Market Pulse · Daily Scan
Daily Market Pulse
Jul 21, 2026 · sentiment/flow window ≈ 48h
Pre-marketnot publishedPost-market
TLDR;
Day two of a semiconductor reversal, on a tape that is fighting a two-front war. The chip complex extended its rebound from last week's rout after strong Taiwan and South Korea export data, with a key sector gauge up 5.2% and the Nasdaq 100 up 1.9%; all three headline indexes snapped three-session losing streaks. Against that: the US-Iran war escalated over the weekend with fresh strikes, Brent crude reached nearly $92 for the first time in five weeks (up from under $72 early this month), Treasury yields rose, and the administration unveiled 50% tariffs on a range of Canadian goods. Earnings are carrying the fundamental load — roughly 88% of the ~66 S&P 500 names reported so far have topped bottom-line estimates, and blended Q2 growth stands at 24.7%. The setup into Jul 22-23 is unusually binary: Alphabet, Tesla and IBM report Wednesday and Intel Thursday, which the market is treating as the first real test of whether AI capex is generating returns. Breadth: BROAD — a meaningful change from the narrow leadership of recent runs. The Russell 2000 (+1.5%) outpaced the S&P 500 (+0.89%), and FinViz showed only ~37.8% of US holdings declining at midday. Industrials and energy led; communication services and consumer defensive lagged. Note the contrast with Jul 20, when decliners beat advancers 1.72-to-1 on the NYSE..A · Daily Pulse
What moved — and why
▲ Rising — dailycomposite
| # | Ticker | Move | Driver | Score |
|---|---|---|---|---|
| 1 | INTC | +8.6% | Day two of the semiconductor reversal. Reported +8.6% (Trading Economics close basis) on an RBC Capital Markets preview note flagging a likely Q2 revenue beat, plus a fresh round of Data Center & AI Group job cuts and reported Xeon/Google Cloud/18A foundry items. Highest options score in the list: the options market is pricing a ~15% post-earnings swing into the Jul 23 AMC print, with elevated put premiums. I infer the move is a positioning reset, not a re-rating — INTC fell ~13.5% last week and remains ~28% off its one-month high. | 85.35 |
| 2 | NBIS | +18.78% | Largest confirmed percentage gain in the eligible universe. Tier-1 driver: Nvidia filed a Schedule 13G with the SEC on Jul 20 disclosing ~9.3% of Nebius's Class A ordinary shares (22,256,412 shares). This is a filed fact, not a rumour. Reported as a passive stake. Directly relevant to prior runs: NBIS was the standing Track B downside WATCH candidate on a customer-becomes-rival neocloud thesis — see track_b_watch for the resolution. ↔ deep-dive | 80.1 |
| 3 | MU | +12.2% | Co-leader of the memory-led rally. AP named Micron and Nvidia the two strongest forces lifting the market. Driver is sector-wide rather than name-specific: strong Taiwan/South Korea export data revived the semis bid, and the Roundhill Memory ETF (DRAM) rose ~11% with WDC and STX up ~12% each. Important context — this is a bounce, not a breakout: MU fell 13.3% over Jul 13-17 as the SOX entered a technical bear market. Reported HBM demand/yield improvements are aggregator-sourced and not independently confirmed, so catalyst is scored mid. ↔ deep-dive | 74.35 |
| 4 | SNDK | +14.3% | Largest gainer among mega-cap memory names. CATALYST FLAGGED UNVERIFIED: an aggregator (TradingKey, AI-generated) attributes the move to a comprehensive earnings beat with gross-margin expansion on enterprise SSD mix — but SNDK does not appear on any Jul 21 earnings calendar I could confirm, and no Tier-1/Tier-2 outlet corroborates a print. I therefore do NOT treat the earnings claim as fact and score catalyst below neutral. The defensible read is sector beta: SNDK is the highest-beta liquid memory name and moved with the DRAM complex. unverified | 71.5 |
| 5 | AMD | +8.1% | Second session of gains extending the reported AMD-Microsoft AI arrangement, with the Helios chip entering Microsoft Azure — read by the market as a dual-track AI supply strategy that reduces hyperscaler dependence on a single vendor. Like the rest of the complex this is partly mean reversion: AMD fell 11.1% over Jul 13-17. | 70.7 |
| 6 | MMM | +7.6% | Top Dow gainer on a confirmed, already-resolved catalyst: Q2 earnings came in better than expected (reported Jul 21, per CNBC and the Schwab earnings calendar). A clean industrials beat — one of the non-tech data points behind the session's broad breadth. | 65.7 |
