🧪 TEST — First test dispatch after service restoration

This report is a test dispatch after the server came back online; please read it primarily as a verification of the system pipeline.

⚠️ Important limitation: this issue lacks recap-based retraining

This system is designed as a cognitive closed loop of pre-market forecast → intraday tape → post-close recap reconciliation → lessons fed back into the next day's pre-market. However, the server was down from 7/24 (Friday) to 7/27 (Monday), and no recap report was generated during that period, therefore:

• This report has no recap reconciliation from the preceding days as input, and cannot use the recent "called it right / called it wrong" hit rate to calibrate this issue's judgments;
• The lessons from the previous US recap (which kinds of recommendations failed, which themes were overrated) could not enter this issue's scoring;
• Consequently, the stock scores and labels in this report rest only on current evidence, without correction from recent live feedback, and their confidence is below the level of normal system operation.

Please downweight this list accordingly. The retraining loop will be re-established once this week's recap reports resume.

Coverage window: 2026-07-24 (Friday) 16:00 ET regular-session close → 2026-07-27 (Monday) approx. 09:00 ET Report generated: 2026-07-27 09:00 ET / 13:00 UTC Fundamental data: pulled live from yfinance, retrieved 2026-07-27 12:49–13:20 UTC, prices and financials are all on a 7/24 close basis (the market has not opened today, so there is no live price). QA: audited line by line by risk-auditor and revised per its P0 comments. Content that failed the audit and was pulled has been deleted; figures that could not be verified are now marked "pending verification".

⚠️ Three basis declarations to read before this report

1. "Pre-market change" and "7/24 close change" are two different things, and this report lists them separately. Any move labeled "7/24 close" is a historical fact that already happened last Friday and can no longer be traded today, not an opportunity this morning.
2. All stock prices, market caps and financial ratios are on a 7/24 close basis, not live prices. Today's pre-market quotes are given only where explicitly labeled "pre-market".
3. The consensus figures in this report (consensus / forward EPS / PEG) come from yfinance aggregate fields, a mixed basis across multiple brokers (some non-GAAP), and were not cross-checked item by item against the analysts' original notes. Such numbers are for relative comparison only and cannot serve as an absolute valuation basis.

---

0. Today in one line

1. The strongest catalyst is geopolitics, not earnings: the US and Iran paused mutual strikes over the weekend, Brent crude fell sharply and at one point broke below $90 — a macro positive that clears the entire chain of "inflation expectations → Treasuries → US equity valuations" in one stroke.
2. The strongest theme = second-order beneficiaries of falling oil: airlines / cruise lines / travel (fuel costs fall directly) + tech growth stocks (inflation expectations recede → discount rate falls); energy stocks are the most clearly identified losers.
3. Flow direction: futures broadly higher with the Nasdaq leading (NQ +1.34% > YM +1.11% > ES +0.85%), VIX −3.98% to 17.84, 10Y Treasury 4.64% (−4bp) — a risk-on profile, with money rotating out of energy/defensives into tech and high beta.
4. Driver types: geopolitics (dominant) + macro (FOMC 7/28-29) + earnings (super week: MSFT/META/ARM Wednesday after the close, AAPL/AMZN Thursday after the close) + M&A (argenx buying Forte).
5. Pre-market state: futures broadly up, oil sharply lower, Treasuries stronger, VIX lower, overnight Asia/Europe leaning positive (CXMT's Shanghai debut surge lifting the Asian memory chain). But today looks more like "sentiment repair ahead of the events" than an established trend — this week's information load is heavily concentrated in Wednesday's FOMC and the capex guidance from four trillion-dollar giants on Wednesday/Thursday, and today's rally lacks follow-on ammunition.

⚠️ Self-falsification of point 3: if the FOMC actually hikes on Wednesday, today's main line of "falling discount rate → growth-stock valuation repair" reverses immediately, the Nasdaq-leadership logic does not hold, and high P/E names (ARM, AAPL) are hurt most. Whether this report's first theme holds depends on an event that has not yet happened.

---

1. News overview

⚠️ Timeliness warning on #9 (important): that hike probability is taken from 7/23, i.e. before the oil plunge. The 7/26 geopolitical de-escalation and falling oil should logically reduce the Fed's urgency to hike, so the probability has almost certainly been repriced — but this report could not obtain a primary snapshot of CME FedWatch this morning, so the true current probability is pending verification. Do not treat 38% as today's market state.

