| Component | Current RS | % 1D | % 1W | % 1M | % off 52W H |
|---|---|---|---|---|---|
| Broad Indices | |||||
| SPYS&P 500 | 50 | +0.34 | +1.1 | +0.17 | −1.76 |
| QQQNasdaq 100 | 47 | +0.08 | +0.55 | −5.13 | −8.1 |
| IWMRussell 2000 | 53 | −0.47 | +0.01 | −2.71 | −3.81 |
| DIADow 30 | 55 | +0.54 | +1.07 | +0.37 | −1.54 |
| Market Segments | |||||
| IWFLarge-cap Growth | 34 | +0.76 | +0.57 | −3.82 | −8.38 |
| IWDLarge-cap Value | 66 | +0.44 | +1.44 | +3.26 | −0.36 |
| MTUMMomentum | 55 | +0.27 | −2.22 | −8.69 | −13.31 |
| SPHBHigh Beta | 50 | +0.05 | −0.92 | −8.17 | −11.25 |
| SPLVLow Volatility | 45 | −0.2 | −1.24 | +1.32 | −3.29 |
| IWOSmall-cap Growth | 47 | −0.75 | +0.12 | −5.22 | −6.56 |
| IWNSmall-cap Value | 58 | −0.25 | −0.08 | −0.21 | −1.45 |
| Equal-Weight Sectors | |||||
| RSPS&P 500 Equal Weight | 56 | −0.17 | +0.67 | +0.75 | −1.42 |
| RSPTTechnology (EW) | 79 | +0.57 | +1.84 | −4.66 | −10.48 |
| RSPFFinancials (EW) | 69 | −0.28 | +1.3 | +4.67 | −1.75 |
| RSPNIndustrials (EW) | 44 | +0.53 | −1.45 | −1.54 | −2.82 |
| RSPHHealth Care (EW) | 84 | −0.43 | +2.84 | +3.24 | −1.92 |
| RSPDConsumer Discretionary (EW) | 46 | −0.47 | +3.86 | +1.41 | −5.82 |
| RSPSConsumer Staples (EW) | 49 | −0.66 | +1.25 | +2 | −7.45 |
| RSPUUtilities (EW) | 24 | −0.78 | −4.18 | −1.67 | −6.66 |
| RSPMMaterials (EW) | 33 | −1.9 | −0.9 | +0.58 | −5.98 |
| RSPGEnergy (EW) | 29 | +1.45 | −0.44 | +10.38 | −6.29 |
| RSPRReal Estate (EW) | 50 | −0.04 | −0.95 | +1.05 | −2.48 |
| RSPCCommunication Svcs (EW) | 16 | +0.12 | +2.16 | −0.97 | −14.23 |
| Sector SPDRs | |||||
| XLKTechnology | 70 | −0.22 | −0.3 | −5.53 | −11.76 |
| XLFFinancials | 68 | −0.11 | +1.12 | +3.94 | −1.15 |
| XLIIndustrials | 47 | +0.81 | −1.54 | −1.92 | −3.54 |
| XLVHealth Care | 75 | −0.59 | −0.01 | +1.89 | −3.55 |
| XLYConsumer Discretionary | 30 | +3.29 | +6.11 | −1.69 | −7.14 |
| XLPConsumer Staples | 40 | −0.49 | +1.09 | +2.1 | −5.65 |
| XLUUtilities | 19 | −0.69 | −4.19 | −0.94 | −7.22 |
| XLBMaterials | 30 | −2.34 | −1.62 | −1.16 | −6.85 |
| XLEEnergy | 36 | +1 | −0.12 | +12.76 | −6.16 |
| XLREReal Estate | 42 | −0.51 | −1.92 | +2.01 | −2.98 |
| XLCCommunication Svcs | 13 | +1.56 | +1.83 | −1.37 | −10.1 |
| Commodities | |||||
| GLDGold (GLD) | 0 | −1.49 | −0.1 | +0.25 | −27.11 |
| USOWTI Crude (USO) | 0 | +1.33 | −5.5 | +25.08 | −16.17 |
| SLVSilver (SLV) | 0 | −2.13 | −0.44 | −2.28 | −52.33 |
| Bitcoin | |||||
| BTC-USDBitcoin | 0 | +1.29 | +0.66 | +2.45 | −48.75 |
On the headline, 2026-07-31 was a non-event: SPY closed 744.19, up 0.34%; QQQ added 0.08% to 684.10; IWM went the other way, off 0.47% to 291.20. Two of three green, one red, nothing that changes a chart. Underneath, the dispersion was enormous — AMZN +15.3%, GOOGL +6.7%, META +3.3% against AAPL -7.4%, RBLX -26.9%, RDDT -21.0%, COIN -10.6%. This was a day of enormous single-name variance netting to almost exactly zero at the index level. That is a specific market condition, and it is not the same thing as calm.
The more interesting tape was in bonds. The Fed held at 3.50%–3.75% on Wednesday in a 9-3 vote, and per CNBC the dissents leaned hawkish, not dovish — Cleveland's Beth Hammack said plainly that now is the time for the Committee to act to speed inflation back to 2%. Friday's curve reflected it: the 2-year eased about 2bp to ~4.27% as the market took a near-term hike off the table, while the 10-year rose ~8bp to ~4.69% and the 30-year rose ~12bp to ~5.21%, the highest since 2007. Core PCE is running 3.3% year over year. A bear steepener into a cycle-high long bond is a term-premium and inflation-persistence event, not a growth event — and equities absorbing it with a shrug is the single most notable thing about the session.
