American Express (AXP)
American Express Company (NYSE: AXP)
American Express (AXP) is one of the world's leading financial-services companies, operating uniquely as both the lender and the payment network for an affluent, loyal base of cardmembers who spend several times the industry average. With a rare closed-loop model and a premium brand recognised worldwide, American Express has established itself as one of the most durable franchises and most closely followed businesses in global equity markets.
However, a compelling story is not the same as a sound investment. The real question is not where the price sits on any single day, but what this business is truly worth across the full range of outcomes ahead, and whether that value offers a margin of safety.
That's the question this analysis sets out to answer. We begin by understanding American Express's business model, competitive position and financial performance before valuing the company through a complete discounted cash flow model built entirely from its own financial statements. By the end of the analysis, the reader will not only understand how American Express generates value, but also whether its current valuation represents an attractive long-term investment opportunity.
American Express Company (NYSE: AXP)
INTRODUCTION TO AMERICAN EXPRESS
American Express is one of the oldest and most durable payment franchises in the world. Founded in 1850 as an express-shipping business, it invented the traveller’s cheque in the 1890s and the plastic charge card in 1958. Today it operates the only large closed-loop card network in the United States, which means it owns both the relationship with the cardholder and the relationship with the merchant. Understanding AXP requires first understanding what a closed-loop network is, why it can charge merchants more than the open-loop networks that compete with it, and how the same company also runs a lending business on top of the card franchise.
The Seven-year-old Explanation
American Express runs a credit-card system where it owns both ends. When a Card Member swipes an American Express card, the merchant pays a small fee to American Express, and American Express also collects the annual fee the Card Member pays to have the card. Almost every other credit-card company only owns one end of the deal.
The customers American Express wants are the ones who spend heavily. Business travellers, affluent leisure travellers, and small businesses that put every expense on a card. That is why American Express cards come with lounges at the airport, hotel credits, and points that can be swapped for flights.
American Express also lends money. When a Card Member carries a balance on the card from one month to the next, American Express charges interest, exactly the way a bank does. It funds those loans by taking deposits from savings customers. So it is in substance a payment network and a bank, running side by side under one roof.
On 24 July 2026, American Express had a very good quarter. Revenue grew 10%, earnings per share beat what analysts expected, and management raised its forecast for the full year. Younger customers (Gen-Z and Millennials) are joining faster than any other age group. The Platinum card, which costs $795 a year, is the fastest-growing part of the U.S. Consumer business.
The reason to look at American Express carefully today is that the stock went down about 4% right after that very good quarter. Investors were hoping for even more. The small drop turned a fairly-priced business into a slightly-cheap one for anyone patient enough to wait for the growth to compound.
This one paragraph is the whole investment thesis. Every pillar below either confirms or challenges the claim that a wide-moat 175-year franchise trading modestly below intrinsic value, growing 10%, and accelerating in premium cohorts is worth owning at $326.17.
The Investor Explanation
American Express is a closed-loop, four-party payment network combined with an issuing bank, headquartered in New York and founded in 1850. It reports four business segments and the way those four contribute to revenue and profit tells the whole story of the current phase of the company. U.S. Consumer Services is the largest at 48% of FY25 revenue and it is accelerating (Q2 2026 billed business +11% versus +9% in Q1). Commercial Services is 23%, growing mid-single-digit steady through the cycle. International Card Services is 18%, the acceleration line that added Swisscard consolidation in January 2026. Global Merchant and Network Services is 11% but carries the highest margin at roughly 70% pretax because it earns pure network fees with no credit exposure.
On the funding side, AXP operates as a bank does. It has $157 billion of customer deposits funding a $229 billion Card Member loan portfolio, earning a net interest yield of 8.1%. Regulatory capital ratios are best-in-class: CET1 at 10.4%, Tier 1 Leverage at 9.6%, both well above regulatory minimums. Stephen Squeri has been Chief Executive Officer since February 2018 and is the primary architect of the current Membership Model strategy.
American Express runs the only large closed-loop card network combined with an issuing bank. The 175-year premium franchise compounds through Card Member fees, discount revenue on every transaction, and net interest income on loans. At $326.17 the market trades 18% below Base intrinsic value but the cushion still sits below the 30% hurdle for a wide-moat compounder.
PILLAR 1: CIRCLE OF COMPETENCE
Before a company can be valued, its sources of revenue must be understood, along with who its competitors are and what forces shape the industry it operates in. This pillar builds that understanding for American Express.
Revenue Mix
Revenue mix is the share of a company’s total revenue that comes from each of its business lines. Understanding it matters because a shift in the mix changes the whole character of a business. A company that sells commodity products at thin margins looks very different from a company that sells premium products at fat margins. Watching a company’s mix shift over time is one of the earliest signals an investor gets about which direction the business is heading.
