Bull Run
Version 0.1.0

Capital One Financial Corporation (COF)

Rapport d'investissement · Date du rapport : 2025-12-31

Généré le 2026-08-13 15:04

1. Fiche d'identité

NomCapital One Financial Corporation
SecteurFinancial Services
IndustrieFinancial - Credit Services
BourseNYSE
Pays de cotationUnited States
Référence

1.1 Modèle d'affaires

Capital One Financial Corporation is a financial services holding company whose subsidiaries include Capital One Bank (USA), National Association, and Capital One, National Association. It operates through three main divisions: Credit Card, Consumer Banking, and Commercial Banking, offering deposit products, card services, and various lending solutions. The company earns money from interest income on loans and securities, fees and commissions, and payment/network-related revenues, supported by its nationwide branch network, digital platforms, and specialty service channels.

Référence

1.2 Données de marché

Prix actuel220.00 USD
Capitalisation boursière135.0 G USD
Référence

2. Diagnostic du Moat

2.1 Résumé du moat

Capital One has a real but not dominant moat. Its top‑tier scale in U.S. credit cards and sizeable, largely deposit‑funded balance sheet give it advantages in marketing, data, and funding cost that most smaller competitors cannot match. However, returns are highly cyclical and have recently deteriorated sharply, indicating that these advantages do not fully shield it from competition, credit losses, or regulatory pressures. The moat is therefore best characterized as moderate and execution‑dependent, likely sustainable over a 3–7‑year horizon but not yet proven as a 20‑year wide moat.

Référence

2.2 Type de moat identifié

The primary moat elements are

These are partly offset by the absence of evidence that COF earns structurally superior risk‑adjusted returns versus major peers, so we cannot claim a wide, impregnable moat; at best, the moat is moderate and visible primarily when compared with smaller banks and fintechs rather than with Chase, AmEx, or BoA.

Référence

2.3 Preuves concrètes du moat

Linking each moat element to evidence:

1) Scale and network position in U.S. credit cards (proven) Capital One’s own filings and multiple independent Nilson‑based summaries confirm that it is one of the largest U.S. credit card issuers. 2023–2024 purchase‑volume rankings show Chase and American Express first and second, with Citi, Capital One, and Bank of America clustered around 8–10% each of U.S. general‑purpose purchase volume. Capital One’s 2025 10‑K MD&A further shows that its Credit Card segment provides the bulk of consolidated net revenue and that card loans grew sharply in 2025 (helped by the Discover acquisition), with Domestic Card accounting for more than 90% of card revenue.

Scale matters in cards because it spreads fixed technology, compliance, and marketing costs over a very large receivables base, and it improves bargaining power with networks and partners. A sub‑scale issuer cannot easily match COF’s national reach or marketing spend. This is a clear, evidenced moat component: COF is entrenched in the top tier of U.S. card issuers.

2) Cost advantages via deposit funding and digital operating model (plausible) Financial data show that Capital One is largely funded by deposits, with a debt‑to‑assets ratio of only ~7.6% and debt‑to‑equity ~0.45x in 2025. Equity has grown steadily over the decade, and cash and short‑term investments represent roughly 9% of assets. External data on U.S. bank deposit shares place Capital One among the top 10 by deposits, with roughly 3–4% market share and a rising deposit base over the last decade.

This suggests COF enjoys relatively low funding costs compared with monoline card issuers that must rely more on wholesale funding or securitizations. Combined with a heavily digital delivery model (online bank, limited traditional branch footprint, cafés), per‑customer operating costs are likely lower than for more branch‑intensive competitors. However, the evidence is indirect: we do not have explicit cost‑of‑funds or efficiency ratios vs Chase, Citi, BoA or Discover, so we classify this moat as plausible rather than proven.

3) Switching costs / data & underwriting know‑how (plausible) Capital One has been in mass‑market cards and consumer lending for decades, accumulating large proprietary datasets on consumer behavior, FICO scores, and repayment patterns. Its 2025 10‑K MD&A discusses detailed credit metrics, including net charge‑off rates, delinquency trends, and provisions, emphasizing the role of its risk‑management models.

These data and models create a barrier to entry: a new fintech or small bank cannot easily replicate decades of granular credit‑performance data, which underpin COF’s ability to price risk across millions of accounts. Customers also face soft switching costs: closing long‑tenured cards can reduce their credit scores and forfeit accumulated rewards; moving to a new issuer entails application friction and underwriting. Nonetheless, 2025 results show that COF’s credit costs can spike, and we lack an industry‑wide NCO/delinquency benchmark to prove that its risk performance is consistently superior. Hence, this moat is plausible but unproven in terms of superior economics.

4) Brand and customer‑acquisition engine (plausible) Capital One is a highly visible consumer brand in the U.S., with heavy national advertising and sponsorships. The company’s market share in card purchase volume implies a very large active customer base. Its 10‑K and profile describe multi‑channel distribution (digital, branches, cafés), which broadens reach and reinforces brand recognition.

From a moat perspective, this brand strength likely reduces customer‑acquisition costs and supports the ability to launch new products and cross‑sell (e.g., from cards to deposits and auto loans). Smaller banks or fintechs cannot easily match this awareness and marketing spend. However, we do not have quantitative evidence on customer retention, NPS, or unit customer‑acquisition costs vs peers, nor is there clear pricing power (e.g., consistently higher yields at similar risk). Thus, brand is a moat component but measured as moderate rather than dominant.

5) No proven wide moat Finally, the financial track record warns against overstating COF’s moat. Despite strong revenue growth (~11% CAGR since 2016) and excellent free‑cash‑flow generation, profitability has been volatile: ROE peaked above 20% in 2021, then fell into the single digits in 2023–2024 and to ~2% in 2025, as net margin compressed to ~3.5%. Operating margins and ROIC followed similar patterns. These swings show that competition, credit cycles, and regulatory capital requirements can materially erode returns. Without segment‑level, peer‑adjusted return metrics, we cannot prove that COF earns sustained excess returns on capital vs top peers.

Overall, the evidence supports a moderate moat built on card scale and deposit‑funded, digital operations, reinforced by data and brand. But the moat is not so strong that it guarantees high returns through cycles, and it has not yet shown the resilience required to be labeled wide and long‑lasting.

Référence

2.4 Durabilité du moat

Sustainability and direction of the moat

Time horizon: medium term (3–7 years)

The structural elements of Capital One’s moat—national card scale, accumulated data, and sizable deposits—are likely to persist over at least a 3–7‑year horizon. Being a top‑tier issuer with millions of active accounts and a large deposit base is not easily or quickly replicated by new entrants. The 2025 acquisition of Discover further strengthens COF’s scale in both cards and direct banking, which should, in principle, widen its advantages in marketing, data, and funding if integration is well executed.

However, several factors limit long‑term durability:

• Cyclicality of returns: The sharp deterioration in 2025 ROE and margins, despite continued growth, shows that credit cycles and funding‑cost changes can overwhelm scale advantages. This cyclicality suggests the moat does not guarantee stable excess returns.

• Intense competition: Chase, AmEx, Citi, and Bank of America are all large, sophisticated competitors with comparable or greater scale and, in some cases, stronger brands or network economics. COF’s advantages are relative to smaller players, but not clearly superior to the top tier.

• Regulatory and technology change: Regulatory pressure on interchange fees, credit‑card pricing, and capital requirements can compress economics for all large issuers. At the same time, technology platforms (e.g., embedded finance, alternative credit‑scoring, BNPL) threaten to erode the traditional revolve‑based card model over longer horizons.

• Execution risk in Discover integration: While the acquisition increases scale, it also adds operational, technology, and regulatory complexity. Integration missteps could destroy some of the expected benefits and even damage the franchise.

Given these dynamics, the moat is best described as durable but modest—likely to sustain above‑average scale and relevance in cards and digital banking, but not guaranteed to deliver superior through‑cycle economics without strong ongoing execution.

Référence

2.5 Menaces au moat

Key threats to Capital One’s moat

1) Credit‑cycle and macro shocks: As a card‑centric lender, COF is highly exposed to consumer credit cycles. Recessions, rising unemployment, or inflation spikes can drive higher charge‑offs and provisions, compressing returns even when revenue is growing. 2025’s margin collapse illustrates this vulnerability.

2) Regulatory and political risk: Large card issuers face scrutiny over fees, interest rates, interchange, and credit‑allocation practices. Regulatory changes could cap revenues (e.g., on late fees or APRs) or raise capital requirements, reducing the profitability of cards and unsecured lending, where COF is most exposed.

3) Competitive intensity from megabanks and fintechs: Chase, AmEx, Citi, BoA, and others have deep pockets and strong brands. They can outspend Capital One on rewards, marketing, and technology, pressuring yields and rewards economics. Fintechs and large tech platforms may also siphon off prime customers via BNPL, digital wallets, and super‑apps.

4) Discover integration and operational risk: Integrating Discover’s card portfolio, network, and deposit franchise into Capital One carries execution, systems, and cultural risks. Any major IT or operational issues could harm customer experience, invite regulatory scrutiny, and weaken the brand.

5) Funding and liquidity competition: Although COF has a solid deposit base, competition for deposits has intensified in a higher‑rate environment, with online banks and money‑market funds offering attractive yields. If COF must pay more for deposits or rely more on wholesale funding, its cost‑advantage narrative weakens.

6) Reputation and conduct risk: As a consumer‑facing financial brand, COF is vulnerable to reputational damage from data breaches, customer‑service failures, regulatory settlements, or perceived unfair practices. Such events can erode trust and increase customer churn, raising acquisition costs and weakening pricing power.

Référence

2.6 Comparaison avec les concurrents

Comparison with direct competitors

Direct competitors considered: JPMorgan Chase, American Express, Citi, Bank of America, and Discover (now acquired).

Scale and market share Independent summaries based on Nilson Report data show Chase as the clear leader in U.S. credit‑card purchase volume, followed by American Express. Citi, Capital One, and Bank of America form a second tier, each with roughly 8–10% market share of U.S. general‑purpose purchase volume in 2023. Discover was smaller, in the low‑single‑digit range (3–5%). COF is therefore a top‑tier but not dominant issuer, roughly comparable in scale to Citi and BoA, but materially smaller than Chase and AmEx.

Profitability and returns Financial data indicate that Capital One’s group‑level ROE and ROIC were attractive in 2021–2022 but have since compressed sharply, reaching ~2% ROE and ~1% ROIC in 2025. Public commentary and historical figures (outside this data set) point to Chase, AmEx, and BoA sustaining higher through‑cycle ROE—often in the low‑ to mid‑teens—suggesting stronger or more diversified moats. Citi’s ROE has also been modest but generally above COF’s most recent levels.

We lack a full 2019–2025 panel of segment‑level card returns, NCOs, and cost‑of‑funds vs each peer, so we cannot conclusively rank moats by risk‑adjusted unit economics. Nonetheless, the available evidence suggests:

Deposits and funding Deposit‑market‑share data place Capital One in the U.S. top 10, but below the largest universal banks (Chase, BoA, Citi, Wells Fargo). COF’s deposit share is higher than Discover’s standalone share and comparable to or somewhat above online‑bank specialists like Ally. This positions COF with a solid but not dominant funding advantage: cheaper than monoline or wholesale‑funded issuers, but at a disadvantage to the mega‑banks’ massive low‑cost deposit bases.

Customer franchise and brand Chase and AmEx possess some of the strongest consumer‑finance brands globally, visible in premium card offerings and loyalty ecosystems. Capital One is well‑known and heavily advertised, but its brand sits more in the mass‑market space. Citi and BoA have broad brand recognition tied to their universal‑bank franchises. Discover was strong in cash‑back cards and online savings but smaller. COF’s brand and marketing engine clearly outclass regional banks and many fintechs, but among the top‑tier issuers its brand moat is mid‑pack.

Net result Relative to smaller banks and fintech upstarts, Capital One has a meaningful moat in cards and digital banking. Relative to the largest U.S. card issuers, its moat is narrower: scale and deposits are competitive but not dominant, and profitability has been more volatile. The Discover acquisition improves relative scale but introduces new risks. Overall, COF’s moat is moderate, roughly on par with Citi and BoA in cards but weaker than that of Chase and AmEx.