⚠️ Numerical discrepancies on #12 and #13: MXL's decline is reported as 16% (Seeking Alpha), 17.9% (GuruFocus) and 18.65% (Tickeron); BE's down day is reported as Thursday (Benzinga headline) and Friday (TIKR). This report confirms only the fact of "a sharp drop" and does not endorse any single figure or single date.

⚠️ Note on #10: +38% YoY earnings growth is far above historical norms, with no explanation of base effects or blended/reported basis, and it is a CNBC paraphrase rather than the original FactSet Earnings Insight; use as background only.

Pre-market index snapshot

Data taken 7/27 approx. 08:10–08:50 ET (spanning about 40 minutes, not a single-moment cross-section); sources: Yahoo Finance quote pages and TradingEconomics. Precise per-item timestamps could not be obtained.

⚠️ The crude data bases conflict, and this report endorses no single figure: Bloomberg (7/26) records Brent "briefly −7.4% and below $90, then recovering about half the drop"; TradingKey (7/27 pre-market) reports "Brent −13% to $85.23, WTI −8% to $83.08"; WaPo reports the September contract at $92.02 and the October contract at $87.48. These four figures belong to different moments, different contracts and possibly different benchmarks, and cannot be assembled into one range. This report confirms only that the direction is down; for magnitude, refer to live quotes. Also note: Brent is still up more than 50% year to date (Bloomberg); this is a pullback from highs, not a turn into a bear market.

---

2. Strongest themes, descending

⚠️ Methodological warning on CXMT: a +466% A-share first-day pop is primarily a local liquidity and retail pricing phenomenon, and elevating it into "a global re-rating of memory assets" is narrative extrapolation, not current fundamental data. Also: CXMT is the fourth-largest DRAM maker globally (after Samsung, SK Hynix and Micron); its "market cap of RMB 3.3 trillion ≈ $487 billion" conversion implies an exchange rate of about 6.78, which is inconsistent with the current RMB exchange rate in order of magnitude; that conversion comes from the original source and has not been verified by this report.

---

3. Overall stock strength ranking

Sorted by total score, descending. "7/24 close change" and "7/27 pre-market change" are strictly listed separately — the former is history that already happened, the latter is this morning's state. All fundamental data is on a 7/24 close basis.

🟢 Bullish zone

🔴 Bearish zone

⚠️ CAPR's −40% vs −65% is not a data error but a difference in basis: a 25-percentage-point gap for the same name over the same window more likely comes from different baselines ("pre-market vs previous close vs two-day cumulative"). This report cannot determine which basis is correct, so both figures are marked pending verification. Also: the exact release time of that FDA briefing document is unconfirmed (pre-advisory-committee briefings are usually published about 2 business days ahead, so both 7/24 and 7/27 are possible) — if it was released on 7/24, this item should not be listed as "today's new news".

---

4. Stock scoring model (100 points total)

⚠️ Declaration on the nature of the scores (must read): the scores below are a subjective ordinal scale, not a precise measurement. The 2-point gap between 78 and 76 carries no ranking meaning and should be read only as "the same tier". In particular:

• "Source authority" varies only within the 11–15 range across the 12 names, so its discriminating power is close to noise.
• Giving "catalyst directness 18/20" to five companies that have not yet reported (MSFT/META/AMZN/AAPL/ARM) is a prediction about events that have not happened, not a realized fact.
• Read by the three tiers 75+/60-74/50-59 rather than by exact scores.

Sub-score detail (all 12 names, fully auditable)

Note on the revision of MU's score: the first draft gave MU "earnings elasticity 15/15", but MU's FY26Q4 guidance midpoint of $50B is below the consensus $50.84B. Under this report's own iron rule that "guidance matters more than a beat in the current quarter", that score is cut to 11, the risk deduction increased to −14, and the total lowered from 62 to 57. The earlier scoring applied a double standard relative to MXL (a double beat judged bearish because of guidance / expectation issues), and this has been corrected.

