Market Pulse · Daily Scan
Daily Market Pulse
Jul 24, 2026 · Post-market · Claude · sentiment/flow window ≈ 48h · Backdated run. Track A reflects the Fri 2026-07-24 regular-session close and the 48h window Wed 07-22 -> Fri 07-24, which captures the second-order reaction to Alphabet/Tesla's 07-22 AMC earnings (INTC's own 07-23 AMC print reacting on 07-24; continued memory-sector unwind; a dense Friday BMO earnings batch) plus the close of a genuinely volatile week. Track B carries forward the 20 names established on 2026-07-22 and updated once on 2026-07-23 (GOOGL, TSLA); no further Track B changes were warranted for this run.
TLDR;
The week of July 20-24, 2026 closed as a narrow, tech-led pullback inside a still-resilient broader tape: the S&P 500 and Dow essentially round-tripped by Friday (weekly S&P -0.6%, Dow -0.4%) while the Nasdaq took the brunt of the damage (weekly -2.1%, -0.64% on Friday alone) after Alphabet and Tesla's earnings reignited an AI-capex-versus-profitability debate that erased roughly $797-800B in Magnificent-Seven value in a single Thursday session (07-23). Intel's post-earnings -6.5% to -7.89% drop despite a beat (revenue +25% YoY, fastest growth in ~15 years) reinforced that this was a valuation/positioning digestion in megacap tech and semis (semis -4.25% on the week, SanDisk -10.79%-11% alongside peers) rather than a fundamentals shock. Iran-related oil-supply fears pushed WTI to seven-week highs mid-week and drove the 10-year Treasury yield to 18-month highs (~4.68-4.69%) ahead of the July 28-29 FOMC, before crude retreated sharply into Friday's close (reportedly aided by mediation efforts, single-sourced). Elevated but not panicked volatility (VIX in the high-18s, easing from a 07-23 spike into the 18.7-19.2 range) and Bank of America's Bull & Bear Indicator sitting at its highest (most complacent) level since 2021 together suggest a market rotating out of richly-valued AI/megacap tech into defensives (REITs, financials, staples) rather than de-risking broadly. Breadth: Narrow and rotational for 07-24 itself - megacap tech/semis lagged sharply (Nasdaq -0.64%, semis -4.4% intraday) while cyclical/defensive sectors (REITs, financials, staples) and small-/mid-cap breadth (Dow +0.46%, Russell 2000's milder -0.35%) held up comparatively better. For the week overall, Monday-Thursday were the negative-breadth days; Friday itself was closer to flat/rotational than outright risk-off..A · Daily Pulse
What moved — and why
▲ Rising — dailycomposite
| # | Ticker | Move | Driver | Score |
|---|---|---|---|---|
| 1 | LMT | +10.54% | $35B THAAD missile-defense award lifted backlog to a record $230.4B alongside a beat-and-raise. Carried in this window from the 07-23 session; still the strongest single Tier-1 catalyst of the trailing three days. | 63 |
| 2 | SLB | +11.01% | Q2 adjusted EPS $0.55 vs $0.51 consensus with North America revenue +36% YoY. WORTH FLAGGING EXPLICITLY: SLB sits on the Track B ONE-YEAR DOWNSIDE list (carried forward this run) on a thesis that it 'loses in both oil regimes' because its book is international/offshore-weighted. This +11% North-America-driven beat does not resolve that debate -- if anything the strength came from the domestic book, not the Middle East exposure the bear case targets -- but it is a live data point worth weighing against the 12-month short. | 61 |
| 3 | DHR | +7.46% | Rebound after the -10.99% post-earnings drop on 07-21 (outside this window): raised guidance and bioprocessing-demand-normalization commentary drove a two-stage recovery. A genuine round-trip name across the two runs -- flagged so the net picture is clear rather than treating this as a fresh, independent catalyst. | 54 |
| 4 | VZ | +5.84% | Beat on wireless service revenue and broadband subscribers, with fixed-wireless-access acceleration -- also read partly as a defensive/dividend bid amid the broader tech selloff this week. | 52 |
| 5 | UNP | +4.02% | Q2 beat plus continued Norfolk Southern merger-progress momentum; a second consecutive session where rail names confirmed freight-volume recovery. | 50 |
| 6 | GEV | +4.69% | Rebound after the -8.69% post-earnings drop on 07-22 (Q2 EPS miss despite a revenue beat and raised FY guidance). NET across the 3-day window is still roughly -4.4% (-8.69% then +4.69% does not fully retrace), so this name's inclusion on the rising list reflects a single strong session, not a full recovery. Same order-book-vs-multiple tension flagged in the 07-22 Track A/B research. | 45 |
| 7 | T | +2.35% | Q2 beat plus an announced ~$10B buyback expansion -- a capital-return catalyst layered on the earnings beat, the second straight session of telecom strength alongside VZ. | 43 |
| 8 | HCA | +1.67% | Q2 EPS/revenue beat despite payer-mix cost concerns. One secondary source headlined a larger +4.2% move; the more internally-consistent close-to-close figure (+1.67%) is used here and the discrepancy is flagged rather than silently picking the bigger number. | 39 |
| 9 | AZN | +1.3% | INCLUDED WITH AN EXPLICIT CAVEAT rather than silently dropped: this is a marginal mover added only because the window's rising list otherwise stops at 8 genuinely defensible large-cap names (see methodology and the guardrail note below). AZN itself does not report until 07-27, inside the forward catalyst calendar; Track B carries an unchanged upside thesis for it. | 32 |