Référence

2.7 Verdict 3.0/5

Capital One clearly has competitive advantages in U.S. credit cards and scaled consumer banking, but the evidence points to a cyclical, execution-dependent franchise rather than a deep, impregnable moat. On the positive side, COF ranks among the top 4–5 U.S. general‑purpose card issuers with roughly a 10% share of purchase volume; its 2025 10‑K shows very large, growing domestic card loans and strong free‑cash‑flow generation. Deposits place it in the top‑10 U.S. banks by deposit share, suggesting a solid funding base. However, pre‑tax profitability and ROE have fallen dramatically since 2021 (ROE down from ~20% to ~2% in 2025), and credit/funding costs are highly cyclical. We lack full segment‑level and cross‑peer unit‑economic data, so we cannot prove that COF earns structurally superior risk‑adjusted returns versus Chase, Citi, AmEx or BoA. The card scale and data/underwriting capabilities support a real but not dominant moat. Hence a mid‑range score of 3: advantages are durable but not bulletproof, and recent results show that competition, credit cycles and regulation can compress returns.

Force du moatmoderate
Durabilité estiméemedium term (3–7 years) with potential to extend if Discover integration succeeds
Référence

3. Anomalies détectées

3.1 Résumé des anomalies détectées

The main anomalies in COF’s 2025 filings stem from the Discover acquisition and its accounting

Liquidity and capital metrics are robust and there is no sign of hidden forbearance programs, but the mix of acquisition accounting complexity, partial control coverage, and optimistic discourse around subdued GAAP results supports a moderate anomaly-risk verdict.

Référence

3.2 Anomalies comptables

Accounting anomalies center on credit loss accounting and purchase accounting for Discover rather than classic revenue/expense manipulation:

• Credit loss provisioning and ACL: The ACL on loans HFI rises from $16.2B (2024) to $23.4B (2025), and provision for credit losses jumps from $11.7B to $20.7B. Notes and the FINANCIAL supplement explain that roughly $8.8B of this provision spike is a day‑one CECL build on non‑PCD Discover loans, and ACL growth is driven primarily by the initial allowance booked on ~$108B of Discover loans, not a broad‑based deterioration of COF’s legacy portfolios. This is an unusual magnitude but is consistent with CECL purchase accounting for a large acquisition and is well‑disclosed.

• Net charge‑off metrics flattered by acquisition accounting: 2025 card net charge‑offs are disclosed at $11.6B, 5.09% vs 5.88% in 2024 and 4.57% in 2023, and management highlights that Discover’s lower‑loss portfolio reduced the reported card NCO rate by ~35 bps and 30+ day delinquencies by 44 bps. Simultaneously, COF fully charged off $19.4B of acquired Discover loans at closing and booked $3.3B of expected recoveries as a benefit to ACL; these write‑offs are excluded from NCO statistics. This mechanical exclusion flatters ongoing loss metrics and is a form of acquisition‑accounting distortion investors need to normalize, though it is not hidden.

• CECL methodology and judgment: Notes state CECL models (multi‑scenario lifetime loss with overlays) are broadly unchanged, but 2025 required new segmentation and overlays for Discover portfolios and emphasizes heightened model‑risk judgment. There is no disclosed wholesale change in scenario weights or estimation techniques that would independently drive a large ACL change, suggesting the big ACL/provision step‑up is consistent with Discover’s consolidation. This limits evidence of opportunistic reserve management, though the scale of judgment around acquired pools is a monitoring risk.

• Income vs. cash flow: 2025 income quality is extremely high (OCF/net income >10x) because heavy non‑cash credit provisions suppress GAAP profit while OCF jumps to $27.7B. This is economically understandable—large CECL builds and purchase‑accounting charges—but it creates a stark disconnect between earnings‑based ratios and cash generation that could be misinterpreted if investors or compensation plans rely on adjusted metrics.

• Goodwill and intangibles: Goodwill and intangible assets rise from ~$15.1B (2024; goodwill only) to ~$45.1B (2025), about 6.7% of total assets—still moderate for a large bank. There is no disclosed impairment or qualitative signal of impairment risk beyond the inherent uncertainty of a major card/network acquisition.

• Receivables and loan quality: Reported nonperforming assets stay low (~0.6% of loans plus foreclosed assets), and financial‑difficulty modifications are modest relative to total loans, with no evidence of large forbearance programs. The anomaly here is more optical: underlying credit stress is partially managed via restructurings, and Discover’s excluded write‑offs and lower‑loss profile obscure legacy portfolio trends, but there is no sign of aggressive under‑reserving or concealed delinquencies.

Net: 2025 accounting is heavily distorted by Discover CECL and purchase accounting, which boosts ACL and depresses earnings while flattering reported loss rates. These effects are disclosed and broadly consistent with GAAP, so the anomaly is complexity and optics, not clear manipulation.

Référence

3.3 Anomalies dans la dette et la liquidité

Debt and liquidity show few red flags and are a relative strength, though earnings coverage is weak:

• Funding and leverage: Total debt is about $51B against $669B of assets and $113.6B of equity, yielding a debt‑to‑equity ratio of ~0.45x and debt‑to‑assets ~7.6% (fmp_balance_sheet; fmp_ratios). Liabilities are dominated by deposits and other current banking liabilities (~$476B other current liabilities). Equity/asset ratio improves to ~17%, up over the decade; leverage has been trending down even as assets grow.

• Liquidity: Cash and short‑term investments rise to ~$61.7B (~9% of assets) from ~$47.1B in 2024 (fmp_balance_sheet). Net debt is negative (~–$6.4B), meaning cash exceeds wholesale debt at the consolidated level. Cash‑flow‑to‑debt is very strong at ~0.54x in 2025 (fmp_ratios), signaling ability to de‑lever using internal cash if needed.

• Interest coverage vs. capital strength: On an EBIT basis, interest coverage is extremely low (~0.14x in 2025 vs 0.40x in 2024 and 2.24x in 2022; fmp_ratios), driven by low GAAP operating income after large provisions and higher interest expense. This is an earnings‑based anomaly: a bank with ample liquidity and capital shows bond‑like interest coverage that would be alarming in a non‑financial. Here it mostly reflects the income-statement effects of Discover‑driven credit provisioning and higher funding costs, not imminent inability to service debt, but it underscores sensitivity to sustained low profitability.

• Regulatory capital and Discover: Item 7 and the FINANCIAL supplement note that Discover represented about 21% of total assets and 20% of net revenues in 2025, and that pro forma Discover results were deeply loss‑making due to the CECL build and write‑offs. The large, front‑loaded provision suggests a cautious stance on loss recognition. At the same time, continued dividends and buybacks during a low‑ROE, high‑provision year dilute the apparent conservatism from a prudential standpoint.

Given high cash, conservative structural leverage, and improving equity ratios, there is no direct liquidity or solvency anomaly. The main tension is between strong balance‑sheet metrics and weak interest coverage/GAAP earnings, amplifying the need to understand acquisition accounting and credit cycles when assessing debt risk.

Référence

3.4 Anomalies dans la rentabilité

Profitability patterns in 2025 are highly unusual and largely unattractive on a GAAP basis:

• Sharp divergence between revenue and earnings: Revenue grows ~28% (from ~$53.9B to ~$69.3B) while net income falls by ~48% (from $4.75B to $2.45B), compressing net margin from ~8.8% to ~3.5% and ROE from ~7.8% to ~2.2% (fmp_income_statement; fmp_ratios; calc_python). Operating margin drops to ~3.3% from ~11%. This extreme negative operating leverage is a clear profitability anomaly.

• Drivers: The income statement shows cost of revenue rising by about $9.9B and operating expenses by ~42% year‑on‑year (from ~$21.5B to ~$30.5B; fmp_income_statement), while the FINANCIAL supplement attributes the bulk of the profit collapse to Discover’s day‑one CECL provision build ($8.8B) and immediate write‑offs, alongside higher interest expense and integration costs. Underlying pre‑acquisition businesses face cyclical credit and funding headwinds but not of a magnitude to explain the entire earnings collapse.

• Adjusted vs. GAAP framing: Management, in the proxy, emphasizes “strong adjusted earnings and capital generation” and celebrates stock‑price performance, even though GAAP profitability is poor and interest coverage is anaemic. This creates an anomaly between economic risk (heightened) and the upbeat framing of performance.

• Cash vs. earnings: Operating cash flow surges to $27.7B (vs $18.2B in 2024), giving free cash flow of ~$26.1B (fmp_cash_flow_statement). Income quality (OCF/net income) exceeds 10x (fmp_key_metrics). This reflects large non‑cash provisions and purchase‑accounting items; underlying cash generation remains strong, but GAAP metrics alone understate current earnings power absent the acquisition hit.

Overall, 2025 profitability is cyclically and transaction‑depressed, and earnings‑based multiples (e.g., very high P/E and very low interest coverage) are misleading. Profitability anomalies are explainable by Discover accounting but still signal higher risk until integration is proven and normalized credit costs are evident.

Référence

3.5 Anomalies dans le discours de la direction

Discourse anomalies relate to how management narrates 2025 results versus the underlying GAAP economics and control coverage:

• Optimistic performance narrative vs weak GAAP: The 2026 proxy’s CD&A describes 2025 as a year of “strong adjusted earnings and capital generation,” highlighting revenue growth, “solid credit performance,” and superior stock‑price performance (edgar_search_proxy). This sits awkwardly against GAAP net margin of ~3.5%, ROE ~2.2%, and interest coverage ~0.14x, with net income almost halved year‑on‑year (fmp_income_statement; fmp_ratios). While Discover‑related non‑cash charges genuinely depress GAAP earnings, the emphasis on adjusted metrics and TSR risks downplaying the extent of short‑term profitability strain.

• Presentation of credit quality: MD&A and notes explicitly acknowledge the $8.8B Discover CECL build and $19.4B write‑off, and they quantify how Discover’s portfolio lowers reported card loss and delinquency rates (edgar_search Item 7; FINANCIAL supplement). However, management also describes credit as “solid” without heavily stressing that a large slug of acquired losses is excluded from NCO statistics, which could leave casual readers with an overly benign impression of ongoing loss rates.

• Internal controls scope: Item 9A confirms disclosure controls are effective, but management explicitly excludes Discover—which accounts for ~21% of assets and 20% of revenue—from the 2025 internal‑control assessment, as allowed by SEC rules for recent acquisitions (edgar_search Item 9A). This is clearly disclosed but is a material caveat: a fifth of the company, including the newly consolidated card/network business driving many anomalies, is outside the tested control environment. That gap is not meaningfully reflected in the upbeat tone about risk management in the proxy.

The overall discourse is candid about the mechanics of Discover accounting but selectively emphasizes adjusted success and strategic momentum, under‑weighting the risks implied by weak GAAP profitability and partial control coverage.

Référence

3.6 Anomalies stratégiques ou opérationnelles

Strategic and operational anomalies arise from the scale and timing of acquisitions and ongoing expansion:

• Transformational Discover acquisition: In May 2025 COF completed the Discover Transaction, adding ~21% of assets and 20% of revenues, ~$108B of loans, and an $8.8B ACL, plus a massive day‑one write‑off of $19.4B of loans (FINANCIAL supplement; Item 7/Note 2). This is a very large bet on credit cards and payment networks, amplifying cyclicality and regulatory risk. The step‑change in goodwill/intangibles and ACL shows the balance sheet is being reshaped.

• Integration and control risk: Item 9A notes Discover is not yet embedded in internal‑control testing, and CECL disclosures point to heightened judgment and model risk for Discover portfolios (edgar_search Item 9A; FINANCIAL supplement). Operationally, management is simultaneously integrating Discover, launching a $265B Community Benefit Plan, and, per the proxy, preparing a further acquisition (Brex) (edgar_search_proxy). This stacking of large, complex initiatives in a single period increases execution and operational‑risk exposure.

• Operating cost trends: SG&A to revenue creeps up and absolute non‑interest expenses rise sharply in 2025 (operating expenses up from ~$21.5B to ~$30.5B; fmp_income_statement; fmp_key_metrics), pressured by integration, technology, and marketing investments. While some of this is strategic (platform and AI investments), the magnitude relative to current earnings is high and erodes operating leverage.

