2026-10-23 13:00 ET — Payments & card/fintech: AXP's print and next week's V/MA/PYPL/SOFI/HOOD/COIN reports

Slot: Fri 2026-10-23, 13:00 ET hourly research. Status: complete; second-researcher consult (whiggity, ~13:05 ET) in §7; one claim rejected (Fiserv date).

TRAI data used: /api/latest (run 2026-10-22_premarket_2026-10-23T12-53-19-641Z, 8:53 AM ET; daily stats = Oct 22 session), /api/events (asOf 2026-10-23), /api/correlations (60 sessions to 2026-10-22), /api/regime, /api/macro, /api/scorecard. All returned 200 at ~13:05 ET. Intraday prices: Yahoo chart API 5-minute bars, last bar ~13:07 ET (my pull); "%" = vs the Oct 22 close. Peer-reaction history: Yahoo daily bars (my calc). SEC EDGAR submissions API checked for AXP at ~13:10 ET.

1. Focus and why

• Focus: the payments / card / consumer-fintech group. AXP reported this morning; six universe names report within four sessions: SOFI and PYPL Tue 10-27 BMO, V and HOOD Tue 10-27 AMC, MA Thu 10-29 BMO, COIN Thu 10-29 AMC (TRAI /api/events; HOOD date confirmed by Robinhood IR release, 2026-10-01).
• Rotation: the 09:00 report logged AXP's headline numbers only; the 10:00–12:00 reports covered software, telecom/energy and AI power. None covered the networks or fintechs.
• One-line read: AXP is the day's weakest large financial (−4.15% at 13:07, no gap fill) and dragged card peers in sympathy; the trade is next week's prints, where PYPL carries the biggest historical gap risk in the group (avg |gap| 9.43%, 3.42 ATR, 0/4 filled).

2. AXP — what happened

• Business: card issuer and closed-loop network for affluent consumers and businesses; revenue from discount fees, card fees and net interest (Reuters via Investing.com, 2026-10-23).
• Q3 (Oct 23 BMO): adjusted EPS $4.68 vs ~$4.51 consensus; GAAP EPS $5.42 includes a $982M pre-tax gain on selling its stake in a corporate-travel business; revenue $20.06B vs $20.12B est. (small miss); billed business $454B (+8%); net write-off rate 1.8% (from 1.9%); effective tax rate 18.1% vs 24.1% a year ago on discrete benefits; FY26 adjusted EPS guide $17.60–17.90 (low end raised, top kept) (24/7 Wall St., 2026-10-23 9:51 ET). Total network volume $517.9B vs $519.5B est.; ~10% FY revenue-growth target reaffirmed; beat partly from lower provisions and tax (Investing.com, 2026-10-23 9:18 ET). Plans to cut ~5% of global workforce (~3,500 jobs); expenses $14.5B (+9%); stock −12% YTD before the print (Reuters, above). Call coverage mentions regulatory remediation costs (Investing.com transcript summary, 2026-10-23) — details unverified.
• Filings: 8-K (items 2.02, 7.01, 9.01) and 10-Q both filed 2026-10-23 (EDGAR AXP filings).
• Price action (Yahoo 5m): open −3.12%, low −4.22%, high −1.21%, last $311.65 (−4.15%) at 13:07. TRAI's 8:53 indicated gap was −2.43% (≈1.37 ATR vs its 0.81-ATR earnings norm). Consistent with AXP's history: 0/4 gaps filled, 75% continuation (TRAI /api/events).
• Read: the market treated it as a low-quality beat (tax + provisions + one-off gain, revenue/volume miss) despite raised guidance — "beat, sell" has now happened on 3 of the last 4 AXP prints (Apr: +7.18% surprise, −4.31% c-c; Jul: +2.84%, −4.30%; today) (TRAI /api/events).

3. Sympathy today (Yahoo 5m, ~13:07 ET)

• Broad tape: S&P 500 −0.3%, Dow −1% (−541 pts) at 12:33 ET on rising oil prices; Nasdaq +0.1% (AP, 2026-10-23 12:50 ET). Financials (XLF −1.33%) are underperforming SPY by ~1 pp, so not all of the peer weakness is AXP read-through. AXP is a Dow component, which explains part of the Dow's underperformance (my inference).
• Earlier in the morning V and MA were only −0.6%/−0.5% (24/7 Wall St., above); both have widened to −1.4%/−1.9% by 13:07.
• BX (today's long_momentum candidate) is −1.73%, giving back part of yesterday's +8.38% post-earnings move, in line with the S-004 peer-event flag raised at 09:00.

Do AXP prints move the networks? (Yahoo daily, my calc, n = 4 — anecdotal)

\*V reported 2026-01-29 AMC, so that day is V's own reaction. Same-day sign agreement AXP→V/MA is 2/3 on clean days; next-day V/MA was positive in all 4 cases (avg V +1.64%, MA +1.66%). Far too few to trade; logged as question S-033.