▼ Falling — dailycomposite
| # | Ticker | Move | Driver | Score |
|---|---|---|---|---|
| 1 | TSLA | −14.5% | Carried forward from the 07-23 session into this window; still the largest and highest-conviction single-name move of the trailing three days. See the 07-23 run for full detail. | 90 |
| 2 | INTC | −7.89% | Revenue +25% YoY to $16.1B was the fastest growth in roughly 15 years and easily beat consensus -- but capex guidance raised from ~$18B to >$20B (with 2027 guided 'significantly above' that) triggered the same capex-as-cost repricing that hit GOOGL/TSLA. The stock's own post-earnings after-hours pop to ~$110 reversing into a -7.89% regular-session close is the cleanest single-name illustration this week of 'a beat is not enough if the capex number is scarier than the revenue number.' | 76 |
| 3 | GOOGL | −7.13% | Carried forward from 07-23; see that run for full detail. The capex guide ($195-205B) is the trigger; Track B (upside, unchanged rank 2) treats the drawdown as the entry, not the exit. | 72 |
| 4 | SNDK | −10.79% | Broad memory-stock selloff on rising Chinese competition (CXMT, YMTC) cooling the NAND-pricing rally that had driven the SAME NAME up ~14% just three days earlier (07-21). This is the mirror image of the 07-22 run's rising-list entry for SNDK -- a name flagged there as PUMP_RISK (medium) on a vertical move with no company-specific catalyst is now giving essentially all of it back. The round-trip itself is the evidence that the earlier move was fragile. pump risk | 67 |
| 5 | MU | −6.99% | Same memory-sector selloff as SNDK on rising Chinese competitive pressure -- MU rode the identical 07-21 rally (+12.17%, this run's PRIOR Track A #3) and is now giving a meaningful chunk of it back. This is the round-trip the 07-22 run's methodology section explicitly flagged as a forward risk to calibrate against ('the signal's half-life in this regime was short') -- it played out within three trading days. Citi reportedly framed it as a dip-buying opportunity (Tier-2 quoted), a direct rebuttal worth weighing against the late-cycle Track B read. ↔ deep-dive | 60 |
| 6 | AMZN | −4.57% | Pure spillover from the AI-capex-fear narrative -- a ~$59B capex figure was cited in coverage raising cash-flow concerns, but AMZN itself does not report until 07-30 (inside this run's forward catalyst window). No AMZN-specific news; this is sympathy selling ahead of its own print. | 52 |
| 7 | AXP | −4.3% | Beat EPS ($4.53 vs $4.40) but expenses rose 12% YoY and the tax rate jumped to 24% from 19%; the same-day announced TheFork acquisition was read by the market as reinvestment over payout. A secondary source headlined a larger -6% move; the close-to-close figure (-4.30%) is used here with the discrepancy flagged. | 48 |
| 8 | ORCL | −4.21% | Continuing slide on AI-datacenter-capex-ROI skepticism (down ~9.6% cumulative over the 4-day window) despite a $7B Pentagon contract win that only briefly lifted shares -- the same capex-fear narrative hitting a fourth major AI-infrastructure name this week (with GOOGL, TSLA, INTC). | 48 |
| 9 | TXN | −3.13% | Fell despite an upbeat forward forecast -- a 'sell the news' move. Notable given TXN is explicitly named in the 07-22 Track B research as one of three analog/auto names (with ADI, NXPI) whose 'structural loser' thesis had broken down on rising estimates; this week's price action is a modest near-term counterpoint to that 12-month read, though the magnitude here is small. | 43 |
| 10 | LUV | −3.84% | Fuel-expense pressure drove a reduced full-year adjusted EPS outlook. This -3.84% figure (from a source specific to this window) differs from a -6.19% figure carried in the 07-23 run for the same date and ticker; both are reported across the two artifacts and the conflict is flagged rather than silently resolved. | 41 |
B · One-Year Conviction
Asymmetric 12-month skew
Free sources only. Valuation multiples and market caps are Tier-2 aggregator data (stockanalysis.com, sourcing S&P Global Market Intelligence / TipRanks) - NOT SEC-verified - and most quotes carry a 2026-07-24 to 2026-07-26 vintage, i.e. AFTER the nominal run date. Post-run-date PRICE REACTIONS were excluded from scoring. FY2027 consensus is paywalled for many tickers; where it was unavailable, 'estimate revisions' is inferred from price-target revision direction and current-year growth rather than from an observed estimate series, and the subscore is capped accordingly. True estimate-revision history (dollar revisions over time) is not available on any free source. TRACK B STABILITY NOTE: per the project's own instructions, Track B is a slow-moving signal and should not churn day to day; large day-to-day churn is itself a red flag. This run carries forward the same 20 names, ranks and sub-scores established in the 2026-07-22 run and updated once on 2026-07-23 (GOOGL, TSLA) -- no further changes were warranted for 2026-07-24, since neither name had a fresh thesis-relevant catalyst in the 07-23-to-07-24 session. This is the intended, correct behavior, not a shortcut.