• Optical credit quality: The combination of Discover’s lower‑loss portfolio, exclusion of acquired charge‑offs from NCO stats, and modest levels of reported financial‑difficulty modifications makes asset quality appear better than it would absent the acquisition; yet pro forma disclosures show Discover itself generated a large net loss in 2025 (FINANCIAL supplement). This underlines operational challenge: absorbing a weaker target while maintaining COF’s legacy standards.

These are not hidden, but they create a more fragile operating backdrop than the high‑level growth narrative might suggest.

Référence

3.7 Anomalies dans la rémunération et la gouvernance

Compensation and governance anomalies are meaningful:

• Pay vs performance: The proxy’s CD&A emphasizes CEO leadership in completing Discover and delivering TSR ~37.7% vs lower index returns, framing 2025 as a strong year (edgar_search_proxy). Yet GAAP ROE is ~2.2%, ROIC ~1.1%, and margins are thin (fmp_key_metrics; fmp_ratios). Incentive decisions appear to lean heavily on adjusted earnings, revenue growth, and stock performance, not on conservative GAAP profitability. This weakens pay‑for‑performance alignment and increases the risk that future decisions continue to prioritize growth and share price over balance‑sheet resilience.

• Equity plan and potential dilution: Proxy disclosures reference substantial equity plan capacity (over 52M shares available under various plans as of year‑end; management_output context), roughly high‑single‑digit percentage of diluted shares, representing notable potential dilution if fully used.

• Capital returns into a downturn: FMP data show a 2025 payout ratio over 70% and continued share repurchases (~$4.6B) and dividends (~$1.77B) despite depressed GAAP earnings and heavy provisioning (fmp_cash_flow_statement; fmp_key_metrics). While cash and capital are strong, maintaining high payouts into a stress and major integration period raises governance questions about prudence and regulatory posture.

• Governance and regulatory scrutiny: Proxies and regulatory commentary highlight CFPB and other regulatory actions and significant shareholder proposals (including one with ~44% support regarding golden parachutes or governance terms) (edgar_search_proxy; web_search context). These indicate elevated regulatory and governance scrutiny even as the board authorizes a transformational deal and aggressive capital returns.

There is no evidence of undisclosed related‑party transactions or outright governance abuses in the obtained text, but the pattern of generous incentives, high payout, and optimistic framing in a weak GAAP year indicates misalignment risk.

Référence

3.8 Conclusion et niveau de risque Modéré

Final verdict: COF’s 2025 anomaly risk is MODERATE. Justification anchored to Sections 3.1–3.7: • Accounting (3.2): The most striking 2025 anomalies—the jump in ACL to $23.4B and provision to $20.7B, and the $19.4B write‑off of acquired Discover loans—are quantitatively large but are explicitly explained as CECL purchase accounting for Discover and are consistent with stable core CECL methodology. Acquisition accounting mechanically flatters ongoing card loss rates by excluding acquired charge‑offs, but the mechanics are disclosed. This increases complexity and optical risk, yet there is no clear sign of earnings manipulation or under‑reserving. • Debt and liquidity (3.3): Structural leverage is conservative (debt/equity ~0.45x; equity/assets ~17%), liquidity is strong (cash and short‑term investments ~9% of assets; net debt negative), and cash‑flow‑to‑debt ratios are robust. These factors materially lower the probability that accounting or profitability anomalies translate into near‑term solvency stress. • Profitability (3.4): GAAP profitability is weak—net margin ~3.5%, ROE ~2.2%, interest coverage ~0.14x—and net income falls ~48% despite ~28% revenue growth. This negative operating leverage and dependence on non‑cash acquisition charges make earnings‑based metrics look alarming. However, very strong operating and free cash flow, and the fact that most of the earnings hit is tied to Discover’s initial provisioning and write‑offs rather than underlying business collapse, mitigate the implication of immediate financial distress. • Discourse (3.5): Management’s emphasis on adjusted earnings, “solid credit performance,” and strong TSR in the proxy and MD&A creates a narrative gap relative to weak GAAP results and partial control coverage (Discover excluded from 2025 internal‑control testing). This raises concern about communication candor and investors’ ability to gauge true risk, but core disclosures about CECL, Discover’s losses, and control scope are present and reasonably detailed. • Strategy/operations (3.6): The Discover transaction is transformational, adding ≈21% of assets and 20% of revenues and materially reshaping COF’s risk profile. Integration, heightened CECL judgment, and stacked growth initiatives (Discover plus prospective Brex) introduce meaningful execution and model risk. Yet these moves are strategic choices, not concealed problems; the 10‑K and proxy acknowledge the scale and complexity of integration. • Compensation/governance (3.7): Incentive design that rewards adjusted metrics and TSR, substantial equity plan capacity, and high payout/buybacks in a low‑ROE, high‑provision year signal misalignment between shareholder risk and management rewards. Governance is under visible regulatory and shareholder scrutiny, but there is no concrete evidence of undisclosed related‑party abuse or governance breakdown. Balancing these elements, the evidence supports a moderate anomaly risk: accounting and profitability are heavily distorted by a major acquisition but transparent; liquidity and capital are strong; yet discourse, control scope for Discover, and compensation/governance choices indicate non‑trivial agency and execution risks that investors must monitor closely.

Catégories d'anomalies :

Référence

4. Analyse financière

4.1 Croissance et qualité des revenus

Revenue has grown materially over the past decade. Reported revenue increased from about $27.5B in 2016 to $69.3B in 2025, implying an estimated ~10.8% CAGR (calc_python). This reflects expansion of the loan book and higher interest income in a higher-rate environment. Growth has been particularly strong post‑2020: revenue rose from $31.6B (2020) to $38.4B (2022), then to $49.5B (2023), $53.9B (2024), and a step up to $69.3B in 2025 (fmp_income_statement ).

Revenue quality, however, is mixed

On balance, revenue growth is strong and diversified across interest and fee income, but quality is somewhat cyclical and currently pressured by funding and credit conditions, as highlighted by the falling margins despite strong top‑line expansion in 2025.

Référence

4.2 Rentabilité et efficacité opérationnelle

Profitability has moved through distinct cycles:

- Net margin and ROE: Net margin peaked in 2021 at ~38.7% with ROE ~20.3%, reflecting unusually low credit losses and reserve releases post‑COVID. Since then, margins and returns have eroded: net margin fell to ~19.2% in 2022, ~9.9% in 2023, ~8.8% in 2024, and just 3.5% in 2025 (fmp_income_statement, fmp_ratios ). ROE similarly declined from ~20.3% (2021) to ~14.0% (2022), ~8.4% (2023), ~7.8% (2024), and ~2.2% (2025) (fmp_key_metrics ). The 2025 collapse in profitability despite high revenue underscores severe credit and cost headwinds.

- Operating efficiency: Operating profit margin was robust in the mid‑20% range in 2018–2019, surged to ~49% in 2021, but has since dropped to 10–12% in 2023–2024 and only ~3.3% in 2025 (fmp_ratios). Sales, general and administrative expense intensity is relatively high and rising: SG&A to revenue has trended from mid‑20s% (e.g., ~25.5% in 2020, ~25.7% in 2018) to ~26.5% in 2025 (fmp_key_metrics). Combined with higher credit costs, this has compressed operating leverage.

- Returns on capital: ROIC was quite strong before the recent downturn (e.g., ~6.0% in 2022, ~3.8% in 2023–2024) but fell to ~1.1% in 2025 (fmp_key_metrics). Return on tangible assets has similarly dropped to roughly 0.4% in 2025 from 1.7% in 2022.

Overall, Capital One has demonstrated that it can earn attractive returns in benign credit conditions, but 2025 profitability is weak. The combination of elevated credit costs, higher funding expenses, and growing operating costs has driven a severe squeeze in margins and returns, making current profitability well below the company’s historical potential and below what Buffett would consider a consistent high‑quality business.

Référence

4.3 Flux de trésorerie et conversion des profits

Cash generation remains a relative strength even as earnings have weakened.

- Operating cash flow (OCF) is consistently above net income, indicating strong profit‑to‑cash conversion and sizeable non‑cash charges (provisions, depreciation, etc.). For example: 2022 net income $7.36B vs OCF $13.8B; 2023 $4.89B vs OCF $20.6B; 2024 $4.75B vs OCF $18.2B; 2025 $2.45B vs OCF $27.7B (fmp_cash_flow_statement ). In 2025, income quality (OCF/net income) is over 10x (fmp_key_metrics), reflecting heavy provisioning and other non‑cash items.

- Free cash flow (FCF) is strong and stable: FCF was ~$11.1B in 2016, rising to $12.1B in 2018, $15.8B in 2019, $16.0B in 2020, $11.6B in 2021, $12.9B in 2022, $19.6B in 2023, $17.0B in 2024, and $26.1B in 2025 (fmp_cash_flow_statement). FCF conversion (FCF/OCF) is typically in the low‑ to mid‑90s%, indicating modest capital intensity; capex has generally been 5–7% of OCF and ~2–3% of revenue (fmp_ratios, fmp_key_metrics).

- Leverage of cash flows to debt: Cash‑flow‑to‑debt ratios are healthy in most years (0.28–0.41 from 2019–2024), reflecting the ability to de‑lever if necessary (fmp_ratios). 2025 shows an even higher metric (0.54x), aided by strong OCF and only modest balance‑sheet debt.

Net result: While accounting earnings are cyclically depressed in 2025, the underlying cash generation of the franchise remains robust, giving management flexibility to absorb credit cycles and continue investing and returning capital. The caveat is that very high OCF in down years often reflects reserve builds and working‑capital swings, so not all of it is distributable on a sustainable basis.

Référence

4.4 Solidité du bilan

For a bank, balance sheet quality is primarily about funding stability, asset quality, and capital, but using the available high‑level data we can still draw several conclusions:

- Size and composition: Total assets grew from ~$357B in 2016 to ~$669B in 2025 (fmp_balance_sheet ). A large portion is classified as long‑term investments (~$518B in 2025), representing loans and securities. Goodwill and intangibles are modest at ~$45.1B (about 6.7% of total assets), suggesting that tangible asset backing remains meaningful.

- Funding and leverage: Liabilities are predominantly customer deposits and other current banking liabilities (other current liabilities of ~$476B in 2025), with comparatively modest wholesale debt: total debt $51.0B and long‑term debt $49.9B (fmp_balance_sheet). Debt‑to‑equity has fallen over time from >1.0x pre‑2020 to 0.45x in 2025 (fmp_ratios, fmp_key_metrics). The debt‑to‑assets ratio is only ~7.6% in 2025, reflecting a conservative funding structure relying more on deposits than bond markets.

- Liquidity: Cash and short‑term investments were ~$61.7B in 2025 (~9% of total assets), up from ~$47.1B in 2024 and much higher than in 2018–2019. Net debt is actually negative in 2025 (about -$6.4B), meaning cash and equivalents exceed financial debt at the consolidated level (fmp_balance_sheet). Liquidity therefore appears ample.

- Capital adequacy proxy: Reported common equity rose from ~$47.5B in 2016 to ~$113.6B in 2025, even after large buybacks (fmp_balance_sheet). The equity/asset ratio improved from ~13% in 2016–2018 to ~17% in 2025, another sign of a more conservatively capitalized balance sheet. Return on equity is low in 2025 because of depressed earnings, not because of excessive leverage.

Risks remain typical of a credit‑card‑centric bank—credit quality, regulatory capital, and interest‑rate risk—which require deeper regulatory metrics (CET1, risk‑weighted assets) beyond this data set. But based on the high‑level financials, the balance sheet appears stronger and less levered than a decade ago, with robust liquidity and growing common equity providing a solid loss‑absorbing buffer.

Référence

4.5 Allocation du capital

Capital allocation has been generally shareholder‑friendly but somewhat pro‑cyclical.

- Dividends: Capital One has paid regular dividends throughout the period with a moderate payout in normal years: payout ratios of ~16–27% of earnings from 2016–2024 (fmp_ratios, fmp_key_metrics). Dividend yield has usually been in the ~1.6–3.1% range, providing modest but consistent cash returns. In 2025, the payout ratio spikes above 70% because earnings fell sharply while the dividend was only modestly adjusted, which is not sustainable if low earnings persist.

