| Component | Current RS | % 1D | % 1W | % 1M | % off 52W H |
|---|---|---|---|---|---|
| Broad Indices | |||||
| SPYS&P 500 | 50 | −1.24 | −2.4 | −1.56 | −4.07 |
| QQQNasdaq 100 | 44 | −1.72 | −6.18 | −8.61 | −11.61 |
| IWMRussell 2000 | 60 | −1.64 | −1.78 | −3.48 | −4.67 |
| DIADow 30 | 58 | −2.18 | −1.16 | −1.2 | −3.22 |
| Market Segments | |||||
| IWFLarge-cap Growth | 28 | −2.09 | −5.68 | −6.5 | −11.68 |
| IWDLarge-cap Value | 75 | −1 | +0.81 | +2.79 | −1.27 |
| MTUMMomentum | 50 | −3.15 | −9.91 | −15.74 | −18.08 |
| SPHBHigh Beta | 50 | −3.15 | −6.96 | −12.1 | −14.55 |
| SPLVLow Volatility | 57 | −0.55 | +1.76 | +2.58 | −1.62 |
| IWOSmall-cap Growth | 53 | −2.14 | −3.32 | −6.8 | −8.32 |
| IWNSmall-cap Value | 65 | −1.12 | −0.2 | +0.14 | −1.29 |
| Equal-Weight Sectors | |||||
| RSPS&P 500 Equal Weight | 67 | −0.9 | +1.42 | +1.26 | −1.09 |
| RSPTTechnology (EW) | 85 | −1.38 | −2.94 | −7.56 | −13.35 |
| RSPFFinancials (EW) | 76 | −1.17 | +2.63 | +7.87 | −1.18 |
| RSPNIndustrials (EW) | 52 | −2.97 | −0.4 | −1.62 | −3.3 |
| RSPHHealth Care (EW) | 94 | −0.11 | +5.62 | +5.05 | −0.82 |
| RSPDConsumer Discretionary (EW) | 56 | −0.02 | +4.96 | +2.31 | −4.64 |
| RSPSConsumer Staples (EW) | 66 | +0.32 | +3.26 | +3.83 | −4.97 |
| RSPUUtilities (EW) | 36 | −1.46 | −2.45 | −2.93 | −5.53 |
| RSPMMaterials (EW) | 45 | −1.32 | +2.29 | +3.1 | −4.08 |
| RSPGEnergy (EW) | 40 | +1.85 | −2.05 | +6.63 | −8.01 |
| RSPRReal Estate (EW) | 64 | +0.12 | +2.62 | +1.57 | −0.43 |
| RSPCCommunication Svcs (EW) | 28 | +0.15 | +2.05 | +1.47 | −12.57 |
| Sector SPDRs | |||||
| XLKTechnology | 60 | −2.64 | −7.6 | −10.16 | −16.18 |
| XLFFinancials | 73 | −1.6 | +1.12 | +5.51 | −1.6 |
| XLIIndustrials | 53 | −3.19 | −1.22 | −3.34 | −5.25 |
| XLVHealth Care | 94 | −0.61 | +4.27 | +3.42 | −1.36 |
| XLYConsumer Discretionary | 26 | −0.77 | −2.11 | −4.7 | −10.72 |
| XLPConsumer Staples | 59 | +0.34 | +3.53 | +3.54 | −3.08 |
| XLUUtilities | 32 | −1.34 | −2.22 | −2.41 | −6.05 |
| XLBMaterials | 44 | −1.15 | +1.81 | +2.13 | −4.43 |
| XLEEnergy | 47 | +1.88 | −0.93 | +9.46 | −7.58 |
| XLREReal Estate | 57 | −0.11 | +2.11 | +2.32 | −1.07 |
| XLCCommunication Svcs | 24 | −0.15 | +0.28 | +1.51 | −9.05 |
| Commodities | |||||
| GLDGold (GLD) | 2 | +0.46 | −2.12 | +0.68 | −27.2 |
| USOWTI Crude (USO) | 17 | +7.32 | −1.8 | +20.76 | −16.08 |
| SLVSilver (SLV) | 0 | +0.14 | −3.99 | −1.73 | −52.86 |
| Bitcoin | |||||
| BTC-USDBitcoin | 0 | +0.24 | −1.85 | +0.83 | −49.45 |
The FOMC left the funds rate at 3.50%–3.75% for a fifth consecutive meeting, but the vote was 9–3 — and all three dissents were for a hike, not a cut (CNN Business and Kiplinger live coverage). That is the most contentious decision of this cycle, and the bond market read it as a committee at risk of falling behind re-accelerating inflation rather than one comfortably parked.