▲ Upside skew — 1yrconviction
| Ticker | Thesis | Key risk | Score |
|---|---|---|---|
| TSM | The unavoidable chokepoint of the AI buildout and the cheapest way to own the cycle. Every other semiconductor name in this research - NVDA, AVGO, AMD, MRVL, CRDO, ALAB - is a TSMC customer, so it monetises the compute buildout regardless of whether merchant GPUs or custom ASICs win the architecture war. In a regime that has begun punishing capex spenders and de-rating high-multiple beneficiaries, TSM offers ~31% forward-year revenue growth at 19.05x forward earnings with $79.70B of net cash. Explicit forward-year consensus was obtainable here, which is rare in this dataset: 2026 revenue NT$5.38T (+41.4%), EPS NT$106.35 (+60.5%), 37 analysts; 2027 revenue NT$7.05T (+30.9%), EPS NT$136.05 (+27.9%). July PT raises: Barclays $625->$650 (07-16), TD Cowen $400->$440 (07-17). The persistent discount is geopolitical, not operational. | Taiwan geopolitical tail risk - unhedgeable, and the sole reason for the multiple discount; TSM bears the capex burden of the entire cycle (leading-edge plus CoWoS packaging), so a hyperscaler digestion phase hits its own capital intensity; reported 'TSMC stake sale' among June 2026 supply-chain worries - UNVERIFIED | 90 |
| GOOGL | The only megacap that is simultaneously the AI model layer, the distribution layer and the cloud layer, and the only one in this set with ACCELERATING revenue growth. The Q2 print released AMC on the run date itself delivered revenue $119.8B (+24%) with Cloud +82%; the prior quarter delivered EPS $5.11 against $2.63 consensus with Cloud +63% on a $460B backlog. 2026E revenue $489.12B (+21.4%), accelerating from +15.1% in 2025; 2026E EPS $14.26 (+31.9%). Of 64 covering analysts, ZERO rate it Sell. The capex raise disclosed on the run date is precisely the regime change described above - the market repricing capex as cost - and that repricing is what creates the entry rather than invalidating the thesis. UPDATE (2026-07-23, unchanged through 07-24): the capex-driven drawdown described above as 'what creates the entry' happened on 2026-07-23 -- GOOGL fell 6.9%-7.13% to ~$317.69 despite the Q2 beat, precisely on the raised capex guide. Retail sentiment trackers notably stayed net-bullish through the drop, and options positioning stayed net call-leaning post-print (P/C 0.48) rather than showing capitulation. No further GOOGL-specific news moved the thesis between 07-23 and 07-24; the price and this write-up carry forward unchanged into this run. | FY26 capex of $195-205B plus $811B of contracted future commitments materially compresses FCF - and the market has just demonstrated it will punish that; antitrust remedy overhang unresolved; search-query cannibalisation by AI assistants is a genuine long-term unknown that no current number resolves | 86 |
| NVDA | Carries the lowest forward multiple of any high-growth name in the AI complex (20.76x) while generating the highest free cash flow ($119.08B), having already de-rated relative to its growth while froth migrated into memory and optics. Its 52-week range ($164.07-$236.54) is the NARROWEST in the complex - roughly 44% peak-to-trough versus SNDK's reported +3,240% - meaning it is the least-crowded, most-owned-on-fundamentals name in a group where positioning is the primary risk. FY2027 consensus: revenue $393.6B, EPS $8.99, 53 analysts as of 07-22. The ~$500B NVIDIA-SK Group AI initiative locks down HBM supply - NVDA using its balance sheet to secure the one input that could constrain it. | ASIC substitution at the margin (Broadcom/Google TPU, Trainium) - the live architectural debate; hyperscaler concentration: any one of the big five cutting capex is the direct offset, and MSFT (07-29), AMZN and META are the swing events; China export restrictions remain an unquantified overhang | 84 |
| VST | The AI-power theme is fundamentally intact and broken on multiple, and that gap is the trade. Order evidence is stronger than at any point in the cycle (GE Vernova orders +88% YoY on a $176B backlog; Eaton Electrical backlog +29% with 1.1 book-to-bill; Quanta $48.5B backlog) while merchant IPPs sit 20-33% below their 52-week highs. The de-rating driver is the DISCOUNT RATE, not demand - and Vistra is where that dislocation is widest at 17.44x forward, 10.79x EV/EBITDA and a PEG of 0.38. All six Jun-Jul analyst actions maintained Buy; Wells Fargo moved its target $152 -> $259 on 07-03. FY2026E revenue $23.28B (+31.23%), EPS $8.45 (+77.55%). ERCOT + PJM merchant nuclear/gas fleet with direct hyperscaler exposure, plus the June 2026 KKR/Nvidia/Vistra 'Helix Digital Infrastructure' $10B venture. Backdrop: Morgan Stanley models US data-centre power demand at 74 GW by 2028 against ~25 GW available today. | D/E 3.55 and $19.93B of debt into a probable September hike - the most rate-exposed name on this list; merchant power spreads are unhedged commodity exposure, so a mild weather year is an earnings miss; the source statistics page is cached 2026-05-29 - multiples are approximate | 83 |
| MSFT | A SELLER of the agentic tools disrupting the rest of software, trading at the cheapest forward multiple in years (20.61x) and ~31% below its high with estimates still accelerating. Azure + Copilot + the OpenAI stake is the widest moat in software, and unlike app-layer SaaS it sits on the correct side of the seat-pricing question. FY26E revenue $329.55B (+17.0%), accelerating from +14.9%; FY26E EPS $16.77 (+23.0%). 42 Strong Buy / 13 Buy / 3 Hold of 58. ↔ deep-dive | the 07-29 capex guide is the sector's key swing factor - MSFT is NOT exempt from the capex-as-cost repricing that hit Alphabet, and it lands INSIDE the catalyst horizon of this run; sources conflict materially on MSFT's own 2026 capex - $120B+ (Futurum, Forbes) vs ~$190B (Tier-3 aggregator) - UNVERIFIED; Copilot seat monetisation is the same seat model being questioned elsewhere | 82 |