- Share repurchases: The company has been an active buyer of its stock, especially in strong years. For instance, it repurchased roughly $3.7B–4.9B of stock annually in 2018–2019 and 2022, and nearly $1.0B in 2021 after temporarily issuing stock during the pandemic (fmp_cash_flow_statement). Even in 2023–2025, with rising credit concerns, the company continued buybacks in the $0.7B–4.6B range per year while also servicing debt and dividends. These buybacks, conducted when shares often traded below book value (PB ratios in the 0.7–1.1x range over 2018–2024), are likely value‑accretive in hindsight.

- Debt management and liquidity: Management has opportunistically raised and repaid wholesale debt, but the overall trend is toward lower leverage: total debt declined from ~$60–61B in 2016–2019 to ~$51B in 2025, even as assets and equity grew (fmp_balance_sheet). Cash holdings increased markedly since 2020, and net debt turned negative by 2025. This indicates a cautious stance on funding risk.

- Reinvestment: Capital expenditure has been steady at ~2–3% of revenue and ~5–7% of OCF, supporting technology and infrastructure investment without straining cash flows (fmp_cash_flow_statement, fmp_key_metrics). ROIC in the mid‑single‑digits to low‑double‑digits in good years suggests that incremental investments have historically earned reasonable returns, though 2025 returns are weak.

Overall, management has blended dividends, buybacks, and balance‑sheet strengthening in a way that is usually rational and shareholder‑oriented. The main concern from a Buffett lens is that share repurchases and payouts have sometimes remained high into the downturn (e.g., 2025), rather than being flexed down more aggressively to preserve capital ahead of, or early in, a credit cycle.

Référence

4.6 Résumé 3.5/5

Capital One exhibits solid long‑term growth and strong cash generation, but 2025 shows a sharp cyclical downturn in profitability and asset quality. Over 2016–2025, revenue grew at ~11% CAGR and interest-earning assets expanded substantially, while pre‑2022 returns on equity and ROIC were attractive. However, net income fell by ~48% in 2025 vs 2024 despite 28% revenue growth, with net margin compressing to 3.5% and ROE to 2.2%, driven primarily by much higher credit costs and operating expense growth. The balance sheet is conservatively funded with deposits (loan/deposit style funding), relatively low debt-to-equity (0.45x in 2025), ample liquidity (cash and short‑term investments ~9% of assets) and net cash position at the holding-company debt level, but regulatory capital and credit risk are key (not fully evaluated here). Free cash flow is consistently strong and typically exceeds net income (OCF-to-net income often >2x), supporting dividends and buybacks, though buybacks have at times been aggressive near cyclical peaks. Capital allocation is generally shareholder-friendly but 2025’s very high payout relative to depressed earnings and low interest coverage (0.14x on an EBIT basis) highlight elevated risk. Overall, the franchise is sound with good cash conversion and adequate leverage, but the current earnings power and risk-adjusted returns look middling; hence a mid‑3s score rather than high‑quality 4–5.

Référence

5. Évaluation de la Gestion

5.1 Qualité et stabilité de l'équipe de direction

Team quality and stability

Référence

5.2 Historique de performance de la direction

Performance history and track record

Référence

5.3 Alignement avec les actionnaires

Shareholder alignment

Référence

5.4 Transparence et qualité du discours

Transparency and discourse quality

Référence

5.5 Discipline opérationnelle

Operational discipline

Référence

5.6 Gouvernance et risques de conflits d'intérêts

Governance and conflict-of-interest risks

Référence

5.7 Points forts et points faibles de la gestion 3.0/5

Management has an experienced, stable leadership team that delivered strong long-term revenue growth (~10.8% CAGR 2016–2025) and excellent cash generation, but 2025 exposed material weaknesses: a ~48% drop in net income, ROIC collapsed to ~1.1%, interest coverage fell to ~0.14x, and management maintained high payout/repurchase activity into the downturn. Proxy disclosures and shareholder proposals show elevated governance/regulatory scrutiny. Overall: competent operators with historically shareholder-friendly capital allocation, but recent risk management, capital-flexing and governance concerns justify a mid–low score.

Points forts :

Points faibles :

Référence

6. Valeur Intrinsèque

Résultats synthétiques des trois scénarios

ScénarioValeur d'entrepriseValeur des actionsValeur par actionPrix de marché actuelMarge de sécurité
Prudent151,890.5 M158,324.5 M250.87 USD220.00 USD12.3%
Réaliste231,635.1 M238,069.1 M377.23 USD220.00 USD41.7%
Optimiste332,997.8 M339,431.8 M537.84 USD220.00 USD59.1%

6.1 Méthode utilisée

Method: I used a 10-year explicit DCF (2026–2035) with three historical years (2023–2025) as the starting base and a perpetual terminal value (Gordon). Each scenario was built from the audited 2025 financials and quarterly TTM data (FMP statements) and the specialist outputs (IDENTITY, FINANCIAL, MOAT, MANAGEMENT, ANOMALIES). The discount rate is explicitly the investors 15% hurdle (so outputs are framed relative to a 15% required return). All non-trivial inputs (revenue step changes, FCF-margin paths, and share-count trajectories) are justified from the financial statements (FMP income, cash-flow and balance-sheet calls) and the MOAT/ANOMALIES/ MANAGEMENT analyses.

Référence

6.2 Hypothèses communes aux scénarios

Shared assumptions across all scenarios: discount rate 15% (investor hurdle capturing cyclical credit and integration risks from MOAT and ANOMALIES), terminal growth rate 2.0% (long-run U.S. nominal GDP/inflation proxy), net debt = -$6.434B (net cash on the 2025 balance sheet). Sources: discount-rate framing tied to investor requirement and risk signals (MOAT and ANOMALIES); terminal-growth anchored to long-term GDP expectations; net-debt from 2025 balance sheet (fmp_balance_sheet ).

Hypothèses générales de valorisation

HypothèseValeurJustification
Taux d'actualisation15.0%Investor target hurdle is 15%. Given Capital One's moderate moat, strong balance sheet but elevated execution and acquisition risks, I used the investor's 15% required return without further upward adjustment. The 15% hurdle embeds a meaningful risk premium for cyclical credit risk, integration execution risk, and governance/compensation risks flagged in MANAGEMENT and ANOMALIES.
Taux de croissance terminal2.0%2.0% perpetual growth approximates long‑run nominal U.S. GDP/inflation and is appropriate for a mature large U.S. financial-services firm with a moderate moat. It is conservative relative to pre-acquisition historical revenue CAGRs and consistent with long-run real growth expectations.
Dette nette-6,434,000,000.00Net debt of -$6.434B (net cash) is taken from the consolidated 2025 balance sheet (FMP balance-sheet ), and was applied across all scenarios as a stable, observable balance-sheet input.
Prix actuel220.00
Référence

6.3 Scénario prudent

Conservative (Prudent) scenario result and assumptions: Value per share = $250.87; Enterprise value = $151.89B; Equity value = $158.32B. Key assumptions: modest 5% revenue growth in 2026 then decelerating to low-single-digits through 2035 due to cautious underwriting and integration drag (supported by FINANCIAL and MOAT notes); FCF margin normalizes from 2025’s elevated levels toward ~23% long-term because of higher provisioning and integration costs (ANOMALIES and cash-flow history). Shares remain near the 2025 level early then drift slightly upward from limited dilution. Inputs and the full 3 historical + 10 projected year-by-year schedule were included in the DCF run (valuation_scenario_compare ).

Projection DCF – Prudent (USD)

Taux d'actualisation
15.0%
Croissance terminale
2.0%
Dette nette
-6,434,000,000.00
nAnnéeChiffre d'affairesMarge FCFFCFFacteur d'actualisationFCF actualiséJustification – chiffre d'affairesJustification – marge FCFActions en circulation
-2202349,484.0 M39.7%19,625.4 M——382.4 M
-1202453,938.0 M31.4%16,958.1 M——382.7 M
0202569,252.0 M37.7%26,128.8 M——631.1 M
1202672,714.6 M30.0%21,814.4 M0.87018,969.0 MOrganic revenue growth slows post-acquisition as credit normalization and funding costs limit absoption of Discover scale; modest 5% growth reflects conservatism given 2025 spike.FCF margin contracts from 2025’s elevated level toward a normalized lower level due to higher provisions, integration costs, and reduced one-time invest/cash benefits (ANOMALIES: 2025 CECL/discover effects); 30% reflects conservative normalization.631.1 M
Management pauses large buybacks during integration; outstanding shares remain at 2025 year-end level.
2202775,895.2 M29.0%22,009.6 M0.75616,642.4 MGrowth decelerates to ~4% as card spend and deposit competition moderate post-integration; reflects mid-cycle performance.Gradual margin pressure from slower revenue mix and sustained elevated credit costs; small annual margin erosion to 29%.635.0 M
Limited dilution from equity compensation and modest share issuance for integration costs; slight rise in share count.
3202878,972.4 M28.0%22,112.3 M0.65814,539.2 MRevenue up ~3.5% as growth returns to structural levels; continued deposit competition and conservative underwriting cap near-term upside.Margins compress to 28% as operating expenses remain elevated during integration and credit normalization continues.638.0 M
Incremental dilution from equity plans and modest issuance; buybacks remain muted given focus on capital preservation.
4202981,341.6 M27.0%21,962.2 M0.57212,557.0 M3.0% growth driven by modest card portfolio seasoning and deposit repricing stabilizing.Further normalization to 27% as incremental revenue is lower-margin and provisioning remains slightly elevated.640.0 M
Share count edges up slightly due to continued small equity issuance and limited repurchases.
5203083,781.9 M26.0%21,783.3 M0.49710,830.1 M3.0% growth driven by steady but unspectacular organic gains as competitive dynamics limit pricing power.FCF margin stabilizes near 26% as operating leverage is slow to re-emerge while credit environment remains cautious.642.0 M
Shares modestly higher as capital is conserved for balance-sheet resilience rather than repurchases.
6203186,186.2 M25.0%21,546.6 M0.4329,315.2 M2.8% growth as the business approaches long-run potential absent major market share gains (industry growth + modest market share).Margins decline to 25% reflecting continued conservative provisioning and higher baseline operating costs post-integration.645.0 M
Small additional dilution from equity comp; no major buyback resumption planned in this scenario.
7203288,379.3 M24.5%21,652.9 M0.3768,140.1 M2.6% growth reflecting macro-driven topline expansion and stable card volumes.Gradual margin recovery stalls; margin at 24.5% as higher funding costs persist for longer.648.0 M
Shares slightly increase as capital allocation remains conservative and some equity dilution occurs.
8203390,540.8 M24.0%21,729.8 M0.3277,103.5 M2.5% growth as the business finishes integration and competes more effectively, though moat remains moderate.Marginal improvement to 24% as operating efficiency slowly improves but elevated credit costs limit upside.650.0 M
Steady share count with marginal dilution from ongoing equity programs.
9203492,693.0 M23.5%21,782.9 M0.2846,192.0 M2.3% growth reflecting low-single-digit structural market expansion and modest share gains.FCF margin near 23.5% as long-term normalization keeps margins below historic peaks due to competitive pressure.652.0 M
Shares modestly higher from continued equity issuance; limited buybacks.
10203594,649.9 M23.0%21,769.5 M0.2475,381.1 M2.1% growth approaching terminal assumptions as the company reaches steady-state post-integration.Terminal-year margin of 23% reflecting conservative through-cycle cash-generation capacity for a moderate-moat bank.655.0 M
Shares slightly higher from cumulative dilution; no aggressive repurchase program in this conservative scenario.
t——22,204.9 M0.24742,220.8 M—
Valeur d'entreprise151,890.5 M
Valeur des actions158,324.5 M
Valeur par action250.87 USD
Référence

6.4 Scénario réaliste

Realistic scenario result and assumptions (base case): Value per share = $377.23; Enterprise value = $231.64B; Equity value = $238.07B. Key assumptions: 2026 revenue +10% (realized Discover synergies), high FCF conversion in early years (mid-30% FCF margin) that normalizes toward low-30s by the terminal year as provisioning normalizes and operating leverage is realized (justified by strong historical FCF and the 2025 cash-flow profile; see fmp_cash_flow_statement ). Share count begins at 631.1M (2025 actual; yahoo_shares ) and declines gradually as buybacks resume. DCF inputs and outputs are from valuation_scenario_compare.

