The 30-Year Yield Hits a 24-Year High Anyway

The 30-year Treasury yield closed at 5.64% on September 30, the highest since 2002, even after a soft reading on the Fed’s preferred inflation gauge.

The two-year yield, which tracks the policy path, finished at 4.88% after trading down to 4.82% on the report. That left 76 basis points between the front end and the long bond, a gap that points to supply and a thin bid for duration, not the next rate hike.

Spending Jumped While the Inflation Gauge Cooled

The Bureau of Economic Analysis said personal consumption expenditures jumped 0.9 percent in August, a $190.8 billion rise and the strongest monthly gain in the revised recent series. Real spending, after prices, rose 0.6%, or $92.8 billion.

Goods did most of the work. Outlays on goods rose $114.1 billion and services added $76.7 billion. Personal income increased only $66.6 billion, or 0.2%. After-tax income rose 0.3% in current dollars and was unchanged in real terms, so households spent while real paychecks did not.

The saving rate was still 4.1%, with saving of $990.2 billion. That mix (a spending surge on a flat real income print) is what bond desks mean when they call demand resilient. It is also why the long end refused to treat the inflation miss as a cue to buy duration.

AUGUST INCOME, SPENDING AND PRICES

From a year earlier, the PCE price index was up 3.4% and the core index was up 3.0%. Both year-over-year rates came in below the forecasts that had been in the market, and both remain above the Fed’s 2% target. Core is one full point above that target.

The monthly price prints were less friendly. Headline PCE rose 0.3% from July, and core rose 0.2%, after 0.1% on both in July. The release also carried the annual update of the national accounts, with revisions back to January 2021, so older July comparisons from before September 30 are not the same series. Traders still used the year-over-year miss to cut the odds of an October hike. They did not use it to bid for 30-year paper.

A 5.64% Close and the 2002 Marker

On the Treasury daily par yield curve, the 30-year finished September 30 at 5.64%. The 10-year closed at 5.29% and the 20-year at 5.68%. A year earlier those three yields were 4.73%, 4.16% and 4.71%.

THE SEPTEMBER 30 CLOSE

• 30-year: 5.64%, the highest since 2002.
• 10-year: 5.29%, with 41 basis points over the two-year.
• Two-year: 4.88%, after an intradaily low of 4.82%.
• 2s30s spread: 76 basis points, a steepener on the day of the inflation print.

The 30-year constant-maturity series was stopped on February 18, 2002, and brought back on February 9, 2006, so the “since 2002” line is the clean historical marker rather than a continuous tape. From 4.95% on June 1, the 30-year is up 69 basis points. A global government-debt index is down 2.1% since June, the largest three-month drop since late 2024, when investors first priced a more expansionary fiscal path after the presidential election.

John Briggs, head of U.S. rates strategy at Natixis Corporate & Investment Banking, said it “feels like a buyers’ strike, really,” and that “the price action overall remains terrible.” Dan Carter, a senior portfolio manager at Fort Washington Investment Advisors, called it “a continuation of bearish momentum,” adding that many accounts were already set for lower rates and may not have much cash left to add.

Why Two-Year Yields Fell as Long Bonds Sold

Two-year yields fell as much as five basis points after the PCE report because that maturity is a Fed-path instrument, and the path for October got cheaper. Traders cut the chance of a hike at the October 27-28 meeting to about 36 percent after the data, from around 70 percent earlier in the week. The two-year then pared the drop and closed at 4.88%, one basis point below the prior session.

The 30-year does a different job. It prices inflation that lasts, the stock of Treasuries that must be absorbed, and the extra yield investors want for locking money up for three decades. August spending at 0.9%, a heavy corporate calendar, and another long-end buyback that the market already treats as optional all hit that instrument. They barely touch the two-year.

Gennadiy Goldberg, head of U.S. interest-rates strategy at TD Securities, said the inflation print let markets “breathe [a] sigh of relief.” He also said they remained hesitant to rally hard, in part because of the changes in how the price data are put together. John Canavan at Oxford Economics said the firmer activity data keep the curve steeper and “long-end rates a touch higher,” and that traders may wait for the October 2 employment report before making large bets.

That is the second move, and it is the one the headline 24-year high conceals. A skip in October, if it comes, would land on the front end. It would not, on its own, create a bid for 30-year bonds while households are still spending and the Treasury still has to place duration.

Bessent’s $6 Billion Cap Still Falls Short

Treasury Secretary Scott Bessent widened regular long-end buybacks, lifting the cap from $2 billion to $6 billion, and has described the program as a way to improve trading in older, less liquid issues. The operations are not billed as a yield cap. The market still reads the size, and the take-up, as a test of whether official demand can offset a weak private bid.

On September 24, in a 20- to 30-year operation, dealers offered $10.468 billion. Treasury took $4.078 billion against the $6 billion maximum. Earlier long-end windows also printed below the new cap, and purchases have bunched in a few issues. Padhraic Garvey, ING’s regional head of research for the Americas, noted that the department is free to buy less than the cap if it dislikes the terms, and that it can always buy more later.

LONG-END BUYBACKS, CAP VERSUS TAKE-UP

The scheduled Treasury securities buyback operations list the October 1 10- to 20-year window with offers from 1:40 to 2:00 p.m. Eastern, settling October 2. Officials had said they would take up to $6 billion, the same language used at the prior two long-term operations, which then accepted less. Angelo Manolatos, a rates strategist at Wells Fargo Securities, said the buybacks are doing the most work in liquidity and spreads, and that “a negative growth catalyst will likely be needed to spark a sustained move lower in the longer-term rates.”

