American Express Company

American Express (AXP, USD): the economic engine is strengthening

American Express is a closed-loop premium payments franchise: it issues cards, signs merchants, runs the network, earns lending income, and layers on annual fees plus travel/lifestyle services. That matters because Amex captures more of the economics of a transaction than Visa/Mastercard-style open networks. Its core revenue streams are merchant discount revenue, net interest income on card balances, and card fees.

This is not a commoditized card lender. The DNA is: premium spend-centric customers + strong brand + rewards/benefits + merchant acceptance + data from both sides of the transaction. That data advantage improves underwriting, fraud control, marketing, and customer retention.

The direction is still good. As of Q2 2026, revenue rose 10 percent, billed business rose 9 percent, network volumes rose 9 percent, and EPS rose 11 percent year over year. Credit stayed excellent: write-offs were 2.2 percent, unchanged, and 30+ day past due balances improved to 1.2 percent from 1.3 percent. That is what a strengthening engine looks like: more spend, more balances, stable credit.

The model is mostly win-win. Cardmembers get rewards, service, status, and financing; merchants get affluent customers and higher spend; Amex earns premium economics. The tradeoff is that merchants pay higher discount rates, so the model only works if Amex keeps delivering better customers and spend productivity.

No major deterioration is obvious today. Watch three things: discount revenue growth, net card fee growth/retention, and credit metrics. If those hold, Amex is winning. If spend shifts away from premium cards or credit losses rise materially, the story weakens fast.

American Express operates in the global electronic payments and short-duration consumer/commercial credit market, and that market is still a tailwind. The core evolution is unchanged: cash and checks keep losing share, commerce keeps digitizing, and affluent consumers and SMEs increasingly want rewards, financing, security, and travel benefits bundled into one payment relationship.

The good news is that AmEx sits in a durable growth market. The harder truth is that it is not in a winner-take-all market: competition for premium customers is fierce, and regulators can pressure economics. Still, over the next few years the market backdrop looks more like a tailwind than a headwind.

Conclusion: American Express has a real moat, and it is modestly widening. The edge is not “payments exposure”; it is AmEx’s closed-loop model wrapped in a premium brand. Owning the cardmember, the merchant relationship, and the network gives AmEx better data, tighter rewards economics, stronger fraud controls, and more targeted merchant offers than pure networks or commodity issuers. That is hard to replicate.

Most recent financial data used: FY2025.

What is not a moat: favorable credit, travel recovery, or generous rewards by themselves. Those can reverse. The durable edge is the premium closed-loop ecosystem.

Conclusion: American Express’ financial strength is strong, not pristine: returns are elite, funding is conservative for a lender, and credit is excellent, but the model is still balance-sheet-levered and its cash-flow statement is less informative than for a non-financial company.

Using the most recent official data, ROE was 35.9% for the first half of 2026 and 36.4% in Q2 2026, far above any reasonable cost of equity and above large-card-issuer norms. That said, part of the brilliance is structural leverage; for a bank-card model, ROE matters more than classic ROIC. Debt looks prudent rather than desperate: long-term debt was $57.0 billion against $45.2 billion of cash and a large $157.0 billion deposit base, while CET1 was 10.4%. In a downturn, Amex should still be serviceable unless credit losses spike far beyond normal recession ranges.

Earnings quality is good. Net write-offs stayed at 2.0% and 30+ day delinquency improved to 1.2%, so profit is not being propped up by visibly deteriorating credit. For a lender, FCF conversion is not a decision-useful metric because receivables, deposits, and funding flows swamp industrial-style free cash flow.

Conclusion: American Express still has a solid reinvestment runway, but it is not an infinite one; this is a high-return franchise with enough internal uses for capital to support durable growth, yet mature enough that buybacks remain a necessary outlet for surplus cash. Using fiscal year 2025 data, the best opportunities are still organic: adding premium consumer and SME accounts, deepening loan and receivable balances, expanding merchant acceptance, and monetizing its closed-loop data through better rewards, pricing, and cross-sell. With worldwide billed business at 1670000000000 and 86600000 proprietary cards in force, the addressable base is still growing, especially outside the U.S. and in commercial spend.

Implied organic growth still looks like high single digits to low double digits, not because AmEx can endlessly add capital, but because it can keep redeploying into affluent customer acquisition and spend share at attractive economics. For a financial company, incremental ROE / ROTCE is the right lens, and AmEx still looks comfortably above its cost of capital.

American Express is a premium-card leader, not the category leader: against domestic issuers it has better brand, spend quality, and fee economics; against Visa and Mastercard it loses on acceptance and global ubiquity, but earns more per relationship because it owns the customer, the merchant link, and the lending spread. Using the most recent filed data: AmEx FY2025, Visa FY2025, Mastercard FY2025.

AmEx is likely gaining share in premium U.S. card spending, driven by younger affluent customers, SMB/corporate strength, and brand-led pricing power. But it is not taking global network share from Visa/Mastercard in any meaningful way; those rails remain the default. The outlook is favorable if AmEx sustains high-spend customer acquisition without sacrificing credit discipline. Valuation usually sits between banks and pure networks, which is fair: better business than most issuers, worse moat than Visa/Mastercard.

Conclusion: American Express looks shareholder-aligned, but not owner-operated. The culture is disciplined, capital return is excellent, and governance is solid for a large regulated financial, but insider ownership is modest and the insider-trading signal is weak.

Berkshire’s roughly one-fifth stake matters: it is a strong external vote for capital discipline and franchise durability. The main caveat is that recent insider activity has generally looked more like routine vesting/sales than conviction buying, so I would not use insider trading as a bullish signal here. Net: good stewards, excellent capital return, only average insider ownership.