Projection DCF – Réaliste (USD)

Taux d'actualisation
15.0%
Croissance terminale
2.0%
Dette nette
-6,434,000,000.00
nAnnéeChiffre d'affairesMarge FCFFCFFacteur d'actualisationFCF actualiséJustification – chiffre d'affairesJustification – marge FCFActions en circulation
-2202349,484.0 M39.7%19,625.4 M——382.4 M
-1202453,938.0 M31.4%16,958.1 M——382.7 M
0202569,252.0 M37.7%26,128.8 M——631.1 M
1202676,177.2 M36.0%27,423.8 M0.87023,846.8 M2026 revenue benefits from Discover synergies and higher card spend; 10% top-line growth reflects immediate scale benefits and management execution.FCF margin mildly below 2025’s elevated level as one-time provisioning effects subside but operating leverage and strong cash-conversion sustain high FCF; 36% reflects a balanced view.631.1 M
Management uses excess cash for a mix of buybacks and integration funding; share count initially unchanged at 2025 level.
2202782,395.0 M35.5%29,250.2 M0.75622,117.4 M8% growth as synergies and cross-sell from Discover accelerate customer monetization and deposits; growth driven by higher card volumes and deposit yields.Marginal margin improvement to 35.5% as integration costs decline and operating leverage starts to appear.625.0 M
Modest buybacks begin funded by strong FCF, reducing shares outstanding from the 2025 level.
3202888,246.5 M35.0%30,886.3 M0.65820,308.2 M6% growth as scale benefits continue and market share stabilizes; revenue expands through fee income and loan growth.FCF margin improves to 35% as recurring synergies materialize and provisioning normalizes.615.0 M
Share repurchases accelerate moderately as earnings and cash flows normalize, reducing diluted share count.
4202993,061.8 M34.5%32,106.3 M0.57218,356.9 M5% growth driven by steady card-loan growth and continued deposit expansion; revenue reflects healthy consumer spending and modest pricing power.Margins inch down slightly to 34.5% as a larger revenue base absorbs fixed costs; provisioning remains stable.605.0 M
Continued opportunistic buybacks reduce share count further as valuation discipline permits.
5203097,513.8 M34.0%33,154.7 M0.49716,483.7 M4.75% growth reflecting slowing post-synergy expansion but continued cross-sell and deposit monetization.FCF margin at 34% reflecting stable cash-conversion from core banking operations and normalized capex.595.0 M
Buybacks funded by excess FCF continue, reducing share count in line with management’s historically shareholder-friendly policy.
62031101,571.8 M33.5%34,026.5 M0.43214,710.6 M4.0% growth as business transitions toward secular growth driven by deposits and card spend; competitive pressures limit faster expansion.Margins slowly normalize to 33.5% as the business matures and provisioning stabilizes.590.0 M
Share count falls further as buybacks persist at a moderate pace and dilution from equity comp is offset.
72032105,415.3 M33.0%34,787.0 M0.37613,077.7 M3.5% growth driven by stable core lending volumes and modest fee expansion; assumes steady macro environment.FCF margin at 33% reflecting continued cash conversion strength in a normalized cycle.585.0 M
Ongoing buybacks reduce shares gradually; management remains disciplined on capital returns.
82033108,986.0 M32.5%35,420.5 M0.32711,579.0 M3.2% growth as momentum slows toward longer-term comp rates; competitive dynamics cap upside.Margins slightly moderate to 32.5% as growth slows and provisioning returns toward average levels.580.0 M
Share count benefits from continued repurchases and limited dilution.
92034112,649.6 M32.0%36,047.9 M0.28410,247.1 M3.4% growth from ongoing deposit monetization and modest loan growth; reflects convergence toward stable mid-single-digit growth.FCF margin at 32% as steady-state cash generation is achieved.578.0 M
Buybacks continue at a disciplined pace; share count edges down slightly.
102035115,628.1 M32.0%37,001.0 M0.2479,146.1 M2.7% growth as growth approaches terminal rate; revenue stabilizes as Discover benefits are fully realized.Terminal-year FCF margin of 32% reflecting sustainable through-cycle cash conversion for a mature financial franchise with a moderate moat.575.0 M
Share count reflects cumulative repurchases and limited dilution; management maintains shareholder-return focus.
t——37,741.0 M0.24771,761.6 M—
Valeur d'entreprise231,635.1 M
Valeur des actions238,069.1 M
Valeur par action377.23 USD
Référence

6.5 Scénario optimiste

Optimistic scenario result and assumptions: Value per share = $537.84; Enterprise value = $333.00B; Equity value = $339.43B. Key assumptions: rapid, low-friction integration with catch-up synergies driving 15% revenue growth in 2026 and sustained high growth for several years, FCF margins rising toward ~42% steady state and aggressive buybacks pushing shares materially lower. This is a high-execution, low-friction outcome and represents the upside if MANAGEMENT achieves swift integration without persistent provisioning (contrasts with ANOMALIES warnings). DCF run: valuation_scenario_compare.

Projection DCF – Optimiste (USD)

Taux d'actualisation
15.0%
Croissance terminale
2.0%
Dette nette
-6,434,000,000.00
nAnnéeChiffre d'affairesMarge FCFFCFFacteur d'actualisationFCF actualiséJustification – chiffre d'affairesJustification – marge FCFActions en circulation
-2202349,484.0 M39.7%19,625.4 M——382.4 M
-1202453,938.0 M31.4%16,958.1 M——382.7 M
0202569,252.0 M37.7%26,128.8 M——631.1 M
1202679,639.8 M39.0%31,059.5 M0.87027,008.3 M15% growth driven by rapid Discover integration, cross-sell, and favorable retention; assumes execution risk is managed and synergies are realized quickly.High FCF margin near 39% reflecting strong operating leverage and one-off cash benefits normalizing into sustainable productivity gains.631.1 M
Management focuses on aggressive buybacks funded by large FCF; initial year retains 2025 count until repurchase authorization executed.
2202791,535.8 M39.5%36,156.6 M0.75627,339.6 M12% growth as synergies and product cross-sell accelerate customer monetization and card spend; Discover integration proves accretive.Margins expand toward 40% as fixed-costs are spread across higher revenue and provisioning normalizes.610.0 M
Aggressive repurchases reduce share count materially as management prioritizes buybacks given strong cash flow and returns.
32028100,204.3 M40.0%40,081.7 M0.65826,354.4 M10% growth as sustained market-share gains and product innovation lift revenue; optimistic scenario assumes limited competitive repricing.FCF margin reaches 40% as operating leverage and efficiency measures generate structurally higher cash conversion.595.0 M
Continued heavy buybacks funded by excess FCF; share count declines further.
42029108,226.8 M40.5%43,831.8 M0.57225,061.0 M8% growth reflecting continued premium performance, higher-fee products and deposit monetization; competitive advantages translate into above-market growth.Margins tick higher to 40.5% as scale and improved underwriting drive superior cash conversion.580.0 M
Share count continues to decline materially as buybacks outpace dilution.
52030114,631.4 M41.0%46,998.9 M0.49723,366.7 M6% growth as expansion moderates but remains above peers thanks to a strengthened franchise and product mix improvement.FCF margin expands to 41% reflecting durable operating improvements and controlled provisioning.560.0 M
Aggressive buybacks reduce shares further; company redeploys cash to repurchases after investment and integration needs met.
62031121,509.1 M41.5%50,426.3 M0.43221,800.7 M5% growth as business matures at an elevated growth profile relative to peers due to sustained market-share gains and product leadership.Marginal margin improvement to 41.5% as scale benefits accumulate and provisioning normalizes.545.0 M
Buybacks continue, lowering share count as excess cash is returned to shareholders.
72032126,584.6 M42.0%53,165.5 M0.37619,986.9 M4% growth as the firm transitions toward sustainable above-market expansion supported by deposit monetization and data-driven underwriting.FCF margin reaches 42% as operating efficiency and lower credit volatility boost cash generation.530.0 M
Share count falls further due to sustained repurchases funded by healthy FCF.
82033131,628.8 M42.0%55,284.1 M0.32718,072.5 M3.5% growth as elevated growth begins to normalize; company maintains higher-than-peer growth due to scale and brand.Margins steady at 42% as structural improvements fully manifest.515.0 M
Continued reduction in shares outstanding through disciplined repurchase programs.
92034135,588.7 M42.0%56,947.2 M0.28416,188.0 M3.0% growth reflecting normalization toward long-run sustainable growth but remaining above GDP due to scale.FCF margin held at 42% reflecting steady-state improved cash generation.505.0 M
Share count gradually declines as buybacks continue at a slower but steady pace.
102035139,178.6 M42.0%58,455.0 M0.24714,449.2 M2.5% growth approaching terminal assumptions while preserving above-market structural advantages.Terminal-year FCF margin of 42% reflecting optimistic structural improvements in profitability and cash conversion.500.0 M
Cumulative repurchases reduce share count to 500M as buyback programs continue in this optimistic case.
t——59,624.1 M0.247113,370.6 M—
Valeur d'entreprise332,997.8 M
Valeur des actions339,431.8 M
Valeur par action537.84 USD
Référence

6.6 Comparaison avec le prix actuel

Comparison vs current market price: Current share price used = $220.00 (IDENTITY market-data snippet). Versus the realistic/base-case intrinsic value of $377.23, the stock trades at a ~41.7% discount to the base-case intrinsic value (margin of safety ≈ 41.68%). Put differently: realistic intrinsic $377.23 vs market $220 implies margin-of-safety = (377.23-220)/377.23 ≈ 41.68% (calc_python ). If an investor requires a 15% hurdle, buying here at $220 with the ~41.7% MOS multiplicatively increases the prospective outcome: (1+0.15)*(1+0.4168)-1 ≈ +62.93% gross improvement relative to the no-MOS baseline. If that uplift compounds over 5 years the implied annualized return ≈ 10.26% p.a.; over 10 years it ≈ 5.00% p.a. — these annualized figures illustrate that timing matters and that the 15% hurdle is the discount-rate input to the DCF rather than an expected realized CAGR over a fixed calendar horizon (see calc_python ). Note: the DCF results show intrinsic values at the investor's 15% hurdle; buying materially below those intrinsic values increases expected return and margin of safety, but the realized CAGR depends on how quickly the company executes and how soon cash flows normalize.

Référence

6.7 Conclusion

Conclusion and sensitivity: My base (realistic) intrinsic value per share for COF as of 2025-12-31 is $377.23 under a 15% discount-rate framework, with a conservative scenario of $250.87 and an optimistic scenario of $537.84.

The stock trades at $220 today, implying a ~41.7% margin of safety vs the realistic case.

That margin is large and attractive on paper, but it must be judged against real execution risks: the valuation is most sensitive to

Small changes in the terminal-growth assumption or persistent elevated provisioning materially swing value because a large fraction of enterprise value resides in the later years/terminal value.

Investors should: (a) watch credit-loss trends and CECL-related disclosures closely (ANOMALIES), (b) monitor Discover integration progress and whether internal controls are broadened into the acquired business (EDGAR Item 9A notes described in ANOMALIES), and (c) track managements capital-allocation discipline (buybacks vs. conservation) as reported in the cash-flow statements and proxy.

If the investors real required discount rate were to rise above 15% to reflect deeper concerns about governance or persistent credit weakness, intrinsic values would fall materially; conversely, if integration proves swift and provisioning normalizes faster than assumed, upside could be material (optimistic scenario).

Overall verdict: at $220 the market price offers a meaningful margin of safety vs my realistic DCF, but that margin must be validated by monitoring credit trends and Discover-integration evidence.

The recommendation is therefore conditional: the price appears attractive relative to DCF central-case results, provided the investor is comfortable with the integration and credit-cycle risks and monitors the specific indicators named above.