Amazon and Apple told the same story from opposite ends of the same supply chain, and the market paid for one and charged for the other.
Amazon raised 2026 capex roughly 10% to about $220 billion and explicitly named higher memory costs as the driver, per CNBC. It got away with it because AWS grew 37% to $42.2 billion with a 39.4% operating margin and a contracted backlog reported near $496 billion. Apple beat on its quarter but guided September revenue growth to 9%–11% against roughly 12% consensus, blaming supply constraints — specifically DRAM reallocation toward AI data centers and advanced-node foundry capacity sold out well into 2027.
Context from the trade press: DRAM spot pricing is up several hundred percent year over year, Q3 contract prices are still guided up 13%–18% for DRAM and 10%–15% for NAND, and memory has gone from roughly 15%–18% of a PC bill of materials to about 35%. So the same input cost is revenue for the hyperscaler and margin compression for the device maker. What makes the day genuinely odd is that the memory suppliers themselves were sold — MU -5.9%, SNDK -5.1% — on a day when their pricing power was cited as the reason two of the world's largest companies guided the way they did. That is a peak-pricing argument starting, not a demand argument. Worth watching whether it holds.
GLD fell 1.49% and SLV dropped 2.13%. Precious metals have been the cleanest expression of "cuts are coming," and they are unwinding as the market repositions toward "they might hike." Meanwhile USO gained 1.33%, with front-month WTI futures reported up roughly 2.3% to around $85 on a ~7.2 million barrel US inventory draw and continued disruption around both the Hormuz and Bab el-Mandeb chokepoints. Crude capped a July gain north of 20%, its strongest month since March.
Bitcoin slid under $63,000, down about 3% on the day, though still up roughly 7.5% for July; COIN dropped 10.6% on soft trading revenue. CoinDesk's read — that leverage was already flushed in late June, so there was little forced selling left — is the constructive spin, but the tape is consistent with a simple story: assets that need falling real rates got hit, assets levered to physical supply shock got bid.
That combination is the uncomfortable one. Energy leading while the long end steepens is precisely the sequence that makes the next inflation print worse, which makes the hawkish dissents louder, which makes the long end steepen further. Nobody traded that loop today. Somebody will trade it in August.
The day's worst outcomes weren't AI-capex skeptics — they were platforms sitting downstream of someone else's algorithm. Roblox fell 26.9%, its worst session ever, after a recommendation-algorithm change hurt in-app spending and management guided to the first negative bookings growth in company history. Reddit dropped 21% despite 61% revenue growth and guidance above consensus, because the CEO called search referrals "choppy" and no new AI licensing deals were announced.
Both were beats. Both were destroyed. The market is drawing a hard line between owning a piece of the AI stack and renting traffic adjacent to it — and it is willing to ignore a good quarter entirely to enforce it.
The group split cleanly along a cost-vs-content line: design, analog and power-management names traded well while memory and legacy-node exposure was sold. That is not a demand verdict — it is the market beginning to argue about how much longer memory pricing power lasts, on a day when two mega-cap guides named memory inflation by name.
The picks-and-shovels layer — power distribution, thermal management, switching and networking silicon — was the day's most consistently green cohort, and it was validated by a hyperscaler raising capex roughly 10% rather than trimming it. This remains the cleanest read-through of the capex cycle, and it now has a fresh multi-quarter spending commitment behind it.
Datacenter interconnect and high-speed transceiver names caught a bid as the scale-out story survived earnings intact. The theme's advantage over compute is that it consumes bandwidth budgets rather than memory budgets, which insulates it from the input-cost squeeze that pressured hardware guidance elsewhere.
Compound-semiconductor substrates and photonic components saw outsized moves, reflecting a tight upstream materials chain feeding the same optical build-out. Small float and thin supply mean this group amplifies the AI-infrastructure signal in both directions — treat the magnitude of the moves as leverage, not as incremental information.
Power scarcity got fresh fuel this week: a step-up in announced 2026 datacenter capex lands directly on a grid that cannot deliver it, and SMR developers have already been the reflexive expression of that gap. The theme is now trading as a duration bet on the AI power deficit rather than on any near-term reactor economics, which makes it vulnerable if the long end keeps steepening.
After years as a funding source for growth allocations, oilfield services and offshore-levered names produced some of the day's largest advances, with crude closing out a July gain above 20% on genuine supply disruption rather than demand optimism. The early tell is that these moves came on a flat index day — the group led rather than followed.
Boutique advisory and M&A-levered financials showed unusual strength in an otherwise defensive tape, consistent with a deal pipeline reopening after a long drought. It is a fragile signal while the long end is at cycle highs, since financing costs are the input that stalled this theme in the first place.
A deeply out-of-favor group produced several of the session's largest gainers even as the sector's large-cap complex was sold, which is the shape of early rotation rather than a broad re-rating. Read it as selective and idiosyncratic — the dispersion within healthcare today was as wide as anywhere in the market.
Non-US internet platforms firmed against a soft domestic-platform tape, extending a quiet recovery in a cohort that has been uninvestable-by-consensus for years. Worth monitoring as a relative-value pressure valve if US mega-cap leadership stays this concentrated.