Both headline averages set records — the Dow closed above 54,000 for the first time at 54,085.88 (+1.71%) and the S&P 500 took out its June high at 7,736.49 (+1.79%) — on a macro input rather than an earnings one: Brent’s 6.08% collapse to $78.68 as Washington signalled a Strait of Hormuz reopening within days. The bond market ratified the disinflation read, with the 10-year falling 6.8 bps to 4.616% and the 2-year 6.2 bps to 4.194%, a second consecutive session of stocks and bonds rallying together. But the construct rests on an unsigned agreement whose June 17 predecessor collapsed over the same routing question still unresolved, and Philadelphia Fed’s Paulson became the third official in three sessions to keep a hike in the distribution. Breadth narrowed rather than broadened: Technology’s 4.25% and two industrial sectors did nearly all the work while five advancing sectors managed under 0.6%.

JOLTS Job Openings Fall to 7.359 Million in June, Missing Estimates as Labor Market Cools (InvestingLive/BLS, August 4, 2026)

What they’re saying:Job openings fell to 7.359 million in June, down from a revised 7.537 million in May and below the 7.4 million consensus estimate. The decline was concentrated in healthcare (-147K), leisure and hospitality (-86K), wholesale trade (-74K), and business services (-71K), while federal government openings surged to their highest level since October 2024.

The context:The miss reinforces signs the labor market is losing momentum even as hires and quits held firmer, a mixed internal picture that still supports the case for a more accommodative Fed path. Treasury yields eased modestly on the print as traders leaned further into a dovish rate outlook.

What to watch:Friday’s (Aug 7) nonfarm payrolls report, expected to show 80K jobs added and the unemployment rate holding at 4.2%.

US Factory Orders Unexpectedly Fall 0.3% in June, Core Orders Miss Badly (Investing.com/Census Bureau, August 4, 2026)

What they’re saying:Factory orders fell 0.3% in June to $656.5 billion, missing the +0.2% consensus and extending May’s revised 1.1% decline. Core orders (ex-transportation) dropped 0.4%, dramatically missing expectations for a 0.4% gain.

The context:The soft print contrasts with July’s ISM Manufacturing PMI hitting its highest level since May 2022, underscoring a bifurcated industrial picture. Equities largely shrugged off the miss, with the S&P 500 trading near record highs on continued AI infrastructure capex.

What to watch:Wednesday’s (Aug 5) ISM Services PMI for confirmation of whether the June softness is isolated to manufacturing or broadening.

US Trade Deficit Narrows to $73.3 Billion in June as Imports and Exports Both Decline (Bloomberg/Census Bureau, August 4, 2026)

What they’re saying:The trade deficit narrowed 5.6% to $73.3 billion in June, roughly in line with the $73B consensus and down from $77.6B in May. Exports fell 0.9% to $314.7B and imports fell 1.8% to $388.0B, with the goods deficit narrowing $3.9B to $102.1B.

The context:The improvement was driven by falling trade volumes on both sides rather than export strength — crude oil exports alone fell $5.7B as prices dropped to $95.82/barrel — pointing to softer global demand rather than a genuine competitiveness gain.

What to watch:July trade data and any tariff-driven shifts in import volumes given the pending Section 301 litigation.

EchoStar’s Hughes Satellite Systems Files Chapter 11 With $1.5 Billion in Debt (Law360/GlobeNewswire, August 3, 2026)

What they’re saying:Hughes Satellite Systems Corporation and certain U.S. subsidiaries filed voluntary Chapter 11 petitions on August 2 in the Southern District of Texas, facing a $1.5 billion note maturity against just $102 million in cash. The filing arrived without a pre-negotiated restructuring plan.

The context:This marks the second Chapter 11 filing from the EchoStar family in weeks, following DISH DBS Corporation’s prepackaged bankruptcy on June 30 — a sign of continued balance-sheet stress across the group as satellite broadband faces intensifying competition from Starlink. EchoStar’s non-Hughes operations, including DISH TV, Sling TV, and Boost Mobile, are unaffected.

What to watch:Progress of the restructuring and whether further EchoStar-affiliated entities face similar liquidity pressure.

25 States Sue to Block Section 301 Tariffs Covering 99.4% of US Imports (CBS News/Boston Globe, August 3, 2026)

What they’re saying:A coalition of 25 states, led by New York, filed suit in the US Court of International Trade against new 10-12.5% tariffs imposed July 23 under Section 301, covering goods from 59 countries and the EU that together account for 99.4% of US imports. The states argue the “forced labor” justification is a pretext to revive tariffs the Supreme Court already struck down.

The context:The suit seeks to have the tariffs declared unlawful, halted, and refunded — an outcome that would remove a meaningful cost overhang for importers if successful, but leaves near-term policy direction unresolved and follows a pattern of legal challenges dogging the administration’s trade agenda.