In AXP’s case the revenue mix has been remarkably stable, which is itself a positive signal. U.S. Consumer Services generates 48% of revenue. Commercial Services 23%. International Card Services 18%. Global Merchant and Network Services 11%. What is changing is the growth mix inside those segments: Q2 2026 printed USCS billed business at +11% year-on-year (up from +9% in Q1), with Gen-Z spend +40% and Millennials +14%. The premium franchise is accelerating in its most valuable demographic, and that shift matters more than the segment-share number alone would suggest.
For context, Visa and Mastercard earn only network fees. No lending, no card fees. JPMorgan Chase’s card business is buried inside its Consumer & Community Banking segment and does not report separately. Capital One, post-Discover, now has closed-loop scale but a very different customer mix that skews toward subprime. AXP is the only large public company with the exact combination of premium consumer cards, corporate cards, and a wholly-owned closed-loop network.
The Four Business Segments
The first segment, U.S. Consumer Services, is AXP’s earnings engine. It contains the Platinum ($795 annual fee), Gold ($325 annual fee), and Centurion ($5,000 annual fee) consumer cards along with the deposit and lending relationships those cards create. Q2 2026 billed business grew 11% year-on-year, up from 9% in Q1. The Platinum portfolio is now the fastest-growing part of the USCS book, and the Gen-Z age cohort spent 40% more year-on-year (against Millennials at 14% and Baby Boomers at 5%). This is the single most important data point in the whole quarter: the premium cohort is expanding, not contracting.
The second segment, Commercial Services, is AXP’s small-business and large-corporate card book. It represents 23% of revenue and grows in the mid-single-digit range through the business cycle. Q2 2026 billed business grew 5% FX-adjusted, up from 4% in Q1, on U.S. small-business strength. This is a steady, cash-generative segment with limited cyclicality (business travel and B2B spend recover faster than consumer discretionary in a slowdown).
The third segment, International Card Services, is the acceleration line. It represents 18% of revenue and delivers premium consumer and small-business cards outside the United States. Q1 2026 billed business grew 20% on the Swisscard joint-venture consolidation; Q2 2026 normalised to 13% as the Swisscard first-quarter effect anniversaried. The underlying international premium card market is under-penetrated, which gives ICS the longest growth runway of the four segments. This is why the My Case scenario in Pillar 6 tilts ICS to Bull.
The fourth segment, Global Merchant and Network Services, represents 11% of revenue but generates roughly 70% pretax margin because it carries no credit risk. It is the pure network layer of the closed-loop system: every time an AXP card swipes at a merchant anywhere in the world, GMNS collects the discount fee. The proposed TheFork acquisition (a European restaurant booking platform with 50,000 restaurants across 11 countries) would expand the merchant footprint into European dining, one of the most durable spending categories.
How AXP Earns a Dollar of Profit
Not every dollar of AXP’s revenue is equally profitable. The highest-margin dollar comes from a merchant discount fee on a Platinum swipe by a wealthy customer, because it combines the pricing power of the closed-loop network (2.3% discount rate versus 1.5-1.8% for the open-loop networks) with the spend intensity of a premium cardholder (roughly $25,000 per year on AXP against an industry average near $4,000). The lowest-margin dollar comes from interest income on a revolving small-business balance during a credit cycle, where credit costs offset most of the yield.
Pretax margin printed 20.7% in Q2 2026, well above the FY25 print of 19.1% and moving toward the historical peak of 22% last reached in FY18 and FY24. This lift is what management calls Variable Customer Engagement (VCE) leverage: as the engaged Card Member cohorts scale, rewards spend grows more slowly than revenue, and the margin gap widens. Whether VCE leverage delivers is the central conviction question on AXP, and this margin trajectory is where the answer will show up.
Market Share
Market share is the percentage of a defined market that a company captures. It matters because it shows how customers are voting with real orders rather than with survey responses. A company that is gaining share against a stronger incumbent is usually improving its underlying economics faster than the headline revenue number suggests. A company that is losing share tends to see profit compression well before the revenue starts to fall.
AXP’s market-share story has two chapters that need to be read together. In premium consumer cards (Platinum tier and above), AXP holds a dominant position: roughly 60% of U.S. premium credit-card spend by some estimates, against JPMorgan Chase Sapphire Reserve and Capital One Venture X. In the broader U.S. credit-card market including subprime, AXP is a small player at roughly 8% of purchase volume, with Visa and Mastercard splitting most of the remainder. AXP is deliberately positioned at the premium end and does not chase mass-market share.
The premium share matters more than the broad share because premium is where the profit sits. A cardholder spending $25,000 per year at a 2.3% discount rate generates $575 of merchant revenue plus roughly $700-800 of net card fees, versus perhaps $60 of profit on a mass-market subprime cardholder. AXP’s premium concentration is the source of both its wide moat and its Buffett-style compounding history.
