A 29K payrolls print took the October hike off the table and the tape bid growth all session. NVDA tagged a record $237.88 intraday, TER +8.0% and SYNA +~14% led semis, and TSLA +~5% on a delivery beat carried discretionary. The other side: WDC −10.1% and STX −10.0% after Toshiba said it will double HDD capacity, and NKE −5.8% on a guide that keeps getting worse. SPX closed +0.73% at 7,723, near the top of the day's range. The catch is that the long end sold off into the close anyway.
SPX closed at 7,722.57, up 56 points on the day. It gapped up on the September jobs report and held: buyers took every dip, and the index finished near the session highs. That said, the gain only got SPX back to flat for the week (+0.01% over five sessions), and it was still a fourth down week in five for the Dow and S&P on Reuters' count. Nasdaq (+1.19%) did the work. The NDX printed new record highs intraday as NVDA cleared its May high to $237.88 and pushed its market cap past $5.7T. AVGO (+3.3%), AMD (+2.9%) and ORCL (+3.1%) came along with it, and the SOX is closing at its highest level since June after clearing its 50-day.
The data was soft on every line. Payrolls came in at +29K against an ~84–90K consensus, the prior two months were revised lower, and unemployment ticked up to 4.2% from 4.1%. CME FedWatch odds of a hike at the Oct 27–28 FOMC fell to roughly 18–24%, versus 64% a week ago. Jamie Cox's line that "September should have been a hold" summed up the desk view. DJIA (+0.49%) lagged as NKE and ACN weighed. RTY gained about 1.0%, its best day in about a month, which is the rate-relief trade at work. Even so, small caps are about 4.4% below where they started the month.
What doesn't fit the story is the bond market. The 10Y rallied on the print, then reversed and closed higher at 5.281% (+~5bp), with the 2Y also +5bp at 4.839%. Equities rallied on "no hike." Bonds are pricing inflation from $91 crude, record diesel and an Iran war in its eighth month. Those two views can coexist for a few sessions, but not for long.
VIX closed around 15.3, down about 1.1 points (−6.6%). Vol sellers had control all day, with no bid for weekend protection despite 9,000 more U.S. troops heading to the Gulf. That complacency is cheap to hedge against at these levels.
| Sector | Session | Read | Notable |
|---|---|---|---|
| Technology | +~1.3% | Leader; SOX highest since June | TER +8.0%, NVDA rec. high; WDC −10.1% |
| Consumer Discretionary | +~1.2% | Tesla-driven | TSLA +~5%; NKE −5.8% |
| Materials | +~1.0% | Miners bid (XME +1.3%) | — |
| Real Estate | +~0.5% | Rate-relief bid faded with the long end | — |
| Communication Svcs | n/c | Mixed | GOOGL +1.6%; PSKY/WBD close targeted 10/6 |
| Industrials | n/c | Mixed | MOD −4.8% post-spin |
| Utilities | n/c | Yield headwind | — |
| Consumer Staples | n/c | Defensive, unloved | PEP reports Thu 10/8 |
| Energy | −~0.1% | G7 reserve release | WTI −$1.76; LNG −4% early |
| Financials | −~0.3% | Curve/credit drag | FICO −~7%; FCNCA +~17% pre-mkt |
| Health Care | −~0.3% | Laggard | — |
Sector moves are SPDR ETF proxies from late-session reads (Benzinga / Reuters). Final closing sector prints were not confirmed at publish. "n/c" means no confirmed print, so those rows are left unranked rather than guessed.
Semiconductors led the tape. NVDA made its first all-time high since May at $237.88 but closed below its prior record close, which matters if the move is to stick. Breadth inside the group was strong. TER +8.0% on a strategic partnership with Tokyo Electron, ALGM +8.5%, MTSI +7.9% and AAOI +7.8%. ON +5.9% even though it is paying up: it switched its SYNA bid to $123/share all-cash (~$5.7B) after a competing third-party proposal, and the market rewarded the clean structure. SMH +2.7% at midday. Cramer-style "logic, accelerators and memory all at once" enthusiasm is in the tape, and we'd stay with it until the minutes.
Storage was the weakest group on the day. A Nikkei report that Toshiba will spend ~¥60B (~$380M) to double HDD capacity by FY2027 pressured the most crowded AI-hardware trade. WDC −10.1% (−12.5% at the lows) and STX −10.0% (−14.8% early) were the worst names in the S&P tech sleeve, and SNDK −3.4% early in sympathy. STX is up ~190% YTD, so this is positioning stress more than a thesis break. Supply discipline is the whole bull case, though, and a credible new capacity number is the bear case. We would not buy it Monday.