| 7 | NVDA | +2% | Small percentage move but enormous index weight — AP named NVDA and MU the two strongest forces lifting the market. Two drivers: the company said its latest chip designs are reaching customers, and it disclosed the 9.3% Nebius stake. Momentum scored moderate because +2% is unremarkable for NVDA; the sentiment and catalyst scores carry the entry. ↔ deep-dive | 65.1 |
| 8 | HAS | +7% | Q2 earnings and revenue beat with full-year adjusted EBITDA guidance raised. Growth led by Magic: The Gathering, whose quarterly revenue topped $500M for the first time in the product's 30-plus-year history — a company-stated figure. Included despite a borderline market cap because the catalyst is confirmed and specific. mcap borderline | 64.6 |
| 9 | AMAT | +8% | Semi-cap equipment participated in the reversal, reported as gaining over 8% alongside Marvell. No name-specific catalyst — this is sector beta off the Taiwan/Korea export data, so catalyst is scored low. Magnitude is a floor figure ('over 8%'), flagged approximate. | 62.4 |
| 10 | GM | +4.9% | Confirmed Q2 beat on both the top and bottom line (CNBC). A second non-tech earnings data point supporting the session's breadth, and part of why the Russell 2000 outpaced the S&P 500. | 61.75 |
▼ Falling — dailycomposite
| # | Ticker | Move | Driver | Score |
|---|---|---|---|---|
| 1 | DHR | −13.7% | Worst large-cap decliner of the session. Confirmed catalyst: Q2 earnings and revenue topped estimates, but the company guided Q3 core revenue growth to 2-3% and cut full-year non-GAAP core revenue guidance to 3-4% from 3-6%, citing soft bioprocessing. This is a guidance reset, not a one-quarter miss — the distinction matters for the durability of the move. Options: 25,397 contracts traded, ~59.7% of the 4.3M average, concentrated in the $185 call expiring Aug 21. | 79.1 |
| 2 | GOOGL | −1.64% | Small decline but the highest-stakes setup on the board, which is why it ranks on options and catalyst rather than momentum. Alphabet reports Q2 AMC Wednesday Jul 22 and the options market is pricing a ~5.4% move in either direction, with call open interest and volume running ahead of puts — the crowd is leaning bullish into the print. It fell as the only red mega-cap in a strong tape, which I read as pre-earnings de-risking layered on the mid-July overhang from the reported Gemini 3.5 Pro delay (GOOGL fell more than 4% on Jul 16 on that report). | 73.75 |
| 3 | MSCI | −10.7% | Reported Q2 results below Wall Street expectations. A notable tell given MSCI's index/analytics franchise is usually treated as a defensive, subscription-like compounder — a double-digit drawdown on an earnings miss suggests the multiple carried little slack. | 71.1 |
| 4 | GPC | −7.4% | Lowered its full-year profit forecast and issued cautious second-half guidance alongside Q2 results. Together with Danaher, the second guidance-driven de-rating of the session — worth watching as a pattern: this tape is punishing forward guidance cuts far harder than backward-looking misses. | 66.95 |
| 5 | SCHW | −2.5% | Fell after missing estimates in its Q2 report. Modest magnitude; included because the catalyst is confirmed and specific. | 58.15 |
| 6 | BA | −2.25% | Worst Dow component. CATALYST UNVERIFIED and, notably, the day's Boeing-specific news flow was POSITIVE — AerCap ordered 15 Dreamliners with GE Aerospace confirming GEnx engines. I could not identify a confirmed negative catalyst, so I infer profit-taking or rotation out of industrials that had led the prior sessions. Catalyst scored well below neutral to reflect the absence of a verified driver. unverified | 48.5 |
| 7 | SHW | −1.87% | Second-worst Dow component with no confirmed single-name catalyst. Consistent with the sector picture: Materials (XLB) was among the weakest sectors on Jul 20 and the rotation on Jul 21 favoured semiconductors and industrials over defensives and consumer-facing cyclicals. unverified | 47.4 |
B · One-Year Conviction
Asymmetric 12-month skew
Free sources only; estimate-revision and valuation rigor limited by paywalls. Directional, not precise. Track B is scored independently of Track A — a name appearing in both carries two unrelated scores.