---

5. Detailed analysis of top names

① MSFT 微软 Microsoft — 78 pts | Priority deep-dive

• Related news: reports FY2026 Q4 on 7/29 (Wednesday) at 16:00 ET after the close. Consensus EPS $4.240, revenue $87.67B (basis: yfinance aggregate, not cross-checked against the analysts' original notes). The Azure constant-currency growth bogey the market is watching is 36%, with management having previously guided 39-40% (paraphrased from media, the company IR original was not obtained, pending verification).
• Catalyst logic: dual drivers of own earnings plus guidance. The real pricing variable is not the current quarter's EPS but FY2027 capex guidance — it determines both Microsoft's FCF and order visibility for the entire AI hardware chain.
• Theme and stage: AI cloud has entered a verification phase of "show me cash flow, not stories". The stock is already down about a third from its high, which shows the market has long been skeptical — this is not a crowded trade.
• Fundamental verification (yfinance, 7/24 basis): latest quarter revenue $82.89B (+18.3% YoY), net income $31.78B (+23.1%); TTM gross margin 68.3%, operating margin 46.8%, net margin 39.3%; OCF $170.1B (OCF/net income 1.36x, healthy quality); net cash +$21.3B; forward P/E 19.7, PEG 1.18.
• Moat: the twin engines of Azure + Office, with extremely high enterprise switching costs; it is the only name in the group whose consensus maintains mid-double-digit revenue and EPS growth across both years.
• 7/24 close and technical level (not live): $381.70, −31.3% from the 52-week high of $555.45 (set on 2025-07-31, a full year ago).
• ⚠️ Biggest flaw: TTM capex has reached $97.2B (versus only $28.1B in FY2023), capex/OCF 57%, compressing FCF to $72.9B. P/E 22.7 vs P/FCF 38.9 — that 16x gap is the price of AI capex.
• Judgment: the best match of valuation to growth in the group, with the thickest expectation-gap cushion. But the gap between "profit" and "distributable cash" is widening. No single-stock catalyst today; the answer comes Wednesday after the close.

② META — 76 pts | Priority deep-dive

• Related news: earnings 7/29 after the close. Revenue expected at $60.26B (+26.8%).
• Fundamental verification: latest quarter revenue $56.31B, +33.1% YoY (fastest of the giants); gross margin 81.9%; OCF $124.0B; PEG 0.87 (lowest in the group).
• ⚠️ Two distortions that must be unpacked separately:
  1. TTM net income is contaminated by one anomalous quarter: net income over the last four quarters was 26.77 / 22.77 / 2.71 / 18.34 (USD billions). The quarter ended 2025-09-30 was only $2.71B. That quarter contains a large one-off item whose nature (tax / regulatory fine / impairment) cannot be confirmed from the three statements — no reliable data available, the quarter's 10-Q must be checked. Normalizing to about $20B would put TTM net income at roughly $88B, taking trailing P/E from 21.6 down to about 17x. ⚠️ That "about 17x" rests on an unverified normalization assumption and must not be used as a conclusion.
  2. The forward cheapness is "borrowed": consensus FY0 EPS is +40.7%, but FY+1 drops sharply to +6.4%. The assumption behind PEG 0.87 is that "EPS barely grows next year".
• 7/24 close: $595.19, −25.3% from the 52-week high.
• Risks: capex is 35.2% of revenue; total debt $86.8B, and it has moved into net debt of −$5.6B for the first time.
• Judgment: the low PEG and "no growth next year" must be viewed together. The object of the deep-dive is Wednesday's capex and 2027 guidance.

③ AMZN 亚马逊 Amazon — 71 pts | Watch closely

• Related news: earnings 7/30 after the close, expected EPS $1.815, revenue $196.18B (+17.0%).
• Fundamental verification: latest quarter revenue $181.52B (+16.6%), net income $30.25B (+76.7% YoY); TTM net margin 12.2% (vs 5.3% in FY2023), the income statement is improving rapidly; P/S 3.4 (lowest in the group).
• ⚠️ The single most glaring number: TTM capex $151.0B vs TTM OCF $148.5B → TTM free cash flow = −$2.47B (now negative). capex/OCF 102%. Annual FCF trajectory: FY2023 $32.2B → FY2024 $32.9B → FY2025 $7.7B → TTM −$2.5B.
• Balance sheet: total debt $209.9B, net debt −$66.8B (heaviest in the group). (Note: yfinance's Total Debt for AMZN typically includes lease liabilities; the basis needs to be checked in the 10-Q.)
• Judgment: the "cheapness" currently shows up only in sales and book profit, not in free cash flow. A P/E of 27.8 currently has no FCF support.