The Competitive Landscape
AXP competes against three specific companies in three specific ways.
Against Visa and Mastercard, AXP is a closed-loop premium player against open-loop mass-market networks. Visa and Mastercard collect roughly 1.5 to 1.8% of every transaction (interchange plus network fees combined), versus AXP’s 2.3%. AXP earns more per transaction but has smaller acceptance footprint internationally. Visa and Mastercard have essentially universal acceptance; AXP has excellent U.S. acceptance and improving international acceptance but still trails in certain emerging markets. This gap is narrowing every year but does not close.
Against JPMorgan Chase, AXP faces the most direct premium-card competitor. Sapphire Reserve at $795 annual fee (matching Platinum) plus the Sapphire Lounge network are aimed squarely at AXP’s core cohort. AXP differentiates on deeper brand equity (175 years versus 25 years for Sapphire Reserve), closed-loop economics (JPM pays interchange to Visa/Mastercard on every swipe; AXP does not), and the wider Membership Model network of lounges and partnerships. JPM’s advantage is the 80 million U.S. retail bank account distribution that lets it cross-sell Sapphire Reserve to existing depositors.
Against Capital One, post-Discover, AXP faces the only U.S. peer with closed-loop scale. Capital One is aggressive in Venture X and Savor cash-back cards and now owns Discover’s network. AXP’s differentiation is the Centurion-Platinum premium cohort, which Capital One does not serve. Capital One’s advantage is subprime lending reach, which AXP deliberately does not pursue.
Customer Concentration
Customer concentration is the share of a company’s total revenue that comes from a small number of large customers. Investors watch this metric because losing a single large customer can rebase the whole revenue trajectory of a business, and a very concentrated business tends to trade at a lower valuation multiple to compensate for that binary risk.
AXP’s customer concentration is exceptionally low compared with most large-cap financials. There are approximately 155 million cards in force serving roughly 50 million cardholders. The single largest customer represents a fraction of one percent of revenue. The top 10 customers combined would not reach 5%. This is the opposite of the hyperscaler-concentration risk that overhangs semiconductor companies like AMD or Nvidia.
The customer concentration is not on the demand side but on the merchant side, and even there it is modest. The largest merchant categories (travel and entertainment, and general goods and services) each represent roughly 30 to 40% of billed business, but no single merchant is remotely material. This diversity is a Buffett-style moat feature.
Top 3 Customers
The Wealthy U.S. Consumer is AXP’s primary customer archetype. This customer holds a Platinum or Gold card, spends approximately $25,000 per year on the card, and values the Membership Rewards points programme, the Centurion Lounge network, and the concierge services. This customer is deeply sticky: retention on Platinum-tier cards exceeds 98% after the third year, and the annual fee is rarely negotiated. Q2 2026 data shows this cohort spending 11% more year-on-year, with the Millennial and Gen-Z sub-cohorts spending 14% and 40% more respectively.
The Small and Mid-size Business customer is AXP’s second most important archetype. This customer holds one of the Commercial Services SME cards, uses it for expense management and 30-day float, and typically spends higher per-transaction than the consumer segment because business purchases are larger. This customer is less cyclical than consumer discretionary, which stabilises the whole business through downturns.
The Large Corporate customer, served by Global Commercial Services, buys AXP for cash-management software, ERP integration, and global expense-reporting infrastructure. These relationships often extend more than 20 years, with multi-decade contracts. This is not a card sale; it is a workflow software sale attached to a card. The switching costs for a large corporate to move from AXP to a competing programme are measured in months of migration work.
The Economics of This Industry
The card-network-and-issuing-bank hybrid industry has a distinctive economic profile. Five features are worth understanding before moving on.
First, cyclicality is moderate to high. Consumer discretionary spending is procyclical, and AXP’s Card Member loan portfolio is exposed to unemployment cycles through the Net Charge-Off rate. Q2 2026 printed 2.0% NCO principal-only, which is best-in-class among major card issuers and flat year-on-year despite an economy that some observers describe as late-cycle. Historically NCO can rise to 3.5 to 4.0% in a serious recession.
Second, capital intensity is low. FY25 CapEx was $2.4 billion at 3.4% of revenue, almost entirely on technology and digital servicing. Unlike traditional retail banks, AXP has no physical branches; unlike semiconductor companies, it has no factories. The capital structure is dominated by regulatory capital held against the Card Member loan portfolio, which is more efficient than most bank models.
Third, customer concentration is exceptionally low as already discussed above.
Fourth, supplier concentration is not meaningful because the business is not physical-goods dependent. Funding comes from a diversified deposit base ($157 billion, 54% of total assets) plus long-term debt ($57 billion, 20%) and other sources. No single funding source dominates.