▲ Upside skew — 1yrconviction
| Ticker | Thesis | Key risk | Score |
|---|---|---|---|
| MU | Memory super-cycle carry, held since inception of this position. The Jul 13-17 drawdown (-13.3%) and the Jul 21 rebound (+12.2%) are both within the expected volatility of a cyclical name and neither triggers a defined invalidation condition. Structural driver intact: HBM/DRAM demand tied to AI infrastructure, with the memory complex now trading as its own factor (the DRAM ETF moved ~11% in a single session). ↔ deep-dive | Cyclical memory pricing rollover; 2027 oversupply as capacity lands; HBM competitive qualification: aggregator reports that rival HBM products are qualifying with Tier-1 customers would erode first-mover pricing power (UNVERIFIED — not Tier-1 confirmed); Valuation compression after a large multi-quarter run | 82.2 |
| NVDA | AI-infrastructure capex remains the dominant secular driver. Structural score raised this run on a Tier-1 datapoint: the Schedule 13G disclosing a 9.3% Nebius stake shows NVDA extending from merchant silicon into equity positions in the neocloud layer that consumes it — ecosystem entrenchment that deepens the moat and secures demand channels. Offsetting, and the reason fundamental_valuation stays in the low 60s: the bear case got sharper this month. ↔ deep-dive | Model-efficiency risk — Moonshot's Kimi K3 revived the DeepSeek thesis that advanced models may need less compute, breaking the link between AI adoption and endless chip demand; Capex-cycle digestion: TSMC's raised 2026 capex guide of $60-64B was read as cost inflation rather than demand strength; Valuation and crowding — the SOX fell into a technical bear market on Jul 17, more than 20% off its late-June record; Customer concentration in hyperscalers now dual-sourcing (AMD Helios into Azure) | 75.1 |
| AVGO | Custom silicon (XPU/ASIC) plus networking leverage to the AI build-out. Fundamental_valuation nudged up this run on a mechanical point rather than a new thesis: AVGO is reported up only ~4.15% year-to-date in 2026 and fell 7.3% during the Jul 13-17 rout, so the entry multiple has compressed materially while the structural case is unchanged. It also showed relative outperformance versus peers through the rout, which is mild evidence the position is less crowded than NVDA. thesis-level | Customer concentration in a small number of hyperscaler ASIC programs; Cycle timing — custom-silicon revenue is lumpy and program-gated; Same model-efficiency risk that hit the whole complex in mid-July; Thesis-level position: no Jul 21 close-basis print was confirmed for AVGO on the free tier | 70.6 |
▼ Downside skew — 1yrconviction
| Ticker | Thesis | Key risk | Score |
|---|---|---|---|
| NKE | Carried unchanged, no-churn. Multi-year turnaround remains unproven; the June 30 print was a low-quality, tariff-refund-driven beat with China sales down 12% and cautious guidance, which was assessed at the time as reinforcing rather than invalidating the thesis. NO NEW INFORMATION this run — no NKE-specific news flow surfaced in the Jul 20-21 window and no catalyst falls inside the 5-day horizon. Scores held flat deliberately: absence of news is not evidence, and manufacturing score drift without a datapoint would be fabrication. | A clean beat with a credible guide would invalidate quickly; Consumer-discretionary rotation could lift the name independent of fundamentals; Position is stale — no fresh datapoint since the June 30 print | 55.75 |