What to watch:Court of International Trade scheduling and any interim injunction request; a ruling could materially reprice import-cost assumptions across consumer and industrial sectors.

12. Palantir Technologies (PLTR): +29.45% | US Commercial Revenue Up 149% Drives the Single Largest Mega-Cap Move of the Session

The Numbers:Released: AMC, Monday August 3. Total revenue of $1.935 billion against roughly $1.8 billion consensus, up 93% year-over-year and 19% sequentially. US commercial revenue of $764 million, up 149% year-over-year and 28% sequentially; US government revenue of $809 million, up 90%; international government $181 million, up 42%; international commercial $182 million, up 26%. Adjusted gross margin 86%, adjusted operating income $1.194 billion for a 62% adjusted operating margin, GAAP net income $1.062 billion for a 55% margin. FY26 revenue guidance raised to $8.15-$8.158 billion from $7.65-$7.66 billion, US commercial guidance raised to above $3.424 billion representing at least 134% growth, adjusted operating income guidance raised to $4.889-$4.897 billion and adjusted free cash flow to $4.5-$4.7 billion. Shares closed at $162.66.

The Problem/Win:The win is the commercial franchise reaching escape velocity. Palantir has spent years being valued as a government contractor with an interesting commercial option; at $764 million growing 149%, total commercial revenue of $945 million is now within $45 million of total government revenue of $990 million. The company is roughly one quarter away from a balanced revenue mix, and it got there by growing the commercial side rather than shrinking the government side, which grew 90% itself. A 62% adjusted operating margin on 93% revenue growth is the combination that justifies the multiple — most software businesses buy that growth rate with margin.

The Ripple:Palantir was the largest single contributor to the Nasdaq 100’s 3.32% advance and the day’s biggest mega-cap gainer by a wide margin. The print drew same-session upgrades to Buy from both Citi, which raised its target to $235 from $210, and Deutsche Bank. It also supplied the enterprise-software leg of a session in which Technology led all sectors at +4.25%, reinforcing the pattern this report has tracked of capital moving toward the applied-AI layer rather than only the silicon underneath it.

What It Means:The AI software layer is now demonstrating the monetisation that the infrastructure layer has been funding, and Palantir is the cleanest listed expression of it. The risk is entirely valuation and comparison base — 149% commercial growth cannot be annualised indefinitely, and a 29.45% single-session move prices a great deal of the next several quarters.

What to watch:Whether US commercial revenue holds above 100% year-over-year growth next quarter as the comparison base steepens — that is the single number the raised $3.424 billion full-year guidance depends on. Watch whether the two upgrades draw further sell-side revisions or mark the top of the ratings cycle.

13. Vertex Pharmaceuticals (VRTX): +1.75% | Revenue Beat and a Guidance Raise Outweigh a Fractional EPS Miss

The Numbers:Released: AMC, Monday August 3. Revenue of $3.33 billion against $3.22 billion consensus, up 12.4% year-over-year and beating by 3.43%. Adjusted EPS of $4.73 against $4.75 expected, a miss of 0.32%. TRIKAFTA/KAFTRIO contributed $2.50 billion and ALYFTREK $573.6 million as global reimbursement expanded to 25 countries. JOURNAVX, the non-opioid pain franchise, generated $50 million against $12 million a year earlier, up 70% sequentially with prescription growth of 45% and more than 535,000 prescriptions in the quarter. CASGEVY revenue rose 151% year-over-year and 78% sequentially to $76 million. Full-year 2026 revenue guidance was raised to $13.10-$13.20 billion from $12.95-$13.10 billion.

The Problem/Win:The win is that the diversification thesis is finally producing measurable revenue rather than milestones. Vertex has been a cystic fibrosis monopoly searching for a second act for a decade; JOURNAVX at $50 million growing 70% sequentially and CASGEVY at $76 million growing 78% sequentially are both small against a $3.33 billion quarter, but both are compounding at rates that make them material within eight quarters. The fractional EPS miss is noise — 0.32% against a $4.75 estimate is a rounding difference, and the guidance raise is the management signal that matters.

The Ripple:Healthcare was effectively flat at +0.03%, the second-weakest advancing sector on a day the S&P gained 1.79%, so Vertex’s 1.75% gain represented meaningful relative outperformance within a sector that did not participate in the rally. The JOURNAVX trajectory is the read-through with the widest reach: a non-opioid analgesic scaling past half a million prescriptions is a direct commercial validation of a category the entire pain-management complex has been unable to crack.

What It Means:Vertex is transitioning from a single-franchise business to a multi-product one on schedule and is raising guidance while doing it. The valuation still embeds the cystic fibrosis annuity, so the incremental franchises are largely unpriced — the risk is execution and reimbursement rather than science.