Fifth, the regulatory environment matters more than in most industries because AXP is a bank holding company supervised by the Federal Reserve. The single largest regulatory risk is the Credit Card Competition Act (bills S.1838 in the Senate and H.R.3881 in the House), which would require AXP and Visa to enable alternative network routing for merchant transactions. If enacted, the effective merchant discount rate on the closed-loop network could fall from 2.3% toward the open-loop level of 1.5-1.8%. This is the single largest tail risk in Pillar 10.
Verdict
AXP’s circle of competence is closed-loop premium card network plus issuing bank, anchored by U.S. Consumer Services and Global Merchant and Network Services. The competitive position is structurally durable on every front. The business is understandable and its revenue drivers are visible in every 10-Q. The two points to carry forward are the Platinum acceleration in USCS (positive for the Base case margin trajectory) and the Credit Card Competition Act regulatory tail (the single largest downside risk).
U.S. Consumer Services generates 48% of AXP revenue and is accelerating on Platinum acceleration and Gen-Z spending. International Card Services is the second growth engine at 18% of revenue and low double-digit growth. Regulatory risk from the Credit Card Competition Act is the single most important thing to keep in mind for every pillar below.
PILLAR 2: BUSINESS QUALITY AND DURABLE MOAT
Every serious equity investor is trying to answer the same question: does this business have some durable advantage that lets it earn returns above what an ordinary business earns, and will that advantage still be there in ten years? That advantage, if it exists, is called an economic moat. This pillar walks through the concept of the moat, the specific measurements that reveal whether it exists, and how AXP scores on each of them.
Return on Tangible Common Equity versus the Hurdle
For a bank or a financial holding company like AXP, the cleanest test of business quality compares two numbers: return on tangible common equity (ROTCE) and cost of equity. Tangible common equity strips out goodwill and intangibles from the equity base, giving the honest capital that shareholders have actually put into the business. ROTCE measures the profit earned on that tangible base.
If ROTCE is above the cost of equity, the business is creating value on every retained dollar of earnings. Its owners are getting more than the market would give them elsewhere at similar risk. If ROTCE is below the cost of equity, the business is destroying value on every dollar it reinvests. The gap between ROTCE and cost of equity, measured in percentage points, is the single most important measurable quality metric of a financial business.
AXP’s Q2 2026 ROTCE was 37.8%. Cost of equity is 9.58% (we build this number in Pillar 6). That is a spread of 28.2%age points per dollar of tangible equity. It is not a good result; it is a great result. It is the mathematical signature of a wide-moat compounder that earns significantly above its cost of capital every single year.
For context, JPMorgan Chase’s consolidated ROTCE runs near 20%. Capital One’s runs near 12%. Bank of America’s runs near 14%. AXP’s 37.8% is roughly double the next-best major bank. This gap is not new: AXP ROE has printed 31.7%, 31.4%, 34.5%, and 33.9% for FY22 through FY25 respectively. Durable through the rate cycle without compression.
ROTCE clearing cost of equity by 28%age points is the single most important measurable in the whole thesis. It is the reason AXP compounds book value at double-digit rates without needing to deploy any incremental external capital, and it is the reason buybacks executed below intrinsic value keep creating shareholder value quarter after quarter.
Net Interest Yield
Net interest yield (NIY) is the interest AXP earns on its Card Member loan book, expressed as a percentage of average loans, minus the interest it pays to fund those loans. It is the second-cleanest measure of pricing power for a lender. Businesses with strong deposit franchises sustain high NIY because their funding costs stay low even when rates change; businesses with weaker deposit franchises see their NIY compress when funding markets tighten.
AXP’s Q2 2026 NIY was 8.1%, up from 6.7% in FY22 as the rate cycle lifted asset yields on the Card Member loan book faster than funding costs on the deposit base. The gap is a durable feature of the AXP model: deposits are sticky Card Member relationships rather than rate-shopping cash, so they do not reprice as fast as the loan book. For comparison: Capital One NIY ~6.5%, JPMorgan Chase consumer bank ~5%, Bank of America ~2.2%.
The Four Kinds of Moat, Applied to American Express
The financial analyst Pat Dorsey codified competitive advantage into four categories, and this framework is a useful lens through which to view any business. The four categories are intangible assets, switching costs, network effects, and cost advantage. A business with none of these is a commodity. A business with one strong category has a moat. A business with two or more strong categories is a wide-moat business of the type Buffett prefers to own.
Intangible assets. Wide. The American Express brand has compounded for 175 years and commands premium pricing. The Platinum annual fee has been raised twice under Squeri (from $550 to $695 in 2021, from $695 to $795 in May 2025) each time with no measurable churn. Centurion sits at $5,000 with a waiting list. Retention exceeds 98% for tenured cardholders. This is the strongest single moat AXP has.
