2026-10-29 17:00 ET: Postclose movers (CVNA, CMG, MO, DASH; GE/LOW/HD/CRM/GILD fades) and Friday gap risk from tonight's AMC prints

Slot: Thu 2026-10-29 17:00 ET hourly research. Status: written 17:00–17:30 ET while after-hours trading was still going. Screening/research, not investment advice.
TRAI data used:

• /api/latest postclose run 2026-10-29_postclose_2026-10-29T20-34-14-137Z (4:34 PM ET; regime transitional, VIX 17.20, risk/trade 0.25%).
• /api/events (generated 2026-10-29 11:42 UTC).
• /api/correlations (60d to 2026-10-22, still stale, see S-068).
• /api/scorecard replay trades 2026-04-23 → 10-21 (asOf 10-22).

After-hours prices: pulled from the Yahoo chart API (5-min bars incl. post-market) at ~17:02 ET. They are thin, preliminary and still moving.
Previous slot: 16:00 report (VRT/ORCL/STX/NOW/META, INTC/AMZN).

1. Focus and why

The 4:34 PM postclose run changed which names TRAI is looking at. Every name in long_momentum is an earnings-reaction trade: CVNA (+15.68%) and CMG (+8.88%) reported Wed 10-28 AMC, and MO (+5.15%) and DASH (+5.61%) reported this morning BMO. Short_momentum is empty. Mean_reversion has fade-shorts in GE, LOW, HD and GILD and a fade-long in CRM.

GILD is a special case. It is a fade-short candidate and it reported at 4:02 PM tonight, beating estimates and raising guidance. This hour's questions:

• What drove each mover?
• Which of tonight's AMC prints (SBUX, PINS, RBLX, COIN, MSTR, RIVN, MDLZ, GILD, plus the INTC/AMZN follow-up) and Friday's BMO prints (XOM, CVX, ABBV, REGN) change gap risk for those candidates?

2. Businesses (one line each; detail in company notes)

• CVNA: online used-car retailer with in-house reconditioning, financing and logistics. note
• CMG: Chipotle, a fast-casual Mexican restaurant chain (company-owned). note
• MO: Altria, US cigarettes (Marlboro), oral nicotine (on!) and an ABI stake. note
• DASH: DoorDash, local-commerce delivery marketplace (restaurants, grocery, retail; owns Deliveroo). note
• GE: GE Aerospace, commercial and defense jet engines plus aftermarket services. note
• HD / LOW: the two largest US home-improvement retailers. HD · LOW
• CRM: Salesforce, CRM/enterprise SaaS (Agentforce AI). note
• GILD: Gilead, HIV franchise (Biktarvy, Descovy, Yeztugo/lenacapavir), liver, oncology (Trodelvy, cell therapy). note
• Tonight's AMC names:
• SBUX (coffeehouses) note
• PINS (visual-search ads) note
• RBLX (user-generated gaming platform) note
• COIN (crypto exchange/custody) note
• MSTR (Strategy, a bitcoin treasury company) note
• RIVN (EV maker) note
• MDLZ (global snacks) note
• Friday BMO: XOM, CVX (integrated oil) XOM · CVX; ABBV (immunology/neuroscience/aesthetics pharma) note; REGN (Dupixent/Eylea biotech) note.