What to watch:JOURNAVX sequential growth next quarter — sustaining anything close to 70% would put the franchise on a path to $500 million annualised and materially change the sum-of-the-parts. Watch whether the raised $13.10-$13.20 billion range is lifted again at the third quarter, which would signal the new products are running ahead of internal plan.

14. Caterpillar (CAT): +5.60% | A Record $20.5 Billion Quarter and a 92% Backlog Surge on Data-Center Power Demand

The Numbers:Released: BMO. Revenue of $20.54 billion against $19.34 billion consensus, a 6.25% beat and up 24% from $16.6 billion a year earlier — the first quarter in company history above $20 billion. Adjusted EPS of $8.17 against $6.22 expected, a 31.42% beat and up from $4.72 a year ago. Backlog of $72.1 billion, up 92% year-over-year with all segments contributing and orders extending into 2030. The Power & Energy segment grew 17% to $8.2 billion, with power generation sales — predominantly data-center related — climbing 29%, and power generation up 72% on data-center demand specifically. Full-year revenue growth guidance was raised to the mid-to-high-teens percentage range from a prior low-double-digit target, and the tariff cost estimate was narrowed to roughly $2.2 billion from $2.2-$2.6 billion. Market cap $403.77 billion.

The Problem/Win:The win is the backlog, not the quarter. A 92% year-over-year increase to $72.1 billion with orders booking into 2030 converts Caterpillar from a late-cycle machinery business into something closer to a contracted infrastructure provider, and it removes the single biggest analytical objection to owning the name at this valuation — that data-center demand is a guided number rather than an order book. It is now an order book. The company is also resuming production of its 10-megawatt gas reciprocating engine platform, which is a capacity decision rather than a forecast.

The Ripple:Caterpillar was the largest point contributor to a Dow that gained 907.47 points and closed above 54,000 for the first time, and it underpinned Industrials’ 2.71% advance as the third-best sector. This report has repeatedly used industrial and electrical order books as the cleanest available confirmation that hyperscaler capital spending is landing in the physical economy rather than remaining a guided figure. Today supplied that confirmation at scale, and it arrived alongside yesterday’s Prysmian-Atkore transaction and the Eaton and Ingersoll-Rand upgrades in the same electrical-infrastructure complex.

What It Means:The AI capital-expenditure cycle now has a contracted, multi-year physical order book behind it, which materially lowers the probability that the capex narrative unwinds abruptly. The offsetting exposure is the $2.2 billion tariff cost, which is a live number given three multi-state legal challenges to the Section 301 duties now pending.

What to watch:Whether backlog growth holds above 50% year-over-year next quarter — deceleration there is the first place a hyperscaler capex slowdown would appear, well ahead of revenue. Watch the tariff cost estimate for revision if the Court of International Trade acts on the Section 301 challenges.

15. Merck & Co (MRK): +0.18% | Revenue Guidance Raised and Profit Guidance Cut in the Same Release

The Numbers:Released: BMO. Revenue of $16.61 billion against $16.37 billion consensus, a 1.45% beat and up 5.1% year-over-year. Adjusted EPS of -$0.13 against -$0.27 expected, a 51.94% beat, while GAAP EPS came in at -$0.54 against -$0.37 expected, a 44.97% miss. KEYTRUDA and KEYTRUDA QLEX sales were $8.4 billion, up 5% and 4% ex-FX, including $463 million from the new subcutaneous formulation. Winrevair generated $588 million, up 75% year-over-year. Full-year 2026 revenue guidance was raised to $66.3-$67.3 billion from $65.8-$67.0 billion, while adjusted EPS guidance was cut to $2.66-$2.76 to absorb a one-time $5.7 billion charge, equal to $2.31 per share, related to the Terns Pharmaceuticals acquisition. Market cap $316.14 billion.

The Problem/Win:The win is the diversification away from Keytruda, which is the entire investment question for Merck. Winrevair at $588 million growing 75% and the subcutaneous Keytruda conversion at $463 million in its early quarters are the two products that determine whether the patent cliff is a step-down or a fall, and both are tracking well enough for management to raise revenue guidance. The problem is the $5.7 billion Terns charge, which is real cash spent on a pipeline asset that will not generate revenue for years and which mechanically converts a revenue beat into a profit guidance cut.

The Ripple:Merck finished up 0.18%, essentially flat, in a Healthcare sector that gained 0.03% and was the second-weakest advancing sector on the day. The Terns charge is the more significant sector signal: it is another large-cap pharma buying pipeline rather than building it, which is the same conclusion the market drew yesterday from the reported AstraZeneca-Bristol Myers merger talks. Two sessions, two confirmations that the sector’s incumbents believe organic growth will not close the patent gap.

What It Means:The underlying business is performing better than the headline EPS suggests, but Merck is paying substantial cash to acquire a future it cannot generate internally. The flat share price is the market declining to take a view until it can assess what $5.7 billion bought.