| AZN | The only large-cap pharma compounding EPS at a double-digit rate THROUGH its own IRA repricing, without needing an obesity story or a patent cliff to defend. Farxiga was in IRA Round 1 with negotiated prices effective 2026-01-01 - and AZN is still guiding to ~12% EPS growth in both 2026 and 2027. Verified forward years: 2026 revenue $63.21B (+7.62%), EPS $10.25 (+11.94%); 2027 revenue $66.79B (+5.66%), EPS $11.51 (+12.25%). JPMorgan moved $165 -> $213 on 2026-07-21 - a 29% target increase one day before the run date. Consensus PT implies the widest upside of any large-cap pharma in this research. | drug-pricing policy is now a REALISED earnings event rather than an overhang - which cuts both ways; extreme target dispersion: Deutsche $153 vs Jefferies $233, a 52% bull/bear gap; HSBC downgraded at $183 (07-13) and Morgan Stanley cut $220->$204 (07-16); ADR structure adds FX translation risk | 78 |
| MRK | Merck is pre-empting its own 2028 Keytruda cliff rather than waiting for it - systematically approving KEYTRUDA QLEX (subcutaneous) in lockstep with every new IV indication (TNBC 1L with Trodelvy 06-25; muscle-invasive bladder cancer with Padcev 07-10). Converting a franchise to a patent-protected subcutaneous formulation ahead of IV biosimilars is the entire LOE mitigation thesis, and it is executing on schedule. It separately opened a new market on 2026-07-16 when the FDA approved LIPFENDRA (enlicitide), the first and only once-daily ORAL PCSK9 inhibitor. READ THE VALUATION CAREFULLY: 2026 EPS consensus is $2.74 (-69.43%), a one-off charge year, versus 2027 EPS $9.66 (+251.83%) - roughly 13.6x 2027 earnings. Any screen ranking MRK on 2026 P/E produces a false signal. | +56% over 52 weeks against a consensus PT of $134.04 (+2.27%) - the re-rating has largely happened; the QLEX conversion rate is NOT disclosed - it is the single most important number for this thesis and it is UNVERIFIED; $43.4B net debt; MK-1167 Alzheimer's Phase 2 terminated ~07-19; source data conflict: quote page shows PT $131.70 vs forecast page $134.04, unreconciled | 78 |
| LLY | The single most important structural fact in mid-2026 healthcare is that US GLP-1 NET PRICING has collapsed by government design while VOLUME has been unlocked - a volume-for-price trade. TrumpRx (Tier 1, US government) now sells Zepbound at $299/mo from a $1,087 list, and Medicare covers obesity drugs for the first time at a ~$50/mo copay. That asymmetry rewards whoever takes enough share to more than offset price. Lilly is the only company owning BOTH the injectable incumbent (Zepbound/Mounjaro) AND the oral disruptor (Foundayo/orforglipron, FDA-approved 2026-04-01), with retatrutide behind it - and the TRIUMPH Phase 3 read out on 2026-07-23 at up to -20.8% body weight. MFN price compression transfers share TO Lilly rather than away from it. 2026 revenue $85.35B (+30.95%), EPS $34.62 (+43.01%); 2027 revenue $98.74B (+15.69%), EPS $44.87 (+29.59%). July PT raises: Citi $1,500->$1,600, Bernstein $1,300->$1,385, UBS $1,250->$1,425. | at 33.5x forward with a consensus PT implying only ~+6%, the sell side sees the story as fully priced - the weakest valuation support of any name on this upside list; MFN/TrumpRx is structurally deflationary even for the winner; Novo sued Lilly on 2026-07-21 and sought a preliminary injunction on 07-24 to block Lilly's DTC advertising - a live legal risk to its key launch channel | 78 |
| AEM | Essentially unlevered equity on the gold price, with the best margin structure and balance sheet in this research. Gold at ~$4,124 on the run date (+21.4% YoY) against a largely fixed cost base is direct margin expansion - and unlike oil, gold's drivers here (geopolitical instability PLUS an inflationary oil shock) are reinforced rather than undermined by the Iran situation. D/E of 0.01 means shareholders capture the full operating leverage with essentially no financial risk: 39.46% profit margin, 7.31x EV/EBITDA, 11.62x forward earnings, $4.54B TTM FCF, total debt $319M against $3.12B cash. Hold ratings fell 6 -> 4 over six months. | a durable US-Iran settlement removes the geopolitical bid - reports of Pakistan/China mediation already knocked crude ~3% on 07-24; a September rate hike raises real yields, gold's primary headwind; FY27 EPS growth decelerates sharply to +5.67% - the easy comps end, and the analyst PT range ($87-$300) is extraordinarily wide, signalling low conviction on the gold deck | 76 |
| NOC | The defense cycle is intact and the trade has inverted: money moved out of narrative (PLTR -20.5%, TDG -22.6%, AXON -28.8%, LDOS -31.6%, AVAV -45.2% over 52 weeks) and into appropriated backlog (HWM +55.2%, LMT +38.9%, RTX +36.0%, GD +22.0%). NOC is the ONLY large prime down over 52 weeks (-4.62%) - and it beat and raised on 07-21, one day before the run date: EPS $7.68 vs $6.82, FY26 EPS guidance raised to $28.60-29.10 from $27.40-27.90, a record $105B backlog and $20B of net awards (~1.9x quarterly book-to-bill). The FY2027 budget request (released 2026-04-03) is $1.5T total defense, +24% real discretionary, weighted to hardware, with ~$55B for a Defense Autonomous Warfare Group where NOC has the deepest franchise among primes. EV/EBITDA has compressed from 21.38 in FY2023 to 14.92. | THE UNRESOLVED QUESTION: six firms cut price targets IMMEDIATELY AFTER the beat-and-raise (Susquehanna $785->$655, Baird $700->$500, DB $645->$615, JPM $635->$600, TD Cowen $580->$550), and only UBS raised. B-21 initial LRIP lots are fixed-price - historically NOC's charge mechanism - and current charge status is UNVERIFIED. This is the single most important unresolved item in the defense work and the reason the estimate-revisions subscore is capped at 68.; Sentinel ICBM cost/schedule breach risk; the FY27 budget is a REQUEST, not law - discretionary spending needs 60 Senate votes | 74 |
▼ Downside skew — 1yrconviction
| Ticker | Thesis | Key risk | Score |
|---|---|---|---|
| ARM | Grows revenue ~21-23% and trades at ~120x forward earnings - a PEG above 5 - with EV/EBITDA at 238.99. The structural problem the AI cycle does not fix: royalty revenue scales with UNITS, not with silicon value, and units are falling (smartphones -12.9%, PCs -11.3%, per IDC). This is the clearest negative revision cluster in the semiconductor complex: HSBC downgraded to Hold from Buy on 2026-07-14 citing foundry capacity constraints limiting earnings growth; UBS cut $470->$360 on 07-20 (-23%); Wells Fargo cut $410->$350 on 07-22, the run date itself (-15%). Only Susquehanna raised. FY2027 consensus: revenue $5.97B (+21.3%), EPS $2.17 (+22.5%), 37 analysts. | datacentre CPU attach (Grace/Graviton/Axion) could inflect royalty RATES rather than units; any AI-licensing announcement re-rates it instantly; SoftBank ownership concentration makes the float squeezable - this is the real hazard in shorting it; earnings 2026-07-29, INSIDE the catalyst horizon | 81 |
