AVGO
$365.10
+1.18%
sympathy showing
A $105 Brent print pushed the entire Treasury curve to 52-week highs and put a Fed hike back on the table. Then oil paused, and everything bounced.
§0 — The print · updated 9:25 AM ET
| August CPI | Actual | Expected | Prior |
|---|---|---|---|
| Headline MoM | +0.4% | +0.4% | +0.1% |
| Headline YoY | +3.4% | +3.4% | +3.4% |
| Core MoM | +0.3% | +0.3%* | +0.2% |
| Core YoY | +2.4% | +2.4% | +2.5% |
*Core MoM consensus was published as 0.2%, 0.3% and 0.4% across different outlets before the print. Treat "in line" loosely — and note that anyone claiming a clean beat or miss on core is picking whichever consensus suits them.
Gasoline rose 3.9% on the month and accounted for more than a third of the entire headline increase on its own. Energy was +2.1% MoM and +16.3% YoY, gasoline +27.4% YoY. Food added 0.1%. This is the same energy print PPI gave yesterday, arriving one rung further down the chain.
The line worth watching is shelter: +0.3% after +0.1% in July. Shelter is the sticky component. One month is not a trend, but if energy is starting to leak into services, that is where it shows up first — and it is the difference between an oil shock the Fed can look through and one it cannot.
Hike odds are still roughly 70% — unchanged from yesterday. This print moved the Fed calculus by approximately nothing. Which means the bounce in futures is not a CPI relief rally.
It's oil. Brent −3.61% to $103.70, WTI −3.35% to $99.05, long-end yields easing off multi-year highs, and futures halting a four-session losing streak. The entire move this week has been one input in both directions: oil went up and everything broke; oil paused and everything bounced. CPI was the scheduled event, not the cause.
AVGO now sits between Thursday's 360.83 close and the 369.74 shelf. That shelf is the whole test at the open — §6 has the full ladder.
§1A — Scoreboard · Thursday's close
WTD is measured from the Friday Sept 4 close. vs peak is from the Sept 3 closing high — the session this entire slide began. The S&P also sits −2.88% below the 7,816.70 all-time high set in August.
§1B — Session ladder
| Index | Sep 3 | Sep 4 | Sep 8 | Sep 9 | Sep 10 |
|---|---|---|---|---|---|
| S&P 500 | +1.06% | −0.38% | −0.58% | −0.48% | −0.58% |
| Dow | +1.18% | −0.51% | −1.18% | −0.77% | −0.60% |
| Nasdaq | +1.40% | −0.29% | −0.32% | −0.64% | −0.65% |
| Russell 2000 | +0.51% | +0.25% | −0.43% | −1.37% | −1.02% |
Monday Sept 7 was Labor Day — no session.
This is a bleed, not a crash. Four consecutive days of roughly half a percent on the S&P, with no capitulation session anywhere in the sequence. Which means no washout low has been established yet.
Small caps are leading the downside and accelerating. The Russell was still green on Sept 4, then −0.43%, −1.37%, −1.02%. The most rate-sensitive part of the market cracked first and is deteriorating fastest. That is the tell worth watching, and the scoreboard above cannot show it.
The Nasdaq is decaying, not dumping. −0.29 → −0.32 → −0.64 → −0.65 is steady erosion in the highest-multiple part of the market as yields climb — the signature of a discount-rate problem rather than a growth problem.
§1C — Rates, commodities, dollar
| Rates | Level | Change | Note |
|---|---|---|---|
| US 2-Year | 4.567% | +14.0 bp | 52-week high |
| US 10-Year | 4.954% | +11.4 bp | 52-week high |
| US 20-Year | 5.390% | +10.2 bp | 52-week high |
| 2s10s spread | 38.7 bp | −2.6 bp | bear flattener |
The entire curve is at 52-week highs and flattening. A front end rising faster than the long end is the market pricing tightening — a policy scare, not a growth scare. Print only the 10-year and this disappears.
§1D — Breadth
Roughly 32.5% of US issues were advancing at midday Thursday, after about 60% were declining early in the session. An index down 0.6% with only a third of issues green is broad de-risking. Down 0.6% with 55% green would be two mega-caps dragging an index — same headline, opposite meaning.
Against that, VIX at 17.89 says nobody is panicking yet. Broad selling without fear pricing is what a repricing looks like before it becomes an event.
§1E — Overnight & pre-market
| Futures | Level | Note |
|---|---|---|
| ES (S&P) | 7,602.00 | ≈ +10 vs cash · flat on the session |
| NQ (Nasdaq 100) | 29,108.25 | −0.09% |
| YM (Dow) | 52,105.00 | flat |
| Globex range | 7,585.50 – 7,666.25 | 81 pts · 1.1% |
| Overnight low | 7,585.50 | first level to defend |
Futures as of approximately 10:00 PM ET Thursday. Globex range covers the Thursday session.
▲ ORCL $159.26, +4.13% on a clean beat — detail in §5.