| PLTR | Carried, no-churn. Valuation-driven de-rating thesis on a crowd-favourite. Supporting evidence this run: continued insider selling (CTO Shyam Sankar $24.05M on Jul 2; Director Alexander Moore $2.14M on Jul 15, both under pre-arranged 10b5-1 plans — the plans mean these are not signals of intent, but the optics compound the positioning problem), and a Jefferies price target of $70 against a Jul 8 close of $128.63. HONEST COUNTER-EVIDENCE: Citi named PLTR a top software pick ahead of the Aug 3 Q2 print, shifting from a Neutral/High Risk stance, and BofA and DA Davidson are in the bull camp; the stock rebounded ~4.84% over the past week. That is a genuine erosion of the sentiment leg, reflected in the sub-score, but it does not meet the defined invalidation condition, so the position is retained rather than churned. retail-heavy | Re-acceleration in government and commercial bookings; Sell-side re-rating already underway (Citi upgrade) — the sentiment leg is weakening; Aug 3 Q2 print sits OUTSIDE the 5-day catalyst horizon; if revenue growth holds above 80% YoY with a firm guide, the valuation debate resets and this thesis is in trouble; Short-side crowding in a retail-heavy name carries squeeze risk | 54.6 |
C · Options Heat
Where the premium is paying up
| Ticker | Signal | Read | Implication | Lag |
|---|---|---|---|---|
| TSLA | vol_oi | 244,000 calls vs 116,000 puts traded through midday Jul 21; calls accounted for more than two-thirds of total premium. Top three most active contracts by volume were all calls, with the most premium in the 380-strike calls expiring Friday. ATM straddle implies a 5.76% post-earnings move — the largest implied move since a ~6% pricing in October 2025. | BULLISH positioning into the Jul 22 AMC print, med-high confidence on the direction of the positioning itself. Separately: an implied move at a one-year high means event premium is rich, so long-premium expressions carry material IV-crush risk even if directionally right. | Free-tier/EOD sources only; no live block or sweep tape. Figures are delayed and may be partial. Not a real-time institutional flow feed. Volumes are midday figures, not final session totals; the implied move will have shifted into the print. |
| INTC | iv_spike | Options market pricing a roughly 15% swing following the Jul 23 AMC report. Put premiums reported as elevated. | UNCERTAIN direction, high confidence that event premium is extreme. A 15% implied move on a large-cap is exceptional and reflects genuine two-sided disagreement about whether a 163% rally is justified by the numbers. Defined-risk structures strongly favoured over naked long premium. | Free-tier/EOD sources only; no live block or sweep tape. Figures are delayed and may be partial. Not a real-time institutional flow feed. Implied-move figure is a Jul 20 aggregator estimate and will have moved. |
| GOOGL | iv_spike | ~5.4% implied move in either direction into the Jul 22 AMC print, referencing a $348.60 share price. Open interest and volume metrics show call activity running ahead of puts. | BULLISH crowd lean, medium confidence. Worth noting the tension: the options crowd is positioned bullish while the stock closed red on a strongly green day, which is an unusual divergence and argues the cash tape is more cautious than the options tape. | Free-tier/EOD sources only; no live block or sweep tape. Figures are delayed and may be partial. Not a real-time institutional flow feed. Aggregator estimate as of Jul 21 midday. |
| DHR | vol_oi | 25,397 contracts traded, ~59.7% of the 4.3M average, concentrated in the $185 call expiring Aug 21. | UNCERTAIN, low confidence. Upside-call concentration after a 13.7% guidance-driven drop is consistent with either dip-buying or covered overwriting, and free-tier data cannot distinguish the two. Reported here because the concentration is specific, not because the direction is readable. | Free-tier/EOD sources only; no live block or sweep tape. Figures are delayed and may be partial. Not a real-time institutional flow feed. EOD aggregator log; volume is below the name's average, so this is a concentration signal, not an unusual-volume signal. |
| ORCL | vol_oi | 293,974 contracts traded, reported as 73.4% of a 40.1M typical daily figure, with the most volume in the $130 call expiring Jul 24. | UNCERTAIN, low confidence — and reported with an explicit caveat: as stated, volume was BELOW the typical daily figure, so this does not meet the vol-to-OI test for new positioning. It is included for completeness and transparency, and should not be read as an unusual-activity signal. | Free-tier/EOD sources only; no live block or sweep tape. Figures are delayed and may be partial. Not a real-time institutional flow feed. Source's own volume baseline appears internally inconsistent; treated as low-quality and down-weighted accordingly. |