| NVO | Triple structural price compression on the same molecule, all three confirmed. (a) TrumpRx MFN pricing: Ozempic $199/mo from $1,028, Wegovy Pen $199 from $1,349 (Tier 1, trumprx.gov). (b) IRA Round 2: Ozempic AND Wegovy negotiated at -44% net, effective 2027-01-01 - i.e. INSIDE the 12-month window. (c) Competitive share loss: Lilly's orforglipron launched 2026-04-01 and retatrutide read out 2026-07-23. Novo eats the price cut without the offsetting share. The tell is mechanical: forward P/E (16.00) is HIGHER than trailing (11.48) - the market is explicitly modelling earnings decline. 2026 consensus revenue DKK 293.81B (-4.94%), EPS 20.97 (-8.93%), 28 analysts, and the consensus PT sits BELOW the market price. Guidance was cut 2026-02-04 citing pricing pressure, patent expirations and competition; CagriSema missed and requires a new study. Behavioural tell: Novo sued Lilly on 2026-07-21 and sought a preliminary injunction on 07-24 to block a competitor's advertising - not the posture of a company winning on product. | semaglutide has real volume upside from Medicare's first-ever obesity coverage at a ~$50/mo copay; 9.1x EV/EBITDA on a 37.2% net margin is genuinely cheap - this is a crowded, already-repriced short; HSBC raised its target to DKK 300 on 07-21; Q2 earnings 2026-08-05, same day as LLY | 79 |
| TSLA | Volumes are recovering while economics collapse. The Q2 print released AMC on the run date itself: revenue $28.236B BEAT consensus $26.4B and deliveries grew 25% YoY to 480,126 - but non-GAAP EPS was $0.33 against ~$0.51 consensus, operating margin 1.4%, operating income -57% to $398M, and free cash flow -$1.092B with capex +142% to $5.8B. Tesla has converted from an auto manufacturer earning 15%+ operating margins into a capex-heavy AI/robotics R&D vehicle funded by a barely-profitable car business. At ~162x forward earnings with negative FCF, essentially the entire market cap is the robotaxi/Optimus option - and management offered limited new concrete milestones on either on the call. UPDATE (2026-07-23, unchanged through 07-24): the thesis materialized almost exactly as described -- Q2 non-GAAP EPS $0.33 missed the ~$0.49-0.51 consensus by roughly 35-38%, and the stock fell 14.4%-14.5% to ~$319.69, a realized move roughly 2.4x-2.5x what options had priced pre-print (~6%). UBS cut its price target to $385 on the print; at least five other analysts cut targets the same day. Because so much of the bear case is now realized in the price, the fundamental-valuation subscore is trimmed slightly (as of the 07-23 update) to reflect a less attractive NEW entry for a short, even though the structural critique is unchanged. No further TSLA-specific news moved the thesis between 07-23 and 07-24. | a credible, dated, revenue-generating robotaxi rollout at scale - the option is real, not fictional; deliveries are GROWING 25%, not shrinking: the post-credit-expiry demand cliff bears expected did not materialise; net cash of ~$27B means no financing risk and no forced-deleveraging catalyst - this can stay expensive indefinitely; data conflict: Q2 deliveries reported as 480,126 (Electrek) vs 489,126 (Teslarati), unreconciled | 76 |
| QCOM | The ONLY large cap in the entire semiconductor universe screened with declining consensus revenue AND declining EPS. FY2026 consensus: revenue $42.58B (-3.86% YoY), EPS $10.81 (-10.14% YoY), 31 analysts. Three independent structural pressures compound at once: losing its largest socket to Apple modem in-sourcing, paying more for inputs (memory crunch - it signed long-term automotive memory supply agreements with Micron ~07-17, which is sensible AND confirms where the pricing power sits), and selling into a shrinking unit market (smartphone units -12.9%, IDC). UBS cut $235->$190 around 07-21. INFERENCE, FLAGGED AS MINE: a 16.94 forward P/E at $166.97 implies forward-year EPS of ~$9.86 - BELOW the FY2026 estimate of $10.81 - meaning consensus appears to model two consecutive down years. This is my arithmetic from the aggregator's own figures; the FY2027 consensus itself was paywalled. | 16.9x with $12.5B of FCF and ~32% consensus upside is not an expensive short; auto/IoT diversification is real and growing; already ~36% below its 52-week high, so much of the modem cliff is priced; FQ3 earnings 2026-07-29, INSIDE the catalyst horizon | 76 |
| MRNA | The cleanest negative asymmetry in this research on the numbers. The stock is +87.5% over 52 weeks while revenue FELL 30% to $2.23B TTM, gross margin is NEGATIVE (-69.9%), operating margin is -113.9%, TTM FCF is -$1.60B and EPS is -$8.16. Consensus PT of $45.85 sits ~44% BELOW spot, rated Hold by 24 analysts. Equity eroded from $13.85B to $7.41B while debt went from ~zero to $1.25B. Management reorganised into three franchises explicitly preparing for launches in 2027 and 2028 - the company itself does not expect 2026 to be the revenue trough. RUNWAY (my arithmetic from two sourced figures, flagged as inference): net cash ~$3.96B against -$1.60B annual FCF implies roughly 2-2.5 years; company runway guidance is UNVERIFIED. | A NEAR-TERM BINARY SITS DIRECTLY ON TOP OF THE RUN-UP: the mRNA-1010 flu vaccine PDUFA is 2026-08-05 and the adcomm voted 9-0 in favour on 2026-06-18. An approval is a real, dated, high-probability-looking catalyst against this short.; Merck-partnered intismeran autogene cut melanoma recurrence/death risk 49% at 5 years (2026-06-01); Q2 earnings 2026-07-31 | 76 |