▼ ADBE roughly −2% on guidance · AVGO $359.43, −0.39% · NVDA $217.76, −0.27%.
Note what is not happening: Oracle printed an exceptional quarter and the semis are not catching a bid on it. That is the first read available on how much sympathy actually exists in this tape.
Dow futures were green before Thursday's bell. PPI landed in line at 8:30. The market still opened down 0.5–0.8%, because oil kept climbing regardless of the data. Fed hike odds walked 60% → 65% by 9:22 AM → 71% by the close.
A reader watching only the futures screen got the day exactly backwards. Futures direction is not the signal in this regime. Oil and the 10-year are.
§1F — The driver
Three separate scales, one shared timeline. The alignment is the point: the S&P peaked on Sept 3, the exact session the 10-year yield bottomed. Everything since has been one trade running in reverse.
| Date | Brent $ | US 10Y % | S&P 500 |
|---|---|---|---|
| Sep 1 | 94.65 | 4.796 | 7,631.47 |
| Sep 2 | 95.63 | 4.794 | 7,666.60 |
| Sep 3 | 95.52 | 4.762 | 7,747.71 |
| Sep 4 | 96.28 | 4.784 | 7,718.60 |
| Sep 8 | 99.85 | 4.804 | 7,673.52 |
| Sep 9 | 101.71 | 4.837 | 7,636.36 |
| Sep 10 | 105.37 | 4.970 | 7,591.79 |
Yields from Investing.com, which prints Sept 10 at 4.970% against TheStreet's 4.954% — the chart uses one source throughout. Sept 9 Brent is derived from the reported +3.6% Sept 10 move.
Since Sept 3: Brent +10.3%, the 10-year +20.8bp, the S&P −2.01%. Note the shape difference — oil rises in a straight line while the S&P declines in small steady increments. That is a market being repriced by an external input, not one discounting its own deterioration. It also means the selling stops when oil stops, not when equities find a level.
§2 — What happened
Oil. Brent closed $105.37, WTI above $100 for the first time since May. This is not a demand story. Iran struck ten ships near the strait on Sept 9 following US strikes on five Iranian tankers, and Houthi forces seized Yemen's port of Mocha on Sept 10. Ceasefire talks never produced an agreement. Brent is up more than 30% from its early-August low.
Rates. That move went directly into the bond market — the whole curve to 52-week highs, as §1C shows. The chain is short and mechanical: energy inflation → rate-hike repricing → higher discount rate → the longest-duration equities take the most damage. That is why semiconductors led the downside while staples led the upside, and why the Russell is cracking fastest.
PPI confirmed it. August final demand came in +0.4% MoM and +5.4% YoY; core +0.3% / +4.7%. Goods rose 1.1% while services rose 0.1% — a pure energy print. Diesel jumped 24.1% in a single month and accounted for more than a third of the entire goods increase on its own.
▲ Leaders — consumer staples, communication services.
▼ Laggards — technology, materials, both off more than 1%.
| Ticker | Move | Why |
|---|---|---|
| AAPL | +3.53% | product announcements |
| RDDT | +4.7% | — |
| CHTR | +4.5% | — |
| ELV | +4.1% | — |
| COO | −14% | — |
| FCX | −7.3% | materials selloff |
| INTC | −5.7% | semis broad weakness |
§3 — The main event
| Series | Expected | Prior |
|---|---|---|
| Headline MoM | +0.4% | +0.1% |
| Headline YoY | +3.4% | +3.4% |
| Core MoM | +0.4% | +0.2% |
| Core YoY | +2.4% | +2.5% |
Note the split: headline is expected to stay pinned at 3.4% while core eases to 2.4%. That gap is the oil story. The market has already decided to forgive headline. The live question is whether energy pass-through has begun leaking into core services.
Why this print carries more than usual: the FOMC meets Sept 15–16 with a decision Wednesday at 2:00 PM ET — and it now carries a genuine rate hike debate. The current target range is 3.50–3.75%. CME-implied odds of a 25bp hike reached roughly 71% by Thursday's close, up from 61% before that morning's PPI — entirely on the oil move.
A Fed that markets believe may raise rates in the middle of an AI-capex bull market is not a configuration with a recent playbook. After the print, watch the 10-year before you watch the index — yields are leading equities in this regime, not the other way round.
§4 — Scenarios
Hike odds fall back under 50%, the 10-year comes off 4.95%, and the beaten-up growth complex gets the sharpest bounce because it is the most oversold. Watch semis for confirmation rather than mega-cap.
Confirms: 10Y < 4.90% within the first hour.
In-line does not kill the hike thesis, so the 10-year stays bid and the tape chops. Expect a gap up on "no bad news," then a fade as Sept 16 reasserts itself. Fading the first 30-minute pop has worked all week.
Confirms: first-hour high holds as resistance into midday.
The 10-year through 5.00% is the single most consequential thing that could happen Friday — a psychological line and a technical one at once. High-multiple AI names take the worst of it.
