| Component | Current RS | % 1D | % 1W | % 1M | % off 52W H |
|---|---|---|---|---|---|
| Broad Indices | |||||
| SPYS&P 500 | 50 | −0.38 | +0.11 | +0.21 | −1.18 |
| QQQNasdaq 100 | 42 | +0.04 | +0.35 | +0.6 | −3.97 |
| IWMRussell 2000 | 52 | +0.28 | +0.09 | −0.75 | −3 |
| DIADow 30 | 51 | −0.53 | −0.18 | −0.76 | −2.32 |
| Market Segments | |||||
| IWFLarge-cap Growth | 37 | −0.02 | +0.54 | −0.07 | −4.44 |
| IWDLarge-cap Value | 65 | −0.67 | −0.27 | +0.59 | −0.98 |
| MTUMMomentum | 33 | +1.82 | +1.72 | −1.07 | −11.79 |
| SPHBHigh Beta | 47 | +1.46 | +1.48 | +1.92 | −5.83 |
| SPLVLow Volatility | 41 | −0.66 | −0.45 | −1.99 | −5.18 |
| IWOSmall-cap Growth | 46 | +0.12 | −0.36 | −1.76 | −4.83 |
| IWNSmall-cap Value | 59 | +0.4 | +0.66 | +0.35 | −1.42 |
| Equal-Weight Sectors | |||||
| RSPS&P 500 Equal Weight | 53 | −0.48 | −0.77 | +0.19 | −1.98 |
| RSPTTechnology (EW) | 49 | −0.16 | −0.94 | +1.07 | −4.93 |
| RSPFFinancials (EW) | 76 | −0.95 | +0.07 | +0.87 | −0.95 |
| RSPNIndustrials (EW) | 36 | +0.2 | −1.61 | −4.88 | −6.52 |
| RSPHHealth Care (EW) | 95 | −1.11 | +0.03 | +6.6 | −1.94 |
| RSPDConsumer Discretionary (EW) | 42 | −0.71 | −2.17 | −3.83 | −7.93 |
| RSPSConsumer Staples (EW) | 48 | −0.74 | −1.47 | −0.31 | −7.43 |
| RSPUUtilities (EW) | 23 | +0.14 | +0.28 | −1.31 | −9.84 |
| RSPMMaterials (EW) | 51 | −0.27 | −1.03 | +0.77 | −3.14 |
| RSPGEnergy (EW) | 71 | −0.94 | +2.03 | +10.88 | −1.96 |
| RSPRReal Estate (EW) | 25 | −0.48 | −1.14 | −2.19 | −5.75 |
| RSPCCommunication Svcs (EW) | 39 | −1.61 | −1.48 | +1.37 | −10.67 |
| Sector SPDRs | |||||
| XLKTechnology | 48 | +0.7 | +0.86 | +1.05 | −5.76 |
| XLFFinancials | 72 | −0.79 | 0 | +0.5 | −0.85 |
| XLIIndustrials | 37 | +0.41 | −1.06 | −5.14 | −6.87 |
| XLVHealth Care | 75 | −1.04 | +0.17 | +4.26 | −2.91 |
| XLYConsumer Discretionary | 35 | −1.33 | −1.96 | −2.7 | −8.08 |
| XLPConsumer Staples | 40 | −0.8 | −1.02 | −0.62 | −6.17 |
| XLUUtilities | 26 | +0.12 | +0.82 | −0.69 | −9.87 |
| XLBMaterials | 47 | −0.34 | −1.39 | +0.52 | −3.23 |
| XLEEnergy | 72 | −0.87 | +2.2 | +10.14 | −2.23 |
| XLREReal Estate | 30 | −0.72 | −1.24 | −1.96 | −5.44 |
| XLCCommunication Svcs | 36 | −1.19 | −0.85 | +0.76 | −6.96 |
| Commodities | |||||
| GLDGold (GLD) | 44 | −0.84 | −0.52 | +4.39 | −20.19 |
| USOWTI Crude (USO) | 58 | −0.09 | +9.45 | +19.42 | −7.87 |
| SLVSilver (SLV) | 26 | −1.21 | −0.33 | +7.11 | −45.53 |
| Bitcoin | |||||
| BTC-USDBitcoin | 100 | +5.12 | +3.87 | +28.13 | −35.66 |
On the surface 2026-09-04 was a nothing day: SPY closed 770.19, down 0.38%, QQQ finished 717.95, up 0.04%, and IWM actually led at 296.01, up 0.28%. Three indices, three directions, none of them far from unchanged. A one-line close-of-day summary would call it "mixed" and move on.