| DE | A peak-quality multiple on trough-and-still-falling fundamentals: 31.65x forward earnings on a third consecutive year of revenue decline. Revenue $61.25B (FY23) -> $51.72B (FY24) -> $45.68B (FY25); net income $10.17B -> $7.10B -> $5.03B; operating margin 25.26% -> 24.27% -> 20.64%. That is -25% revenue and -51% net income since FY2023. FY26E revenue $41.5B (-9.15%), FY26E EPS $18.06 (-2.36%). THE STRUCTURAL POINT IS TIER-1 AND UNDERAPPRECIATED: USDA ERS forecasts 2026 net farm income of $153.4B (-2.6% inflation-adjusted), and the support is TRANSFER PAYMENTS, not demand - government payments +45.2% to $44.3B, 'primarily due to higher anticipated payments from Farm Bill programs that trigger payments when commodity prices fall', while crop cash receipts rise only +1.2% (a decline in real terms). Farmers receiving deficiency payments do not buy $800,000 combines; equipment capex tracks crop receipts, not total farm income. Quality-of-earnings flag: Q2 FY26 margins were partly boosted by a $272M one-time tariff refund. Citi raised its PT to $610 on 07-13 but kept Neutral - and $610 is BELOW the market price. | the trough-buying argument is live and has mainstream backing (Barron's, 2026-07-09: 'Farming Is Terrible Right Now. That's Good for Deere and AGCO Stocks'); the 2026-06-12 White House ag-equipment tariff cut is a direct cost tailwind; FY2027/28 estimates are being revised UPWARD, implying the street already treats FY2026 as the trough; precision-ag software and the dealer network are genuine moats | 73 |
| NET | The single most unsupportable large-cap multiple found in this research: ~202x forward earnings for 31.6% revenue growth, unprofitable (EPS -0.25), with the stock near its 52-week high AND a consensus price target of $259.84 that sits BELOW the current price. It is the only name in the entire cross-desk dataset where the sell side has marked its target under spot while the stock sits near its high - complete capitulation on upside with no valuation cushion. The broader structural point: the market spent 2026 punishing app-layer SaaS on terminal-value fear while data/infra/security names saw market caps rise 31-99% on 20-31% growth at 80-200x forward with GAAP losses. That cohort - NET, CRWD (140x fwd), SNOW (127x fwd), DDOG (100x fwd), PANW (82.5x fwd on EPS -34.1%) - is where the crowding, and therefore the unpriced downside, actually sits. In a probable-hike regime these are the highest-duration assets in software. | Cloudflare's edge/inference positioning is genuinely strategic and could see an AI-workload inflection that resets the growth rate upward; shorting momentum at a 52-week high is how people lose money; no debt distress; this is a MULTIPLE short, not a fundamental one - revenue is compounding at 31.6% | 72 |
| SLB | The cleanest structural short in energy because it loses in BOTH states of the world. The central paradox of mid-2026: crude is up ~38% YoY while SLB's earnings estimates are FALLING. Its book is international/offshore-weighted and concentrated in exactly the Middle East geographies now disrupted by the Iran conflict and Hormuz throttling - war raises the oil price but halts the drilling SLB is paid for (Saudi output already fell 7,010 -> 6,637 kbpd May->June 2026). Then, if the conflict resolves, prices fall toward EIA's $65 Brent 2027 case and operators cut budgets. FY2026 EPS $2.53, revised DOWN 13.73% from $2.93. FY2027 consensus assumes a +30.15% EPS rebound to $3.29 - an aggressive recovery resting on the same de-escalation that E&P bulls fear. THE TWO CONSENSUS VIEWS ARE MUTUALLY INCONSISTENT, and that inconsistency is the thesis. Corroborated intraday by Track A: HAL fell 5.47% on a double beat on 07-21 specifically on Middle East revenue -10.7%. | a durable peace could unleash pent-up Middle East reconstruction spend; FY27 consensus already assumes recovery, so a beat re-rates fast; consensus is still Buy with ~17% PT upside - this is a contrarian position; data flag: shown EV/EBITDA of 4.52 is irreconcilable with a 25.42 trailing P/E - flagged UNVERIFIED and not used; FCF listed as n/a | 72 |
| INFY | The purest expression of AI disrupting the labour-arbitrage pyramid, and one of very few names in software/services with genuinely FALLING forward estimates. Trailing P/E 13.17 vs forward P/E 13.26 - forward ABOVE trailing, the mechanical signal of negative revisions. Revenue TTM $20.30B, +4.1%: the weakest growth of any name examined across five research desks. Consensus rating Hold, only 14 analysts, PT implying ~+6.7% - by far the weakest analyst support in the dataset. The mechanism: services revenue is billable hours x headcount, precisely what agentic AI compresses. The sector repriced violently on 2026-06-18 when Accenture cut guidance from 3-5% to 3-4% local-currency growth, missed on revenue and posted bookings -2%; ACN fell ~18%, its worst single-day drop on record, with Capgemini -8.4%, CTSH -5.8%, INFY -3.2%. | THERE IS A REAL HOLE IN THE CONSENSUS STORY: Accenture's HEADCOUNT ROSE during the quarter it missed, CTSH's bookings rose 21%, and IBM tripled entry-level hiring despite claiming developers are 40% more productive. Revenue deceleration is documented; the causal attribution to AI is NOT.; at ~13x with ADR/FX effects, much is priced; if AI INTEGRATION demand materialises, services firms capture it; related names ACN (10.25x fwd, 4.44% yield) and CTSH (7.86x fwd, FCF +31.6%) are explicitly NOT recommended as shorts - too cheap; express as underweight | 70 |
| HWM | SHORT THE MULTIPLE, NOT THE BUSINESS - and the low composite reflects exactly that. The most extreme multiple expansion found anywhere in this research: forward P/E 24.29 (FY21) -> 26.72 (FY23) -> 36.19 (FY24) -> 48.99 (FY25) -> 54.80 now; EV/EBITDA 16.69 -> 45.56, a 173% expansion in five years; P/S 2.70 -> 13.42. A reversion to even the FY2025 EV/EBITDA of 36.35x is roughly a 20% drawdown with NO change to earnings. Consensus PT implies only ~+7.75% even at a Strong Buy rating. Q2 guidance ($2.39-2.41B revenue, $1.22-1.24 EPS, reporting 08-06) is a narrow band with little beat room. In a probable-hike regime, 45x EV/EBITDA is the most rate-sensitive kind of asset there is. | THIS IS NOT A FUNDAMENTAL SHORT AND THE ESTIMATE-REVISIONS SUBSCORE OF 40 SAYS SO EXPLICITLY: estimates are RISING. FY26E revenue $9.76B (+18.24%), EPS $5.06 (+34.31%); the company RAISED FY26 EPS guidance to $4.88-5.00 from $4.35-4.55; Jefferies, TD Cowen, Morgan Stanley, Citi and UBS all raised targets; consensus is Strong Buy.; genuine oligopoly in investment-cast turbine airfoils; the hyperscaler/industrial-gas-turbine demand leg is a new, uncorrelated vector the market may still be underwriting; 737 rate increases are direct volume | 60 |
