The bond market is arguing with itself.

Traders spent the past month unwinding bets on a Federal Reserve rate hike. They did not, however, buy long-dated Treasuries. The result is one of the widest gaps in months between what the front end of the yield curve expects from policy and what the long end demands as compensation for holding government debt.

According to the Federal Reserve’s H.15 Selected Interest Rates release, whose latest published data runs through August 13, the 30-year Treasury constant maturity yield closed that day at 5.21%, after finishing as high as 5.25% on August 10. That is barely off the 5.27% close on July 31 — the highest level since July 2007. The 10-year settled at 4.63%, down from 4.72% on August 10. The 2-year ended at 4.15%.

Meanwhile, the odds of a quarter-point increase at the September 15-16 FOMC meeting have fallen to roughly 34%, according to fed funds futures. On July 13 those odds were above 75%.

Why the Hike Trade Fell Apart

Two inflation prints did the damage. The Bureau of Labor Statistics reported on August 12 that the Consumer Price Index rose 0.1% in July and 3.4% from a year earlier, down from 3.5% in June. Core CPI rose 0.2%. Producer prices were flat on the month.

Then came the demand shock. The Census Bureau’s advance report, released August 14, put July retail and food services sales at $763.6 billion, down 0.6% from June — against a consensus calling for a small gain. Sales were still up 5.0% from a year earlier, but the monthly drop was the sharpest signal yet that the consumer is pulling back.

Sentiment confirmed it. The University of Michigan’s preliminary August index fell to 51.0 from 55.2 in July, snapping two months of improvement.

For rate-hike bulls, that combination is fatal. You do not tighten into a consumer that is already retreating.

So Why Won’t the Long End Fall?

Because the long end is not really trading monetary policy anymore. It is trading supply.

Primary dealers project a federal budget deficit of roughly $1.95 trillion for the fiscal year ending in September, widening toward $2 trillion in 2027. Every dollar of that has to be funded in a market that already absorbed record issuance. Investors are responding the way any buyer responds to an oversupplied seller: by demanding a discount. That discount shows up as term premium, and it is why the shape of the curve has steepened rather than flattened even as hike expectations died.

The arithmetic is worth spelling out. With the effective federal funds rate at 3.63% and the target range at 3.50%-3.75%, the 30-year yields about 158 basis points more than overnight money. The 2s10s spread sits near 48 basis points, up from 46 a week earlier.

Note what is not driving this. The 10-year TIPS real yield closed at 2.39% on August 13, implying a 10-year breakeven inflation rate of roughly 2.24% — close to the Fed’s target and nowhere near a panic. Consumers disagree: the Michigan survey put year-ahead inflation expectations at 4.3%, up from 4.2%, with the five-to-ten-year figure steady at 3.3%. But the bond market itself is pricing fiscal risk far more aggressively than price risk.

Oil is the wildcard that keeps the hike trade alive at 34% rather than zero. Crude topped $100 a barrel last month on supply concerns tied to the conflict with Iran, and an energy shock is precisely the kind of tail event that could force the Committee’s hand. Three FOMC members already dissented in favor of a 25-basis-point hike at the July meeting, which ended in a 9-3 vote to hold.

What It Means for Stocks and Mortgages

Higher long rates raise the discount rate applied to future corporate earnings, which is why long-duration growth stocks tend to wobble when the 30-year pushes toward 5.25%. They also set the floor under household borrowing costs.

Freddie Mac’s Primary Mortgage Market Survey put the 30-year fixed-rate mortgage at 6.67% for the week ended August 13. “The 30-year fixed-rate mortgage averaged 6.67 percent as of August 13, 2026, down from last week when it averaged 6.69 percent,” the agency reported. “A year ago at this time, the 30-year FRM averaged 6.58 percent.” The 15-year averaged 5.96%.

Read that year-over-year comparison again. Mortgage rates are higher than they were twelve months ago despite five consecutive FOMC holds. That is the long end at work: mortgage pricing tracks the 10-year Treasury, not the fed funds rate, which is why a Fed on pause has not translated into cheaper housing.

What to Watch This Week

  • Monday, August 17: Empire State manufacturing survey at 8:30 a.m. ET and the NAHB housing market index at 10:00 a.m. ET. The NAHB print is the cleanest read on whether 6.67% mortgages are still choking builder confidence.
  • Tuesday, August 18: July housing starts and building permits. Permits are the forward-looking half.
  • Wednesday, August 19: Minutes of the July 28-29 FOMC meeting, 2:00 p.m. ET. This is the week’s main event. Read them for how close the three hawkish dissenters came to persuading the rest of the Committee, and for any discussion of the balance sheet — the pace of runoff matters directly to long-end supply, a mechanism explained in our guide to tapering.
  • All week: the 30-year yield itself. A decisive break above the July 31 close of 5.27% would mark a fresh post-2007 high and put renewed pressure on rate-sensitive equities. A retreat back below 5.10% would suggest the fiscal-risk trade is finally easing.

The September meeting also carries a Summary of Economic Projections, meaning the dot plot arrives four weeks from now. Between here and there, the minutes are the only direct window into a Committee that is more divided than its 9-3 vote suggests.

Bottom line: the front end has stopped fighting the Fed. The long end has not stopped fighting the Treasury.