④ AAPL 苹果 Apple — 69 pts | Watch closely

• Related news: earnings 7/30 (Thursday) after the close, expected EPS $1.89 (vs $1.57 a year ago), revenue $108.86B (+15.8%). Analyst consensus is Buy with an average target of $325.36 — below the current price; dispersion is extreme: Morgan Stanley $364, HSBC $366, but KeyBanc $250. (The ratings and targets above are paraphrased from media roundups; the original broker notes were not obtained, pending verification.)
• Catalyst logic: the market is repricing Apple's "light AI" strategy — not pouring money into infrastructure, but monetizing a huge installed base. Against the backdrop of a capex arms race, that has instead become a differentiating advantage.
• Fundamental verification: FCF $129.2B, FCF/net income 1.05x, TTM capex only $11.1B (2.4% of revenue, capex/OCF only 8%) — the only mega-cap whose FCF has not been eroded in the AI capex wave. Growth is accelerating: FY2025 full-year revenue rose only +6.4%, while consensus FY0 revenue is +15.0%.
• ⚠️ Basis reminder: ROE of 141.5% cannot be read as a quality indicator — the denominator, shareholders' equity, has been compressed to $106.5B by buybacks, with equity/assets at only 28.7%. ROA 34.9% is the clean basis. Also: net debt −$16.2B, so it is no longer a "net cash company".
• Pre-market and technicals (both bases given):
  • 7/27 pre-market: $334.00, +0.29% (source Benzinga Pro, via chinatechnews)
  • 7/24 close: $333.02; 52-week high $334.99 (set on 2026-07-17)
  • Distance from the 52-week high: −0.30% on the pre-market price, −0.59% on the closing price. Both bases are correct; the baselines differ.

  ✅ On the "+3.53%" shown by one quote source: that figure does not match the arithmetic. $333.02 (7/24 close) × 1.0029 = $333.99 ≈ $334.00, precisely matching the pre-market price reported by Benzinga, confirming that +0.29% is the real pre-market move against Friday's close. A +3.53% move would correspond to about $344.8, above the 52-week high of $334.99, contradicting the fact that the stock is still below its prior high. Therefore +0.29% is accepted.

• Risks: PEG 2.68, highest in the group; hugging the 52-week high with elevated IV into earnings, so the IV crush risk after earnings is the largest; the expectation-gap score is only 3/10.
• Judgment: best fundamental quality in the group, worst position in the group. The two do not conflict, but what a buyer takes on today is the latter.

⑤ BA 波音 Boeing — 63 pts | Watch closely

• Related news: Q2 earnings 7/28 (Tuesday). Expected EPS −$0.28 to −$0.34 (still a loss), revenue $24.05-24.27B.
• Positives: commercial aircraft deliveries +14% YoY to 171.
• Risks: the 787 engine bottleneck; total debt $47.2 billion; free cash flow expected to be negative.
• Judgment: the turnaround story is progressing but not delivered; cash flow is the only number worth watching. Note that it is the first heavyweight report of the week and may set the tone for the whole week.

⑥ ARM — 60 pts | Watch only

• Related news: earnings 7/29 after the close, expected EPS $0.402, revenue $1.265B.
• Fundamentals: FY2026 revenue $4.92B (+22.7%); latest quarter $1.49B (+20.1% YoY), net income $0.31B (+49.0%). Consensus is accelerating: FY0 revenue +21.3% → FY+1 +35.4%.
• ⚠️ 97.5% gross margin vs 18.5% operating margin: the licensing/royalty model produces near-100% gross margin, but R&D plus stock-based compensation eat nearly 80 percentage points. ROE only 12.0%, ROA 9.2% — lowest in the group. High gross margin, low return on capital.
• ⚠️⚠️ Valuation-basis trap (the one item in this report to watch most closely): P/E (TTM) 302.3 uses GAAP EPS of $0.84; forward P/E 84.5 uses forwardEps of $3.076, and that value happens to equal the non-GAAP consensus for FY+1 (ending 2028-03). Referencing FY2026 non-GAAP EPS of $1.77 vs GAAP $0.85 (a 2.1x gap, mainly SBC), the true forward GAAP P/E would be far above 84.5x. Inside that "collapse" from 302x to 84.5x, part is growth and part is purely a change of accounting basis plus a jump of two fiscal years.
• 7/24 close: $260.01, −42.6% from the 52-week high of $452.70 (set on 2026-06-18, just 5 weeks ago).
• Judgment: P/S 56.4, P/B 33.4 and PEG 2.14 are all the highest in the group. Watch only.

