the return on the 10-year would be 14%. On the 30-year, more than 20%.

“If / when we get that next ‘acute rate shock’ catalyst [whether] hot data or the dreaded ‘US diesel export ban’ headline, many in the market are going to look to reprice for the ‘growth shock’ via demand destruction [from] higher prices and higher rates,” McElligott said, in the same note mentioned above. “But man, the same underlying disease seemingly remains in bonds, which just cannot squeeze or hold [a] meaningful rally.”


 

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

4 thoughts on “For Treasurys, A Month To Remember. And Forget.”

  1. Two lines really stood out to me:

    “an economy that arguably depends on stock prices staying buoyant” and “BofA’s Michael Hartnett said. “The Trump administration knows macro and stocks are ‘too big to fail.’”

    Nor is it just the US economy. In today’s FT there is a story which makes sense and supports that notion. “Foreign capital flows into US stocks hit record as appetite for debt fades:
    Overseas purchases of US equities topped $940bn in the year to July, coinciding with strong gains in the S&P 500.” So even foreigners would also suffer if US equities crater.

    Does this mean that US stocks rather than US Treasuries are now the riskless asset? That’s what “too big to fail ” implies, no?

  2. Normally, a bear flattener means the short-end is pricing in inflation and rate hikes, and the long-end is witnessing a shift from stocks to bonds as long-term growth expectations diminish. The difference this time is that the long(er)-end is also being pressured by massive corporate issuance which is also driving-up rates.

    For example, this past week, SoftBank issued $4.5 billion in 7-½ year debt at an interest rate of 9.75%, as part of a much larger $11.1B debt offering to fund AI data centers. I believe SoftBank’s bond rating is currently “junk,” so I don’t think that counts as a traditional “flight to safety,” but at least for now the effect on the long end may be rather similar.

    I heard or saw somewhere this week that at rates of 8-9% the calculus for profiting from the AI boom begins to break down very quickly, and at rates of 10% and above many of these companies enter into the “danger zone” (as in “Danger Will Robinson, danger!”) It would appear now that we are almost there.

    As to investing in (U.S.) bonds, if you plan to hold on to maturity, the yields on the 5 and 7-year maturities also look pretty sweet (not investing advice). Perhaps the biggest risk is the opportunity cost of those yields going even higher.

10th Anniversary Boutique

Coming Soon

Create a free account or log in

Gain access to read this article

Yes, I would like to receive new content and updates.