Switching costs. Medium. Membership Rewards points balances, status benefits, statement credits, and habit all create switching friction. A Platinum holder can move to a Sapphire Reserve, but the accumulated Rewards balance, the Centurion Lounge access, and the year-count-to-status create meaningful friction. Less than 2% of tenured cardholders switch to a competitor in any given year.
Network effects. Strong. The closed-loop is the structural moat. More premium Card Members attract more merchant acceptance; more merchant acceptance attracts more premium Card Members. This two-sided reinforcement has resisted every open-loop challenger for 50 years. No fourth global card network has emerged in that window despite repeated attempts (Discover has never scaled beyond the U.S., JCB is regional to Japan, UnionPay is regional to China).
Cost advantage. Not applicable. AXP competes on revenue per Card Member, not on unit cost. The Membership Model deliberately pays back roughly 40% of revenue in rewards, which is the price of the premium franchise. This is a design choice, not a weakness.
Moat Trajectory
The direction of a moat matters as much as its current width. AXP’s moat is widening. The Platinum refresh raised the annual fee 14% in May 2025 with no measurable churn, which is a direct market test of the pricing power. Premium card share within the AXP book rose from 64% in FY22 to 72% in FY25. ICS billed business grew 13% reported in Q2 2026 against a global card market growing at low-to-mid single-digit rates. Gen-Z spend +40% is the leading indicator of the next-generation cohort adopting the premium franchise, which extends the moat by another decade of Card Member relationships.
Verdict
AXP is a wide-moat business. Strong intangible assets (175-year brand, Platinum pricing power) plus strong network effects (closed-loop economics). The required margin of safety in Pillar 7 is the standard 30% quality-compounder threshold, not the 35-40% required for narrow-moat businesses. Widening moat plus Gen-Z acceleration could compress the required cushion toward 25% on Q3 confirmation.
AXP is a wide-moat business. The closed-loop network is a durable network effect, the Membership Model is a durable intangible-assets moat, and ROTCE at 37.8% runs 28%age points above cost of equity of 9.58%. This is the mathematical signature of a wide-moat compounder. Pillar 7 will require the standard 30% margin of safety.
PILLAR 3: MANAGEMENT QUALITY AND CAPITAL ALLOCATION
Two factors separate a great business from a merely good one over decades: the durability of its competitive advantages, and the quality of the people running it. Pillar 2 covered the first. This pillar covers the second.
Operational Track Record
When investors talk about management quality, what they usually mean is track record: the sum of the decisions a CEO has made over her or his tenure and how those decisions played out against the promises made at the time. Track record is the closest thing to hard evidence of management quality, and it matters enormously because a great operator can compound value over decades while a mediocre one in the same seat can destroy it in a few years.
Stephen J. Squeri became Chief Executive Officer of American Express in February 2018. He was an internal hire after 18 years running the company’s technology and B2B commercial functions. He inherited an AXP that was mid-transition after the Costco co-brand loss in 2016, with revenue of roughly $34 billion and a strained margin. Over eight years he has taken revenue from $40 billion to $72 billion, diluted EPS from $7.91 to $15.79, and pretax margin from a low near 15% to a Q2 2026 print of 20.7% that is heading toward the historical peak of 22%.
For contrast, JPMorgan Chase’s Consumer & Community Banking segment (run by Marianne Lake) has grown card revenue at low-single-digit compound rates over the same window. Capital One’s domestic card business (under Richard Fairbank) has grown mid-single-digit. AXP’s revenue-per-share growth of roughly 10% CAGR is roughly double the peer average.
Four Operational Wins Across Eight Years
The track record breaks down into four specific wins, each of which would be a career achievement in its own right.
Revenue doubling. Revenue grew from $40 billion in FY18 to $72 billion in FY25, an 8.7% compound annual growth rate. Nearly twice the industry average.
EPS doubling. Diluted EPS grew from $7.91 in FY18 to $15.79 in FY25, a 10.4% CAGR. Powered by revenue growth, margin recovery, and disciplined buybacks below intrinsic.
COVID resilience. FY20 revenue fell 17% as travel spend collapsed. Squeri suspended buybacks, sustained the dividend, held marketing spend, and prepared the business for recovery. AXP regained pre-COVID billed business by Q2 2022, one full year ahead of Mastercard.
Membership Model launch. The Membership Model repositioning turned AXP from a payment product into a lifestyle brand. Platinum fee raised twice under his watch (from $550 to $695 in 2021, from $695 to $795 in May 2025) each time with no measurable churn.
Capital Allocation
Capital allocation is how management deploys the cash the business generates. Every dollar of cash a business produces has to be spent somewhere: reinvested in the business, used to buy another business, paid out as a dividend, used to buy back shares, or held on the balance sheet. Capital allocation is the clearest read on whether management is serving owners or diluting them.