3. Today's long_momentum movers: catalysts

_(TRAI /api/latest 4:34 PM run.)_

CVNA: Q3 2026, reported Wed 10-28 after the close

• Results vs consensus:
• Revenue $8.04B (+42%) vs $7.67B consensus.
• Retail units 217,385 (+39%), the first quarter above 200k.
• Adj. EBITDA a record $800M.
• Adj. EPS $0.39 vs $0.44 (a miss).
• Source: Investing.com, 2026-10-29; StreetInsider, 2026-10-29.
• Guidance: FY26 adj. EBITDA now expected at the high end of $2.7–3.0B. Q4 retail units guided up sequentially. Net debt/EBITDA is 0.8x, and refinancing $1.7B of secured notes saves about $45M a year (Motley Fool transcript, 2026-10-29).
• Weak spot: adj. EBITDA margin was about 10% vs 11.3% a year ago (same transcript).
• Analysts: Needham target cut to $100 from $120, BTIG cut to $82 from $87, Jefferies reiterated $88 (Investing.com, above). Morgan Stanley named it a "Top Pick", target $88 (TipRanks, 2026-10-29).
• After hours ~17:00: −0.13% (Yahoo).
• TRAI data gap: /api/events shows CVNA with next = null, and its history ends at 2026-07-29. The 10-28 print is not recorded, the same failure mode as NOW in S-068. CVNA's last 4 prints: avg |gap| 8.54%, fill 25%, avg day-1 in gap direction −1.25%.

CMG: Q3 2026, reported Wed 10-28 after the close

• Results vs consensus:
• Comparable sales +2.1% vs ~1.5%; transactions +0.8%, the third straight positive quarter.
• Revenue $3.29B (+9.6%), in line.
• Adj. EPS $0.27 vs $0.29 (a miss).
• Restaurant-level margin 23.3% (−120bp).
• Source: CNBC, 2026-10-28; Bloomberg via CNBC TV18, 2026-10-29.
• Guidance: Q4 comps of about 2%. October was "choppy", with brisket launching Nov 5. About 3% menu pricing in Q4. FY26 outlook held. BMO, RBC, Guggenheim, KeyBanc and Jefferies all cut price targets (Investing.com, 2026-10-29).
• After hours: −0.58%.
• TRAI data gap: CMG also shows next = null, with history ending 07-29. The 10-28 print is not recorded. TRAI has only 3 CMG prints; day 2–5 drift after the July print was −12.49%.

MO: Q3 2026, reported Thu 10-29 at 7:00 AM ET

• Results vs consensus: adj. EPS $1.52 (+4.8%) vs $1.51 consensus. Net revenues $6.132B, and revenues net of excise taxes $5.441B (+3.6%).
• Guidance and capital return: FY26 adj. EPS raised to $5.66–5.74 (prior $5.61–5.72, MO note). The board authorized a $1B expansion of the buyback, to Dec 2027 (Altria IR, 2026-10-29; 10-Q).
• Midpoint vs consensus: the $5.70 midpoint compares with $5.68 consensus (Investing.com, 2026-10-29).
• Call details:
• Domestic cigarette volume −2.9% (−4% adjusted).
• on! shipments +13.3% to 47.8M cans, reversing Q2's −4.2%.
• A $124M duty-drawback revenue benefit.
• Source: Motley Fool transcript, 2026-10-29.
• Reaction: close-to-close +5.15% with gap +3.15%, and it closed at the high (rangePos 0.98). After hours +0.21%.
• History: TRAI's last 4 MO prints show fill 0/4, continuation 75%, so MO's gaps tend to extend.
• Sympathy: MDLZ is +3.15% after hours on its own beat (ρ 0.60 to MO, cluster 4), a mild tailwind.

DASH: Q3 2026, reported Thu 10-29 before the open

• Results vs consensus:
• Revenue $4.62B (+34%; +23% ex-Deliveroo) vs $4.56B.
• EPS $1.06 vs $0.81.
• Orders 993M (+28%), Marketplace GOV $33.9B (+36%), adj. EBITDA $1,060M.
• Q4 guide: GOV $35.4–36.4B, adj. EBITDA $0.9–1.1B.
• Source: DoorDash IR, 2026-10-29; MarketWatch, 2026-10-29; 8-K.
• Instacart deal: DoorDash agreed to acquire Maplebear (Instacart) for $65.00/share, at least 75% cash. MarketBeat values it at about $17B, with close expected late 2027 or early 2028 (MarketBeat, 2026-10-29; the deal value and close timing are secondary-source figures).
• Reaction: opened 179.13 vs a 169.49 prior close and closed 179.00 (gap +5.33%, so roughly flat open→close). After hours +0.58%.
• History: TRAI's last 4 DASH prints show avg day-1 in gap direction −1.04% and continuation 25%.

