Why Softer PCE Did Not Budge 5.29% Long Yields

The August core PCE reading cooled to 3.0% year over year, under the 3.3% consensus, and long Treasury yields barely blinked, holding at 5.29% for the 10-year, its highest since the early 2000s. The softer print did not loosen the market's grip because the numbers that actually drive the long end pointed the other way: the moderation is mostly a statistical re-benchmark, and the real economy beneath it is running hot.

Start with the inflation mechanics, because they matter. The BEA released the August PCE report on September 30 showing headline inflation at plus 3.4% year over year and core at plus 3.0%, both below consensus. The softness, however, arrives through the annual index revision, which rewrote the reference base back to 2021 and pushed the July core reading down by roughly 0.36 percentage points to about 3.0%. A lower year-over-year number produced largely by a rewritten base is not the same thing as disinflation taking hold.

What the same data run showed underneath the index was a demand picture that is not decelerating. Q2 gross domestic product was revised up to plus 2.2% annualized from plus 1.5% in the second estimate. Real personal consumption expenditures rose 0.6% month over month in August, the best gain since March 2025, and real final sales to private domestic purchasers climbed 4.6%. The Fed's cleanest gauge of demand is accelerating, not slowing, and that is the reason the market handed the soft headline no concession at the long end.

Liquidity compounds the pressure. M2 money supply rose for a fourth straight month to $23,342.8 billion in August, up 0.53% month over month and re-accelerating from the February trough near $22.6 trillion, while reserve balances held steady at $6.74 trillion per the Fed's H.4.1. Money is available. That is supportive of risk and of an economy that keeps spending, which is precisely the combination that keeps real yields high and caps the appeal of duration. Cash is the competitor, and cash continues to look fine.

Widening about +46 basis points from roughly +25 on September 22, the curve has stopped listening to the PCE headline. The 10-year sits at 5.29%, the 2-year at 4.88%, and the 30-year at 5.64%, with the 10-year minus 2-year spread widening to about +46 basis points from around +25 basis points on September 22. A steepening curve into a data print is the bond market saying the Fed is not restrictive enough, or at least that the term premium deserves compensation for a policy path that still points up, just later.

Federal Reserve messaging has resolved the timing question for many observers without touching the direction. Williams said on September 29 there is no need for urgency and that a further hike may be appropriate late this year. Jefferson told the committee it will need to take more time. Only Logan leaned firmly hawkish, calling September's move a first step and arguing more is needed.

Kalshi now prices an October hold near 66% and a 25 basis point hike near 30%, down from roughly a 56% hike probability a week earlier. That is a repricing of when, not of whether. The Fed's own September projections still put the year-end policy rate at 4.1%, one more hike above the current 3.75% to 4.00% range.

The bull case for risk assets rests on the soft core and a revised-down history giving the Fed cover to skip October, easing front-end pressure, while a real economy that is accelerating on a re-expanding money supply fits a soft landing. A soft payrolls print on top of that would validate the no-October-hike repricing and could squeeze yields and duration-sensitive assets sharply. The bear case is that the moderation is methodological, energy keeps re-inflating the headline, and strong real demand argues the Fed stays restrictive, keeping the 10-year pinned and putting continued discount-rate pressure on long-dated equities and real-yield headwinds on gold.

Today's September employment report, due at 12:30 UTC, is the swing variable. Consensus has settled near plus 84,000 to plus 100,000 against August's plus 162,000, with unemployment expected to hold at 4.1% and a fat left tail lower. A soft print near or below 50,000, the tail some forecasters flag, would validate the skip-October repricing and likely compress the front end, delivering relief to duration and gold, because long yields above 5% leave little room before the Fed is read as restrictive. A hot print above consensus would do the opposite, reigniting the October-hike trade and pushing the 10-year decisively through 5.3%. With long yields pinned near 24-year highs, there is little in the way of a cushion: the move is binary, and the screen will be set by whether the two-sided tail resolves soft or hot.

This content is for informational purposes only and does not constitute financial advice. Past performance is not indicative of future results. Consult a qualified financial advisor before making investment decisions.