⑦ UAL 联合航空 United Airlines — 58 pts | Watch closely

• Related news: oil falls, UAL +3.5% pre-market, LUV +3.1%, DAL/AAL strengthening in tandem.
• Catalyst logic: fuel is an airline's largest variable cost, and falling oil flows straight into the income statement, with a short and certain transmission path. The hardest-logic second-order beneficiary today.
• Reverse verification (important): when oil spiked on 7/8 on the US-Iran conflict, AAL fell 5%, UAL fell 4%, and DAL/JBLU fell 3%. The two-way sensitivity of this same set of stocks to oil has been repeatedly verified by the market — this beta is real, not imagined.
• ⚠️ Risks: this is a "pause" in strikes, not a ceasefire agreement, and its duration and the schedule of follow-up talks have not been publicly disclosed. Oil already recovered about half of the −7.4% drop on Sunday. A single geopolitical headline could give back all of today's gains.
• Judgment: hardest logic, weakest persistence (persistence only 5/10). Watch closely.

⑧ MU 美光 Micron — 57 pts | Watch only

• Related news: CXMT surged 466% on its Shanghai debut, the memory chain is broadly higher this morning, MU +2.5%.
• ⚠️ This news is a double-edged sword and can flip direction at any time: the same event on 7/16 (when CXMT priced its $8.6 billion IPO) sent MU down 5% and SK Hynix and SanDisk down 7% each, on the rationale of "Chinese DRAM capacity expansion → oversupply". Today it has become bullish, on the rationale that "the listing valuation re-anchors global memory assets". Same fact, two narratives; today the market is running the optimistic one.
• ⚠️ Note the freshness of the news: the FY26Q3 financials below were released on 2026-06-24, already a month ago, and are not a catalyst today.
• Fundamental verification (cross-checked item by item against Micron's official press release; the data matches): FY26Q3 (ended 5/28) revenue $41.46B, +73.7% QoQ, +345.7% YoY; quarterly GAAP net income $28.24B; a fifth consecutive record revenue quarter. TTM gross margin 72.6%, operating margin 65.7%; net cash +$19.6B, total debt only $6.4B.
• ⚠️ New red flag — an abnormally high quarterly net margin: $28.24B / $41.46B = a net margin of 68.1%, only 4.5 percentage points below its TTM gross margin of 72.6%, implying that operating expenses plus income tax together are only about 4.5pp of revenue. For a fab with very heavy R&D and depreciation, that can almost only be explained by a large one-off tax item (such as the release of a valuation allowance on deferred tax assets). If so, three metrics — net margin 55.9%, P/E 20.8 and OCF/net income 1.02x — are all contaminated; the nature must be checked in the 10-Q, no reliable data available for now.
• ⚠️ Guidance below consensus (missed in this report's first draft): Q4 guidance midpoint $50B, below the consensus $50.84B. Under this report's iron rule that "guidance matters more than a beat in the current quarter", this is a deduction and the score has been lowered accordingly.
• ⚠️ Cash-flow quality flaw: TTM OCF/net income is only 1.02x. For a fab with very heavy D&A, that ratio should normally be well above 1.5x, and being compressed to 1.0 indicates heavy working capital absorption (receivables + inventory).
• ⚠️⚠️ The core valuation contradiction — the low-PE trap: P/E (TTM) 20.8 looks inexpensive, forward P/E 6.0, PEG 0.14. But forward EPS of $153.74 is the FY2027 consensus, implying revenue rising again from $129.8B to $238.8B. For a highly cyclical stock, the lowest PE appears precisely at peak earnings — the denominator of 20.8x is peak EPS of $44, and the denominator of 6.0x is an assumption of "doubling again". PEG 0.14 has no analytical meaning here.
• 7/24 close: $920.95, 52-week range $103.38-$1255.00, −26.6% from the high.
• Judgment: the fundamentals are real, the valuation metrics are fake. A net margin of 55.9% is not a normal state but a cyclical extreme. Watch only.

⑨ DAL 达美航空 Delta Air Lines — 56 pts | Watch closely

• Same logic as UAL, with the best margins and balance sheet within the industry. The exact move this morning is pending verification (TradingKey only says "up in tandem"). The risks are identical to UAL's, and it shares the same falsifiable anchor (the ceasefire duration is unknown).

⑩ FBRX Forte Biosciences — 54 pts | Watch only

• Related news: pre-market on 7/27, argenx announced an all-cash acquisition at $77/share, about $2.2 billion in total, a premium of about 40% to the previous close, and an 86% premium to the average price since the 7/9 vitiligo Phase Ib data release.
• Asset value: FB102, a first-in-class anti-CD122 antibody, with clinical proof of concept achieved in vitiligo and celiac disease. argenx intends to build a second growth curve beyond Vyvgart.
• ⚠️ The arithmetic of the remaining arbitrage spread (the first draft overstated it by about 4x, now corrected): backing out from the $77 consideration and the 40% premium gives a previous close of about $55.00; pre-market +39% → about $76.45, only about 0.72% away from $77, not the "several percent" claimed in the first draft. If closing takes 1-2 months (the company expects Q3), the annualized return is about 4-9%, and it carries the risk of failure from HSR antitrust review and the majority-tender condition.
• Judgment: watch only. The real significance is that it lifts M&A expectations across the entire autoimmune / clinical-stage biotech space.