TRAI replay read on this bucket (my calc, /api/scorecard replay trades, net of 5bp + impact)

• long_momentum signals on an earnings-reaction day: n 48, −0.56%/trade, hit 44%, t −1.18, vs +0.05% for the whole bucket (n 385). That is S-066's evidence.
• All four of tonight's long_momentum names are reaction-day signals. CVNA is also a ≥+10% day, which falls under the S-066 guard (§6.2 of the 16:00 report: −1.00% open→close since 2024 for big AI-infra names).

4. Mean-reversion bucket: what drove the moves

• GE +3.96% (fade-short; z1d 2.53, relVol 1.12). I found no company-specific catalyst. JKN reports the move came "without a specific confirmed company catalyst" and notes peers diverged (FTAI surged; Honeywell was pressured by supply constraints) (JKN, 2026-10-30 KST). GE last reported Q3 on 10-20 BMO (GE Aerospace, 2026-10-20). Next print 2027-01-21 (TRAI). After hours −0.30%.
• LOW +4.48% and HD +3.60% (fade-shorts). No company news; this was a sector move. Freddie Mac's 30-year rate fell to 7.24% from 7.32% (Freddie Mac, 2026-10-29). MarketWatch shows both outperforming a down tape (HD, LOW). A TradingKey market-movers blurb credits LOW's "early Black Friday promotions"; that is low-quality and unverified. HD–LOW ρ 0.92 (cluster 12), and TRAI already flags them "Consumer Disc. ×2", so they are one position. Next prints: HD 11-17 and LOW 11-18 (TRAI). After hours flat.
• CRM −4.63% (fade-long). This was NOW sympathy. StreetInsider ties the drop to ServiceNow's Q4 cRPO guide (18.5% vs ~21% expected) hitting the sector. The article also cites Agentforce monetization and the pending $5.66B Braze deal (StreetInsider, 2026-10-29). NOW–CRM ρ 0.75 (cluster 6). After hours −1.14%. In the replay, CRM's last 4 prints were 4/4 positive day-1, but CRM is not reporting until ~12-02 (estimate).
• GILD +4.03% (fade-short; z1d 2.28, relVol 1.44, closed at the high, rangePos 0.98). GILD ran up into its own 4:02 PM print. The print was a beat-and-raise (§5). After hours +2.17% (169.20), which is more than GILD's historical avg |gap| of 1.68% (TRAI events).

5. Tonight's AMC prints: results and Friday gap set-up

After-hours moves are Yahoo quotes at ~17:02 ET. Base rates are TRAI /api/events, last 4 prints.

GILD: Q3 2026 (4:02 PM)

• Results vs consensus:
• Product sales $8.29B (+13%); ex-Veklury $8.2B (+16%).
• HIV ~$6.2B (+18%); Biktarvy ~$4.1B (+10%); Descovy ~$1.0B (+49%); Yeztugo $305M; Trodelvy $506M (+42%).
• Non-GAAP EPS $2.55 vs $2.17 consensus (Investing.com).
• Source: Gilead release via StockTitan, 2026-10-29; Business Wire; Investing.com, 2026-10-29.
• FY26 guide:
• Product sales raised to $31.0–31.3B (from $30.1–30.4B) vs $30.77B consensus.
• Non-GAAP EPS guidance moves to a loss of $(1.20)–(0.85), from $(0.65)–(0.30), reflecting about $12.4B of acquired IPR&D (Arcellx, Tubulis, Ouro, PepStar).
• Investing.com compares the EPS guide with a −$0.45 consensus. Because of the IPR&D charges, that headline "EPS guide miss" is not about the operating business.
• Other items:
• Medicare's Biktarvy Maximum Fair Price takes effect 2028-01-01.
• Cash and equivalents $6.4B.
• Read: the fade-short thesis (an ordinary-volume move that should revert) has been overtaken by new fundamental information. The after-hours gap is in the opposite direction to the short.
• Base rates: GILD's last 4 prints show day-1 in gap direction −0.68% and fill 50%. Day-1 drifts were +3.21%, +4.63%, +1.02% and −3.12%, so mixed.
• Replay: zero mean_reversion trades entered into the same name's own print (my calc). TRAI has never been tested on this situation (S-071).