Référence

7. Score d'attractivité Buffett

7.1 Tableau

| Critère | Score | Justification | |---|---:|---| | Prévisibilité | 2.5/5 | Les flux futurs sont moyennement prévisibles. D’un côté, les revenus ont crû à ~10,8 % CAGR 2016–2025 avec un FCF très régulier et largement supérieur au résultat net (OCF >2x le net income sur plusieurs années, FCF positif et en hausse jusqu’à 26,1 Md$ en 2025 – FINANCIAL 4.3), appuyés sur un modèle bancaire diversifié (cartes, dépôts, prêts – IDENTITY). De l’autre, la rentabilité est très cyclique: marge nette passée de ~38,7 % en 2021 à 3,5 % en 2025, ROE de ~20,3 % à 2,2 % (FINANCIAL 4.2), avec une forte sensibilité au cycle du crédit et aux effets CECL/Discover (ANOMALIES 3.2–3.4). Cette cyclicité et la transformation Discover limitent la visibilité à long terme. | | Modèle d'affaires | 3.5/5 | Le modèle bancaire orienté cartes de crédit/banque de détail/commerciale, financé majoritairement par les dépôts, est économiquement solide: croissance d’actifs de 357 Md$ à 669 Md$ sur 2016–2025, forte génération de trésorerie, capex faible (5–7 % de l’OCF; FINANCIAL 4.3–4.4). La base de revenus est diversifiée entre intérêts, commissions et services de paiement (IDENTITY), avec un bilan bien capitalisé (equity/assets ~17 %, dette/actifs ~7,6 %, net cash – FINANCIAL 4.4). Cependant, le modèle reste exposé au cycle du crédit et aux régulations sur les cartes, comme le montrent l’effondrement de ROE/ROIC en 2025 et la forte hausse de provisions (FINANCIAL 4.2, ANOMALIES 3.2–3.4). D’où un bon modèle, mais pas exceptionnel. | | Moat | 3.0/5 | Le spécialiste MOAT conclut à un moat modéré: Capital One est un des top 4–5 émetteurs de cartes aux États‑Unis (~10 % de parts de volume) et un top‑10 en dépôts, avec avantages d’échelle, données de crédit et marque nationale (MOAT 2.1–2.3). Toutefois, la rentabilité est très volatile (ROE ~20 % en 2021 puis ~2 % en 2025) et il n’y a pas de preuve qu’il génère des rendements ajustés du risque structurellement supérieurs à ceux de Chase, AmEx ou BoA (MOAT 2.3, 2.5–2.6). Le moat est donc réel mais dépendant de l’exécution et du cycle, pas « wide ». | | Gestion | 3.0/5 | MANAGEMENT souligne une direction expérimentée et stable ayant livré une croissance du CA d’~10,8 % CAGR 2016–2025 et une forte génération de FCF historiquement, avec dividendes réguliers et rachats souvent créateurs de valeur (buybacks à PB 0,7–1,1x; FINANCIAL 4.5, MANAGEMENT 5.1/5.5). Mais 2025 révèle des faiblesses: net income –48 %, ROIC ~1,1 %, couverture des intérêts ~0,14x (MANAGEMENT 5.2), maintien d’un payout ~72 % avec rachats importants malgré des résultats déprimés (MANAGEMENT 5.3/5.5, ANOMALIES 3.7) et une gouvernance sous pression (propositions actionnariales à 44 %, litiges régulatoires – MANAGEMENT 5.4/5.6). Cela justifie une note médiane plutôt que élevée. | | Risques | 2.5/5 | ANOMALIES conclut à un risque d’anomalies MODÉRÉ: forte complexité comptable liée à Discover (provision CECL +8,8 Md$, write‑off de 19,4 Md$ exclu des NCO, Discover représentant ~21 % des actifs mais hors périmètre de test de contrôle interne 2025 – ANOMALIES 3.2, 3.5–3.6). La rentabilité 2025 est très faible (marge nette 3,5 %, ROE 2,2 %, interest coverage 0,14x – ANOMALIES 3.3–3.4). En contrepartie, le bilan est solide: equity/assets ~17 %, dette/équité ~0,45x, liquidités élevées, cash‑flow‑to‑debt 0,54x et net cash (ANOMALIES 3.3, FINANCIAL 4.4). Les risques de solvabilité sont donc contenus, mais les risques de cycle du crédit, d’intégration Discover, de gouvernance et d’opacité partielle restent significatifs. | | Valorisation | 4.0/5 | Le spécialiste VALUATION estime la valeur intrinsèque réaliste à 377 $/action avec un taux d’actualisation de 15 %, contre un cours actuel de 220 $, soit une décote d’environ 41,7 % (valuation_output 6.4–6.6). Même le scénario prudent donne 250,9 $, au‑dessus du cours, alors que le scénario optimiste atteint 537,8 $. Cette marge de sécurité est importante pour un profil de risque/moot modéré, mais l’analyste insiste sur la sensibilité aux hypothèses de normalisation des provisions, de synergies Discover et de croissance terminale (VALUATION 6.7), ce qui empêche d’attribuer la note maximale. |

Grille Buffett

CritèreScoreJustification
Prévisibilité2.5/5Les flux futurs sont moyennement prévisibles. D’un côté, Capital One a affiché une croissance de revenu d’environ 10,8 % par an entre 2016 et 2025 et une génération de trésorerie très régulière, avec un FCF positif et en hausse (de ~11,1 Md$ en 2016 à 26,1 Md$ en 2025) et un OCF systématiquement supérieur au résultat net (OCF/net income souvent >2x – FINANCIAL 4.1 et 4.3). De l’autre, la rentabilité est fortement cyclique: la marge nette est passée de ~38,7 % en 2021 à 3,5 % en 2025 et le ROE de ~20,3 % à 2,2 % (FINANCIAL 4.2), principalement à cause de la hausse des coûts du crédit et des effets CECL/Discover (ANOMALIES 3.2–3.4). La transformation liée à l’acquisition de Discover (ajout d’environ 21 % des actifs, 20 % des revenus et forte hausse de provisions – ANOMALIES 3.1–3.2) ajoute une couche d’incertitude. Cela justifie une prévisibilité intermédiaire, pas faible mais loin d’être « linéaire » au sens Buffett.
Modèle d'affaires3.5/5Le modèle d’affaires repose sur trois segments (Credit Card, Consumer Banking, Commercial Banking) monétisés via revenus d’intérêts, commissions et services de paiement, appuyés sur une base de dépôts nationale et une plateforme digitale (IDENTITY). Sur 2016–2025, les revenus sont passés de ~27,5 Md$ à 69,3 Md$ (~10,8 % CAGR – FINANCIAL 4.1) et les actifs de ~357 Md$ à ~669 Md$ (FINANCIAL 4.4), avec un FCF robuste et peu capitalistique (capex 5–7 % de l’OCF, 2–3 % du CA; FINANCIAL 4.3). Le bilan est financé surtout par les dépôts avec faible dette de marché (dette/actifs ~7,6 %, dette/fonds propres ~0,45x, equity/assets ~17 %, net cash – FINANCIAL 4.4), ce qui soutient la durabilité économique du modèle. Néanmoins, la forte sensibilité au cycle du crédit et aux régulations (chute du ROE de 20 % à 2 %, explosion des provisions en 2025 – FINANCIAL 4.2, ANOMALIES 3.2–3.4) montre que ce n’est pas un modèle « quasi‑obligataire » robuste en toutes circonstances, d’où une bonne note mais pas maximale.
Moat3.0/5Le spécialiste MOAT conclut à un moat modéré. Capital One figure parmi les 4–5 premiers émetteurs de cartes de crédit généralistes aux États‑Unis avec ~10 % de part de volume et fait partie du top‑10 américain en dépôts (MOAT 2.1–2.3), ce qui lui confère des avantages d’échelle (répartition des coûts IT/compliance/marketing), un accès à des données de crédit massives, une marque nationale reconnue et une base de financement à faible coût via les dépôts. L’acquisition de Discover renforce encore l’échelle cartes/dépôts (MOAT 2.4, comparatif Discover). Cependant, la rentabilité a été très volatile: ROE ~20 % en 2021 puis ~2 % en 2025, avec des marges compressées par le cycle du crédit et les coûts de financement (MOAT 2.3, 2.5). MOAT souligne aussi l’absence de preuve de rendements ajustés du risque durablement supérieurs à ceux de Chase, AmEx ou Bank of America et la forte intensité concurrentielle (MOAT 2.5–2.6). Le moat est donc réel mais dépendant du cycle et de l’exécution, ce qui justifie 3/5.
Gestion3.0/5MANAGEMENT décrit une équipe dirigeante expérimentée et stable ayant piloté une croissance du chiffre d’affaires d’environ 10,8 % par an entre 2016 et 2025 (MANAGEMENT 5.1–5.2) et une forte génération de cash, avec FCF important et utilisation active de dividendes et de rachats, souvent à des multiples de valeur comptable attractifs (PB 0,7–1,1x, suggérant des rachats créateurs de valeur – FINANCIAL 4.5). La structure de capital s’est renforcée (baisse du leverage, hausse des capitaux propres – FINANCIAL 4.4), signe d’une certaine prudence. Toutefois, 2025 met en lumière des faiblesses importantes: chute du net income de ~48 %, ROIC tombé à ~1,1 %, couverture des intérêts à ~0,14x (MANAGEMENT 5.2), tout en conservant un payout ratio >70 % et des rachats significatifs (MANAGEMENT 5.3/5.5, ANOMALIES 3.7), ce qui questionne la flexibilité de la politique de capital en bas de cycle. S’ajoutent une faible détention d’actions par les dirigeants (~1,9 % – MANAGEMENT 5.3) et une gouvernance sous contrainte (litiges CFPB, propositions actionnariales dont une à 44 % de soutien – MANAGEMENT 5.4–5.6). Globalement, la gestion est compétente mais pas exemplaire du point de vue Buffett, d’où 3/5.
Risques2.5/5ANOMALIES attribue à Capital One un risque d’anomalies MODÉRÉ en 2025. Sur le plan comptable, la hausse spectaculaire de l’ACL (16,2 à 23,4 Md$) et des provisions (11,7 à 20,7 Md$) ainsi que le write‑off de 19,4 Md$ de prêts Discover sont expliqués par la comptabilisation CECL et bien divulgués (ANOMALIES 3.2), mais ils complexifient la lecture des pertes réelles (NCO flattés). Les résultats GAAP sont très faibles (marge nette 3,5 %, ROE 2,2 %, interest coverage 0,14x – ANOMALIES 3.3–3.4), tandis que la direction met l’accent sur des métriques ajustées et un récit optimiste (ANOMALIES 3.5, 3.7). L’intégration Discover concentre des risques d’exécution et de contrôle, l’entité représentant ~21 % des actifs mais étant exclue des tests de contrôle interne 2025 (ANOMALIES 3.5–3.6). En contrepartie, le bilan est robuste (equity/assets ~17 %, dette/équité ~0,45x, liquidity ~9 % des actifs, cash‑flow‑to‑debt ~0,54x, net cash – ANOMALIES 3.3, FINANCIAL 4.4), ce qui rend peu probable un risque de solvabilité à court terme. Le compromis entre complexité/anomalies de rentabilité et solidité du bilan conduit à une note médiane de 2,5/5 (5 signifiant des risques largement contenus).
Valorisation4.0/5Le spécialiste VALUATION, avec un DCF à 10 ans actualisé à 15 %, estime la valeur intrinsèque « réaliste » à 377,23 $/action et la valeur « prudente » à 250,87 $, contre un cours d’environ 220 $ (IDENTITY, VALUATION 6.3–6.6). Cela implique une marge de sécurité d’environ 41,7 % par rapport au cas réaliste (calc_python cité dans VALUATION 6.6). Le scénario optimiste aboutit même à 537,84 $/action. Cette décote importante, malgré un taux d’actualisation élevé intégrant le risque de cycle du crédit et d’intégration Discover (VALUATION méthodologie et discount_rate_justification), plaide pour une valorisation attrayante au sens Buffett. Néanmoins, VALUATION souligne une forte sensibilité aux hypothèses de normalisation des marges de FCF et de croissance terminale, et à la réussite de l’intégration Discover (VALUATION 6.7). Compte tenu de ces risques mis en avant par MOAT, MANAGEMENT et ANOMALIES, une note élevée mais non maximale (4/5) est appropriée.
Référence

7.2 Score final et interprétation 18.5/30

Score Buffett total: 18,5/30 (soit 61,7 % du maximum). Ce niveau traduit une franchise bancaire de bonne qualité mais clairement cyclique: le modèle d’affaires et la génération de cash sont attractifs, et le moat comme la gestion sont corrects sans être de tout premier ordre. Les risques (crédit, intégration Discover, gouvernance, complexité comptable) restent modérés mais réels, ce qui limite les notes de prévisibilité et de risques. À l’inverse, la valorisation actuelle offre, selon le DCF à 15 %, une marge de sécurité significative. En termes d’attractivité de long terme pour un investisseur « Buffettien », Capital One se situe plutôt dans la catégorie des « bons business à prix intéressant avec risques cycliques à surveiller », en dessous des franchises exceptionnelles mais potentiellement rémunérateur pour un actionnaire patient et discipliné sur le suivi du risque.

