July's core rose 0.246% with 46% of it a quarterly investment-management fee; strip that line and core still runs 3.0%. August PCE (Sep 30) lands with the annual update that replaces how BEA prices that fee — and two weeks after the FOMC has already decided.

Q2 advance — does hard output data finally show the slowdown the surveys keep predicting?

Core capex ran +10.5% y/y — the pillar of the no-growth-crack read. Still standing?

June's slack-ratio rise was labor-force exit, not demand — openings fell. July (Sep 1) shows whether openings keep sliding while the unemployed count shrinks.

July answered the June question: no — prices stalled at 71.1 while activity hit a four-year high. August tests which July was real, ISM's acceleration or S&P Global's slowdown.

July answered June: prices back above 70 (70.3) and employment back under 50. August (Sep 3) shows whether 70+ is the standing level or the oil spike passing through.

July settled it: a stall (three months average +20K, rate falling on exits). August (Sep 4) lands after the Aug 28 preliminary benchmark — does the vintage survive it, and does the labor-force exodus continue?

July's producer core fell to 4.7% over 12 months but is running 5.1% annualized over three, and freight — most of the annual decline — is half base effect. August (Sep 10) tests whether the run rate follows the annual rate down, and lands with no quarterly asset-fee or construction repricing to lift it.

July answered the floor question — all of 2026 has held 4.01–4.19M — on contracts signed at 6.4–6.5% mortgages. August (Sep 10) is the first report to price the climb to 6.69%, with investors down to 14% of buyers (from 20%) and inventory below year-ago.

Two consecutive core readings of 0.2% leave the September hike case resting on PCE alone — and July PCE showed about half of PCE's premium over CPI is financial services. August CPI (Sep 11) lands four days before the FOMC: whether the rent measures finally join the slowdown — owners' equivalent rent has held 3.1–3.3% every month this year — decides how much further core can fall.

July fell 0.6% with the control group's largest drop since January 2025 — the first dent in the consumer that has carried the no-recession case. August (mid-September, the week the FOMC decides) says whether that was Prime-Day timing or a turn.

Three members voted to hike on July 29. The level argument is all that survives of that case, and it rests on core PCE 3.3% — about half of whose premium over CPI's 2.5% is financial services, and 0.3 points of which is a fee line BEA reprices on Sep 30. The Committee sees August CPI (Sep 11) and PPI (Sep 10), not August PCE.

July answered June: the apartment jump reversed, single-family fell to 808K (lowest since November 2022), and the South accounts for the whole annual decline in total starts. August (Sep 17) shows whether 808K is the new level or one month's noise — and whether the rising stack of authorized-but-unstarted homes (+10% y/y) ever breaks ground.

A record 115,000 of the unsold new homes have not been started, and Midwest sales halved to a 2012 low on one month's data. August (Sep 24) shows whether the Midwest was a bad draw and whether builders keep listing plans instead of houses.

Core capex ran +10.5% y/y — the pillar of the no-growth-crack read. Still standing?