Core PCE cool 3.0 YoY. DXY firmed. Brent cracked $98.

Catalyst check. Wednesday September 30. Three scheduled catalysts landed. At 8:30 AM ET, BEA Personal Income and Outlays for August: headline PCE up 0.2 percent MoM, headline PCE up 2.6 percent YoY; Core PCE up 0.2 percent MoM versus consensus 0.3 percent, Core PCE up 3.0 percent YoY versus consensus 3.3 percent (source: BEA Personal Consumption Expenditures release). Also at 8:30 AM ET, BEA Q2 GDP Third Estimate came in on the expected 1.5 percent (source: BEA GDP release). At 12:00 PM ET, USDA NASS Quarterly Grain Stocks: corn stocks 2.10 billion bushels (up 35 percent YoY, 177 million bushels above trade expectations), wheat stocks 1.85 billion bushels (down 14 percent YoY), soybean stocks 315 million bushels (down 3 percent YoY, 9 million below expectations) (source: USDA NASS Grain Stocks September 2026). Verified against BEA schedule and USDA NASS calendar.

Tuesday's piece flagged the asymmetric setup: a Consumer Confidence collapse to 81.9 that the currency ignored, teeing up a Wednesday PCE where a cool print would be the harder test. The print delivered: 0.2 percent MoM (10 basis points below consensus), 3.0 percent YoY (30 basis points below consensus, the largest cool surprise since March). Under a typical Fed reaction function this is a decisively dovish print. Under the Warsh-led reaction function the currency firmed. DXY closed 101.45 (up 8 pips from Tuesday), 10-year yield closed 5.289 (up 3.4 basis points), gold closed $4,184.50 (up $4.80, essentially flat), Brent CFD spot cracked $98.43 (down $4.16 in a single session). The transmission that did not happen (a rate rally on a cool print) is the specific data point.

The print

Headline PCE up 0.2 percent MoM, up 2.6 percent YoY. Core PCE up 0.2 percent MoM (consensus 0.3, prior 0.3), up 3.0 percent YoY (consensus 3.3, prior 3.3). Three-month annualised core sits at approximately 3.0 percent; six-month annualised at approximately 3.1 percent. The specific decomposition matters: goods PCE flat, services PCE up 0.3 percent, and shelter (the sticky component that has driven every hot print this year) softened to plus 0.24 percent MoM from plus 0.29 percent in July. The specific line the Fed's staff watches (supercore, services ex-shelter) undershot the trailing three-month pace.

Real Personal Spending up 0.3 percent MoM, real Disposable Personal Income up 0.1 percent, savings rate 4.4 percent (essentially flat from July). The specific pattern is a household that is still spending in real terms even as sentiment collapsed on Tuesday; the spending-sentiment decoupling that the Sep 25 reference piece described continues to hold.

The tape

• DXY: 101.45, up 8 pips from Tuesday's 101.37 (source: Yahoo Finance daily close, Sep 30). Eighth consecutive daily close above 101. The currency held the range on a print that under a typical reaction function would have delivered a 40 to 80 pip fade. That specific asymmetry, back to back with Tuesday's Consumer Confidence miss, is the two-day signature of a Warsh Fed that discounts both real-side softness and inflation-side softness in favour of the September SEP trajectory.
• 10-year yield: 5.289, up 3.4 basis points from Tuesday's 5.255 (source: Yahoo Finance daily close, Sep 30). Fresh 19-year daily-close high. The specific mechanic is real-yield extension: inflation compensation compressed on the cool print, and nominal yields held their range; the arithmetic delivers a real-yield spike. That real-yield spike is doing the dollar-support work.
• Gold: $4,184.50, up $4.80 from Tuesday. Third consecutive close below $4,200. A cool inflation print with real yields rising is not a gold-bullish combination; the specific pattern is the "positioning cleanout has run, but macro has not yet delivered a reset" signature.
• USD/JPY: 157.32, essentially unchanged from the Tuesday close (source: Yahoo Finance daily close, Sep 30). Quarter-end Japanese repatriation flow through the London 4pm fix (11:00 AM ET) partially offset the dollar bid; the pair traded 157.10 to 157.60.
• Brent CFD spot: $98.43, down $4.16 from Tuesday's $102.59 (source: Yahoo Finance daily close, Sep 30). First close below $100 since August 12. The specific move is OPEC+ signal noise plus Ust-Luga full unwind plus the USDA-driven risk-off in the softs complex spilling into energy futures. The Brent break helps the disinflation narrative on the goods-PCE side into October.