C · Options Heat
Where the premium is paying up
| Ticker | Signal | Read | Implication | Lag |
|---|---|---|---|---|
| INDEX_VOL | iv_spike | VIX closed 18.58 (-0.64%) on 07-24, down from the 18.7-19.2 range on 07-23; chip-sector vol gauge -4.4% intraday on 07-24 as semis dragged the Nasdaq for a second straight session. 10Y at 4.71% (per this pass's source; a separate source in the market_regime section shows 4.679-4.69%, both cited). | an orderly, second-stage unwind rather than a fresh panic -- vol eased from Thursday's spike even as the Nasdaq kept falling on 07-24, consistent with single-name (semis/megacap) dispersion rather than a market-wide re-rating | Cappnotes (Frank Cappelleri) is a single independent Tier-2 commentary source, not the primary CBOE feed. |
| FOMC_PRICING | pc_skew | CME FedWatch-style odds of a JULY FOMC hold slipped to ~64% from ~87% a week earlier (single-sourced, Cappnotes/CNBC-style commentary) | the market is genuinely repricing near-dated FOMC risk into 07-29, not merely drifting -- a material week-over-week shift in probability mass toward a hike, which cuts against several richly-valued names on both Track B lists | Single-sourced probability reading; not cross-verified against a second FedWatch-style aggregator. |
| VIX_TERM_STRUCTURE_07_22 | iv_spike | carried context from 07-22 (one session prior to this window's start): VIX1D 10.14 (-24.5%), VIX9D 15.48 (-12.9%) - a 5.34pt gap between 1-day and 9-day tenors, explicitly said by the source to span both the FOMC decision (07-29) and the PCE print. SKEW 151.66 (+3.8%) - elevated OTM downside-put pricing already building. | event-vol was being front-loaded for FOMC week even before this window began; the pattern is consistent with the elevated SKEW persisting through 07-23/07-24 | Saxo desk commentary (Tier 2), figures as-of 07-22 intraday, not refreshed for 07-24. |
| TSLA | sweep | carried from 07-23: post-print call volume 770,000 (~2x avg), put volume 695,000 (~2x avg); most-active contract the weekly 7/24 $330 put with new opening put buying that WOULD HAVE EXPIRED inside this window | the weekly $330 put positioning built on 07-23 was directly live during the 07-24 session covered by this run | Schaeffer's-derived day volume, not vol-vs-prior-OI; no consolidated tape. |
| MSFT | iv_spike | options market pricing an ~8% expected move by the 07-31 weekly expiry (implied range ~$364-$431) ahead of the 07-29 AMC print; IV term structure clearly front-loaded (31-Jul weekly/earnings expiry ~65% IV -> 21-Aug ~45% -> further-out ~37%) | the source explicitly notes this term structure 'argues against buying naked front-week premium' given the expected post-earnings crush -- a specific, actionable observation about the MSFT print six trading days out (Track B upside rank 5, unchanged) | Saxo desk note dated 07-23; a fresh 07-24 options snapshot for MSFT specifically was not located. |
| CBOE_EQUITY_PC | pc_skew | STILL UNRESOLVED, third consecutive run. Convex's series: 07-22 1.60, 07-23 1.95, 07-24 1.26 (a sharp one-day reversal). CBOE.com's own live snapshot and YCharts continue to show materially different baselines with no historical archive reaching this window. | No level is quoted. The one directional note: Convex's own series shows the 07-23 spike (1.95, coinciding with the Mag-7 selloff) partially reversing by 07-24 (1.26) -- consistent with the VIX easing over the same two sessions, though again this is offered as color from a possibly-mislabeled series, not a verified market-level P/C reading. | Same unresolved cross-source conflict flagged in both the 07-22 and 07-23 runs; no new reconciliation found this pass. |
D · Catalyst Calendar
Upcoming events
| Date | Event | Tickers | IV move |
|---|---|---|---|
| 2026-07-27 | Q2 2026 earnings (BMO) | AZN | unknown |
| 2026-07-27 | Durable Goods Orders (June), 08:30 ET; Dallas Fed Manufacturing; $69B 2-yr and $70B 5-yr Treasury auctions | MACRO | normal |
| 2026-07-27 | Q2 2026 earnings - day placement per a single aggregator, UNVERIFIED | BKRNUECLSAPLDWHR | unknown |
| 2026-07-28 | Q2 2026 earnings (BMO 08:30 ET call) - Tier-1 confirmed via Coca-Cola IR press release on release timing | KO | unknown |
| 2026-07-28 | Q2 2026 earnings | PYPLBAUPSRCLSTXENPHVFKLAC | unknown |
| 2026-07-28 | Advance Goods Trade Balance; Case-Shiller 20-City Home Price Index (May); Richmond Fed Manufacturing + Conference Board Consumer Confidence (July); $44B 7-yr Treasury auction | MACRO | normal |
| 2026-07-29 | FOMC INTEREST RATE DECISION 14:00 ET + Fed Chair Kevin Warsh press conference 14:30 ET. Tier-1 confirmed via federalreserve.gov press release naming Warsh as FOMC Chairman. | MACRO | high |
| 2026-07-29 | FY26 Q4 earnings (AMC) | MSFT | high |
| 2026-07-29 | Q2 2026 earnings (AMC) | META | high |
| 2026-07-29 | FQ3 2026 earnings (AMC) | QCOM | high |
| 2026-07-29 | FQ1 FY2027 earnings | ARM | high |
| 2026-07-29 | Q2/FQ4 2026 earnings | PGSBUXBSXADPORLYHOODLRCXCMGFTNT | unknown |
| 2026-07-30 | FQ3 2026 earnings (AMC) | AAPL | normal |
| 2026-07-30 | Q2 2026 earnings (AMC) | AMZN | high |
| 2026-07-30 | Q2 GDP advance estimate; Personal Income/Spending + PCE and Core PCE Price Index; weekly jobless claims; China Manufacturing/Non-Manufacturing PMI overnight - all the day after the FOMC decision | MACRO | high |
| 2026-07-31 | Employment Cost Index Q2 2026 (BLS-confirmed 08:30 ET); Chicago PMI; University of Michigan Consumer Sentiment (final); Baker Hughes weekly rig count | MACRO | normal |
| 2026-07-31 | Q2 2026 earnings (BMO) | XOMCVXABBVCLD | unknown |
E · Trade Ideas
What the data implies
SNDKneutral · track A
Rationale A textbook realized pump-risk unwind: flagged PUMP_RISK (medium) in the 07-22 run on a +14.27% vertical move with no company-specific catalyst; gave back -10.79% on 07-24 on a sector-wide de-rating (rising Chinese NAND competition). The round-trip is the confirmation that the earlier move was fragile, not the start of a new trend either way.