A $6 billion pass at off-the-run notes does not retire the stock of duration the market is being asked to hold. When offers come in around $10 billion and the desk takes $4 billion, the rejected paper still has to live on someone’s book. That is how a liquidity tool can coexist with a 5.64% 30-year: the window is real, and it is small next to the selling.

A 7.03% Mortgage Is the Household Rate

Households do not borrow at the two-year yield. They borrow off the long end, with a spread. Freddie Mac’s weekly survey, dated September 24, said the 30-year fixed-rate mortgage averaged 7.03%, up from 6.95% the prior week and 6.30% a year earlier, a 0.73-point rise on the year. The 15-year averaged 6.42%, up from 6.26% and from 5.49% a year ago.

THE FREDDIE MAC WEEKLY READ

• Survey date: September 24, covering applications from the prior Thursday through Wednesday, before the 5.64% close.
• 30-year fixed: 7.03%, up 0.08 point on the week and 0.73 point on the year.
• 15-year fixed: 6.42%, up 0.16 point on the week and 0.93 point on the year.
• What it is: Conventional, conforming purchase loans with 20% down and strong credit, not a quote for every borrower.

Sam Khater, Freddie Mac’s chief economist, said the housing market “remains supported by a solid labor market and an economy that is growing at a healthy rate.” That support is the same spending the BEA just printed. It is also the reason mortgage rates can keep climbing while the Fed debates whether October is live: if consumers are still buying goods, the long-term rate that sets the monthly payment does not have to fall with the two-year.

Funds that hold mortgage bonds get squeezed when long yields jump, because the value of those bonds drops and some accounts then sell Treasuries to raise cash or cut duration. That selling can push long yields higher again. A 25-basis-point move in the funds rate, even if it arrives in December, does not break that loop. A 7.03% survey print that is already a week stale versus a 5.64% 30-year is the channel through which the long-end selloff shows up in household budgets.

No Need for Urgency After the September Hike

The Federal Reserve raised its target range by 25 basis points this month, to 3.75% to 4%, the first increase since 2023. New York Fed President John C. Williams, a permanent voter and vice chair of the rate-setting committee, spoke at the University at Buffalo on September 29 and pulled October off autopilot without taking a late-year hike off the table.

With the policy action we took at our September meeting, there is no need for urgency after September, and we have time to gather more information. The accumulation of more data should provide greater clarity on the underlying trends in the economy and the associated risks to achieving our goals, and thereby the appropriate setting of monetary policy.

John C. Williams, President, Federal Reserve Bank of New York, University at Buffalo, September 29, 2026

If the economy tracks his forecast, he said, “one further upward adjustment of the federal funds target range may be appropriate late this year” to bring inflation back to 2% on a faster clock. He added that this is his forecast, and that “time, and the totality of the data, will tell.” The remaining meetings are October 27-28 and December 9. Evercore ISI read the Buffalo remarks as most consistent with skipping October and hiking in December.

Williams was talking about the overnight rate. The 30-year is not the overnight rate. A committee that has “time to gather more information” can wait on funds. Accounts that have to mark 20- and 30-year paper every day cannot wait on a $6 billion window that has already printed light. The 36 percent October-hike price after PCE is a two-year story. The 5.64% close is what is left when that story is stripped out.

Friday’s Payrolls Meet a Steeper Curve

The next hard print is the October 2 employment report. Canavan’s caution about “major adjustments” before that release still holds. Private payroll figures for September, from ADP Research, had already come in above forecasts, which is one reason the long end treated Wednesday as a growth day rather than an inflation holiday. Index-related buying around the monthly bond-index rebalancing, due at 4 p.m. New York time on September 30, was another possible bid. The 30-year still closed at 5.64%.

WHAT WE KNOW

• The close: 30-year 5.64%, 10-year 5.29%, two-year 4.88% on September 30.
• The inflation mix: PCE up 3.4% from a year earlier and 0.3% from July; core up 3.0% and 0.2%.
• The last long buyback: $4.078 billion taken against a $6 billion cap on September 24.

WHAT IS UNCONFIRMED

• October 1 take-up: Whether the 10- to 20-year window fills the $6 billion cap or prints light again.
• October 2 payrolls: Whether the jobs report is weak enough to be the growth shock Manolatos said the long end still needs.
• October 27-28: Whether the committee skips, as Williams’s “no urgency” line invited, or delivers the remaining hike of his forecast.

The 30-year at 5.64% is the rate the Treasury pays to borrow for three decades, the reference that feeds mortgage quotes, and the price of a buyers’ strike that a soft inflation print did not end. Friday’s payrolls can still reprice the two-year. They have to do much more than that to reprice the bond that just made a 24-year high.

Disclaimer: This article is news reporting and market analysis for information only. It is not investment advice, a recommendation to buy or sell Treasuries, mortgage products, or any other security, and it is not a forecast you should trade on. Speak with a licensed financial adviser, broker, or tax professional who knows your situation before you make any investment or borrowing decision. Yields, mortgage averages, buyback results, and policy odds are those published by the cited agencies and firms as of the dates given and will change with new data, auctions, and Federal Reserve decisions.