That summary would miss the most violent single-day factor rotation of the week. Underneath a flat print, capital moved out of software and consumer and into the physical layer of the AI build with real force. This was not a risk-off day and it was not a risk-on day. It was a re-underwriting day — the market re-priced *which* part of the AI trade actually earns money, and it did so with the index barely twitching.
The August payrolls report landed before the open and it was not close. CNBC and Reuters both put nonfarm payrolls at +162,000 against a Dow Jones consensus near 53,000, with the unemployment rate holding at 4.1% and prior months revised up 55,000. That is the strongest monthly gain since March. Per CNBC's rates desk, the 2-year yield rose about 5.3bp to 4.39% — its highest since January 2025 — and the 10-year added roughly 1.8bp to 4.782%. Fed funds futures moved the odds of a quarter-point hike at the September meeting from about 52% to 59%.
Sit with that. The debate in this market is no longer how fast the Fed cuts. It is whether the Fed *hikes* in two weeks. And equities responded by falling a third of a percent.
That is the tell, and it is the contrarian read of the day: equities are treating a hawkish labor surprise as a growth confirmation, not a discount-rate shock. The cleanest evidence is IWM finishing green while SPY finished red. Small caps are the most rate-sensitive cohort in the index complex. If the tape were trading the rates channel, they would have been the worst thing on the screen. They were the best. The market chose nominal growth over the discount rate — a regime signal, not a one-day quirk.
The precious complex agreed, and it is the honest counterweight: GLD fell 0.84% to 406.77 and SLV dropped 1.21% to 59.82. Higher real yields and a receding cut path do exactly that to metals. If you want a bear case for equities, it is not in the index — it is in the fact that the hedge instrument sold off on the same catalyst.
The leadership board was a single sentence long. Memory and storage went vertical: SNDK +11.9%, SK hynix ADRs +8.1%, KLAC +7.3%, MRVL +7.0%, STX +6.3%, MU +6.1%, WDC +5.9%. 24/7 Wall St. tied the move to widening demand past HBM into the mass-capacity NAND and drive tier, with Seagate's fiscal Q4 revenue up roughly 48% year over year and a guide toward $4.1B. The interconnect and light layer followed — COHR +6.6%, GLW +5.7%, AAOI +5.1%, ALAB +9.8% — and the compute-landlord tier ran with it: CBRS +10.3%, NBIS +7.5%, IREN +7.3%, CRWV +5.7%, BE +7.4%.
The other side of the ledger was almost entirely software. GWRE -19.9% on a light fiscal Q1 revenue guide despite a clean quarter, PATH -16.6%, FICO -16.7%, ADSK -8.3%, ADBE -6.7%, SNPS -5.4%, SNOW -5.4%, ZS -4.5%, PLTR -4.5%. Seeking Alpha framed the enterprise-software group as led lower by earnings, and the pattern rhymes with what the sell side has been calling the "SaaSpocalypse" — roughly $2 trillion of software market value erased since late 2025 on the thesis that AI collapses seat-based pricing.
Here is where I part ways with the consensus write-up. The single-day evidence does not support "AI kills software." GWRE and PATH were punished for *guidance*, not for an AI disclosure. What actually happened is a duration and certainty trade: with a hike back on the table, the market paid up for revenue you can count in units — wafers, drives, transceivers, megawatts — and defunded revenue that requires a multi-year renewal assumption. Same AI thesis. Different collateral.