AXP reinvests approximately 12% of revenue back into technology, marketing, and Card Member benefits. That is a heavy reinvestment rate for a mature financial services company, and it is the right choice for a business whose product is the Membership Model itself. Any less would risk falling behind on card benefits (which is how JPMorgan gained ground with Sapphire Reserve). Any more would compress current margins without a commensurate return.
The buyback record is disciplined. FY22 through FY25 repurchases were $4.4B, $4.9B, $5.4B, and $5.9B respectively, at a trailing five-year average purchase price near $185. Today’s share price of $326.17 is 76% above that average purchase price. The earnings yield at $185 was 8.5%, comfortably above the 9.58% cost of equity. This is the discipline that generates the roughly 3% per year share-count shrinkage which lifts EPS growth above revenue growth.
M&A has been conservative. Swisscard was consolidated as a wholly-owned subsidiary in January 2026, small-ticket and immediately accretive. TheFork acquisition was proposed in Q2 2026: a European restaurant booking platform with 50,000 restaurants across 11 countries. No transformational large-cap M&A that would create integration risk. This is Buffett-style discipline about paying the right price.
Insider Ownership and Compensation
Squeri owns approximately 167,000 common shares plus 1.1 million restricted stock units and performance share units, worth roughly $400 million at current spot. Officers and directors combined hold 0.4% of float. This is not the founder-operator ownership structure Buffett prefers (where the CEO holds 10% or more of the business), but it is enough that Squeri’s personal fortune moves meaningfully with the share price.
Executive compensation is 80% variable pay. Of that, 30% is stock options and 50% is performance share units tied to revenue growth, EPS growth, and ROE against a peer set. The gap in the compensation design is the absence of an explicit ROTCE-versus-cost-of-equity hurdle, which would be the ideal alignment for a bank holding company. Return to shareholders is well-aligned; the specific quality measure could be sharpened.
Verdict
Excellent operator with grade-A capital allocation. Eight years of consistent execution on revenue, margin, and per-share value. Squeri continuity is itself a thesis pillar. The principal unmodelled risk is CEO succession: Squeri is 63 years old and there is no publicly identified successor. A poorly managed handover would justify a partial trim of the position.
Stephen Squeri is one of the best financial-services operators of his generation. Eight years of consistent execution on revenue, margin, and per-share value. Capital allocation is grade-A: buybacks below intrinsic, disciplined M&A, growing dividend. The biggest unmodelled risk is his eventual succession.
PILLAR 4: FINANCIAL HEALTH AND BALANCE SHEET (Q2 FY26)
The balance sheet reveals whether a business can survive the downside cases under consideration, and whether it can fund the upside cases in view. Weak balance sheets force companies to issue equity at the worst possible moments, which permanently dilutes existing shareholders. Strong balance sheets do the opposite: they let a company keep repurchasing shares in downturns, when everyone else is a forced seller. AXP is firmly in the second camp.
How Much Capital versus How Much Lending
The starting question for any bank balance-sheet analysis is different from an industrial company. For a bank, the key measure is not net debt versus cash, but capital adequacy: does the bank hold enough Common Equity Tier 1 (CET1) capital against its risk-weighted assets to absorb losses in a stress scenario? The Federal Reserve sets a minimum CET1 ratio for large bank holding companies at approximately 7% including buffers. AXP as of Q2 2026 sits at 10.4%, which means it holds approximately 240 bp of cushion above the regulatory floor.
In dollar terms, this is $27.8 billion of CET1 capital against $268.3 billion of risk-weighted assets. Tier 1 capital adds another $1.6 billion of preferred stock and hybrid instruments, taking Tier 1 to 11.0%. Total capital reaches 13.1% when Tier 2 instruments are included. The Tier 1 Leverage ratio, a simpler denominator-of-total-assets measure, stands at 9.6%, well above the 4% regulatory minimum.
For peer context: JPMorgan Chase CET1 ~15% (G-SIB buffer). Capital One ~13.5%. Bank of America ~11.5%. Wells Fargo ~11%. AXP’s 10.4% is on the lower end of the major-bank peer group by design: AXP has a smaller balance sheet, a more predictable revenue mix (fee-heavy), and lower stress-scenario loss expectations because its Card Member loan book skews to super-prime borrowers.
Credit Performance and Reserve Coverage
Q2 2026 credit metrics were the best-in-class in the major card issuer peer group. Net Charge-Off rate printed 2.0% flat year-on-year. Reserve coverage stood at 2.7% of Card Member balances, down from 3.0% in Q2 2025. The reserve reduction reflects an improving forward credit outlook: AXP’s super-prime cohort continues to perform well and the models are dialling back the provisioning that had been built up in 2023-2024.
