To access the live webcast of this call, please go to the investors section of Capital One's website at capitalone.com. A copy of the earnings presentation, press release, and financial supplement can also be found in the investors section of Capital One's website by selecting financials and then quarterly earnings release. Information regarding Capital One's financial performance and any forward-looking statements contained in today's discussion and the materials speak only as of the particular date or dates indicated in the materials. Capital One does not undertake any obligation to update or revise any of this information, whether as a result of new information, future events, or otherwise.

In the second quarter, Capital One earned $3 billion, or $4.73 per diluted common share. As a reminder, the Brex acquisition closed in early April, and we have provided additional details related to the purchase accounting in the appendix of tonight's presentation. Net of these adjusting items, second quarter earnings per share were $5.81. Relative to the first quarter, revenue increased 4% and non-interest expense grew 7%, resulting in pre-provision earnings growth of 1%.

Our total portfolio coverage ratio decreased 26 basis points and now stands at 5.02%. I'll cover the drivers of the changes in allowance and coverage ratio by segment on slide five. The coverage ratio decreased by 41 basis points and now stands at 6.99%. The allowance build was primarily driven by strong growth in the auto business.

What went well
  • Capital One earned $3.0 billion, or $4.73 per diluted common share ($5.81 net of Discover/Brex adjusting items), up sharply from a $(8.58) net loss a year earlier that had included the Discover day-one credit provision.
  • Credit performance was strong: the domestic card charge-off rate fell to 4.71% (down 39 bps sequentially and 54 bps year over year) and delinquencies to 3.39%, enabling a $662 million allowance release that lowered total coverage to 5.02%.
  • Top-line growth was solid, with domestic card revenue up 30% year over year (9.5% excluding Discover), consumer banking revenue up about 26%, and purchase volume up 26%.
  • Net interest margin rose 14 basis points sequentially to 8.01%, aided by an extra day, lower retail deposit rates, and a decline in average cash.
  • The Discover integration is 14 months into a planned 24 months and going well: debit revenue synergies are fully in the run rate, about one-third of operating-expense synergies are realized, and the full $2.5 billion of announced synergies remains on track.
  • Capital One repurchased $2.7 billion of stock and ended with a 13.7% CET1 ratio, while continuing to invest in its heavy-spender franchise, technology/AI, and the Discover global payments network.
What went wrong
  • The legacy Discover card portfolio remains in a temporary 'brownout,' with loans down 1.5% year over year and continued near-term contraction expected to bottom around the fourth quarter of 2026 as Discover credit-policy pullbacks and personal-loan reductions work through.
  • The CET1 ratio fell 70 basis points sequentially to 13.7%, reflecting $2.7 billion of buybacks, roughly a 40-basis-point Brex impact, and higher risk-weighted assets more than offsetting net income.
  • Non-interest expense grew 7% sequentially and total marketing expense reached about $1.7 billion (up 23% year over year), keeping the efficiency ratio elevated at 57.05% as sustained tech, AI, and network investments continue.
  • GAAP EPS of $4.73 was well below the $5.81 adjusted figure due to acquisition-related adjusting items (including $494 million of acquisition amortization expense).
  • The auto charge-off rate rose 18 basis points year over year to 1.43% as the subprime origination mix returns toward pre-pandemic levels.
  • Management continues to decline formal guidance on the efficiency ratio, citing many moving investment and synergy pieces, which leaves near-term expense trajectory hard for investors to pin down.

Guidance Changes

MetricPeriodCurrent guidance
Post-integration earnings power (ROTCE)After Discover integrationStill expected to be consistent with the deal-announcement thesis, inclusive of Brex and travel insourcing; ROTCE math shown at 12.5% CET1 for comparability (capital need defined as 11%)
Discover synergiesBy H2 2027On track for full $2.5 billion; ~one-third of opex synergies realized to date, remainder by H2 2027
Discover card loan 'brownout'Through 2026Continued near-term contraction, bottoming around Q4 2026 before returning to growth on Capital One's tech stack
Net interest marginNear termBack half of 2025 a reasonable proxy for structural NIM; ~9 bps day-count tailwind in each of the next two quarters; NII roughly neutral to rates
Discover tech migrationQ3 2026 - Q1 2027Front-book originations fully on Capital One tech by end of Q3 2026; back-book conversion waves (July/October/January) completing in Q1 2027

Performance Breakdown

MetricYoYNote
Total net revenue +4% sequentially to $15.9B Revenue grew 4% versus the first quarter; year-over-year comparisons are distorted by the Discover acquisition, which closed May 18, 2025.
GAAP net income / diluted EPS $3.0B / $4.73 (vs $(8.58) loss) Prior-year quarter included the Discover day-one CECL provision; adjusted EPS was $5.81.
Domestic card revenue +30% (9.5% ex-Discover) Largely the addition of a partial quarter of Discover, plus underlying organic growth in legacy Capital One purchase volume and loans; revenue margin 17.4%.
Consumer banking revenue +26% Partial-quarter Discover operations, Discover revenue synergies, and auto loan growth (auto originations +19%).
Net interest margin 8.01% (+14 bps QoQ) One additional day, lower retail deposit rates, and a $5 billion decline in average cash.
Provision for credit losses $3.0B (down $1.1B / 27% QoQ) $662 million allowance release on favorable observed credit; $3.7 billion of net charge-offs.
Efficiency ratio 57.05% (adjusted 51.38%) Sustained investment in technology, AI, network acceptance, and marketing; operating efficiency ratio 46.57%.
CET1 ratio 13.7% (-70 bps QoQ) $2.7 billion of buybacks, ~40 bps Brex impact, and higher risk-weighted assets.