Strip out the factor story and the genuinely new information on 2026-09-04 was demand-side. LULU -17.4% to 100.61 after cutting full-year EPS guidance to roughly $9.48–$9.73 from $10.95–$11.15, with revenue down about 4% and North America comps down 12% (Forbes, Motley Fool). OXM -15.7%. DASH -4.6%. TMUS -3.5%. SPOT -3.2%.
A labor market printing 162,000 jobs is not supposed to coexist with discretionary comps down double digits. One of those two signals is wrong, and the equity market spent the session hedging that question rather than answering it — which is precisely why the index finished flat.
My read: this was a healthy, well-behaved rotation inside a market that has already decided the AI capex cycle is real and is now sorting it by proof of cash. The risk is not that the theme breaks. It is that a hike gets priced, the consumer keeps deteriorating, and the market discovers it owns one very crowded trade in the only place it was willing to hide.
The clear leadership group of the day, and specifically the memory and storage tier rather than the logic tier — a widening of the AI bid out of HBM and into NAND, DRAM and mass-capacity drives. Capital equipment moved with it, which is the signal that the market is underwriting a capacity cycle rather than a single quarter's pricing. This is now the group carrying the index's growth narrative on days when the macro is hostile.
The compute-landlord and neocloud complex advanced firmly against a flat tape, extending a pattern where the market pays for contracted, unit-countable AI revenue. With consensus 2026 hyperscaler capex tracking near $772B and pushing toward $1T for 2027, the group is trading on visible order books rather than on model releases. The tail risk is concentration: much of the revenue traces back to a handful of counterparties.
Interconnect and transceiver names participated strongly, consistent with 800G-plus shipments projected to more than double this year and 1.6T procurement plans reportedly doubling for 2026. This is the bottleneck layer of the build — capacity constrained through at least 2027, which means pricing power. It behaved today as a high-beta expression of the same trade driving memory.
The laser, component and materials layer beneath the transceiver names also caught a bid, with strategic capital from accelerator vendors having already anchored the supply chain earlier this year. The read is that the market is now underwriting the light path as durable infrastructure rather than a cyclical component business. Margin structure, not unit volume, is the thing to watch here.
Not a leadership group today — the power complex participated through fuel-cell and data-center-power adjacencies rather than through reactors themselves. The structural case is intact, with a reported 9.8GW-plus of announced nuclear capacity committed to AI data centers and every major hyperscaler having signed at least one deal. Treat it as a follower of the AI-infrastructure impulse for now, with construction timelines the standing risk.
The most out-of-favor cohort of the last two years closed green while large-cap beta closed red, and it did so on a hawkish rates catalyst that should have hurt it most. That inversion is the first genuinely interesting breadth signal in weeks. One session is not a trend, but systematic strategies have reportedly been rebuilding small-cap exposure from July lows, which gives it a flow underpinning.
A group left for dead through the higher-for-longer stretch produced some of the largest single-day advances on the board. A labor print this strong argues the demand floor under residential and light-commercial construction is firmer than the group's multiple implies. Early, thin, and entirely dependent on the rate path not overshooting.
Another deeply out-of-favor cyclical corner showed outsized strength against a flat index, consistent with a market starting to look past a long destocking cycle. The tell to watch is whether the strength is order-driven or purely a low-expectations reset. For now it reads as the second thing after small caps that a strong labor market genuinely helps.
Modest but positive participation on a day when broad discretionary was being liquidated, which is a meaningful relative signal for a group trading at recession-adjacent multiples. If the payrolls number is real and the consumer weakness is concentrated in goods and apparel rather than experiences, this is the group that benefits from the split. Watch fuel staying flat as a supporting condition.
Pockets of beaten-down consumer lending and point-of-sale credit traded sharply higher against a tape that was punishing consumer names, suggesting the market is separating credit providers from the merchants themselves. A strong labor market is the single best input for this group's loss curves. The counterweight is obvious: a September hike raises their funding cost at exactly the wrong moment.