The reaction was a textbook bear steepener. Per CNBC, the 2-year slipped about 4bp to 4.236% while the 10-year added roughly 5bp to 4.657% and the 30-year jumped more than 9bp to 5.193% — Bloomberg flagged that as the highest long-bond yield since 2007. Swaps closed pricing something near a 60% probability of a September hike under Chair Warsh. Short end easier, long end worse: that is the market repricing term premium and inflation persistence, not policy path.
That 5.19% handle is the number that mattered for equities. It is a discount-rate problem aimed squarely at the longest-duration assets on the board — which is precisely where the damage landed. SPY closed -1.24%, QQQ -1.72%, IWM -1.64%; the Dow shed roughly 1,150 points, or about 2.1%. Growth underperformed value, small caps underperformed large, and the Nasdaq 100 slid further into what several desks are now openly calling correction territory.
The second shock was geopolitical and it arrived pre-open. After IRGC forces launched ballistic missiles at US forces in the Middle East on Tuesday — all intercepted — President Trump told Fox News the US would hit Iran hard. Brent gained 7.9% to settle at $90.74 and WTI advanced 6.6% to $84.46 (CNBC, ABC News). Our own crude proxy closed +7.32%, the single largest move on the board today in either direction.
Here is the uncomfortable interaction the curated feeds tend to under-price: an oil shock on the same afternoon the long bond breaks to 19-year highs is not two independent stories. It is one story about the inflation path. A hawkish hold is survivable. A hawkish hold plus an 8% crude move plus a Middle East escalation risk premium is how a September hike goes from a 60% coin-flip to a base case — and how a market that has spent 2026 discounting an eventual easing cycle has to re-underwrite the whole thing. Gold's quiet +0.46% and bitcoin's flat session near $63.9k both fit that read better than the equity tape does: this was a rates event dressed as a risk-off event.
The internals argue against panic. Decliners beat advancers by only about 1.43-to-1 on the NYSE and 1.39-to-1 on the Nasdaq, and 43% of S&P 500 members still closed green — poor, but nothing like a washout. Ten of thirty Dow components advanced. Caterpillar (-6.9% on our tape, following a Baird downgrade citing data-center investment and regulatory risk) and Goldman Sachs together supplied roughly 72% of the Dow's point loss. Strip those two and the index decline is ordinary.
The sector map is the real tell. Energy was strongly higher as crude ripped; industrials and financials led the downside; and the semiconductor complex was the epicenter, with SOXX off more than 4% against a far shallower XLK decline. Memory and AI-infrastructure names took the worst of it — Micron -9.9%, KLA -10.8%, Applied Materials -8.4%, Arm -8.1%, Vertiv -17.3% on organic growth of 17.8% against a ~23.6% consensus, plus double-digit losses across the neocloud and AI-power cohort. Asia had already telegraphed it: Korea's Kospi fell about 6% overnight on reporting that SK Hynix may slow HBM expansion in favor of commodity DRAM (TheStreet).
And yet nearly every large-cap enterprise software name on our board closed higher: Manhattan Associates +21.3%, Intuit +6.4%, Adobe +5.7%, Datadog +5.3%, Workday +5.2%, Snowflake +4.6%, ServiceNow +4.6%, Salesforce +3.8%. Yahoo Finance's basket work has 44 of 51 software names positive in July with a median gain near 6%, against a median chip stock down close to 20%. On a day the index fell 1.2%, capital did not leave equities. It moved one layer up the stack.
The consensus framing tonight will be "hawkish Fed plus Iran equals risk-off." I think that is half the picture, and the less interesting half.
