Hot Core CPI Flips the Script: Fed Rate-Hike Odds Jump to ~70% Ahead of Sept 16 Meeting

Takeaway

August CPI came in sticky on the core print, and in about an hour the market went from pricing a coin-flip on next week's FOMC decision to pricing a ~70% chance of a rate hike. That's a regime flip for a tape that's spent 2026 trading on AI CapEx, not on the Fed tightening further — and it lands directly on the rate-sensitive, high-multiple end of Mag7.

What happened

The Bureau of Labor Statistics released the August CPI report at 8:35am ET this morning (Sept 11). Headline CPI rose 0.4% month-over-month and 3.4% year-over-year — both in line with LSEG consensus. Core CPI (ex food/energy) rose 0.3% MoM, about 0.1 point hotter than economists expected, with the year-over-year core rate at 2.4%, in line. (CNBC, "Inflation persisted in August, potentially locking in a Fed interest rate hike," published 8:35–8:36am ET; Fox Business, same figures, same timestamp.)

The market reaction was immediate and asymmetric to the "hotter than expected" core beat: the 2-year Treasury yield jumped roughly 15bp to a two-year high, and rates markets moved the implied probability of a hike at next week's FOMC meeting (Sept 16, 2pm ET) to roughly 70–72%, up from essentially 50/50 the day before (TheStreet market-update live blog, updated through midday; Yahoo Finance/ca premarket wrap, 4:22am ET). WTI crude sold off close to 3.4% to just above $99, and equity futures — which had been modestly higher into the print on Oracle's post-earnings pop — were choppier after the data hit.

This is the first CPI print this cycle where hike odds, not cut odds, are the dominant market pricing story — a reversal of the disinflation-and-cuts narrative that's underpinned the AI-capex rally for most of 2026.

Mag7 impact

Rates re-emerging as the de-rating mechanism hits the complex unevenly, and it layers on top of an already-dispersed tape (desk context: 10Y at 4.83%, four of seven Mag7 names above EMA20, no name above 31% conviction pre-print).

• AAPL / MSFT — highest-multiple, longest-duration cash-flow names in the group; a 2-year yield at a two-year high is a direct discount-rate headwind. AAPL in particular has run hard into this (+3.56% 24h, RSI 58.9, IV30 already at a rich 47% vs 23.4% realized) — a hawkish repricing gives the crowded long a reason to de-risk into strength rather than chase.
• NVDA / semis — mixed. Higher real rates pressure the multiple, but NVDA's setup (IV30 18.8pts under realized vol, per desk tape) means a rates-driven vol spike is more likely to be a repricing of cheap optionality than a fundamental re-rate — watch whether it's optionality or spot that moves first.
• META / GOOGL / AMZN — less duration-sensitive on the surface (ad/retail cash generation vs. Mag7 peers), but META's already-extended RSI (66.4) and GOOGL's laggard status (2.4% below EMA20 despite the largest CapEx bill) mean a broad multiple compression hits the group's weakest hands hardest.
• TSLA — high beta to risk sentiment broadly; a hawkish surprise into an already-elevated IV30 (62.8%) raises the cost of carrying the position either way.

Net: this isn't a single-name catalyst, it's a discount-rate shock that reprices the whole complex's terminal-value assumptions simultaneously — the opposite of the idiosyncratic, dispersion-driven tape the desk has been trading into this week.

What to watch next

• CME FedWatch-style pricing into the close — does the ~70% hike probability hold, extend, or fade as the initial reaction is digested? A print this close to the meeting (5 days out) leaves little room for the Fed to talk it back before Sept 16.
• Long-end yields, not just the 2-year — the 10-year was already at 4.83% pre-print; a move meaningfully higher there is what actually pressures Mag7 multiples, more than front-end pricing.
• Fed communications between now and Sept 16 — any pre-meeting Fedspeak (blackout period rules may already be in effect) or dissenting commentary that pushes back on the hike pricing.
• Whether AI-capex names decouple — if NVDA/hyperscaler capex conviction holds up even as multiples compress, that's the bull case (physical demand > discount rate) the desk has been leaning on; if it doesn't, the equipment-complex weakness already in the tape (LRCX, AMAT, SMH) spreads to the mega-caps.

Sources: CNBC (cnbc.com/2026/09/11/cpi-inflation-report-august-2026.html, 8:35am ET); Fox Business (foxbusiness.com/economy/cpi-inflation-august-2026, 8:35am ET); TheStreet market-update live blog (thestreet.com, updated through midday 9/11); Yahoo Finance premarket wrap (4:22am ET); FedRateCalc.com for the Sept 16 FOMC date.