| MARKET | pc_skew | UNVERIFIED — no CBOE total or equity put/call ratio value retrievable on free sources for the Jul 21 session. | None drawn. Per the no-fabrication rule this is recorded as absent rather than estimated, and every name-level options sub-score in Track A that lacked verified data was set to neutral (45-55) rather than inferred from the broad tape. | Free-tier/EOD sources only; no live block or sweep tape. Figures are delayed and may be partial. Not a real-time institutional flow feed. Absence of data recorded explicitly; no value inferred. |
D · Catalyst Calendar
Upcoming events
| Date | Event | Tickers | IV move |
|---|---|---|---|
| 2026-07-22 | Q2 2026 earnings (AMC). First of the mega-cap AI capex tests; cloud backlog and AI monetisation in focus. | GOOGL | high |
| 2026-07-22 | Q2 2026 earnings (AMC). Delivery guidance, FSD/AI progress, margins. | TSLA | high |
| 2026-07-22 | Q2 2026 earnings. Follows the Jul 14 pre-announcement that drove a >25% single-day decline — unusually high uncertainty for a mature name. | IBM | high |
| 2026-07-22 | Q2 2026 earnings (AMC). Analog read-through on industrial/automotive end-market recovery; sector tone-setter. | TXN | unknown |
| 2026-07-22 | Q2 2026 earnings (mixed BMO/AMC). | NOWTCMECSXGEVMCOPMURI | unknown |
| 2026-07-23 | Q2 2026 earnings (AMC). Street ~$0.22 EPS vs -$0.10 a year ago; revenue ~$14.42B, up ~12% YoY. Terafab and Google deal updates in focus. | INTC | high |
| 2026-07-23 | Q2 2026 earnings. XOM and NEM carry direct read-through from Brent near $92 and the gold move. | XOMVZAXPNEENEMDECKURI | unknown |
| 2026-07-23 | Initial jobless claims, 8:30am ET. | MACRO | low |
| 2026-07-24 | New home sales and manufacturing PMI. Light macro week overall — the earnings tape dominates. | MACRO | low |
| 2026-07-27 | Q2 earnings season continues; no confirmed mega-cap prints identified within the horizon for this date. | SPX | unknown |
| 2026-07-28 | Q2 earnings season continues; no confirmed mega-cap prints identified within the horizon for this date. | SPX | unknown |
| 2026-07-29 | FOMC rate decision 2:00pm ET followed by Chair Warsh's press conference. BOUNDARY ITEM — falls on the sixth trading day, just outside the 5-day horizon, but included because it dominates the risk picture immediately beyond it. Warsh's hawkish testimony keeps the rate-hike debate alive; Bloomberg Economics expects no change for the balance of 2026. | MACRO | unknown |
E · Trade Ideas
What the data implies
INTCneutral · track A
Rationale Volatility view, not a directional one. The options market is pricing a ~15% post-earnings swing into Jul 23 — exceptional for a large-cap and a sign of real two-sided disagreement. The stock rallied 8.6% on an RBC beat-preview note but is still down ~13.5% on the week and ~28% over the month, so both a relief re-rating and a disappointment leg are live.
Invalidation n/a for a volatility view. If expressed directionally long, a close back below the Jul 21 reversal low would negate the setup.
Risk Severe IV crush after the print is the base case regardless of direction. Sizing matters more than selection here.
MUlong · track A/B
Rationale The one idea where Track A and Track B agree. Track A: +12.2% on confirmed close basis, co-leader of the memory complex. Track B: highest conviction score in the book at 82.2, with estimate revisions the strongest sub-signal. The Jul 13-17 drawdown of 13.3% reset positioning without any confirmed change to the demand picture.
Invalidation Tier-1 confirmation that competitor HBM has qualified with Tier-1 customers (currently aggregator-sourced and UNVERIFIED), evidence of DRAM price rollover, or a close back below the Jul 17 rout low.
Risk Deeply cyclical. Both the July drawdown and this rebound are within normal volatility for the name; neither is confirmation of anything. Overlaps the deep-dive library.