What to watch:Winrevair’s growth rate next quarter against today’s 75% — that franchise is the largest single non-Keytruda contributor to the raised revenue guidance. Watch management’s business-development commentary for whether Terns is the last acquisition of this size or the first of several.

16. McDonald’s (MCD): +1.17% | US Comps Slip to 0.8% as the Under-$3 Menu Reaches Only Two-Thirds of the System

The Numbers:Released: BMO. Revenue of $7.10 billion against $7.13 billion consensus, a 0.42% miss but up 3.7% year-over-year. Non-GAAP diluted EPS of $3.38 against $3.32 expected, a 1.86% beat and up 6%. US same-store sales rose 0.8%, below expectations; global same-store sales rose 1.3%. Management attributed roughly two-thirds of the US traffic underperformance to value-menu execution: only 60% to 65% of the system has implemented the under-$3 menu, the loose $3 parameter allowed some franchisees to raise prices on items such as small fries, and a pullback on digital offers alienated loyal customers. The June FIFA campaign underdelivered, and too many simultaneous deployments overwhelmed restaurant operations, slowing service times and lowering satisfaction scores. Market cap $190.66 billion.

The Problem/Win:The problem is self-inflicted and management said so directly, insisting the company does not have a strategy problem. That framing is the most useful disclosure in the release: a value menu implemented in two-thirds of restaurants, with a price floor loose enough that some operators raised prices, is not a demand failure — it is a franchise-execution failure, and it is fixable in a way that a consumer pullback is not. The counter-reading is less comfortable. If the low-income consumer were healthy, a partially implemented value menu would not cost two-thirds of US traffic performance, and the reliance on discounting to hold comps at 0.8% is itself the signal.

The Ripple:Consumer Cyclical closed flat at 0.00%, the only sector to register no move on a day the S&P gained 1.79%, and Consumer Defensive managed just 0.37%. McDonald’s is the highest-frequency read on the US low-income consumer available in the large-cap complex, and 0.8% domestic comps landed on the same session that JOLTS job openings missed and JPMorgan cut Nike to Underweight. Three independent consumer signals pointing the same direction on one day is more informative than any of them alone.

What It Means:The share price rose 1.17% because the EPS beat and management framed the shortfall as executional. Investors are being asked to accept that the value proposition is a rollout problem rather than a consumer problem, and the next two quarters will settle which it is.

What to watch:Under-$3 menu system penetration moving from 60-65% toward full implementation — if comps do not improve as coverage rises, the executional explanation fails. Watch US traffic separately from comps, since price alone can hold the comp while traffic declines.

17. Pfizer (PFE): +1.52% | Non-COVID Revenue Up 18% Operationally and the Guidance Midpoint Raised $500 Million

The Numbers:Released: BMO. Revenue of $15.03 billion against $14.40 billion consensus, a 4.41% beat. Adjusted EPS of $0.77 against $0.68 expected, a 12.87% beat, while GAAP EPS came in at -$0.04 against a $0.49 estimate. Revenues from launched and acquired products grew 18% operationally. Full-year 2026 revenue guidance was raised by $500 million at the midpoint to $60.5-$62.5 billion, while adjusted EPS guidance was reaffirmed at $2.80-$3.00. The full-year expectation for COVID products — Comirnaty and Paxlovid — was cut to $4 billion from roughly $5 billion, with Comirnaty revenues declining 34% operationally on a lower favourable returns-provision adjustment and reduced US utilisation. Phase 2b results for the monthly obesity candidate berobenatide, from the $10 billion Metsera acquisition, suggest weight loss comparable to tirzepatide and potentially better than semaglutide. Market cap $144.82 billion.

The Problem/Win:The win is that the post-COVID base business is now growing fast enough to absorb a $1 billion downgrade to the COVID franchise and still support a $500 million guidance raise. Eighteen percent operational growth in launched and acquired products is the number that answers the question Pfizer has faced since 2023 — whether anything replaces the pandemic revenue. The GAAP loss reflects acquisition accounting rather than operations, and the reaffirmed adjusted EPS range confirms management does not view it as recurring.

The Ripple:Pfizer’s 1.52% gain was the strongest of the four large-cap pharma names reporting today, in a Healthcare sector that closed at +0.03%. The berobenatide data is the item with the widest sector reach: a monthly injectable delivering weight loss comparable to tirzepatide would introduce genuine dosing-frequency competition into a market currently split between two weekly incumbents, and it lands the session before Eli Lilly reports.

What It Means:Pfizer has bought its way into the obesity market at a $10 billion price and has early data suggesting it did not overpay. The stock is being valued on the COVID run-off; the guidance raise argues the base business is what should set the multiple.