MDLZ: Q3 2026 (4:05 PM)

• Net revenue $9.967B (+2.3%), organic +2.1% with vol/mix +1.8pp. Adj. EPS $0.78 (+5.5% at constant FX).
• FY26 reaffirmed: organic ≥2%, adj. EPS flat to +5% at constant FX, FCF about $3B (MDLZ via GlobeNewswire, 2026-10-29). Consensus EPS was not verified this hour.
• Sympathy: cluster 4 links to MO (ρ 0.60), PEP 0.68, PG 0.64, KO 0.63.

SBUX: FQ4 2026 (quarter ended Sep 27)

• Results vs consensus: adj. EPS $0.80 vs $0.72. Revenue $9.14B vs $9.23B (−4%, China joint-venture transition). Global comps +7.2% vs 6%, US comps +7.7% vs 6.5%, international comps +0.9% vs 3.2%.
• FY27 guide: adj. EPS $3.15–3.25 vs $3.14 consensus, but revenue −1% vs +1.4% expected (includes a 53rd week and the China joint venture) (CNBC, 2026-10-29; Starbucks IR).
• After hours −1.57%. SBUX–CMG ρ only 0.20, so the read-through to CMG's long is weak. SBUX–MCD is 0.43.

PINS: Q3 2026

• Results vs consensus: revenue $1.17B (+12%) vs $1.21B, adj. EPS $0.46 vs $0.49, MAU 661M.
• Q4 guide: revenue $1.425–1.475B (+8–12%), adj. EBITDA $538–588M (Investing.com, 2026-10-30 IST).
• After hours −12.25%. PINS's last 4 prints: fill 0/4, avg |gap| 16.9% (4.05 ATR).
• Sympathy (cluster 14): UBER 0.61, SNAP 0.59; also AFRM 0.54. PINS–DASH ρ is only 0.34; DASH's top links are AFRM 0.54, NFLX 0.50 and UBER 0.47. SNAP is −2.99% after hours.

RBLX: Q3 2026

• Results vs consensus: revenue $1.57B vs $1.46B, but DAU 138M vs 145.77M. Bookings $1.70B (−12% y/y; above $1.63B est).
• Q4 guide: bookings $1.89–2.00B vs $2.01B, and net loss $363–423M (Investing.com, 2026-10-29).
• After hours −18.74%. Its last two prints went −18.3% and −26.9% close-to-close (TRAI events).
• RBLX is a singleton (cluster 51); top ρ COIN 0.50, AAPL 0.49.

COIN and MSTR (crypto cluster 9: COIN–MSTR 0.87, HOOD 0.73, MARA 0.62–0.73)

• COIN: EPS $2.51 vs −$0.26 estimate, revenue $1.2B (in line), no guidance. Trading-volume share 10.7%, USDC held $22B (Investing.com, 2026-10-29).
• MSTR: EPS $44.75 vs $13.86 est., driven by a $20.91B digital-asset gain. Revenue $125.4M vs $128.0M. 848,517 BTC held at 10-25. The CFO steps down 11-05 (Investing.com, 2026-10-29; Strategy, 2026-10-29).
• After hours COIN −3.48%, MSTR −3.21%, HOOD −3.92%. That move matches BTC weakness (CoinDesk live blog: bitcoin "clings to $80,000", 2026-10-29) more than either print. None of these names is in a TRAI bucket.

RIVN: Q3 2026

• Results vs consensus: adj. loss/share $0.48 vs $0.62 expected, revenue $2.16B vs $1.99B, record 19,248 deliveries.
• FY26 guide: deliveries 67–70k (from ≥65k), adj. pre-tax loss $1.75–1.85B ($100M better at the midpoint), capex $1.4–1.5B (−$300M) (CNBC, 2026-10-29).
• After hours +4.10%. Links: COIN 0.60, TSLA 0.50.