For comparison: Capital One Domestic Card NCO ~5.5% (subprime-tilted). JPMorgan Chase Card Services ~3.0%. Discover ~4.5%. AXP’s 2.0% is not merely best-in-class; it is dramatically better than any peer with comparable scale, and this gap is the direct financial expression of the premium Card Member cohort.
The 30-plus days past due rate on Card Member balances stood at 1.2% in Q2 2026, down from 1.3% in Q1. The 90-plus days past billing rate on corporate cards was 0.4%, flat. These leading indicators are all pointing in the same direction: credit is improving, not deteriorating, despite a broader economic environment some observers characterise as late-cycle.
Funding Structure and Liquidity
AXP funds its $229 billion Card Member loan portfolio primarily through customer deposits. As of Q2 2026 deposits stood at $157 billion, or 54% of total assets. Long-term debt at $57 billion (20% of total assets), short-term borrowings at $2 billion, and other liabilities complete the funding mix.
Customer deposits are the cheapest and stickiest funding source in banking. AXP’s deposits are Card Member relationships rather than rate-shopping cash, which means they do not reprice as fast when interest rates change. This is why AXP’s Net Interest Yield expanded from 6.7% in FY22 to 8.1% in Q2 2026 as the Federal Reserve raised rates: deposit rates lagged loan rates, and the spread widened.
Long-term debt at $57 billion is laddered across a series of maturities running to 2036 and beyond. There is no debt maturity wall before 2028, which means AXP does not have to enter the bond market in any near-term year to refinance. This protects AXP from a spike in interest rates.
The liquidity buffer stands at $45.2 billion of cash and cash equivalents at Q2 2026. This is above both LCR (Liquidity Coverage Ratio) and NSFR (Net Stable Funding Ratio) regulatory minimums with material excess. Enough to fund one full year of operating expenses without any external financing.
Goodwill
When a company buys another company, it usually pays more than the target’s tangible assets are worth on paper. The excess price paid over tangible book value shows up as an intangible line called goodwill. Goodwill sits on the balance sheet indefinitely and is written down only when the acquired business demonstrably underperforms the price paid.
AXP’s balance sheet carries $4.5 billion of goodwill, on total assets of $308.2 billion. That is roughly 1.5% of total assets, which is very low by bank standards. JPMorgan Chase’s goodwill runs near 3.5% of assets. Bank of America’s runs near 7%. Capital One’s runs near 6%. AXP’s goodwill is low because it has grown organically rather than through large-cap M&A. The Swisscard consolidation added some, and the proposed TheFork acquisition will add more, but the goodwill share stays well below peers.
The low goodwill is a Buffett-style feature. It means AXP’s reported equity is honest rather than inflated by acquisition premiums. Tangible common equity as of Q2 2026 was roughly $30 billion, versus reported total equity of $34.3 billion. The gap of $4.3 billion is essentially the goodwill line.
Verdict
Fortress. CET1 at 10.4% gives ample regulatory cushion. Credit metrics are best-in-class. Funding is diversified and cheap. Goodwill is low. Balance sheet supports the thesis rather than constraining it. The Credit Card Competition Act (Pillar 10 Risk 1) is the one line that could rebase the whole revenue trajectory, but the balance sheet itself is not the source of that risk.
AXP’s balance sheet is a fortress by bank standards. CET1 ratio at 10.4% gives 240 bp of cushion above the regulatory minimum. Net Charge-Off rate at 2.0% is best-in-class among major card issuers. Balance sheet supports the thesis rather than constraining it.
PILLAR 5: EARNINGS QUALITY
Not every dollar of reported profit is a real dollar of cash. Companies with weak earnings quality use accounting choices (accruals, reserve reversals, non-recurring gain reclassifications) to make their reported profits look better than the cash they actually generate. Companies with strong earnings quality do the opposite: their cash generation runs at or above their reported profit, because their accounting is conservative. This pillar tests whether AXP’s earnings are the real thing.
The Gap Between Accounting Profit and Real Cash
In FY25, AXP reported net income of $10.8 billion. Over that same period the business actually generated free cash flow of $16.0 billion. The cash produced was roughly 48% higher than the profit reported. That is an unusually large positive gap for a large financial-services company.
The gap exists for a specific and mostly benign reason. AXP’s reserve build under CECL (Current Expected Credit Loss) accounting reduces reported net income even when actual charge-offs are lower than the reserve amount. In FY25 the reserve build was approximately $2.6 billion while actual charge-offs were roughly $4.3 billion, so the accounting charged $2.6 billion to the income statement that never left the business as cash. Additionally, growth in customer deposits added roughly $2 billion of operating cash flow that is not reflected in net income at all.