What to watch:Eli Lilly’s report on Wednesday before the open for any commentary on monthly-dosing competition following the berobenatide data. Watch whether the $4 billion COVID guidance holds or is cut again, since it has now been reduced twice.

18. Space Exploration Technologies (SPCX): -7% AH | Starlink Doubles to 12 Million Subscribers, Revenue Beats by Nearly $1 Billion, and the Stock Falls Anyway

The Numbers:Released: AMC — the company’s first earnings report as a listed company. Revenue of $7.81 billion against $6.83 billion consensus, a 14.46% beat and up 92% year-over-year. GAAP EPS of -$0.09 against -$0.23 expected, a 60.09% beat. Starlink reported 12 million subscribers, double a year earlier and up 17% from the first quarter, against 10.3 million across 164 countries as of March 31. The AI segment moved from a loss into positive adjusted EBITDA for the first time, helped by $14.1 billion in new AI contracts. Shares closed the regular session up 9.43% at $125.33 and fell roughly 7% after hours to about $117. Market cap $1,662.77 billion. The IPO lockup expires August 6; the company listed on June 12 at $135 and peaked at $225.64 on June 16.

The Problem/Win:Every operating metric beat and the stock fell 7%, which makes the sequencing the story rather than the numbers. Yesterday this report flagged that the options market was pricing a post-earnings swing of roughly $204 billion and that a disappointing print followed by the August 6 insider unlock would compress two sources of supply into seventy-two hours. The print was not disappointing — revenue beat by nearly a billion dollars, the loss narrowed by 60%, Starlink subscribers doubled and the AI segment turned EBITDA-positive. The stock fell regardless, which isolates the unlock as the operative variable rather than the fundamentals.

The Ripple:SpaceX was one of the five largest mega-cap gainers of the regular session at +9.43%, contributing to the day’s advance before giving it back after the close. The Starlink subscriber disclosure is the read-through with the longest reach — 12 million subscribers doubling year-over-year is the first audited scale figure the market has had for low-earth-orbit broadband, and it arrives while a satellite-broadband incumbent’s subsidiary sits in Chapter 11 after failing to repay $1.5 billion of bonds that matured August 1.

What It Means:The fundamentals are stronger than the debut price implied and the share price is being set by supply mechanics rather than by the business. That resolves in one direction or the other within days rather than quarters.

What to watch:August 6, when the IPO lockup expires — price action into and through that date is the cleanest available test of whether today’s after-hours decline was anticipatory supply or a verdict on the print. Watch whether the $14.1 billion AI contract backlog is broken out by counterparty in subsequent disclosure.

19. Advanced Micro Devices (AMD): -8% AH | A Beat-and-Raise Undone by Capital Expenditure at Nearly Triple the Modelled Figure

The Numbers:Released: AMC. Record revenue of $11.536 billion against $11.31 billion consensus, a 2.00% beat and up 50% year-over-year. Adjusted EPS of $1.66 against $1.62 expected, a 2.71% beat. Data Center revenue of $6.7 billion, up 107% year-over-year and now 58% of company sales, driven by fifth-generation EPYC processors and Instinct MI350 Series GPUs. Q3 revenue guided to $13 billion plus or minus $300 million against $12.52 billion expected. CEO Lisa Su said Data Center revenue is expected to more than double year-over-year in 2027. Capital expenditure was $808 million against roughly $299 million modelled, and free cash flow fell to $1.56 billion from $2.57 billion in the first quarter as the company buys capacity ahead of the Helios rack ramp. Shares fell roughly 8% after hours, back below $480. Market cap $845.60 billion.

The Problem/Win:The problem is the cash-flow statement, not the income statement. AMD beat on revenue, earnings and margin, guided the next quarter roughly $480 million above consensus, and told the market data-center revenue would more than double again in 2027 — and the stock lost 8%. The reason is $808 million of capital expenditure against a $299 million model, which cut free cash flow by roughly 40% sequentially. Buying capacity ahead of the Helios ramp is defensible and arguably necessary, but it converts a capital-light merchant-silicon story into something that consumes cash to grow, and it does so at a moment when memory and capacity costs are rising across the industry.

The Ripple:The timing is what makes this consequential beyond AMD. The semiconductor complex rallied violently during the regular session — Intel and SanDisk each adding 10.84%, Technology leading all sectors at +4.25% — and then the sector’s most-watched AI reporter beat and sold off on cost. That is the same input-cost inflation that has been lifting memory suppliers arriving on a merchant supplier’s own accounts, and it directly tests the thesis behind today’s rally.

What It Means:Demand is not the question — 107% data-center growth and a raised outlook settle that. The question is what share of AI revenue the merchant suppliers keep after paying for the capacity to deliver it, and today’s answer was worse than expected.