DHRshort · track A
Rationale The distinction that matters: Q2 earnings and revenue beat, but full-year non-GAAP core revenue guidance was cut to 3-4% from 3-6% with Q3 core growth guided to 2-3% on soft bioprocessing. Guidance resets of this shape typically take several quarters to work through estimates, unlike a single-quarter miss. GPC's -7.4% on the same day on the same pattern suggests the tape is systematically punishing forward guidance cuts this season.
Invalidation Reaffirmed or raised full-year guidance, or confirmed bioprocessing order reacceleration, at the next quarterly update.
Risk Crowded post-gap shorts are squeeze-prone, and life-sciences names are recurring M&A and activist targets.
GOOGLneutral · track A
Rationale A genuine divergence worth flagging rather than resolving. The options crowd is leaning bullish into the Jul 22 print — call OI and volume ahead of puts, ~5.4% implied move — while the cash tape sold it as the lone red mega-cap on a strongly green day. One of those two is wrong. The mid-July Gemini 3.5 Pro delay report, which knocked more than 4% off on Jul 16, is the live overhang the print has to address.
Invalidation For a long view: cloud revenue growth deceleration or further AI roadmap slippage on the call. For a short view: a clean beat with cloud backlog conversion evidence.
Risk Direction is genuinely uncertain and I am not resolving it. Alphabet also sets the tone for the entire AI-capex complex, so the read-through risk to other holdings exceeds the single-name risk.
NBISneutral · track B
Rationale Recorded to close the loop on a standing position rather than to open one. The Track B downside watch on NBIS is retired this run — the promotion condition was met but the underlying thesis was falsified by Nvidia's Schedule 13G. No position is taken in either direction. Chasing an 18.78% single-session move on a passive-stake disclosure is not a thesis, and the framework's answer to 'the trigger fired but the reasoning broke' is to take no position, not to invert it.
Invalidation n/a — no position held. A fresh thesis would be required for re-entry in either direction.
Risk The main risk of this decision is being wrong about the significance of a passive stake. See track_b_watch.caveat.
Methodology & limitations
- No live options tape. Every options claim derives from free/EOD aggregators; no block or sweep data.
- CBOE total and equity put/call ratios were NOT retrievable for Jul 21 — recorded as UNVERIFIED, not estimated.
- Name-level options data existed for only 5 of 17 Track A names. The remaining 12 carry neutral (45-55) options sub-scores by rule, which mechanically compresses their composites relative to names with verified options signals. This is a known bias of the free-data regime and it favours names with upcoming earnings.
- Sentiment velocity is inferred from breadth of mainstream coverage, not measured. No Reddit or FinTwit mention-rate data was retrieved this run, so the Tier-3 crowd-positioning input is effectively absent and no PUMP_RISK flag could be computed from mention-spike evidence. No name in either list showed the vertical-move-without-catalyst pattern that would raise the flag on price action alone.
- Market caps: verified figures obtained for SNDK and HAS only; NBIS derived from the 13G share count and flagged approximate. All other names assert eligibility via index membership (Dow 30 / S&P 100 / S&P 500 / Nasdaq-100), which guarantees the $10B floor by index construction but does not provide a figure.
- Several single-name magnitudes are midday or 'over X%' floor figures rather than settlement prints; each is flagged at the entry level (print_intraday, print_approx).
- Brent, WTI and gold levels are narrative figures from close wraps, not settlement prints. Gold carries an unresolved two-source discrepancy ($4,082.73 vs $4,133.10).
- VIX is arithmetically derived, not an independently confirmed settlement print.
- Energy was a live theme (Brent near $92, five-week high) but no single-name energy close was confirmable, so the sector is absent from Track A despite its likely relevance. This is a coverage gap, not a judgment that energy did not move.
- Scores are heuristic and not backtested. Sub-score levels are internally consistent across runs but have no external validation.
- Seven-session gap since the last run (Jul 10). Gap events are reconstructed from weekly wraps rather than daily runs, so single-session attribution within Jul 13-17 is coarser than usual.
Not financial advice. Educational/analytical only — generated by Daily Market Pulse pipeline — run 2026-07-21 (post-close, 7-session gap reconciliation). Do your own diligence.