INTC and AMZN: updates to the 16:00 report

• INTC: Reuters adds that an explosion at the Kiryat Gat (Israel) fab will cost about $600M of revenue over two quarters: Q4 revenue −$300M and gross margin −200bp, with a similar hit in Q1. Q4 PC revenue is expected down mid-20s%. Q4 EPS guide $0.25 vs $0.43 (LSEG) (Reuters via StreetInsider, 2026-10-29). After hours −16.72% at 17:02 (it was −18.1% at 16:12).
• AMZN: after hours −6.28% at 17:02 (−7.6% at 16:12). Seeking Alpha's headline blames "sales and capex guidance" (SA, 2026-10-29; headline only, body not read). Whether 2027 capex was quantified on the call is unverified.
• Cluster-1 after hours at 17:02: VRT −4.44%, ORCL −3.10%, MU −2.81%, ETN −1.45%, NVDA −1.43%, SNDK −1.53%. TipRanks headlines SNDK EPS of $49.51 as a beat (headline only; details unverified).

6. Friday BMO prints (Fri 10-30)

• XOM/CVX: small gaps that usually fill (TRAI). BofA points to refining margins (+60%) and global gas (+38%) offsetting a 7% lower oil price in Q3 (TipRanks, above). Cluster-2 sympathy: XOM–CVX 0.86, COP 0.81–0.82. Dec Brent expiry is also 10-30 (S-059/S-060). No energy names are in today's buckets.
• ABBV: the last 4 prints all gapped down (−2.3%, −7.0%, −2.9%, −1.9%). It is in cluster 3 with GILD (ρ 0.59). Its morning print lands on the same session as GILD's reaction.
• REGN: singleton-ish (cluster 46; PFE 0.54). Watch Dupixent and Eylea HD share; FDA action on cemdisiran is due in November (Investing.com preview).
• Also Friday: Q3 ECI 8:30, Chicago PMI 9:45, month-end, NVT BMO (16:00 report §7).

7. Analysis

1. The long_momentum bucket is 100% earnings-reaction trades. Those trades lost money on average in replay (−0.56%, n 48). Two of the four (CVNA, CMG) were EPS misses that rallied on top-line growth and guidance, and analysts cut targets on both. That is a fragile set-up for day 2: CMG's last day 2–5 drift was −12.5%, and CVNA's day-1 in gap direction averages −1.25%. MO is the strongest of the four on TRAI's own data: its print gaps have filled 0/4 times, it closed at the high, and MDLZ is +3% after hours in the same cluster. DASH is in the middle: a clean beat, but a large acquisition overhang and a day-1 base rate of −1.04%.
2. GILD is the clearest Friday conflict. TRAI's postclose run picked GILD as a fade-short at 4:34 PM, 32 minutes after a beat-and-raise release the engine does not see. The after-hours quote is +2.2%, above GILD's typical |gap|. The fade-short premise is gone. Its cluster-3 peer ABBV reports before Friday's open, which adds a second healthcare catalyst to the same session.
3. HD/LOW is a rates trade on ordinary volume, with no news. These are textbook MR fades. Count them as a single position (ρ 0.92). The 30-year mortgage rate moved by only 8bp, so the catalyst is weak.
4. CRM's fade-long runs into a negative software/megacap after-hours tape. AMZN is −6%, CRM −1.1% after hours, and NOW–CRM ρ is 0.75. On the other side, CRM had no own-news and fell on ordinary volume (relVol 1.06), which suits a fade.
5. The biggest after-hours moves (RBLX −19%, INTC −17%, PINS −12%) are outside TRAI's buckets. Their sympathy links don't touch the candidates directly: PINS–DASH is only 0.34, RBLX is a singleton, and INTC's cluster 1 contains none of tonight's candidates. GE–ETN is only 0.42.

8. Actionable for Fri 10-30

ACTIONABLE: YES, on one name, plus sizing caveats.