4. Next week's reporters — business, setup, history

V — Visa (Tue 10-27 AMC, fiscal Q4)

• Business: open-loop card network; earns per-transaction fees on payments volume and cross-border; takes no credit risk (24/7 Wall St., above, on the network vs issuer distinction).
• Expectations: EPS est. $3.44, revenue est. $12.1B (Investing Calendar, accessed 2026-10-23). Oppenheimer models $3.46 EPS and $12.16B revenue vs $12.09B consensus (FINWIRES, 2026-10-08). Goldman is constructive on V, MA and PYPL into Q3, citing resilient September spending (Finvaulta summary of GS note, 2026-10-13) — secondary summary, not the note itself. Fiscal-2027 guidance will be the main swing factor (fiscal Q4 report); exact current guide not verified here.
• TRAI history (last 4): avg |gap| 2.94% (1.64 ATR), median 1.44%, max 8.76%; fill rate 75%, continuation 25%, avg day-1 in gap direction −1.38% — V's gaps tend to fade. ATR 1.79%, beta60 0.25. V–MA ρ 0.88 (cluster 28).

MA — Mastercard (Thu 10-29 BMO)

• Business: open-loop card network, same model as V.
• Expectations: EPS est. $5.16, revenue est. $9.7B (Investing Calendar); Oppenheimer models in-line EPS ($5.15) and revenue $9.61B vs $9.58B (FINWIRES, above).
• TRAI history: avg |gap| 2.24% (1.06 ATR), max 3.95%, fill 50%, continuation 50%, day-1 in gap dir +0.54%. ATR 1.65%.
• Linkage: V reports two sessions earlier; with ρ 0.88, MA's 10-28 session will carry V read-through (S-004 already covers same-cluster peers).

PYPL — PayPal (Tue 10-27 BMO)

• Business: digital wallet (PayPal, Venmo) and merchant processing (Braintree); revenue driven by total payment volume × take rate.
• Expectations: EPS est. $1.31, revenue est. $8.7B (Investing Calendar). Oppenheimer expects in-line EPS but revenue below consensus ($8.68B vs $8.72B) on a lower take rate (FINWIRES, above).
• TRAI history: avg |gap| 9.43% (3.42 ATR), max 18.05% (2026-02-03), 0/4 gaps filled, continuation 50%. Largest earnings-gap risk among the group. ATR 1.99%, beta60 0.87. PYPL is a singleton cluster (47); highest ρ only SOFI 0.38, V 0.30 — its print is mostly idiosyncratic.

SOFI — SoFi (Tue 10-27 BMO)

• Business: digital consumer lender/bank (personal & student loans, deposits) plus a tech-platform segment.
• Expectations: EPS est. $0.17, revenue est. $1.3B (Investing Calendar).
• TRAI history: avg |gap| 4.26% (0.89 ATR), fill 75%, continuation 75%, day-1 in gap dir −1.58%; post-print drift days 2–5 was large (−7.2%, −8.6%, +12.8%, −19.7%). ATR 4.31%, beta60 3.49. Cluster 11 with AFRM (ρ 0.74) and HIMS (0.70).

HOOD — Robinhood (Tue 10-27 AMC; call 5 PM ET)

• Business: retail brokerage (equities, options, crypto, event contracts); revenue from transaction rebates, net interest and subscriptions.
• Quarter data already public: September crypto notional $21.9B (+21% m/m), event-contract volume +37% m/m; stock volume −2% and options −1% m/m but above consensus (TokenPost, 2026-09-29). August crypto $17.5B was still 38% below August 2025 (Crypto Briefing, 2026-09-28). BTIG raised its PT to $135 from $125 on the metrics (BitNewsBot, 2026-09-29). Note the dates on these secondary crypto sites look inconsistent in places; treat as unverified until checked against Robinhood's own monthly-metrics release.
• TRAI history: avg |gap| 5.96% (0.95 ATR), max 11.90%, fill 25%; 3 of 4 reactions were negative c-c despite 3 EPS beats (−10.8%, −8.9%, −13.2%, −3.6%). ATR 5.49%, beta60 3.24. Fell −5.25% Oct 22. Cluster 9 (crypto) with COIN/MSTR (ρ 0.73).

COIN — Coinbase (Thu 10-29 AMC)

• Business: crypto exchange, custody and stablecoin (USDC) revenue share; tied to crypto prices and volumes.
• TRAI history: avg |gap| 5.30% (0.81 ATR), fill 25%, continuation 75%; drift days 2–5 negative in 3 of 4. ATR 5.98%. HOOD reports two sessions earlier (ρ 0.73) — HOOD's crypto line is a read-through for COIN. BTC-USD was −0.14% at 13:07.
• Data issue: TRAI's COIN EPS-surprise values are −364.89%, −3458.88%, −487.47% — near-zero or sign-flipped estimates make "surprise %" meaningless. See S-034.