Référence

8. Résumé exécutif ACHAT

Executive summary (as of fiscal year 2025‑12‑31):

Business: Capital One Financial Corporation is a U.S. financial‑services holding company concentrated in Credit Card, Consumer Banking and Commercial Banking. The franchise earns interest income, fees and payment/network revenues from a large, digitally oriented card and deposit base and ranks among the top U.S. card issuers.

Financial health: On a balance‑sheet basis the company looks conservative for a card‑centric bank: deposits provide the bulk of funding, debt/equity was low (~0.45x) and cash & short‑term investments were elevated (~9% of assets), producing a net cash position. Cash generation is a structural strength—operating cash flow and free cash flow have been consistently strong (FCF ~ $26.1B in 2025). However, GAAP profitability in 2025 was severely depressed: net income fell ~48% year‑over‑year despite revenue growth, compressing net margin to ~3.5% and ROE to ~2.2%, with interest coverage near 0.14x—a clear earnings‑quality and cycle signal.

Moat: The firm possesses a real but moderate moat: top‑tier scale in U.S. credit cards, significant deposit share and accumulated data/underwriting capabilities give competitive advantages versus smaller peers. The moat is not wide—returns are cyclical and can be eroded by competition, regulation and credit shocks; the Discover acquisition increases scale but also execution risk.

Management quality: Management is experienced and has historically delivered strong revenue growth (~10.8% CAGR 2016–2025) and shareholder returns via buybacks and dividends. 2025 exposed weaknesses in risk and capital‑allocation timing—most notably maintaining high payouts into a pronounced earnings and provisioning stress—and proxy disclosures show elevated shareholder and regulatory scrutiny.

Key risks / anomalies: 2025 results are materially affected by the Discover transaction and related purchase accounting: large CECL day‑one provisioning (~$8.8B of the provision), a $19.4B write‑off of acquired loans excluded from ongoing NCO metrics, and Discover being outside 2025 internal‑control testing complicate credit‑trend assessment. Governance/compensation signals and continued high payout in a low‑ROE year raise agency and prudence concerns. Despite these anomalies, the balance sheet and liquidity metrics do not indicate immediate solvency stress; the primary risks are integration execution, persistent elevated provisioning, and regulatory/governance outcomes.

Valuation versus price: A 10‑year DCF at a 15% discount rate yields a realistic intrinsic value of $377.23 per share (conservative $250.87; optimistic $537.84).

The current market price ($220) implies ~41.7% margin of safety versus the realistic case.

The valuation is sensitive to

Recommendation summary: Given a moderate (but manageable) anomaly risk, a conservative and liquid balance sheet, powerful cash generation, and a substantial DCF margin of safety, recommend ACHAT for investors who (a) accept cyclical credit risk and integration/gov‑risk, and (b) are willing to actively monitor credit‑loss metrics, CECL disclosures, Discover integration progress (including expansions of internal‑control testing), and management capital‑allocation decisions. For more risk‑averse investors or those unwilling to watch these indicators closely, CONSERVER would be more appropriate.

Points forts :

Points faibles :