The more durable development is that the market is now differentiating inside the AI trade rather than treating it as one factor. For eighteen months, the reflex was that anything touching the buildout — chips, power, cooling, optics, neoclouds — traded as a single high-beta instrument. Today it did not. The capex-consuming layer was liquidated while the capex-benefiting layer was bought, and the pivot fact underneath is the suggestion that hyperscalers may have enough GPU capacity to become sellers rather than perpetual buyers. If GPU scarcity has an expiration date, then the entity that captures AI margin is the software company that rents compute cheaply, not the vendor that sells it. That is a rotation with legs, and it does not require an AI-bubble collapse to keep working.
The risk to that thesis is that a 5.19% long bond is not selective. Multiple compression via term premium hits high-multiple software eventually too; today it simply hit chips harder. Watch whether the software bid survives a second consecutive up-in-yield session. If it does, this is a genuine leadership handoff. If it doesn't, today was just the first leg of a broad duration de-rate that happened to start with semis.
The epicenter of the session, with the semiconductor index down roughly four times the broad tech sector and memory/equipment names carrying the worst of it after overnight reporting that HBM expansion plans may be throttled in favor of commodity DRAM. This is a valuation and capex-durability reset rather than a demand miss — the group's earnings have largely been fine, but the multiple assigned to those earnings is being renegotiated in public. With a 30-year yield at 19-year highs, the highest-multiple end of the complex has the least defense.
The power, cooling, and neocloud cohort took the deepest single-name damage of the day, in several cases on results that beat outright but decelerated on organic growth — the market is now punishing the second derivative, not the print. The bear case in circulation is that hyperscalers with excess capacity become sellers rather than perpetual buyers, which converts the group from scarcity plays into cyclicals. A large-cap machinery downgrade citing data-center investment and regulatory risk shows the theme has begun leaking into industrials.
Optical networking traded as a pure AI-beta proxy today rather than on its own fundamentals, giving back a slice of an enormous year-to-date run alongside switching hardware. Positioning here is the problem: the complex is crowded, high-beta, and owned by generalists who bought it as an AI expression, so it de-risks first when the parent theme wobbles. Nothing in the day's news touched interconnect demand itself.
The underlying technology roadmap — co-packaged optics, silicon photonics, optical circuit switching — is unchanged by a rates shock, and multi-quarter backlog and hyperscaler design-in commentary from earlier in the quarter still stand. What repriced is the willingness to pay forward multiples on that roadmap while the long end is repricing term premium. Worth separating the tape action from the physics: copper still runs out of headroom at cluster scale.
The nuclear and next-gen power theme has been derating through July as investors demand licensing and deployment evidence instead of AI-adjacency, and today's broad AI de-risking gave it no relief. Enterprise-value-to-sales multiples in the high tens to low nineties leave no cushion when the demand narrative is questioned. Ironically, a crude shock strengthens the long-run baseload argument even as it does nothing for the near-term tape.
After spending most of 2026 as the market's dogs, large-cap application and data-platform software closed broadly green on a down day — the cleanest relative-strength signal on the board. The emerging thesis is that if compute stops being scarce, margin accrues to the layer that rents it rather than the layer that sells it. The unresolved test is whether this bid survives a second consecutive up-in-yield session, since term premium eventually hits high-multiple software too.
Consulting and IT services names posted some of the largest percentage gains in the market today, an unusually strong showing for a group that has spent two years priced as AI roadkill. Deal-flow and bookings commentary is starting to suggest that enterprise AI adoption needs implementation labor, not just models. Early and narrow, but this is the first session in a while where the group led rather than bounced.
Energy was the day's only meaningfully positive sector, and it got there on a geopolitical risk premium rather than an improvement in the demand picture — an important distinction for anyone tempted to call it a trend change. Still, a sector this under-owned catching a bid on a day when everything else de-rated is worth logging. The setup reverses quickly on any Middle East de-escalation headline.
Healthcare equipment produced one of the day's biggest large-cap gainers on a clean earnings beat, in a group that has been an afterthought for most of the AI cycle. Defensive cash flows are getting a second look now that the rate path has turned two-sided. One session is not a trend, but the sector is cheap enough that a couple of beats can change the conversation.