⑪ SNDK SanDisk — 51 pts | Watch only

• Related news: lifted by the CXMT listing, +3.6% pre-market.
• Fundamentals: latest quarter (ended 2026-03-31) revenue $5.95B, +96.7% QoQ, +251.0% YoY, with quarterly net income of $3.62B. Whereas FY2025 (ended 2025-06) full-year revenue was only $7.36B with a net loss of −$1.64B.
• ⚠️ P/E 49.0 is distorted: EPS over the last four quarters was 23.03 / 5.15 / 0.75 / −0.16 — the TTM figure of 28.77 mixes in a loss quarter. Annualizing the latest quarter gives a P/E of about 15.6x.
• ⚠️ One data point that must be verified before use: TTM capex is only $0.18B, 1.4% of revenue; MU, also a memory maker, is at 28%. The plausible assumption is that wafer capacity is carried through the joint-venture structure with Kioxia and taken off balance sheet via the equity method, but the 10-Q footnotes were not checked to confirm this, so it is not used as a conclusion. Until verified, SNDK's FCF cannot be compared with MU's.
• ⚠️ New red flag — the credibility of the price data itself is in doubt: a 52-week range of $40.10 – $2354.39 (about 58x), alongside a $1436 share price and a share count of roughly 157 million, strongly suggests split adjustments or an incompletely adjusted price series from the early post-spin-off period. (SanDisk was spun off and listed from Western Digital, and price series in the early spin-off period are prone to distortion.) That range data is pending verification, and the conclusion of "−39.0% from the high" is therefore pending as well.
• Judgment: more extreme elasticity than MU, but lower data credibility. Earnings 8/5 after the close — the first new memory reading of this cycle.

---

6. Bearish / avoid list

⚠️ Note on deleted content: the first draft cited "$1.5 million of executive share sales" and "$8.9 million of insider selling" for CAPR and MXL as bearish evidence. Because the original SEC Form 4 filing links could not be obtained for verification, both items have been deleted entirely and are not used as evidence in this report.

Key reminder of this section: MXL is today's most memorable lesson — a "double beat + above-consensus guidance" can still drop 16-18%. This is the real-world footnote to "guidance matters more than a beat in the current quarter, and the expectation gap matters more than guidance". Before four trillion-dollar giants report this week, carve this into your mind. The same yardstick applies to MU — its Q4 guidance is below consensus.

---

7. Intra-theme rankings

Theme 1: the beneficiary chain of falling oil

Theme 2: super earnings week (ranked by "expectation-gap cushion", not by company quality)

Theme 3: the memory chain

---

8. Open-of-session verification signals

Pre-market (now → 09:30 ET)

• Watch the airlines, not tech: if UAL/DAL can hold +3% after the open on expanding volume, it means the "oil → profit" logic is being endorsed with real money; if they open high and fade, the market is treating the ceasefire as a one-day mood.
• Watch whether energy's decline widens: if XOM/CVX narrow their losses from −3% to −1%, it means oil is filling back in and the whole theme is a one-day affair.
• Are MU/SNDK gap-and-go or gap-fill: the memory chain's +2.5%/+3.6% are gaps that follow Asian sentiment, with no domestic US catalyst, and MU's guidance is in fact below consensus — the odds of a gap-fill are not low.

Intraday (09:30-10:00 ET)

• Sector ETF confirmation: for confirmation, watch whether JETS (airline ETF) outperforms SPY and whether XLE (energy) clearly lags. If only individual airline stocks rise while JETS does not move, it is a lone-runner move and not sustainable.
• Volume: today is the Monday before a super earnings week and the FOMC, so volume will most likely be light and wait-and-see. A rally on light volume is far less reliable than one on heavy volume.

Options sentiment

• IV for this week's four giants is already elevated. The IV crush risk after earnings is very high for AAPL/MSFT/META/AMZN — even if you get the direction right, buying options can still lose money.
• VIX has already fallen to 17.84. If it keeps drifting lower into the FOMC, the market may be too complacent about Wednesday.