Confirms: 10Y > 5.00%, VIX > 20, ES loses 7,585.50.
Oracle's print tonight is genuinely excellent AI-capex news. Under scenario B or C it will not matter. Strong sector news has been sold all week on macro — do not let a good headline talk you out of what the tape is actually doing.
§5 — Earnings
| Q1 FY2027 | Actual | Expected |
|---|---|---|
| Revenue | $19.35B | $19.14B |
| Adj. EPS | $1.92 | $1.74 |
| IaaS revenue | $7.4B · +121% | — |
| Total cloud | $11.6B · +62% | — |
| RPO / backlog | $664B | ~$630.6B |
| Backlog YoY | +$209B | — |
| New AI cloud contracts | $30B+ | — |
| GPUs delivered | 300,000+ · ~3× QoQ | — |
| FY27 guide | ≥$90B rev · $8.10 EPS | raised |
| FY27 capex | $90–95B | unchanged |
Two lines carry the print. First, 300,000+ GPUs delivered, roughly triple the prior quarter — backlog is converting into deployed silicon rather than sitting on paper, which answers the bear case head-on. Second, management stated the new contracts require no incremental capital raising, which goes straight at the AI-debt-sustainability worry that has been the sector's undercurrent all week.
And yet. ORCL closed $152.94, −5.38% on the session and recovered only to $159.26 (+4.13%) after hours. Its 52-week range is $114.50 – $345.72 — the stock is down more than half from its high. A clean beat-and-raise buying back four percent is the clearest available evidence of what this tape is doing to AI-capex names. Adobe told the same story: a beat, then a ~2% after-hours decline on guidance.
§6 — Spotlight
Last $360.83 (−0.97%) · after hours $359.43 · market cap $1.72T · P/E 46. Q3 AI semiconductor revenue was +221% year over year to $16.7B. Fundamentals are not the question here.
About $10.60 of cushion sits below the close. Tonight's Microsoft headline — data center capacity to 38 gigawatts by 2032 from roughly 12 GW today, close to a 3× expansion — is the most directly bullish AVGO item on the tape, since Microsoft is a named custom-silicon partner through the Maia program. Separately, the Wall Street Journal reports the Pentagon is in talks over a ~$5B loan to FluidStack, which reads as the US government backstopping AI infrastructure financing.
None of that outranks the 10-year on Friday. If yields break 5.00%, 350.24 gets tested regardless of how good the sector news is.
52-week range $164.27 – $236.54. Jensen Huang told the Goldman tech conference that cybersecurity is likely the next major AI use case, alongside new arrangements with Palantir on sovereign AI, chip startup d-Matrix, and Australian data center operators. Against that: the DOJ is reportedly investigating whether Nvidia structured its Groq licensing deal to sidestep antitrust review.
§7 — The clock
§8 — The week ahead
A live hike debate plus a fresh dot plot is a far larger event than a routine FOMC. The dots are what will tell you whether this is a one-off insurance hike against an oil shock, or the opening move of something longer.
Then two days later, triple witching. Quarterly expiry of index futures, index options and single-stock options all land Friday the 18th. Volume roughly doubles and the daily range expands, with the last hour — the "witching hour" — carrying most of it. Positioning into a Wednesday Fed decision has to be unwound or rolled by Friday, so the two events compound rather than sit side by side.
Expiry-week price action gravitates toward strikes with large open interest, which means chart levels behave differently that week — the AVGO $400 call wall and the 377.5–381 zone become pinning magnets rather than ordinary resistance. Treat any level read from Monday to Thursday as provisional until Friday's expiry clears.
§9 — Scorecard
Every edition logs its calls and the next one grades them — hits, misses and partials alike. Graded at 9:25 AM against the print; the two that resolve at the close stay open until then.
Badges: Open · Hit · Partial · Miss. Misses stay posted.
Triple witching is Friday Sept 18 — the third Friday of the quarter-end month — not Sept 11. Last night's thread had it a week early. The corrected date makes next week heavier, not this one: FOMC Wednesday the 16th and triple witching Friday the 18th fall in the same week, with a live hike debate running through both. Corrections run here, at the top of the scorecard, not quietly in the text.
§10 — From the server
A reader flagged something the brief underweighted: the sympathy plays resolve at 8:30, not at 9:30. AVGO and ORCL were always going to make their move around the CPI print rather than at the opening bell, which compresses the decision window into the pre-market hour and makes the 9:30 open a confirmation rather than an entry.
That held up. ORCL was +4.13% after hours and +7.62% by 9:20 — most of the repricing happened before the bell. AVGO went from −0.39% after hours to +1.18% pre-market. Anyone waiting for the open to decide watched the move happen without them.
The same reader expected chop rather than a clean trend, with the real opportunity in the 3–4 PM window once the morning's positioning washes out. Both points are now standing calls in the scorecard above, and the §8 clock has been adjusted to reflect the pre-market compression.