What to watch:Whether the after-hours decline transmits to Intel, Micron and the broader complex at Wednesday’s open, which would mark today’s 10%-plus semiconductor rally as premature. Watch AMD’s capital-expenditure guidance for the second half — a repeat of the $808 million run rate would make the free-cash-flow compression structural rather than a single-quarter build.

20. Arista Networks (ANET): +11% AH | The First $3 Billion Quarter and Operating Margin Expanding to 49.9%

The Numbers:Released: AMC. Revenue of $3.036 billion against $2.83 billion consensus, a 7.26% beat, up 37.7% year-over-year and 12.1% sequentially — the company’s first quarter above $3 billion. Non-GAAP diluted EPS of $1.02 against $0.89 expected, a 15.14% beat. Product revenue of $2.61 billion drove the bulk of the result, with non-GAAP operating margin expanding to 49.9% from 48.8% a year earlier. Q3 revenue guided to approximately $3.30 billion with non-GAAP diluted EPS of $1.06-$1.08. Growth was broad-based across AI networking, data centers, campus and routing. The company introduced 1.6 Tbps AI fabric platforms, including liquid-cooled options for scale-up, scale-out and scale-across networks. Shares closed the regular session up 4.98% and rose roughly 11% after hours. Market cap $239.89 billion.

The Problem/Win:The win is margin expansion alongside 37.7% revenue growth, which is the exact inverse of what AMD reported an hour earlier. Arista grew faster than the merchant silicon vendors and expanded operating margin to 49.9% while doing it, because networking sits at a point in the AI stack where the customer pays for performance rather than for commodity capacity. Guiding Q3 to approximately $3.30 billion implies continued sequential acceleration rather than a plateau.

The Ripple:Arista and AMD reported within the same hour with opposite outcomes — Arista up 11% on expanding margin, AMD down 8% on capital expenditure — and the contrast is the most useful signal of the evening. It suggests the market is no longer paying uniformly for AI exposure but is discriminating by where in the stack a company can defend margin. That is a refinement of the buyers-versus-sellers divergence this report tracked last week, and it favours networking and interconnect over compute silicon.

What It Means:Arista is capturing AI infrastructure spend without the capital intensity that is currently penalising the chip vendors. That combination — 37.7% growth, 49.9% operating margin, no comparable capex build — is the strongest expression of the AI trade currently available in the large-cap complex.

What to watch:Whether operating margin holds near 49.9% as the 1.6 Tbps platforms ramp, since new product introductions typically compress margin before they expand it. Watch cloud-titan concentration in the revenue mix, which is the single largest structural risk to the guidance.

21. Amgen (AMGN): AH: n/a | Six Growth Drivers Up 26% Carry a Guidance Raise on Both Revenue and Earnings

The Numbers:Released: AMC. Revenue of $10.05 billion against $9.43 billion consensus, a 6.66% beat and up 10% year-over-year, with product sales up 9% on volume growth. Adjusted EPS of $6.29 against $5.62 expected, an 11.97% beat. The six key growth drivers — Repatha, EVENITY, TEZSPIRE, rare disease, innovative oncology and biosimilars — grew 26% year-over-year and generated nearly 70% of second-quarter product sales. Full-year 2026 guidance was raised on both lines, revenue to $37.1-$38.5 billion and non-GAAP EPS to $21.70-$23.10, with management citing growth drivers already running ahead of plan. Repatha growth is coming from both cardiologists and primary care physicians, particularly for high-risk diabetes patients. The MariTide obesity programme continues through Phase 3 across weight management, cardiovascular outcomes and heart failure. Shares closed the regular session up 3.03%; no after-hours figure was available at the time of writing. Market cap $210.50 billion.

The Problem/Win:The win is portfolio composition. Six franchises growing 26% and supplying nearly 70% of product sales means the legacy biologics that have historically defined Amgen no longer set the growth rate, and management raised both revenue and EPS guidance rather than trading one for the other — a distinction worth noting on the same day Merck raised revenue and cut profit. Repatha broadening from cardiology into primary care is the specific detail that matters, because primary-care prescribing is what converts a specialty drug into a volume franchise, and it is holding even after Merck’s oral PCSK9 approval.

The Ripple:Amgen’s 3.03% regular-session gain was among the strongest in a Healthcare sector that closed at +0.03%. MariTide is the sector-level variable: a monthly obesity injectable in Phase 3 alongside Pfizer’s berobenatide Phase 2b data released the same morning means two credible monthly-dosing challengers to the weekly incumbents surfaced within hours of each other, one day before Eli Lilly reports.

What It Means:Amgen is executing on the transition it has been promising, and raising both guidance lines is the cleanest form of management confidence available. The obesity programme is unpriced optionality rather than a near-term earnings contributor.

What to watch:Whether the six growth drivers hold above 20% growth next quarter, since the raised full-year guidance depends on them rather than on the legacy portfolio. Watch MariTide Phase 3 readout timing against the competitive monthly-dosing entrants that emerged today.

