S&P 500
7,585.73
−0.45% Tue
−0.93% week · −2.95% vs ATH
The 10-year went through 5% for the first time since 2007. AVGO lost 350.24 on the way. A hike is 92% priced — so the dots are the only thing left to trade.
FOMC 2:00 PM · Presser 2:30 · Triple witching Friday§1 — The tape
The S&P and Nasdaq both closed Tuesday at their lowest levels in more than a month. Energy was the only sector working — up roughly 2.4% Monday while technology took the damage from both ends: a 5% discount rate and a fresh argument about the AI buildout itself.
§2 — The driver
This is the line the last two editions said to watch, and it broke. The 10-year touched 5.04% — its highest level since 2007 — and sat around 5.00% into Tuesday's close.
Sept 3 is the low at 4.762% — the same session the S&P peaked. Sept 15 is an intraday high; the close was nearer 5.00%.
Trace it back and the shape is the whole story: the 10-year bottomed at 4.762% on Sept 3 — the exact session the S&P made its high — and has gone up almost without interruption since. Everything the market has done in two weeks is this line inverted.
At a 5% risk-free rate, the arithmetic on long-duration cash flows changes materially, and there is nineteen years of muscle memory saying this level doesn't hold. It is also the level above which the equity risk premium stops compensating anyone for owning a 40x multiple. That is why semis are leading the downside and energy is the only green sector.
§3 — This morning
| August retail sales | Actual |
|---|---|
| Headline MoM | +1.2% |
| Ex-autos MoM | +1.4% |
| Headline YoY | +6.0% |
| Ex-autos YoY | +6.9% |
| Total sales | $773.9B |
On the headline this is a strong consumer, and it removes the last argument against hiking. But the category breakdown is where the actual information is:
| Category | YoY |
|---|---|
| Gasoline stations | +21.0% |
| Miscellaneous retail | +14.0% |
| Sporting goods / books | +10.7% |
| Nonstore (e-commerce) | +9.9% |
| Electronics & appliances | +7.8% |
| Food services | +5.8% |
| Motor vehicles | +2.1% |
| Furniture | +1.9% |
Gasoline stations +21% year over year is not consumption, it is price. Retail sales are measured in dollars, not units — so an oil shock mechanically inflates the headline while the consumer is buying the same gallons for more money. Meanwhile the two categories people actually defer when money is tight, motor vehicles (+2.1%) and furniture (+1.9%), are barely growing in nominal terms, which means they are shrinking in real ones.
So the number is hawkish for today's decision and weak underneath. Consumer discretionary led the sector declines on Tuesday, which suggests the market had already worked this out.
§4 — Levels
350.24 had held since spring and anchored every ladder in the last two editions. It is gone. AVGO closed Tuesday at $339.27 (−1.58%), which is 3.13% below the floor and 6.28% below Friday's close.
A level that held for months and then fails does not become meaningless — it flips. 350.24 is now the ceiling, and the first rally back into it is the test that tells you whether this was a flush or a change in trend. Below, there is very little structure between 337.90 and the 52-week low; the range simply hasn't traded there recently.
Pre-market Wednesday: AVGO $342.20 (+0.86%) — back at the underside of the broken floor before the Fed.
| Ticker | Last | Change | Note |
|---|---|---|---|
| ORCL | $144.79 | −3.65% | 52wk $114.50–$329.50 |
| NVDA | $212.17 | +0.57% | held while AVGO broke |
| AVGO | $339.27 | −1.58% | floor gone |
Oracle keeps getting worse. Down another 3.65%, now roughly 56% below its 52-week high — and this is a week after a beat-and-raise with a $664B backlog. The stated reasons: AI-regulation headlines, and the fact that a 5% 10-year directly raises the financing cost of the data-center buildout the entire thesis rests on. Larry Ellison cancelling a planned $7.5B share sale — normally a bullish signal — didn't move it.
NVDA held green while AVGO broke, which is the first divergence in the complex for a while and worth watching rather than explaining away.
§5 — The main event
| Item | Detail |
|---|---|
| Decision | 2:00 PM ET |
| Press conference | 2:30 PM ET · Chair Kevin Warsh |
| Also released | SEP + dot plot |
| Current target | 3.50% – 3.75% |
| Base case | 25bp → 3.75% – 4.00% |
| Priced | ~92% |
| Sunday | ~85–90% |
| A week ago | ~62% |
At 92%, the hike itself carries almost no information. Three things do:
Does the dot plot show a second hike this year, or does it cap tightening at this one move? This is the single most important number released today.
The previous projection had cuts in 2027. If those are erased, the market is repricing far more than one meeting.
His first hike vote as Chair. Whether he calls this insurance against an energy shock or the start of a sequence decides how the curve trades into Friday.
The statement will cross the wire first and algorithms will trade it. The SEP lands at the same moment and takes humans a minute or two to parse. That gap is where the first move gets faded — it has happened at nearly every FOMC where the dots and the statement told different stories.
§6 — Scenarios
25bp, and the 2026 median shows no second move — framed explicitly as insurance against oil. The 10-year comes back under 5%, and the most oversold names snap hardest. AVGO's test is reclaiming 350.24 from below.
Confirms: 10Y < 4.95% by the close, AVGO back above 350.24.
A hike and a dot plot that preserves optionality. Nothing resolves, the 10-year stays pinned at 5%, and positioning into Friday's expiry starts doing more work than fundamentals. The most likely path and the hardest to trade.
Confirms: 10Y holds 4.98–5.05%, AVGO chops between 337.90 and 350.24.
A second hike in the 2026 dots and 2027 cuts erased sends the 10-year decisively above 5.05% and opens the air pocket below 337.90 on AVGO. The mirror case — a hold against 92% pricing — spikes equities, then raises the question of what the Fed sees that the market doesn't.
Confirms: 10Y > 5.05%, VIX > 22, AVGO loses 337.90.
Sunday's edition put the bearish branch at 25% and it is the one that happened — the 10-year through 5%, AVGO testing and breaking 350.24, exactly as written. The scenario text was right and the probability on it was too low. I've widened B at the expense of A today rather than pretend the last set was calibrated.
§7 — The clock
§8 — Friday
Quarterly expiry of index futures, index options and single-stock options lands Friday Sept 18. Volume roughly doubles and the daily range expands, with the last hour carrying most of it.
The compounding matters: every position taken into today's decision has to be unwound or rolled by Friday. If the dots surprise in either direction, that repositioning gets forced into the single highest-volume session of the quarter.
Expiry-week price action gravitates toward strikes with large open interest, so treat anything read off a chart between now and Friday's close as provisional. Round numbers — 350 on AVGO, 7,500 on the S&P — carry more weight than usual this week precisely because that is where the options sit.
§9 — Scorecard
The thesis was right and the weighting was wrong, which is a specific kind of error worth naming rather than smoothing over.
Badges: Open · Hit · Partial · Miss. Misses stay posted.
Free newsletter
Sunday's week-ahead map and Wednesday's midweek check. Free, no spam.