Earnings Call Themes & Trends

TopicPrevious mentionCurrent periodTrend
Discover integration and brownoutIntegration underway14 of 24 months complete; debit synergies fully in run rate, ~one-third of opex synergies realized; Discover card loans in a temporary brownout (down 1.5% YoY) bottoming around Q4 2026, with front-book tech migration done by Q3 2026 and back-book conversions through Q1 2027.
Global payments networkDebit converted to Discover NetworkDebit conversion a 'smashing success' with full-quarter debit synergies now in results; testing origination and migration of Capital One credit accounts onto the Discover Network, and investing in domestic gap-closing and international acceptance (Mexico, Caribbean, Canada, U.K.) via four levers.
Brex acquisitionClosed early AprilOver 100 days in, management is 'as excited as ever'; early tailwinds from Capital One's brand and lower cost of funds, a high-potential lead-sharing program launched, with marketing and travel-portal integration to follow; framed through Capital One's 'horizontal accounting' value-creation lens.
Heavy-spender franchise and marketingOngoing investmentTotal marketing of ~$1.7 billion (+23% YoY) funds heavy-spender acquisition at the top of the card market and national checking growth, with strong new-account originations.
Technology and AI transformationMulti-year tech transformationIn year 14 of its bottom-up tech transformation, investing in foundational capabilities plus AI infrastructure and experiences, alongside efficiency gains from retiring legacy tech.
Consumer credit strengthManagement sees a resilient U.S. consumer and strong portfolio metrics: payment rates above pre-pandemic levels, stabilized revolve rates, and 2024-2025 originations outperforming 2022-2023 vintages, with no observed K-shaped weakness in its book.

Q&A Summary

Terry Ma (Barclays) asked how much of the Brex growth investment is already in the expense run rate and when benefits become visible.
Rich Fairbank clarified Capital One never promised instant growth acceleration but is mobilizing solutions that don't require full integration; over 100 days in, Brex is seeing early brand and cost-of-funds tailwinds and a launched lead-sharing program, while fuller marketing and travel-portal benefits require more technical integration and will come later, so most investment dollars are not yet reflected.
Terry Ma (Barclays) asked how card growth and marketing spend evolve once Discover originations are fully on Capital One's platform.
Fairbank detailed the temporary Discover 'brownout' from prior credit-policy pullbacks (loans down 1.5% YoY, bottoming around Q4 2026), noted 50% of Discover front-book originations are already on Capital One tech (fully by end of Q3), and said the company is leaning into marketing to drive front-book applicant flow as it returns Discover to growth.
Sanjay Sakhrani (KBW) asked whether the liquidity drag means a higher NIM entering the third quarter.
CFO Andrew Young said average cash fell only about $5 billion (ending down $20 billion), so a NIM catch-up should occur in Q3 as average cash aligns with ending cash, plus a ~9-basis-point day-count tailwind each of the next two quarters, while pointing to the back half of 2025 as a proxy for structural NIM.
Ryan Nash (Goldman Sachs) asked whether the combined company is still at least a 20% return business and about capital.
Fairbank reiterated that post-integration earnings power (measured as ROTCE) should be consistent with the deal announcement despite moving variables (Discover brownout, better credit, Brex, insourced travel), declining to brand a precise number; Young added that 11% CET1 is the defined long-term need, not a target, and the company is not racing to drive capital down.
Rich Shane (JPMorgan) asked how Capital One balances Brex's growth orientation against ROTCE and margin.
Fairbank said the objective is not near-term ROTCE maximization; Brex's value creation fits Capital One's founding 'horizontal accounting' approach of investing in long-lived annuities, and analysis shows Brex's investment tranches are value-creating across a large commercial-card, payables, and expense-management market, so Capital One will lean in while rigorously measuring returns.
Erika Najarian (UBS) pressed on unpacking the 'earnings power consistent with announcement' claim against consensus EPS and ROTCE math.
Young corrected several assumptions, explaining the deal used consensus estimates with a Discover loss adjustment, that the ROTCE math uses a 12.5% CET1 weighted-average consensus purely for comparability (not a target), and that share-price and line-item assumptions have moved, which is why the company anchors on ROTCE rather than a specific EPS number.

More on Capital One Financial Corp

Reported 2026-07-21 · figures from the Capital One Financial Corp Q2 2026 earnings call.

See how VectorShift works for your firm

Request Demo