• GILD (mean_reversion fade-short): stand down. GILD reported a beat and raise at 4:02 PM: product sales guide raised to $31.0–31.3B, EPS $2.55 vs $2.17. It trades +2.2% after hours, above its 1.68% average print gap. The engine's fade signal was computed without seeing the print, and replay has no precedent (n 0) for fading into a name's own print (S-071). ABBV, a cluster-3 peer (ρ 0.59), also reports BMO. If the 9:25 indicated gap is still > +1%, skip the short.
• long_momentum (CVNA, CMG, MO, DASH): all are earnings-reaction signals, so apply S-066. Use opening-range confirmation or half size. CVNA is a ≥+10% day and both CVNA and CMG missed EPS; they are the weakest set-ups. Prefer MO: TRAI print history shows 0/4 fills, it closed at the high, and MDLZ is +3.15% after hours.
• HD + LOW fade-shorts: treat as one position (S-001; ρ 0.92).
• CRM fade-long: after hours −1.1% with AMZN −6%. Re-check the 9:25 indicated gap (S-067 logic applied to cluster 6).
• No direct sympathy risk from RBLX, PINS, COIN or MSTR to any candidate.

9. Strategy implications (proposals; see strategy.md)

• S-071: an own-earnings guard for mean_reversion. Skip or flag fade candidates whose own print falls before the entry session (GILD case).
• S-072: a reaction-day concentration flag. When at least half of a bucket is earnings-reaction trades, show a warning and apply S-066 bucket-wide. Also tag prior-evening AMC prints that /api/events missed (CVNA, CMG; extends S-068).
• S-073: post-close after-hours sympathy re-check for MR and momentum candidates across all clusters (MDLZ→MO, AMZN/NOW→CRM). Generalizes S-067.
• S-074 (NEEDS DATA): day 2 after an "EPS miss but top-line/guide beat" gap-up (CVNA, CMG). Is it weaker than after a clean beat?
• S-075 (NEEDS DATA, from whiggity): realized/implied + CLV short, staple day-2 drift, XOM–CVX gap divergence, the COIN/MSTR residual against BTC, and gap retention across 8:30 macro prints (§11).

10. Unknowns and open questions

• Final after-hours and premarket levels. Every after-hours number here is a thin ~17:02 quote.
• MDLZ's consensus EPS; whether AMZN quantified 2027 capex; SNDK's full results.
• Whether Investing.com's GILD EPS-guide "consensus −$0.45" already included the PepStar IPR&D charge.
• A real catalyst for GE's +3.96%: none found.
• The LOW "Black Friday" claim (TradingKey): unverified.
• /api/correlations, /api/scorecard and /api/regime are still at asOf 10-22 (S-068).
• /api/events is missing the 10-28 AMC prints for CVNA and CMG (and NOW, per the 16:00 report).

11. Research partner (whiggity)

_Source: whiggity (Monica custom bot), reply received ~17:10 ET in the chat "Post Earnings Movers and Friday Gap Risk". whiggity did not confirm actual results; its catalysts were asserted, not sourced, and its links were undated quote pages. Each point below was checked against a cited public source or TRAI data at ~17:15 ET. Not independently authoritative._

Verified:

• XOM call at 9:30 AM ET Fri 10-30, overlapping the open. ExxonMobil IR lists the Q3 call for 8:30 to about 9:30 AM CT, i.e. 9:30 to 10:30 ET, with slides posted about 5:30 AM CT (ExxonMobil IR). whiggity cited a Yahoo quote page; the IR page is the primary source.
• Q3 ECI at 8:30 AM ET Fri 10-30 (BLS ECI schedule).
• Chicago PMI at 9:45 AM ET Fri 10-30 (MNI publication calendar).
• Fri 10-30 is the last trading day of October (Oct 31 is a Saturday), so month-end rebalancing and MOC flows fall on Friday. (The 16:00 report §10 had already noted this.)
• CVNA's move beat the options-implied move, but the implied figure was ±13.4%, not ~12%. EarningsWatcher, updated 2026-10-29, gives ±13.4% frozen at the last pre-print reading. CVNA's actual move was +15.68% close to close (TRAI), so realized/implied ≈ 1.17 (EarningsWatcher). whiggity's ~12% came from an undated Perplexity Finance page and is not used.
• Next week's calendar:
• ISM manufacturing Mon Nov 2 and ISM services Wed Nov 4 (ISM release calendar).
• October payrolls Fri Nov 6 at 8:30 (BLS November schedule).
• ADP Wed Nov 4 was not checked.
• Sympathy pairs that TRAI can check (60d ρ, asOf 10-22, stale per S-068):
• MO–PM 0.69 and XOM–CVX 0.86 are strong.
• DASH–UBER 0.47 is moderate.
• CMG–SBUX is only 0.20, so SBUX's print tonight is a weak read-through to CMG.
• KMX, CARG, AN, CAVA, SG, WING, SHAK, CART, LYFT and BTI are not in TRAI's universe.