22. Gilead Sciences (GILD): -0.84% AH | Yeztugo Passes a $1 Billion Run Rate in Its First Full Year on the Market

The Numbers:Released: AMC. Revenue of $7.80 billion against $7.40 billion consensus, a 5.47% beat. Adjusted EPS of -$6.75 against -$7.26 expected, a 7.03% beat, while GAAP EPS of -$8.45 missed a -$7.67 estimate by 10.18%. HIV sales reached $5.7 billion, up 12% year-over-year and 13% sequentially, with Biktarvy at $3.8 billion, up 7% year-over-year and 12% sequentially. Yeztugo generated $232 million, up 40% sequentially and more than double a year earlier, passing a $1 billion annual run rate with strong patient retention; full-year Yeztugo guidance was raised to $1 billion. Base business sales rose 10% year-over-year to $7.6 billion, the company’s strongest second-quarter growth in three years. The FDA accepted a supplemental NDA for Yeztugo as a once-weekly oral HIV pre-exposure prophylaxis with a PDUFA target action date of February 2, 2027. Shares closed the regular session up 3.14% at $135.27 and eased 0.84% after hours to $134.11. Market cap $167.92 billion.

The Problem/Win:The win is Yeztugo reaching blockbuster scale in its first full year, which is fast even by the standards of a well-supported HIV launch. Forty percent sequential growth with strong retention indicates the prevention market is expanding rather than cannibalising Gilead’s existing franchise, and Biktarvy’s 12% sequential growth confirms that. The FDA’s acceptance of the once-weekly oral formulation is the more valuable item on a multi-year view — an oral option addresses the adherence population that injectables reach poorly.

The Ripple:Gilead’s 3.14% regular-session gain, alongside Amgen’s 3.03% and Vertex’s 1.75%, meant the large-cap biotech complex materially outperformed a Healthcare sector that finished at +0.03%. The divergence within healthcare is now well established: biotech with launch-stage products is being rewarded, while large-cap pharma facing patent cliffs is not.

What It Means:Gilead has converted a single approval into a billion-dollar franchise inside a year and has a line of sight to expanding it into oral dosing. The modest after-hours easing after a 3.14% regular-session gain reads as profit-taking rather than a verdict on the print.

What to watch:Yeztugo sequential growth next quarter against today’s 40% — sustaining even half that rate would put the franchise well above the raised $1 billion guidance. Watch the February 2, 2027 PDUFA date for the once-weekly oral formulation.

23. Booking Holdings (BKNG): +5% AH | Gross Bookings Beat a Deliberately Conservative Guide Set for Middle East Disruption

The Numbers:Released: AMC. Revenue of $7.35 billion against $7.19 billion consensus, a 2.23% beat. Adjusted EPS of $2.54 against $2.43 expected, a 4.44% beat. Gross bookings exceeded consensus, which analysts had modelled at approximately $49.42 billion against $46.70 billion in the year-ago quarter. Management had guided Q2 conservatively — room nights up 2-4% and gross bookings, revenue and adjusted EBITDA each up 4-6% — on the explicit assumption that the Middle East conflict’s impact would be more pronounced in the second quarter, following a first quarter in which gross bookings rose 15% to $53.8 billion on 338 million room nights. Shares rose more than 5% after hours. Market cap $150.54 billion.

The Problem/Win:The win is that the conflict discount management built into the guide did not materialise at the scale assumed. Booking told the market in April to expect the Middle East to bite hardest in Q2 and set the bar accordingly; beating it on gross bookings means either the disruption was narrower than feared or travel demand elsewhere absorbed it. Either reading is constructive, and the second is the more valuable one — it would indicate that global travel demand is resilient enough to route around a closed regional corridor.

The Ripple:The result lands directly on the day’s dominant macro theme. Booking is the largest listed proxy for discretionary international travel, and its beat arrived on the session Brent fell 6.08% on Hormuz reopening reporting and the Dow Jones Transportation Average gained 2.58%. Cheaper jet fuel and a beat on travel demand from the same session reinforce each other, and both point at the same forward variable — whether the strait actually reopens.

What It Means:Discretionary travel demand held up against a guide built for disruption, which is a better consumer signal than anything McDonald’s or Nike delivered today. The gap between the low-income consumer read and the international-travel read continues to widen.

What to watch:Q3 guidance on room nights, which will reveal whether management still assumes a Middle East drag now that a Hormuz reopening is being discussed as imminent. Watch whether the beat came from room-night volume or from average booking value, since only the former indicates genuine demand strength.

Q2 2026 earnings season is roughly 61% complete, and Wednesday August 5 carries a dense large-cap slate spanning obesity, media, storage and healthcare services.