Risk signals (contrarian)

1. Reversal after the gap: the whole market's higher open stems from a geopolitical headline at the level of a "pause", the type of catalyst that reverses most easily of all, and the duration of that pause and the schedule of follow-up talks have not been publicly disclosed.
2. Stock-bond divergence: today stocks and bonds are both up, which is healthy. If "stocks up, bonds down" (yields jumping) appears intraday, it means the market is repricing an FOMC hike.
3. ⚠️ Two macro logics are offsetting each other: falling oil → easing inflation pressure → reduces the Fed's urgency to hike (bullish); but a services PMI of 53.6 shows services inflation is still sticky (bearish). And the hike probability this report has in hand is stale data from before the oil plunge, so the true current pricing is unknown.
4. ⚠️ Symmetric scenario analysis of a hike (missed in the first draft): if the FOMC actually hikes on Wednesday, today's main line of "falling discount rate → Nasdaq leadership" reverses immediately, with the damage ordered ARM (P/S 56) > AAPL (PEG 2.68) > AMZN > META > MSFT, while airlines would also be weakened by rising debt costs. Whatever rises most today would fall most then.
5. This week's calendar density: Tuesday BA; Wednesday FOMC + MSFT/META/ARM; Thursday GDP + June PCE + jobless claims + AAPL/AMZN; Friday ECI + month-end rebalancing. The information load is concentrated in the last four days, and today lacks follow-on ammunition.

---

9. Final conclusions

① The 5 stocks most worth watching today

② Today's 3 strongest themes

③ Directions to avoid today + reasons

1. Energy stocks (XOM/CVX/FANG/DVN/APA) — falling oil directly pressures EPS, with no single-stock-level support today.
2. CAPR and names tied to the 7/29 FDA advisory committee — a binary regulatory event with extreme two-way volatility, unsuitable long or short; also note possible sector spillover to comparable DMD names.
3. AI power concept names (BE and others) — permitting and interconnection are external bottlenecks; and this is last week's news, with no incremental catalyst today.
4. Chasing any trillion-dollar giant into earnings — especially AAPL (pre-market price 0.30% from the high). IV crush makes it possible to "get the direction right and still lose money".
5. Chasing the memory chain's gap (MU/SNDK) — the catalyst comes from Asian sentiment with no incremental domestic US news; MU's Q4 guidance is in fact below consensus; and the same CXMT news was bearish on 7/16.

④ Final one-line judgment

Today is a sentiment repair ignited by two words — "strike pause": the direction is right and the logic is hard, but the fuel is shallow. The real pricing power sits with Wednesday's FOMC and the capex guidance of four trillion-dollar giants, and neither of those has happened yet.

This report's center of observation: the hardest logic is the airlines (but recognize their reversibility and unknown duration), the thickest expectation gap is MSFT/META (but you have to wait for earnings), the worst positions are chasing AAPL and energy, and the least complete information is the memory chain that has already gapped. Last Friday's MXL — "double beat + above-consensus guidance and still a sharp drop" — is the most important dress rehearsal of the week: at this level, the expectation gap matters far more than the results themselves.

---

Appendix: data gaps this report actively flags (the no-fabrication principle)

The following 12 items are things this report knows but could not verify. Any conclusion touching these data should be treated as pending verification.

A. Affecting core conclusions (high priority)

1. Missing a primary CME FedWatch snapshot from this morning — the 35-38% hike probability used in the report is taken from 7/23, i.e. before the oil plunge, and has almost certainly been repriced. This directly determines whether the first theme (falling discount rate → growth stocks) holds.
2. The exact crude level — sources conflict ($85.23 / $87.48 / $92.02 / "recovered to about $92"), belonging to different moments and contracts. Only the direction is confirmed, not the level.
3. The one-off item behind MU's FY26Q3 quarterly net margin of 68.1% — only 4.5pp below its TTM gross margin, structurally anomalous, suspected to be a large one-off tax item. If so, net margin 55.9%, P/E 20.8 and OCF/net income 1.02x all become invalid. The 10-Q must be checked.
4. The nature and size of the one-off item behind META's net income of only $2.71B in the quarter ended 2025-09-30 — cannot be confirmed from the three statements. It contaminates four items: TTM net income, TTM EPS, trailing P/E and OCF/net income. The "normalized P/E of about 17x" rests on an unverified assumption.