Références

[1] FMP company profile
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/profile/COF?symbol=COF
[2] EDGAR 10-K Item 1
EDGAR · edgar · 2026-08-06 · Item 1
https://www.sec.gov/Archives/edgar/data/927628/000092762826000024/cof-20251231.htm#i7687b40c0e86405a90048ec815ca20f5_568
Item 1. Business—Overview.” Insider Trading Policy We maintain insider trading policies and procedures governing the purchase, sale and/or other dispositions of our company’s securities by directors, officers, employees and other covered persons that we believe are reasonably designed to promote compliance with insider trading laws, rules, and regulations, as well as New York Stock Exchange (“NYSE”) listing standards. A copy of our insider trading policy is filed as Exhibit 19 to this Annual Report on Form 10-K.
[3] Yahoo Finance quote for COF
Yahoo · api · 2026-08-06
https://finance.yahoo.com/quote/COF
[4] Yahoo Finance summary for COF
Yahoo · api · 2026-08-06
https://finance.yahoo.com/quote/COF
[5] FMP fmp_income_statement
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/income-statement/COF?period=annual&limit=10
[6] FMP fmp_key_metrics
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/key-metrics/COF?period=annual&limit=10
[7] FMP fmp_ratios
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/ratios/COF?period=annual&limit=10
[8] U.S. Credit Card Market Share by Network & Issuer – 2024 Facts & Statistics
Tavily · web · 2026-08-06
https://upgradedpoints.com/credit-cards/us-credit-card-market-share-by-network-issuer
[9] Credit and Debit Card Market Share by Network and Issuer
Tavily · web · 2026-08-06
https://www.fool.com/money/research/credit-debit-card-market-share-network-issuer
U.S. Credit Card Market Share by Network & Issuer https://upgradedpoints.com/credit-cards/us-credit-card-market-share-by-network-issuer Chase also led the way when it came to the total spend on credit cards across the first half of 2023. This purchase volume was more than double all rivals, with the exception of American Express: Chase — $602.1 billion American Express — $547.6 billion Citi — $287.2 billion Capital One — $272.6 billion Bank of America — $244.2 billion Discover — $105.8 billion U.S. Bank — $98.8 billion Wells Fargo — $90.6 billion Source: The Nilson R…
[10] Credit Card Statistics (2025) | SellersCommerce
Tavily · web · 2026-08-06
https://www.sellerscommerce.com/blog/credit-card-statistics
[11] Credit Card Market Share by Issuer (2026)
Tavily · web · 2026-08-06
https://wallethub.com/edu/cc/market-share-by-credit-card-issuer/25530
U.S. Credit Card Market Share by Network & Issuer https://upgradedpoints.com/credit-cards/us-credit-card-market-share-by-network-issuer Chase also led the way when it came to the total spend on credit cards across the first half of 2023. This purchase volume was more than double all rivals, with the exception of American Express: Chase — $602.1 billion American Express — $547.6 billion Citi — $287.2 billion Capital One — $272.6 billion Bank of America — $244.2 billion Discover — $105.8 billion U.S. Bank — $98.8 billion Wells Fargo — $90.6 billion Source: The Nilson R…
[12] Credit Card Market Share (2026): Visa vs Mastercard vs Others
Tavily · web · 2026-08-06
https://capitaloneshopping.com/research/credit-card-market-share-statistics
[13] U.S. Credit Card Market Share by Network & Issuer
Tavily · web · 2026-08-06
https://upgradedpoints.com/credit-cards/us-credit-card-market-share-by-network-issuer
U.S. Credit Card Market Share by Network & Issuer https://upgradedpoints.com/credit-cards/us-credit-card-market-share-by-network-issuer Chase also led the way when it came to the total spend on credit cards across the first half of 2023. This purchase volume was more than double all rivals, with the exception of American Express: Chase — $602.1 billion American Express — $547.6 billion Citi — $287.2 billion Capital One — $272.6 billion Bank of America — $244.2 billion Discover — $105.8 billion U.S. Bank — $98.8 billion Wells Fargo — $90.6 billion Source: The Nilson R…
[14] Nilson in the News - Nilson Report
Tavily · web · 2026-08-06
https://nilsonreport.com/news-events/nilson-in-the-news
[15] US Mastercard/Visa Credit Card Issuers–Midyear 2023 - Nilson Report
Tavily · web · 2026-08-06
https://nilsonreport.com/articles/us-mastercard-visa-credit-card-issuers-midyear-2023
[16] 1249 - Nilson Report
Tavily · web · 2026-08-06
https://nilsonreport.com/newsletters/1249
[17] Capital One-Discover: A Competition Policy and Regulatory Deep Dive - American Economic Liberties Project
Tavily · web · 2026-08-06
https://www.economicliberties.us/our-work/capital-one-discover-a-competition-policy-and-regulatory-deep-dive
[18] Ally Bank vs. Capital One: Which Is Better for You
Tavily · web · 2026-08-06
https://smartasset.com/checking-account/ally-bank-vs-capital-one
[19] Ally Bank vs Discover Bank 2026: Accounts, Fees and Access
Tavily · web · 2026-08-06
https://www.dollarscout.net/banking/compare/ally-vs-discover
[20] Bank Market Share by Deposits & Assets (2026)
Tavily · web · 2026-08-06
https://wallethub.com/edu/bank-market-share-by-deposits/25587
[21] Best Online Banks Of 2026
Tavily · web · 2026-08-06
https://www.forbes.com/advisor/banking/best-online-banks
[22] 2023 Consumer Credit Card Market Report
Tavily · web · 2026-08-06
https://files.consumerfinance.gov/f/documents/cfpb_consumer-credit-card-market-report_2023.pdf
U.S. Credit Card Market Share by Network & Issuer https://upgradedpoints.com/credit-cards/us-credit-card-market-share-by-network-issuer Chase also led the way when it came to the total spend on credit cards across the first half of 2023. This purchase volume was more than double all rivals, with the exception of American Express: Chase — $602.1 billion American Express — $547.6 billion Citi — $287.2 billion Capital One — $272.6 billion Bank of America — $244.2 billion Discover — $105.8 billion U.S. Bank — $98.8 billion Wells Fargo — $90.6 billion Source: The Nilson R…
[23] 1249
Tavily · web · 2026-08-06
https://nilsonreport.com/newsletters/1249
U.S. Credit Card Market Share by Network & Issuer https://upgradedpoints.com/credit-cards/us-credit-card-market-share-by-network-issuer Chase also led the way when it came to the total spend on credit cards across the first half of 2023. This purchase volume was more than double all rivals, with the exception of American Express: Chase — $602.1 billion American Express — $547.6 billion Citi — $287.2 billion Capital One — $272.6 billion Bank of America — $244.2 billion Discover — $105.8 billion U.S. Bank — $98.8 billion Wells Fargo — $90.6 billion Source: The Nilson R…
[24] Citi Prestige vs Capital One Venture X Credit Card - Your Best Credit Cards
Tavily · web · 2026-08-06
https://yourbestcreditcards.com/card-comparison/citi-prestige-vs-capital-one-venture-x-credit-card
Citi Prestige vs Capital One Venture X Credit Card - Your Best Credit Cards https://yourbestcreditcards.com/card-comparison/citi-prestige-vs-capital-one-venture-x-credit-card Citi Prestige vs Capital One Venture X Credit Card - Your Best Credit Cards. BROWSE CREDIT CARDS BY TYPE By Spend Category To find the best cards for your spend By Issuer If you know what you’re looking for By Key Feature i.e. lounge access, 0% intro APR, etc By Hotel & Airline We think you know... * No Annual Fee Credit Cards. * Credit Cards with 0% Intro APR Offers. * Credit Cards with 0% Balance Transfer Offers. * Cre…
[25] Capital One vs. American Express Credit Cards (2026)
Tavily · web · 2026-08-06
https://wallethub.com/answers/cc/capital-one-vs-american-express-credit-card-2140888930
Citi Prestige vs Capital One Venture X Credit Card - Your Best Credit Cards https://yourbestcreditcards.com/card-comparison/citi-prestige-vs-capital-one-venture-x-credit-card Citi Prestige vs Capital One Venture X Credit Card - Your Best Credit Cards. BROWSE CREDIT CARDS BY TYPE By Spend Category To find the best cards for your spend By Issuer If you know what you’re looking for By Key Feature i.e. lounge access, 0% intro APR, etc By Hotel & Airline We think you know... * No Annual Fee Credit Cards. * Credit Cards with 0% Intro APR Offers. * Credit Cards with 0% Balance Transfer Offers. * Cre…
[26] Credit Card Showdown: American Express® Green Card vs. Citi Premier® Card - 10xTravel
Tavily · web · 2026-08-06
https://10xtravel.com/american-express-green-card-vs-citi-premier-card
Citi Prestige vs Capital One Venture X Credit Card - Your Best Credit Cards https://yourbestcreditcards.com/card-comparison/citi-prestige-vs-capital-one-venture-x-credit-card Citi Prestige vs Capital One Venture X Credit Card - Your Best Credit Cards. BROWSE CREDIT CARDS BY TYPE By Spend Category To find the best cards for your spend By Issuer If you know what you’re looking for By Key Feature i.e. lounge access, 0% intro APR, etc By Hotel & Airline We think you know... * No Annual Fee Credit Cards. * Credit Cards with 0% Intro APR Offers. * Credit Cards with 0% Balance Transfer Offers. * Cre…
[27] As Citi and American Airlines Go Exclusive, Their Cards Need Tweaks
Tavily · web · 2026-08-06
https://www.nerdwallet.com/credit-cards/news/citi-american-airlines-cards-could-use-some-tweaks
Citi Prestige vs Capital One Venture X Credit Card - Your Best Credit Cards https://yourbestcreditcards.com/card-comparison/citi-prestige-vs-capital-one-venture-x-credit-card Citi Prestige vs Capital One Venture X Credit Card - Your Best Credit Cards. BROWSE CREDIT CARDS BY TYPE By Spend Category To find the best cards for your spend By Issuer If you know what you’re looking for By Key Feature i.e. lounge access, 0% intro APR, etc By Hotel & Airline We think you know... * No Annual Fee Credit Cards. * Credit Cards with 0% Intro APR Offers. * Credit Cards with 0% Balance Transfer Offers. * Cre…
[28] Costco Anywhere Visa Card by Citi vs Delta SkyMiles® Reserve American Express Card - Your Best Credit Cards
Tavily · web · 2026-08-06
https://yourbestcreditcards.com/card-comparison/costco-anywhere-visa-card-by-citi-vs-delta-skymiles-reserve-american-express-card
Citi Prestige vs Capital One Venture X Credit Card - Your Best Credit Cards https://yourbestcreditcards.com/card-comparison/citi-prestige-vs-capital-one-venture-x-credit-card Citi Prestige vs Capital One Venture X Credit Card - Your Best Credit Cards. BROWSE CREDIT CARDS BY TYPE By Spend Category To find the best cards for your spend By Issuer If you know what you’re looking for By Key Feature i.e. lounge access, 0% intro APR, etc By Hotel & Airline We think you know... * No Annual Fee Credit Cards. * Credit Cards with 0% Intro APR Offers. * Credit Cards with 0% Balance Transfer Offers. * Cre…
[29] FMP fmp_income_statement
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/income-statement/COF?period=annual&limit=5
[30] FMP fmp_balance_sheet
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/balance-sheet-statement/COF?period=annual&limit=5
[31] FMP fmp_cash_flow_statement
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/cash-flow-statement/COF?period=annual&limit=5
[32] FMP fmp_ratios
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/ratios/COF?period=annual&limit=5
[33] FMP fmp_key_metrics
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/key-metrics/COF?period=annual&limit=5
[34] EDGAR 10-K Item 7
EDGAR · edgar · 2026-08-06 · Item 7
https://www.sec.gov/Archives/edgar/data/927628/000092762826000024/cof-20251231.htm#i7687b40c0e86405a90048ec815ca20f5_184
illion in 2025 primarily driven by lower margins. •Non-Interest Income: Non-interest income increased by $111 million to $1.3 billion in 2025 primarily driven by our capital markets business. •Provision for Credit Losses: Provision for credit losses increased by $279 million to $287 million in 2025 primarily driven by a net allowance build compared to a net allowance release in 2024. •Non-Interest Expense: Non-interest expense remained substantially flat at $2.0 billion in 2025 compared to 2024. Loans Held for Investment: •Period-end loans held for investment increased by $2.1 billion to…
[35] EDGAR DEF 14A Director Compensation Director Compensation
EDGAR · edgar · 2026-08-06 · Director Compensation Director Compensation
https://www.sec.gov/Archives/edgar/data/927628/000119312526124131/d701179ddef14a.htm#txa77
connection with the Discover Transaction, the largest acquisition-related community commitment plan in the history of the U.S. banking system   ∎ Associates volunteered hundreds of thousands of hours with nonprofit organizations in the communities where Capital One operates, including over 18,000 hours of pro bono services   ∎ Ranked #36 on Fortune magazine’s 100 Best Companies to Work For ® list in 2025, which marks Capital One’s 14th consecutive year on this list | CEO Contribution to Company Performance In determining the CEO’s 2025 year-end incentive awards, in addition to the Company’s…
[36] EDGAR 10-K Item 9A
EDGAR · edgar · 2026-08-06 · Item 9A
https://www.sec.gov/Archives/edgar/data/927628/000092762826000024/cof-20251231.htm#f-5037
Item 9A. Controls and Procedures Overview We are required under applicable laws and regulations to maintain controls and procedures, which include disclosure controls and procedures as well as internal control over financial reporting, as further described below.  (a) Disclosure Controls and Procedures Disclosure controls and procedures refer to controls and other procedures designed to provide reasonable assurance that information required to be disclosed in our financial reports is recorded, processed, summarized and reported within the time periods specified by SEC rules and forms and …
[37] Python calc
Calcul · calculation · 2026-08-06
calculation://calc_python
Calcul : net_income_2024 = 4750000000 net_income_2025 = 2453000000 revenue_2024 = 53938000000 revenue_2025 = 69252000000 ni_change_pct = (net_income_2025/net_income_2024 - 1) * 100 rev_change_pct = (revenue_2025/revenue_2024 - 1) * 100 ni_change_pct, rev_change_pct — {"code": "net_income_2024 = 4750000000\nnet_income_2025 = 2453000000\nrevenue_2024 = 53938000000\nrevenue_2025 = 69252000000\n\nni_change_pct = (net_income_2025/net_income_2024 - 1) * 100\nrev_change_pct = (revenue_2025/revenue_2024 - 1) * 100\nni_change_pct, rev_change_pct"}
[38] Lessons for Antitrust from the Capital One-Discover Merger: Is There a Subprime Market in Credit Cards? - International Center for Law & Economics
Tavily · web · 2026-08-06
https://laweconcenter.org/resources/lessons-for-antitrust-from-the-capital-one-discover-merger-is-there-a-subprime-market-in-credit-cards
[39] Order Approving the Acquisition of a Bank Holding Company, the Merger ...
Tavily · web · 2026-08-06
https://www.federalreserve.gov/newsevents/pressreleases/files/orders20250418a2.pdf
[40] [PDF] Letter re capital One/Discover merger - Regulations.gov
Tavily · web · 2026-08-06
https://downloads.regulations.gov/OCC-2024-0006-0003/attachment_1.pdf
[41] Capital One Navigates Regulatory Maze in Discover Financial Acquisition
Tavily · web · 2026-08-06
https://www.reportlinker.com/article/9217
[42] Letter to Regulators re CapOne/Discover
Tavily · web · 2026-08-06
http://www.economicliberties.us/wp-content/uploads/2025/04/Letter-to-Regulators-re-CapOne_Discover-4.04.25.pdf
[43] Python calc
Calcul · calculation · 2026-08-06
calculation://calc_python
Calcul : rev_2016=27519 rev_2025=69252 cagr=(rev_2025/rev_2016)**(1/9)-1 cagr — {"code": "rev_2016=27519\nrev_2025=69252\ncagr=(rev_2025/rev_2016)**(1/9)-1\ncagr"}
[44] FMP fmp_cash_flow_statement
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/cash-flow-statement/COF?period=annual&limit=10
[45] FMP fmp_balance_sheet
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/balance-sheet-statement/COF?period=annual&limit=10
[46] EDGAR DEF 14A Pay Versus Performance Pay Versus Performance
EDGAR · edgar · 2026-08-06 · Pay Versus Performance Pay Versus Performance
https://www.sec.gov/Archives/edgar/data/927628/000119312526124131/d701179ddef14a.htm#ixv-33905
nd marketing as well as short-term or cyclical factors such as credit provision or the level of interest rates. As such, the Committee and the Independent Directors believe that net revenue and the growth of net revenue are valuable complements to earnings-driven measures and important indicators of Capital One’s financial strength, growth trajectory, and long-term sustainability.   | CAPITAL ONE FINANCIAL CORPORATION | | 2026 PROXY STATEMENT | | 119 | | --- | --- | --- | --- | --- | Table of Contents EXECUTIVE COMPENSATION     Section VIII - Equity Compensation Plans     Equity Com…
[47] FMP fmp_key_metrics
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/key-metrics/COF?period=annual&limit=6
[48] FMP fmp_ratios
FMP · api · 2026-08-06
https://financialmodelingprep.com/api/v3/ratios/COF?period=annual&limit=6
[49] Python calc
Calcul · calculation · 2026-08-06
calculation://calc_python
Calcul : revenue_2016 = 27519000000 revenue_2025 = 69252000000 years = 2025 - 2016 cagr = (revenue_2025 / revenue_2016) ** (1/years) - 1 shares_out_2025 = 540700000 available_shares = 52043397 dilution_pct = available_shares / shares_out_2025 round(cagr, 4), round(dilution_pct, 4)
[50] Python calc
Calcul · calculation · 2026-08-06
calculation://calc_python
Calcul : ni_2024 = 4750000000 ni_2025 = 2453000000 pct_change = (ni_2025/ni_2024 - 1) * 100 round(pct_change,1) — {"code": "ni_2024 = 4750000000\nni_2025 = 2453000000\npct_change = (ni_2025/ni_2024 - 1) * 100\nround(pct_change,1)"}
[51] Yahoo Finance ownership for COF
Yahoo · api · 2026-08-06
https://finance.yahoo.com/quote/COF/holders
[52] EDGAR DEF 14A Audit Committee Audit Committee
EDGAR · edgar · 2026-08-06 · Audit Committee Audit Committee
https://www.sec.gov/Archives/edgar/data/927628/000119312526124131/d701179ddef14a.htm#txa701179_62
Audit Committee shall not be deemed to be soliciting material or filed with the SEC and is not incorporated by reference into any of Capital One’s previous or future filings with the SEC, except as otherwise explicitly specified by Capital One in any such filing.   | CAPITAL ONE FINANCIAL CORPORATION | | 2026 PROXY STATEMENT | | 127 | | --- | --- | --- | --- | --- | Table of Contents STOCKHOLDER PROPOSALS     Section XII - Stockholder Proposal (Item 4 on Proxy Card)     Stockholder Proposal Requesting Shareholder Approval Requirement for Excessive Golden Parachutes (Item 4 on Proxy …
[53] Yahoo Finance shares for COF
Yahoo · api · 2026-08-06
https://finance.yahoo.com/quote/COF
[54] Python calc
Calcul · calculation · 2026-08-06
calculation://calc_python
Calcul : value=377.22877975776595 price=220.0 mos_pct=(value-price)/value combined = (1+0.15)*(1+mos_pct)-1 annualized_5 = (1+combined)**(1/5)-1 annualized_10 = (1+combined)**(1/10)-1 mos_pct, combined, annualized_5, annualized_10 — {"code": "value=377.22877975776595\nprice=220.0\nmos_pct=(value-price)/value\ncombined = (1+0.15)*(1+mos_pct)-1\nannualized_5 = (1+combined)**(1/5)-1\nannualized_10 = (1+combined)**(1/10)-1\nmos_pct, combined, annualized_5, annualized_10"}
[55] Grille interne de scoring Buffett
Bull Run Internal Scoring · internal · 2026-08-06
internal://buffett-score-grid