Conflicts (whiggity was wrong or out of date):

• PCE "Fri 10-30 8:30 with personal income": wrong. September Personal Income & Outlays (PCE) came out Thu 10-29 (CNBC, see 16:00 report §1). BEA's schedule shows the next release (October data) on Nov 25 (BEA schedule). Friday's 8:30 print is the ECI only.
• "FOMC Nov 5/6": wrong. The FOMC met Oct 27–28 and held at 3.75–4.00% (Fed statement, 2026-10-28). The next meeting is Dec 8–9 (Fed calendar). Nov 6 is payrolls, not the Fed.
• "UMich final 10:00 Fri": wrong. The October final came out Oct 23, and the November prelim is due Nov 6 (UMich 2026 release dates).
• "AAPL/AMZN contagion tonight": only half right. AMZN did report tonight (§5). TRAI events show AAPL on Nov 2 AMC, which repeats the 16:00 consult's error (16:00 report §10).
• CVNA consensus "EPS $0.49 on $7.62B" (unsourced) does not match our sourced consensus of $0.44 EPS on $7.67B (Investing.com, §3). Not used.

Unverified, carried as open questions or proposals (see S-075 in strategy.md):

• (a) Earnings-move and closing-location test. If realized/implied > 1.25 and Thursday CLV < 0.5, does shorting a break of Friday's opening-range low work? This doesn't trigger on any current candidate: CVNA's ratio is ≈ 1.17 with rangePos 0.83, and TRAI has no implied-move data for CMG, MO or DASH.
• (b) Low-beta staple, earnings day > 3× ATR, then day-2 drift. Is day 2 positive, so that buying a VWAP pullback works? This is relevant to MO, and TRAI's 0/4 fill history supports it. It overlaps S-066 and S-074.
• (c) XOM–CVX co-reporting gap divergence. Does the gap between the two (both report BMO Fri) mean-revert 30–50% by 11:00? A related idea, CMG's reaction against SBUX's gap, is weak at ρ 0.20.
• (d) COIN/MSTR residual vs overnight BTC. Does the part of their gap not explained by BTC fade in the first hour?
• (e) Premarket gap retention across 8:30 macro prints (ECI Friday).
• Mechanisms with no source:
• Gap-and-go setups that close near the high get a small follow-through gap, then fade from 10:00 to 11:30.
• Friday weekly-option expiry drives call monetization and delta unwind in CVNA and CMG.
• MO, as a low-beta staple, keeps "continuing steadily".
• Per-name gap-risk metrics. The ones listed (COIN take rate/USDC, MSTR BTC yield, PINS ad guide, RBLX bookings/DAU, RIVN gross profit, SBUX comps, MDLZ cocoa, GILD Biktarvy/lenacapavir, XOM/CVX upstream/refining/buybacks, ABBV Skyrizi/Rinvoq, REGN Eylea HD) are reasonable watch items. Most are already covered in §5–§6 with sourced figures.

Effect on §8: none. Everything verified (ECI, Chicago PMI, month-end, the XOM call timing) was already in §6, or touches only XOM/CVX, which are not in TRAI's buckets. The 9:25 indicated-gap re-checks already come after the 8:30 ECI.