The October Fed Hike Just Died — and It Wasn't the CPI That Killed It

BLS delivered the August CPI on Friday, September 11: headline +0.4% MoM SA and +3.4% YoY NSA, with core +0.3% MoM (+2.4% YoY, down a tenth from July's 2.5%). It was a mixed print rather than a clean miss — core MoM ran a tenth hotter than the +0.2% consensus, but the YoY core rate decelerated and the Kobeissi Letter flagged it as 'the biggest increase [in YoY core decline] since May 2026.' Headline held at 3.4% because a 16.3% YoY surge in energy offset a still-cooling core trend, and the BLS noted lodging-away-from-home rose 2.4% MoM while wireless telephone services jumped 5.9% MoM — the largest single component driver of core services.

The CPI print was ambiguous enough that the Warsh-led Fed still had cover to hike 25bp at its September 16 meeting, which it did 12-0, lifting the target range to 3.75-4.00%. What changed everything was what came nineteen days later. On September 30, the BEA released August PCE alongside its annual methodology update, which revised core PCE lower by roughly 18-36bp on average going back to 2021 — depending on which window you pick — and produced a print that missed cleanly: core PCE at 3.0% YoY versus 3.3% expected, with monthly core at +0.2% versus +0.3% expected.

Right after the August CPI, the front end did its expected thing: rate-hike odds for the September 16 FOMC jumped to roughly 90% (from 70% the day prior), the Fed delivered, and Warsh told the press conference 'inflation is too high' and 'this summer's inflation readings do not indicate underlying trends have meaningfully improved.' The dot plot pinned the median 2026 fed funds rate at 4.1%, which by arithmetic implies one more 25bp hike from the new 3.75-4.00% range — and on the morning of September 25, Kalshi traders had October at 65%.

Then the September 30 PCE hit. Kalshi showed October hold odds jump to 65% with hike odds collapsing to 33% in a single session, and CME FedWatch tracked similarly with the October hike probability at 34.9% post-release. The 10-year Treasury yield, which had spent the first three weeks of September pushing toward 5% (a 100bp rally from the August low per Schwab), eased off the highs — and the long end in particular found a bid after months of relentless selling. The dollar index pulled back from post-CPI highs; SPDR Gold Shares caught a safe-haven bid as the 'higher for longer' narrative cracked.

Chair Kevin Warsh used his Jackson Hole debut on August 28 to lay the hawkish foundation, citing 65 months of 'elevated inflation' and warning the Fed has 'work to do' if it cannot get confidence that inflation is moving to 2%. By the September 16 press conference he was explicit: 'Today's FOMC decided that this standard has not been satisfied,' and the SEP showed total PCE at 3.7% for 2026 falling to 2.3% in 2027 — a glidepath that required at least one more hike to validate.

The September 30 PCE release cut against that narrative more sharply than any single print could. The methodology changes — replacing CPI deflators in portfolio management and legal services with CES-based quantity indices and PPI composites, and shifting computer software deflators to a PPI-CPI blend — collectively shaved roughly 18bp off the annual core PCE pace per RBC's pre-release estimate, with Reuters citing a larger 36bp reduction. That alone was a structural downgrade to the inflation backdrop the Fed has been reacting to, and the August monthly print coming in 30bp below consensus on core meant traders no longer needed to price October as a live meeting.

Bonds are the most direct read. iShares 20+ Year Treasury Bond ETF had been the year's worst-performing major asset class proxy — YTD total return down 8.00% as of September 30 per BlackRock's own factsheet — under relentless long-end selling driven by the 'higher for longer' Fed. A 25-32 point swing in October hike probability is meaningful but not yet a regime change: the 2-year yield closed September 16 at 4.74% and the 20-year at 5.39% per BondSavvy, both still elevated. If October delivers the pause markets now expect, the front end will lead, and iShares 7-10 Year Treasury Bond ETF is the cleanest expression of the move.

Equities are more complicated. Real personal consumption expenditures rose 0.6% in August with nominal spending surging 0.9% (per the BEA release), so the disinflation is not coming from a collapsing economy — it's coming from a slower pass-through. That favors duration-sensitive growth (tech), supports rate-sensitive cyclicals, and relieves pressure on small caps that have been crushed by the long-end sell-off. The savings rate fell to 4.1%, the lowest since 2022 per Verified Investing — consumers are spending into the disinflation rather than pulling back, which is the soft-landing scenario the market wants to underwrite.

Near-term (days to weeks), the catalysts are sequenced and binary. The September CPI is due October 14, the October FOMC lands October 27-28, and the Q3 PCE prints October 31. The setup is asymmetric: any one upside surprise on core CPI or core PCE can revive October hike odds, but it takes a clean chain of three above-consensus prints to put December back in play. Watch the BLS September CPI print especially for any unwind of the August wireless-telephone services spike (which added roughly 7bp to core CPI) and the shelter reacceleration from +0.1% MoM in July to +0.3% MoM in August.

Long-term (one to three years), the structural implications matter more than the October meeting. The BEA's annual methodology revision is a permanent downgrade to the core PCE trajectory that the Fed has been using as its primary inflation yardstick, which means the median dot plot at 4.1% is now probably 25-50bp too high relative to the revised data — and the FOMC's December SEP will be the first formal reset of that picture. If the revised core PCE data continues to track at 2.7-3.0% rather than the 3.0-3.3% range the Fed had been working from, the entire 'higher for longer' framework built under Warsh's first year collapses — and the December 2026 / first-half 2027 meeting sequence becomes a cut sequence, not a pause sequence.

The August inflation data cycle did two things at once. It confirmed that the trend is still decelerating — core CPI at 2.4%, core PCE at 3.0%, both 10-30bp below where the Fed had been pricing the trajectory — and it materially revised the past five years of PCE history lower. The combined effect has knocked roughly 30 percentage points out of October hike odds, easing pressure on the entire rate-sensitive complex. The Fed is not done, but the window for 'one more' has narrowed to December — and December requires a meaningful upside surprise on the September CPI to remain a live meeting.

For positioning, the asymmetry favors duration and rate-sensitive cyclicals over defensives and gold over cash. iShares 7-10 Year Treasury Bond ETF is the cleanest expression of a Fed pause with cuts not yet priced in; iShares 20+ Year Treasury Bond ETF carries the biggest mark-to-market recovery if long-end yields roll over from current 5%+ levels. SPDR S&P Regional Banking ETF and iShares US Real Estate ETF carry the steepest upside on a sustained pause. SPDR Gold Shares benefits from a softer dollar and lower real rates if the 'higher for longer' narrative fully unwinds. The single biggest risk to this whole trade is a hot September CPI on October 14 — that's the print Warsh needs to revive October.