Invalidation n/a
Risk Included specifically to demonstrate the pump-risk flagging methodology working as intended across consecutive runs, not as a trade idea.
INTCneutral · track A
Rationale Revenue +25% YoY, fastest growth in ~15 years, easily beat consensus -- yet the stock fell -7.89% on 07-24 after an after-hours pop to ~$110 reversed, because capex guidance was raised from ~$18B to >$20B. The cleanest single-name illustration this week that a beat is not enough if the capex number scares the market more than the revenue number helps it.
Invalidation a subsequent quarter showing the capex translating into foundry customer wins (external, not just internal, demand) would directly address the market's stated concern
Risk This is the same capex-as-cost repricing pattern that hit GOOGL and TSLA the same week; treat as one instance of a market-wide theme, not an INTC-specific signal.
MUneutral · track A/B
Rationale Rode the 07-21 memory rally (+12.17%, the top Track A rising name in the 07-22 run) and is now giving a meaningful chunk back (-6.99% on 07-24) on the same competitive-pressure narrative hitting SNDK. Citi reportedly framed it as a buy-the-dip opportunity the same day. This is exactly the short half-life the 07-22 run's own methodology section flagged as a forward risk to calibrate against, playing out within three trading days.
Invalidation a reversal in the reported DRAM contract pricing deceleration (+90-95%->+58-63%->+13-18% QoQ) would meaningfully change the 12-month read
Risk The two tracks continue to disagree on this name by design; do not conflate the fast Track A signal with a 12-month view.
ARMshort · track B
Rationale Unchanged from the 07-22 run (Track B downside rank 1), but this run's catalyst-calendar research surfaced a new, more precise data point: options are pricing an ~11% expected move into the 07-29 print -- the LARGEST implied move of that day's five-name earnings batch (MSFT ~6.7-8%, META ~7.9%, QCOM ~8.2%, ARM ~11%). That is directly consistent with the standing SoftBank-float-concentration squeeze-risk caveat already carried on this position.
Invalidation unchanged: datacentre CPU attach (Grace/Graviton/Axion) inflecting royalty rates rather than units, or any AI-licensing announcement
Risk The 11% implied move is the single largest of the week's five-name megacap batch on this run's data -- size any expression accordingly.
Methodology & limitations
- BACKDATED RUN, compiled 2026-07-26 for a 2026-07-24 run date. Track A moves/closes/catalysts are strictly in-window (07-22 to 07-24); market cap figures are aggregator snapshots that may postdate the event by days, flagged per-name where materially borderline.
- RISING list returned only 9 names, not 10, per guardrail #7 - the market did not produce a defensible 10th large-cap mover with a real catalyst this window; a marginal name (AZN) was added at explicitly reduced weight to reach 9 rather than silently padding to 10.
- NVR's move magnitude is UNRECONCILED across sources (-4% to -6% range, direction confirmed) and was excluded from the ranked falling list on that basis rather than scored on a guessed figure.
- LUV's move magnitude conflicts across this run (-3.84%) and the 07-23 run (-6.19%) for the SAME ticker and date; both figures are published across the two artifacts with the conflict flagged rather than silently resolved in either direction.
- HCA and AXP both show a materially larger move in one secondary headline (+4.2% and -6% respectively) than the close-to-close figure used here (+1.67% and -4.30%); flagged per-entry as source_discrepancy.
- CBOE put/call ratio remains unresolved for a THIRD consecutive run: Convex and CBOE.com/YCharts continue to disagree by roughly 2x with no reconciling historical archive found. No level is quoted anywhere in this artifact.
- No contract-level volume-vs-prior-OI is available on any free source for a historical date - the brief's priority-1 options signal was not computable for any name across any of the three runs produced so far.
- SPCX's basic public-listing status remains unresolved on free sources for a third consecutive run; excluded from all ranked lists each time on that basis.
- TRACK B THIS RUN IS UNCHANGED FROM THE 07-23 UPDATE. Per the project's own stability instruction, this is intended behavior: neither GOOGL nor TSLA had a fresh thesis-relevant catalyst between 07-23 and 07-24, and none of the other 18 names had a fresh catalyst across the full three-day span examined by this and the two prior runs.
- Sentiment trackers (AltIndex, ApeWisdom) have no historical mode on free tiers; snapshots used are dated 07-24/07-26 and in-window peaks are inferred from mention-count deltas, not directly measured.
Not financial advice. Educational/analytical only — generated by Daily Market Pulse template v1 - backdated run for 2026-07-24, compiled 2026-07-26 via four fresh research desks (market regime, Track A movers, options+sentiment, catalyst calendar) plus a carried-forward, previously-updated Track B. Do your own diligence.