B. Affecting single-stock judgments

5. The accounting cause of SNDK's extremely low capex (TTM $0.18B, 1.4% of revenue) — the plausible assumption is off-balance-sheet treatment via the Kioxia joint-venture structure, not verified against the 10-Q footnotes. Until verified, SNDK's FCF cannot be compared with MU's.
6. Whether SNDK's 52-week price range ($40.10–$2354.39, about 58x) is fully adjusted — suspected distortion in the early post-listing spin-off data, so "−39.0% from the high" is pending as well.
7. Whether ARM's forward EPS of $3.076 is on a GAAP or non-GAAP basis — judged to be non-GAAP and corresponding to FY2028, and the corresponding GAAP estimate was not found, so the forward P/E of 84.5x is severely understated.
8. CAPR's pre-market decline of −40% vs −65% — this is a difference in basis rather than an error (pre-market vs previous close vs two-day cumulative) and cannot be resolved. Also, the exact release time of the FDA briefing document is unverified (7/24 or 7/27); if it was 7/24, it should not be listed as today's new news.
9. MXL's decline (16% / 17.9% / 18.65%) and BE's down day (Thursday / Friday) differ across sources — only the fact of "a sharp drop" is confirmed.

C. Method and source level

10. The data vendor, GAAP basis and as-of date of the consensus figures (consensus / forward EPS / PEG) — all come from yfinance aggregate fields, a mixed basis across multiple brokers, not cross-checked item by item against the analysts' original notes.
11. The pre-market index snapshot was taken across 08:10–08:50 ET, a 40-minute span rather than a single-moment cross-section; precise per-item timestamps and source links could not be obtained. Several hard numbers (the Azure bogey of 36%, AAPL broker price targets, BA's 33% upside) are media paraphrases, with no original research note / IR document obtained.
12. The duration of the US-Iran "strike pause" and the milestones of follow-up talks — not publicly disclosed. This is the only falsifiable anchor for judging the persistence of the report's first theme, and we do not have it.

Content deleted from the report: the insider-selling amounts for CAPR and MXL ($1.5 million / $8.9 million), because the original SEC Form 4 filing links could not be obtained for verification. Risks this report does not cover: the impact of this week's Treasury auctions and refunding schedule on the 10Y yield; month-end rebalancing flows on 7/31.

---

Data sources

• Bloomberg — Oil Tumbles as US and Iran Pause Military Strikes: Markets Wrap¹
• Washington Post — Oil prices ease after US and Iran pause their attacks²
• Yahoo Finance — Stock market today: Monday July 27¹⁵
• CNBC — Stocks making the biggest moves premarket: MU, FBRX, XOM, BKR⁵
• CNBC — CXMT surges 466% in Shanghai to become China's most valuable company⁴
• argenx — Official press release: acquisition of Forte Biosciences³
• StockTitan — FBRX 8-K filing¹⁶
• Investing.com — Capricor Therapeutics stock tumbles after FDA briefing documents⁷
• Forbes — Markets Price In Rising Odds Of July Fed Rate Hike (7/23, before the oil plunge)⁸
• CNBC — Stock market next week: Outlook for July 27-31, 2026⁹
• Seeking Alpha — RingCentral jumps 25% on Q2 beat, higher outlook¹⁰
• Seeking Alpha — MaxLinear slumped 16%, extending a two-week slide¹¹
• Benzinga — What's Going On With Bloom Energy Shares?¹²
• TradingKey — US Pre-Market: index futures rise as falling oil lifts tech and airlines⁶
• TradingEconomics — US 10 Year Treasury Yield¹³
• TipRanks — Boeing Will Report Q2 Earnings on July 28¹⁷
• ChinaTechNews — What's Going on With Apple Stock Monday (AAPL pre-market +0.29% @ $334.00, citing Benzinga Pro)¹⁸
• Micron IR — Q3 FY2026 official press release (SEC EX-99.1)¹⁹
• 24/7 Wall St — SK Hynix and SanDisk Sink 7%, Micron Falls 5% as China's CXMT Readies an $8.6B Memory IPO (7/16, counter-evidence)²⁰
• 24/7 Wall St — American Airlines Sinks 5%, United Falls 4% as Crude Oil Jumps (7/8, reverse verification of the oil beta)²¹
• Fundamental data: pulled live from yfinance (retrieved 2026-07-27 12:49–13:20 UTC, prices and financials all on a 7/24 close basis)

---

⚠️ Risk disclaimer: this list is only a pre-market information review and watchlist and does not constitute investment advice. US equities carry high volatility and pre-market gap risk, and post-earnings IV crush and guidance reversals are possible; auto-generated content may contain stale information or factual errors. Rely on company disclosures / SEC filings, and do not use this directly as a basis for trading.