Eli Lilly (LLY) — BMO, Wednesday August 5 — consensus $6.01 EPS on $20.69 billion of revenue. The largest report of the week by market value at roughly $1.05 trillion, with options implying a post-earnings move of about 7.4%. Key focus: GLP-1 and obesity franchise demand, tirzepatide volumes, and full-year guidance — sharpened by Pfizer’s berobenatide Phase 2b data and Amgen’s MariTide Phase 3 progress, both of which surfaced today as monthly-dosing challengers to Lilly’s weekly incumbency.

Walt Disney (DIS) — BMO, Wednesday August 5 — consensus $1.86 EPS on $25.39 billion of revenue. Key focus: streaming profitability and subscriber economics, theme park attendance and per-capita spend, and the content pipeline. Parks are the cleanest discretionary-consumer read in the report, and it lands against a Consumer Cyclical sector that closed flat at 0.00% today.

Shopify (SHOP) — BMO, Wednesday August 5 — consensus $0.40 EPS on $3.45 billion of revenue. Key focus: gross merchandise volume growth, merchant additions, and monetisation of AI-powered commerce tools. The most direct listed read on small and mid-sized merchant health, which is where the Section 301 tariff costs land hardest.

Uber Technologies (UBER) — BMO, Wednesday August 5 — consensus $0.80 EPS on $14.24 billion of revenue, against $0.63 on $12.65 billion a year ago. Key focus: ride-hailing demand, delivery growth and the trajectory of segment profitability. Falling crude is a second-order tailwind through driver economics rather than a direct input.

CVS Health (CVS) — BMO, Wednesday August 5 — consensus $1.85 EPS on $100.03 billion of revenue. Key focus: healthcare services performance, insurance segment medical loss ratios, and pharmacy margins. Reports into continued managed-care pressure that pushed UnitedHealth down 1.88% today as one of the session’s largest mega-cap decliners.

Sandisk (SNDK) — AMC, Wednesday August 5 — consensus $34.80 EPS on $8.44 billion of revenue. Closed today at +10.84%, one of the session’s five largest mega-cap gainers. Key focus: NAND flash pricing, consumer storage demand and profitability. The most direct test of whether memory price inflation is reaching supplier margins or being absorbed by capacity spending — the precise question AMD’s capital-expenditure line failed tonight.

Western Digital (WDC) — AMC, Wednesday August 5 — consensus $3.31 EPS on $3.70 billion of revenue. Key focus: AI-driven storage demand, enterprise SSD shipments and margins. Reports the same evening as Sandisk, giving two independent reads on the storage pricing cycle within hours.

AppLovin (APP) — AMC, Wednesday August 5 — consensus $3.76 EPS on $1.94 billion of revenue, with revenue expected to grow roughly 54.6% year-over-year. Key focus: the e-commerce advertising ramp beyond mobile gaming, following a 12%-plus single-day decline on July 13 when industry data suggested e-commerce ad growth was slowing. Q1 delivered $1.84 billion of revenue, $1.21 billion of net income and an 85% EBITDA margin before the e-commerce expansion contributed materially.

Reporters for Thursday August 6 and Friday August 7 were not visible in the current earnings calendar view and will be carried in tomorrow’s report.

1. If a Hormuz agreement is signed this week, does it specify whose territorial waters inbound vessels transit — and if it does not, how much of an 11% two-session decline in Brent survives the first routing dispute?

2. Does AMD’s after-hours decline on a clean beat transmit to Intel, Micron and the rest of a complex that rallied double digits today without reporting — or was the capex line an AMD-specific problem rather than the industry’s?

3. Two sessions of US data have moved yields not at all. When Friday’s payrolls arrive, does the curve finally trade domestic macro — and which way does the accumulated backlog break, given Monday’s 6.2% growth nowcast and today’s softening labour demand point in opposite directions?

Manufacturing employment crossed 50 in July, after 33 months below it — the only one of these four series that changed state rather than merely extended a move. Orders were never the constraint: New Orders added 0.7. What broke loose was the queue behind it. Backlog jumped 4.5 points to 55.0 while customers’ inventories, “too low” for 22 straight months, left nothing to ship from stock — every unit demanded had to be built. Production answered, and only then did hiring follow, because manufacturers do not hire on orders; orders can be met with overtime, with a shift extension, with drawing the queue down. They hire when the backlog outlasts what existing headcount can absorb, and 60% of panelists now say their companies are hiring. But the same congestion that created the job is booked twice. Supplier Deliveries, 58.9 and slowing for an eighth month, enters the composite as a positive — strip it and the other four sub-indices average 54.8 — then lands again in the price index, where raw materials have risen 22 consecutive months. That is cost-push meeting a labor market that just started absorbing again, in the most rate-sensitive, inventory-financed sector in the economy, in front of a Fed that held 9–3 with all three dissents for hikes and September priced near 57%. The queue that finally created the job is the same queue that argues for the hike that ends it.