I'll research Capital One (COF) before writing the preview. Let me pull recent filings, transcripts, news, and price action.I have the recent transcripts. Let me read the most recent earnings call (2026Q1) and release, plus the year-ago quarter (2025Q2) for context, and grab recent news and price data.Let me get the 2026Q1 call transcript, the year-ago Q2 release for context, and recent price action.Let me pull the year-ago Q2 2025 release for the comparison baseline and recent sector/consumer credit news.Let me compute price performance stats and check current analyst estimates.# Capital One (COF) — Q2 2026 Earnings Preview
This is the first "clean" quarter to judge the combined Capital One–Discover franchise on its own merits — and the first quarter that includes Brex and the in-sourced Capital One Travel/Hopper business. The year-ago quarter (Q2 2025) is essentially uncomparable on a GAAP basis: Discover closed on May 18, 2025, and the quarter carried an $8.8 billion initial CECL allowance build that produced a reported net loss of $4.3bn (–$8.58/sh) against adjusted EPS of $5.48. So focus the read on sequential trends vs. Q1 2026 and on adjusted metrics.
For reference, Q1 2026 printed GAAP EPS $3.34 / adjusted $4.42, revenue $15.2bn, and pre-provision earnings of $6.8bn.
Capital One closed Brex on April 7, 2026 (~$4.5bn consideration) and brought Capital One Travel/Hopper technology and talent in-house in April. Management guided that Brex would reduce the CET1 ratio by a little over 40bp in Q2, and said it would break out the Brex purchase-accounting marks on this call since the deal closed too recently to finalize last quarter. Watch for: the size/shape of those marks, incremental integration/investment spend, and any early commentary on Brex growth and revenue trajectory. Note management's framing that Brex is an "enablement," not a rush-to-integrate, strategy — meaning more near-term investment and deferred vertical earnings impact.
Q1's adjusted efficiency ratio came in just under 50%, flattered by seasonally low marketing that was explicitly shifted out of Q1 into Q2 and later quarters. Management said marketing will be "heavier over the course of the year" and that it intends to "increasingly lean into marketing." Expect a step-up in marketing spend and a higher efficiency ratio sequentially — this is the single most-scrutinized line given the ongoing tug-of-war between investment spend and the "earnings power on the other side of integration" promise. Brex and Hopper costs also now enter the run rate for the first time.
Q1 NIM was 7.87%, down 39bp sequentially, hurt by two fewer days, seasonal card paydown, and an elevated cash balance (Discover Home Loans sale, strong deposit growth, tax-season flows). For Q2, management flagged a ~9bp tailwind from one extra day and said the elevated cash position should trend down (~$8bn of Q2 debt maturities, tax payments). The framing: absent a big balance-sheet mix shift, the post-Discover "structural" NIM level should persist, with each quarter subject to seasonality. A clean sequential NIM rebound would be reassuring.
Credit remains the bull case. In Q1, Domestic Card charge-offs were 5.1% (down 109bp YoY, ~half from Discover mix) and delinquencies 3.7% (down 55bp YoY, trending better than seasonal). Auto charge-offs 1.64%, near pre-pandemic levels. Watch whether the delinquency outperformance vs. seasonality continues — management was cautiously optimistic but not ready to "declare" a favorable divergence. Any read on the consumer amid the energy-price / geopolitical backdrop flagged last quarter will matter.
Q1 ended with a $23.6bn allowance, 5.28% coverage, and a modest $230m build — with favorable observed credit offset by qualitative overlays for geopolitical/energy risk. Domestic Card coverage was 7.4%. Watch whether improving observed credit lets coverage drift lower (a release/EPS tailwind) or whether macro caution keeps overlays elevated. Also watch commercial — Q1 had a small real-estate/C&I reserve build management called "lumpy."
Legacy Discover card loans are still contracting (down ~1.2% YoY in Q1) due to prior credit pullbacks; management said the brownout will deepen near-term before growth resumes post-integration. Key integration milestones to track: - Discover debit conversion complete — full revenue synergies from debit should show up in Q2. - New Discover card originations to be fully on Capital One's platform by end of Q3; back-book conversion into Q1 2027. - Early testing of credit-card origination on the Discover network. - Reaffirmation of the $2.5bn total synergy target by mid-2027 (revenue front-loaded, expense synergies back-loaded).
CET1 was 14.4% at Q1 (before the ~40bp Brex drag). The company repurchased $2.5bn in Q1 with ~$12bn authorization remaining. Analysts (Erika Najarian, Saul Martinez) pressed on why COF isn't more aggressive given excess capital; management leaned on "asymmetrical value of capital" and conservatism. Watch for any acceleration in buyback pace now that the regulatory picture (Basel reproposal is a modest net positive ~+20bp) is clearer. Management reiterated the deal's ROTCE earnings-power target normalized to 12.5% CET1, with room to hold even at higher capital levels.
Shares closed around $206.81 on July 20. COF is down roughly mid-teens percent YTD (off a January peak near $258) but has rallied ~16% off the early-March low near $177, and sits roughly flat-to-modestly-higher since the Q1 print in late April. A modest recent insider sale (General Counsel M. Cooper, ~$728k) and a favorable legal outcome — a Maryland federal court dismissed a proposed class action alleging excessive credit-card interest rates on July 20 — are minor housekeeping items into the print.
| Watch item | Bull signal | Bear signal |
|---|---|---|
| Marketing/efficiency | Ramp disciplined; adj. efficiency contained | Spend surges, efficiency ratio deteriorates |
| NIM | Sequential rebound toward structural level | Cash stays elevated, NIM soft |
| Credit | Delinquencies beat seasonality again | Consumer softening / energy drag appears |
| Reserves | Coverage drifts lower (release) | Bigger macro/commercial builds |
| Brex marks | Clean, contained dilution | Larger-than-expected marks/investment drag |
| Capital | Buyback accelerates above $2.5bn/qtr | Pace unchanged despite excess capital |
The core debate remains unchanged: strong, improving credit and top-line momentum vs. a heavy, hard-to-size investment agenda (Discover integration, network build-out, Brex, travel, AI) that clouds near-term earnings visibility. Management's repeated anchor — that post-integration earnings power remains consistent with the original Discover deal model, now inclusive of Brex and Hopper — is the statement to watch for reaffirmation.
Note: Consensus EPS/revenue/NII figures above are from Wall Street estimate aggregators as of the days before the print and are subject to revision; treat GAAP YoY comparisons as distorted by the timing of the Discover close.