Consumer split in two. TSLA +~5% on 486,532 Q3 deliveries against 463K Bloomberg consensus (−2% y/y) and 13.7 GWh of storage deployed, the biggest S&P contributor after NVDA. NKE −5.8% to ~$33.10. FQ1 revenue was $11.21B against $11.33B consensus (−4% y/y), FY27 guidance calls for a high-single-digit revenue decline, layoffs are coming, and China is weak. Gross margin +60bp to 42.8% is the only clean line in the print.
| SYNA | ON flips bid to $123 all-cash | +~14% |
| TER | Tokyo Electron partnership; led S&P | +8.0% |
| AAOI | Optical/chip beta | +7.8% |
| HPE | $1.2B Vultr order, networking guide up, PT hikes | +7.6% |
| SPCX | Three launches in <13 hrs | +7.6% |
| ON | Cleaner SYNA deal structure | +5.9% |
| TSLA | Q3 deliveries 486.5K vs 463K est | +~5% |
| WDC | Toshiba to double HDD capacity | −10.1% |
| STX | Same; worst S&P tech print | −10.0% |
| FICO | FHFA cuts to two bureaus | −~7% |
| ACN | Gives back Thursday's +16% pop | −6.0% |
| NKE | Rev miss; FY27 HSD decline guide | −5.8% |
| APP | New 52-wk low; TRO vs Unity denied | −4.4% |
| MOD | Post-Gentherm spin-off | −4.8% |
Closing prints per 24/7 Wall St. 4:10 PM board where available. "~" = last confirmed intraday print.
NVDA opened at $236.05 and ran to $237.88, its first all-time high since May, before closing below the prior record close. It gained roughly 1.3% on the day. The setup is simple: rate relief plus continued AI capex. The tell is the close. A record intraday high that can't hold into the bell is a flag on the chart. A close above $237.88 next week turns it into a breakout. We'd own it into the minutes but not chase above the high.
Nikkei reported that Toshiba will invest ~$380M to double data-center HDD capacity within FY2027. STX traded as low as −14.8% at the open before recovering a bit, and WDC was −12.5% at its worst. Both names have priced near-perfect supply discipline since data-center demand outran capacity. Toshiba adding capacity on a 2027 timeline doesn't change 2026 numbers, but these multiples were pricing something like a permanent shortage. Expect sell-side "overreaction" notes over the weekend, and wait for them to fade before buying.
onsemi amended its June 25 agreement to buy Synaptics for $123/share in cash (~$5.7B), replacing the all-stock structure after an unsolicited competing proposal. SYNA now trades as a merger-arb spread, with the competing bidder as a free option. ON rallying while paying more tells you the street prefers no dilution and cash certainty. The spread to $123 is the trade now.
Momentum from HPE's networking investor day carried through. The company announced a $1.2B AI systems order from Vultr on 9/30 and raised FY27 networking growth guidance to high-teens to low-20s%, up from 14–17%. Barclays, Truist and Morgan Stanley all raised targets. The stock closed at $69.49, near the session high. DELL also gained ~4% on AI-server backlog read-through.
FQ1 revenue was $11.21B against $11.33B consensus, down 4% y/y, and EPS was $0.48 against $0.49 a year ago. The real damage is the FY27 guide for a high-single-digit revenue decline, plus a restructuring that cuts roles across the company. Gross margin +60bp is the only clean line. The stock is near $33 and the turnaround timeline keeps moving out. It is still a source of funds, not a value buy.
Thursday's numbers were good: FQ4 EPS $3.29 vs $3.18 and revenue $18.68B vs $18.03B, and the stock jumped ~16% on Thursday. Then holders reread the FY27 guide of 3–6% local-currency growth, with pricing pressure and a ~$1B Middle East headwind, and sold ~6% Friday. The stock is still ~50% off its YTD highs, so this was a reassessment of a squeeze, not a new low. AI services demand is broadening, but the tape isn't paying for it yet.
FHFA will direct Fannie and Freddie to have lenders pull from two credit bureaus instead of three, with a formal announcement expected later in October. This adds to the pressure from last week's mortgage-pricing change. The scoring franchise is losing pricing power in stages, and any rally should be sold until the formal rule is out.
The cross-asset message is mixed. A weaker labor market should have bull-steepened the curve. Instead the front end and the 10Y both closed higher, because the market doesn't believe the Fed can ease into $6.37 diesel, $100+ Brent and a war premium. Today's G7 diesel release dealt with a symptom and not the cause. Crude bounced $3 off its lows by the settle, and Brent finished almost unchanged. For equities, that keeps growth duration stocks exposed to a long-end repricing. The rally worked today because the move in yields was small. A 10Y back through 5.34% would undo it.
Gold falling on a soft jobs print tells you the same thing: real yields and the dollar are winning. The cleanest macro hedge for a long-chips book is short duration rather than long gold, at least through CPI on 10/14.
Stocks closed near the highs as a 29K payrolls print took the October hike off the table and money rotated hard back into chips, with NVDA at a record and the Nasdaq leading. The most important development was in bonds, not stocks. The 10Y finished higher on a soft jobs number, which means the market is pricing an oil-and-war inflation problem that one weak payrolls report can't fix. Wednesday's FOMC minutes and the 10/14 CPI decide whether this is a new leg up or a relief rally that hits a yield ceiling. Lean long into Monday: keep the semis overweight, fade any storage bounce, and buy a weekend Iran gap toward 7,666 rather than selling it, with a rates hedge on above 5.34%.