5. Macro context

• TRAI regime: neutral (score 17.5), VIX label transitional, choppy trend (ADX 13.5) (TRAI /api/regime, asOf 2026-10-22). Credit (HYG/IEF) −2.49 pts over ~1 month; 10y 4.96% (TRAI /api/macro).
• Rising oil pressured stocks today (AP, above); October flash PMI was on today's calendar (Investing.com 9:18 ET, above) — reading not checked here.
• Credit-card rate-cap politics: President Trump called for a one-year 10% cap on card interest rates in January, which hit issuers including AXP, COF and AFRM that day (Reuters, 2026-01-12). Current status of that proposal not verified this hour — open question.

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

• S-032 (proposal): treat "beat but low-quality" BMO prints (revenue miss + EPS helped by tax/provisions/one-offs) as continuation candidates on the gap side — AXP today and 2/3 prior AXP "beat, sell" prints never filled. Needs a cross-universe test using revenue surprise, which /api/events does not store yet.
• S-033 (question): do card networks (V/MA) rebound the session after an issuer's (AXP/COF) negative earnings reaction? n = 4, all positive; test across all issuer prints.
• S-034 (data fix): cap or blank EPS-surprise % when |estimate| < $0.05 or the sign flips (COIN).
• S-036 / S-037 (questions, from whiggity): V→MA lead-lag into MA's 10-29 gap; BTC overnight return as a filter on HOOD/COIN post-print trends.
• S-035 (proposal): for Tuesday 10-27, PYPL's 3.42-ATR average gap with 0% fill argues against fading a PYPL gap on day 1; V's 75% fill / 25% continuation argues the opposite for V on 10-28. Both n = 4.
• Existing S-004 (peer-event risk) applies: V → MA on 10-28, HOOD → COIN on 10-28/29, SOFI → AFRM on 10-27.

7. Second researcher (whiggity)

Consulted Kieran's Monica bot whiggity at ~13:05 ET (new chat; reply took ~4 min; its search step showed two "Something went wrong" errors and some irrelevant hits). Its six links were all AXP documents; claims about V, MA, PYPL, HOOD, COIN and politics came without sources. Kept only what I could check; the rest are open questions.

Verified / kept (from whiggity):

• Visa's 10-27 print is fiscal Q4 2026 (Sept 30 fiscal year-end), so it carries the initial FY2027 outlook — the likely driver of the 10-28 gap more than the quarter's beat/miss. Fiscal-Q4 framing is consistent with the Oppenheimer note ("fiscal fourth quarter", FINWIRES, 2026-10-08). Visa's guidance habits (conservative initial guide, client-incentive ranges) are whiggity's claim, unverified.
• Macro collision with MA: BEA's Q3 advance GDP and September personal income & outlays (PCE) are both scheduled for Thu 10-29 at 8:30 ET (BEA release schedule), the same morning as MA's BMO print. Correction: whiggity put PCE on Fri 10-30; BEA lists it with GDP on 10-29.
• AXP low-quality-beat framing (tax rate, provisions, revenue/volume miss) matches §2 sources.

Rejected:

• whiggity said Fiserv reports Tue 10-27 BMO alongside PYPL/SOFI. Fiserv's own release says Tue 11-03 BMO (Fiserv IR, 2026-10-05).

Unverified — open questions (whiggity):

• PYPL: transaction-margin dollars ex-interest, not revenue, as the key metric; Braintree mix as the cause of take-rate pressure.
• V/MA benefit from higher gas prices on nominal volume (it cited 7–9% of card transactions at the pump — no source); MA's larger cross-border mix vs AXP's US skew.
• Rate-cap / Credit Card Competition Act headlines ahead of the Nov 3 midterms as a driver of today's XLF weakness — no source given; today's AP coverage attributes the market drop to oil.
• HOOD/COIN swing factors (event contracts, net interest, COIN retail vs institutional mix, USDC revenue).
• Read-through from SYF/ALLY credit commentary and from big-tech prints (GOOGL/MSFT 10-28, AMZN 10-29 per TRAI /api/events) to payments names.

Test ideas (whiggity), folded into strategy:

1. Issuer→network sympathy drift: after AXP < −2.5% on its print with V/MA underperforming SPY by > 0.75 pp, do V/MA drift up into their own prints? (Extends S-033.)
2. Same-morning BMO cluster (SOFI + PYPL Tue): condition gap continuation on pre-market volume vs its 20-day average (needs pre-market history).
3. V→MA lead-lag: does MA's earnings-day gap load on V's post-print day return minus MA's same-day sympathy return? (S-036.)
4. HOOD/COIN: condition the post-print first-hour trend on BTC's overnight return after the release (S-037).

8. Unknowns / open questions

• AXP regulatory remediation detail and the exact prior FY guide range — not verified.
• Visa's FY2027 outlook expectations (the print sets the initial FY27 guide) — consensus for it not found.
• MA's 10-29 open coincides with Q3 GDP and September PCE at 8:30 ET (BEA).
• Status of any federal credit-card rate-cap action in October 2026 — unknown.
• HOOD/COIN Q3 consensus EPS — not found in a source I could verify this hour.

_Screening/research, not investment advice._