Conclusion: AmEx management looks strong: disciplined on capital, solid on execution, and generally clean on ethics/accounting, with low evidence of governance-driven downside as of FY2025.

Capital allocation has been sensible rather than flashy. Management has favored organic reinvestment, rewards/product refreshes, dividends, and buybacks over empire-building; share count fell from 744192702 (Feb 2023) to 686614005 (Jan 2026), and there is no history of large, obviously value-destructive deals in the current era. Execution has also been good: management’s post-2022 growth framework has broadly been met or exceeded through 2025, helped by steady spend growth and resilient credit.

On ethics/transparency, the record is cleaner than most large financials. The cited filings show no error-correction restatement flags, no auditor disagreement disclosure, and clean SOX 404 attestation. Litigation exists, as it always does in consumer finance, but nothing in the filing appears framed as existential or thesis-breaking.

American Express looks fairly valued: the business is high quality, but at USD 204.47B of market cap the stock already discounts a lot of the good news.

For a lender-network hybrid like AmEx, earnings power plus tangible book is the right lens, not DCF. At today’s price, the stock trades around 18-19x trailing earnings and roughly 6.0x book. That is not distressed, and it is not bubble territory either; it is the kind of price you pay for a franchise with strong brand equity, affluent customers, disciplined underwriting, and persistent buybacks.

Management’s medium-term playbook has been roughly high-single to low-double-digit revenue growth and mid-teens EPS growth. Credibility is decent: from 2022 to 2025, revenue rose from USD 52.9B to USD 72.2B and diluted EPS from 9.85 to 15.38, helped by spend growth, fee income, and share count reduction. If AmEx can still deliver something close to 10% revenue CAGR and 11-13% EPS CAGR through 2031, the stock can work — but mainly through compounding, not multiple expansion.

The current price seems to embed high-single-digit to low-double-digit EPS growth with no major credit accident. That is reasonable, but it leaves limited margin of safety. My intrinsic value estimate is about USD 265B by 2031, implying only a moderate annualized return from here before dividends.

Liquidation value is unexciting. This is a financial company, so “cash minus debt” is misleading; the relevant floor is closer to tangible common equity of about USD 28.5B, and in a stressed wind-down even that could be marked down.

Conclusion: American Express still looks like a durable compounder, but not an open-ended one. This is more likely a 2-3x in 10 years if the premium-membership loop holds than a true multi-bagger.

The moat should last a long time because AmEx owns the cardmember, merchant, and network relationship together. That gives it better data, better underwriting, better rewards targeting, and a brand affluent customers still choose to carry. FY2025 billed business of 1670000000000 and 86600000 proprietary cards show that the loop still has scale.

Reinvestment still appears productive. More spending drives more merchant relevance; more merchants improve acceptance; more relevance supports annual fees, rewards economics, and cross-sell. Buybacks add a second compounding engine, as shares outstanding fell from 743000000 in 2022 to 686000000 in 2025.

What erodes first is not payments volume; it is premium pricing power. If AmEx has to spend materially more on rewards and marketing just to keep billed business growing, incremental returns will compress. In 10-20 years it should still matter competitively, but only if it remains the preferred high-spend wallet, not just another issuer riding generic network rails.

Thesis-break signal: sustained deceleration in card fee revenue and billed business per card, while credit losses and rewards intensity rise together. That would mean the brand flywheel is weakening, not just the cycle.

Risk Assessment

Conclusion: American Express is not a fragile business; most near-term worries are uncertainty, not impairment. The real permanent-risk case is a slow erosion of its premium closed-loop model, not a normal credit cycle.

The single biggest risk is structural compression of merchant discount and rewards economics. If AmEx stops being meaningfully better for affluent cardmembers and acceptable for merchants, its flywheel weakens on both sides. That would hurt returns more durably than a credit downturn.

Financial risk looks contained: FY2025 filings still show strong capital, deposit funding, and access to preferred markets in 2026. Governance risk appears low; I do not see fraud or related-party red flags in the cited filings. Latest official financial data used: FY2025, with an August 2026 capital-markets filing.

Final Verdict: TRACK

American Express is a very good business, but not a very good stock at USD 302.78. The core franchise is real: premium brand, closed-loop data advantage, strong spend from affluent customers, disciplined underwriting, and high returns on equity. But the market already knows this. At roughly USD 204470000000 market cap versus a prior base-case USD 265000000000 by 2031, the expected return looks merely acceptable, not compelling enough to deploy fresh capital today.

The key point is simple: this is not a fragile lender headed for impairment, but it is also not cheap. For a 5-10 year investor, permanent capital-loss risk looks moderate rather than high; the bigger issue is paying a full price for a business whose upside still depends on preserving premium pricing, cardmember loyalty, and merchant economics in a brutally competitive payments landscape.

The inversion case against this verdict is strong: if AmEx keeps winning younger affluent users, sustains double-digit revenue growth, and keeps shrinking share count, today’s price could look fine in hindsight. That is plausible. It is just not enough of a fat pitch.

For new investors, the right move is wait for a better setup rather than force an entry. For existing holders, the right default is HOLD: the business remains attractive enough to own, but not attractive enough to add aggressively. I would only buy in small tranches on meaningful weakness, not load up here.

Is the analysis accurate and complete? Not fully. Research further:

• latest quarterly delinquencies/net write-offs versus affluent-peer cohorts
• evidence that merchant acceptance gains are still improving unit economics
• how much of EPS growth is true operating leverage versus buybacks
• whether younger-cardmember acquisition remains high quality, not just high growth