The transmission that did not happen

Under a typical Fed reaction function (2004-2019 vintage), Core PCE 0.2 MoM and 3.0 YoY versus consensus 0.3 and 3.3 would produce: 10-year yield down 6 to 10 basis points, DXY down 40 to 80 pips, gold up $15 to $30, and a specific compression in hike-implied probabilities across the CME FedWatch curve. The Wednesday tape delivered zero of those. Yield up 3.4bp, DXY up 8 pips, gold essentially flat, hike odds unchanged inside statistical noise.

The specific interpretation matters. This is not a "the market did not believe the print" outcome; the print was clean, the methodology is well understood, and there is no data-quality asterisk. It is a "the market is not pricing off Core PCE the way it used to" outcome. Under the September SEP framework, the FOMC's median participant is anchored on a specific inflation trajectory (3.6 percent 2026 median, moving to 2.8 percent in 2027) that already builds in a modestly softer path than the March SEP. A single cool print inside that trajectory does not change the trajectory; it confirms it. See our reference on a cool print in a hawkish regime for the specific decomposition of why a print that is dovish relative to consensus can be neutral relative to the SEP.

The USDA Quarterly Grain Stocks release added a cross-asset overlay. Corn stocks 177 million bushels above trade expectations is a large bearish surprise for the ags complex, which pulled roughly 4 percent through the front-month CBOT contract in the noon session. The specific transmission to macro is via headline inflation: a lower ag-complex baseline supports the goods-PCE disinflation narrative into October CPI. It does not affect core PCE directly (food is in headline, not core), but it does anchor the specific "Warsh-look-through" framing that a bumpy real-side print does not shift the reaction function.

The read

Wednesday told us three specific things. First, the September SEP trajectory holds through a materially cool Core PCE print; the market's dollar-strength trade does not require the print to confirm inflation-side firmness. Second, real-yield extension is doing the dollar-support work: nominal yields have held their range, inflation compensation is compressing, and the arithmetic delivers a real-yield spike that supports the currency independent of the inflation print itself. Third, the Brent crack below $100 gives the goods-PCE disinflation narrative a fresh leg into October, which reduces the tail risk that the November data cluster (October CPI, October ISM, October payrolls) forces a repricing.

Two observations worth logging for Thursday and Friday. First, the specific asymmetry (cool print, firmer dollar, real yields extending) is the exact mirror of the 26 June Core PCE reaction, when a hot print produced a modest dollar fade because the SEP had priced hot. The specific mechanic is symmetric: the market prices off deviations from the SEP path, not from consensus. Second, the pattern raises the bar for what would count as a genuinely policy-shifting print into October: a Core PCE at 2.9 percent YoY (five basis points below consensus for October) or below with headline services also softening would be the specific combination that would test the trajectory.

Working thesis lightly adjusted. Additional-hike-within-2026 at 70 percent (unchanged; the print did not compress hike odds). Hold-through-2027-with-no-cuts at 26 percent (unchanged). Cuts-restart-inside-2027 at 4 percent. The distribution held through both Tuesday's real-side dovish surprise and Wednesday's inflation-side dovish surprise; the specific data point of the week is the stability of the distribution.

Setup update

Confirmed if: Thursday October 1 ISM Manufacturing PMI prints at 54.0 or above (consensus 53.2) with prices-paid holding above 55, and Friday October 2 September Nonfarm Payrolls prints at 175k or above (consensus 155k) with average hourly earnings at plus 0.3 percent MoM or above. Under that combination, the Warsh reaction function survives a cool Core PCE by pointing to a still-firm labor market; DXY tests 102.00; the additional-hike-within-2026 probability drifts toward 78 percent.

Invalidated if: ISM Manufacturing prints below 51.0 with prices-paid below 52, and NFP prints below 100k with average hourly earnings at plus 0.2 percent MoM or below. Under that combination, the real-side softening signal that ran through Tuesday's Consumer Confidence collapse becomes a labor-market signal, which is the specific transmission the Warsh reaction function does respect. DXY breaks 100.60; gold recovers above $4,250; hike odds compress toward 55 percent.

Watch this week: Thursday October 1 delivers September ISM Manufacturing PMI at 10:00 AM ET (source: ISM Report on Business schedule). Friday October 2 delivers September Nonfarm Payrolls at 8:30 AM ET (source: BLS Employment Situation release schedule). The mid-October CPI print is the second-order test of the Core PCE undershoot signal; check the BLS schedule for the specific date (source: BLS CPI release schedule). Fed blackout for the October 27-28 FOMC opens Saturday October 17; expect FOMC-participant remarks through mid-October that either endorse or hedge the September SEP trajectory. USDA October WASDE lands the second week of October (source: USDA WASDE release schedule).

Nothing on this site is investment advice or a recommendation to trade. Setups published here are falsifiable hypotheses, not signals.