This gap is a structural feature of the AXP model, not a one-off. It has printed above 1.3× consistently for the past five years. As long as AXP grows its Card Member loan book (which requires reserve build under CECL) and its customer deposit base (which generates operating cash flow), FCF will continue to run meaningfully above net income.
How Operating Cash Converts into Free Cash
The other clean read on earnings quality is the ratio of operating cash flow to net income. AXP’s FY25 operating cash flow was $18.4 billion against net income of $10.8 billion. That is a ratio of 1.70×. A ratio above 1.0 is healthy. It says the working capital of the business (receivables, payables, and the reserve accounts specific to a card issuer) is scaling sensibly with growth rather than absorbing cash the way it does in businesses with hidden accrual problems.
For peer context: JPMorgan Chase OCF/NI ~1.4×. Capital One ~1.5×. Bank of America ~1.3×. AXP’s 1.7× is at the top of the peer set, reflecting the fee-heavy revenue mix (discount fees convert to cash immediately, unlike interest income which sits in receivables for 30 days).
The FCF/NI ratio at 1.48× is arrived at by subtracting CapEx ($2.4 billion in FY25, mostly technology and digital servicing) from OCF. AXP has no branches to fund, no factories to build, no equipment to depreciate at industrial scale. CapEx is a small line, and the resulting FCF is close to OCF.
Stock-based Compensation
One accounting item deserves special attention because it has been the largest source of hidden dilution in technology-company earnings: stock-based compensation, or SBC. SBC is the portion of employee pay issued as shares in the company rather than as cash. Because it does not leave the balance sheet as cash, many companies present a version of their earnings that adds SBC back to make cash generation look higher.
This is a mistake, and a serious one. Every share issued to an employee dilutes the ownership stake of every existing shareholder by a small amount. Over time these small dilutions add up to material ownership transfer from existing owners to employees. Any honest valuation charges SBC as a real cost, because it is a real cost, even though it is not a cash cost.
AXP spends about 0.76% of revenue on SBC each year, or $551 million in FY25. This is dramatically lower than the tech-sector norms. Nvidia runs 4 to 5% of revenue in SBC. AMD runs the same. Meta and Google run above 10%. AXP’s 0.76% reflects the fact that it is not competing for engineering talent against Silicon Valley; it is a New York-headquartered financial services company where compensation is dominated by cash bonuses tied to performance.
The Compoundex methodology charges SBC in full in the DCF and never adds it back. For AXP this is a relatively small conservatism because SBC is so low, but the discipline matters as a matter of principle: it represents the true cost of running the business.
Verdict
AXP’s earnings are clean. Real cash generation runs 1.48× reported profit, cash conversion is healthy, and SBC is charged as a real cost. Nothing about the earnings pattern relies on accruals or one-off gains. The cash the valuation discounts is the cash the business actually produces.
AXP’s reported profit is backed by real cash. Free cash flow runs 1.48× net income, operating cash flow runs 1.70×. Stock-based compensation is 0.76% of revenue and charged in full in the DCF, never added back. Nothing in the earnings rests on accruals or one-off gains.
PILLAR 6: VALUATION
This pillar values AXP three different ways: a full discounted cash flow model, a reverse-engineered version of the same model, and a shortcut method based on cash yield. All three agree that today’s price offers a real but thin discount to intrinsic value.
The Headline
My Case intrinsic value stands at $426.50 per share (equity value $290.9 billion divided by 682 million diluted shares) against a market price of $326.17 on 25 July 2026 (Q2 FY26 close). That is a fair-value margin of safety of +23.5%. Base case intrinsic value stands at $397.90 with a fair-value MOS of +18.0%. The market offers a real cushion below both, but the cushion is still short of the 30% required for a wide-moat compounder.
The Four Scenarios
Each row is the DCF-implied share price under its own revenue, margin, and discount-rate switches. Margin of safety is measured against intrinsic value at a spot price of $326.17 as of 25 July 2026. Bear is below spot; Base, My Case and Bull all clear spot. Only Bull clears the 30% wide-moat hurdle.
The Discount Rate (WACC)
The DCF discounts future cash back to today using a rate that represents what AXP’s investors demand at similar risk elsewhere. That rate is called the Weighted Average Cost of Capital, or WACC. For AXP it is 8.40% as of 25 July 2026. This is at the low end of the major financial-services peer range (JPMorgan Chase near 9.5%, Capital One near 10%, Bank of America near 9%) because AXP has a moderate debt-to-equity mix and a mid-range beta of 1.20.
Reverse DCF
The reverse DCF takes today’s price and solves for the growth and margins the market must already be assuming. For AXP at $326.17 with Base scenario cash flows, the implied WACC is approximately 10.50%, which is 210 bp above the peer-honest 8.40% that CAPM produces. In plain words: the market is pricing AXP as if the appropriate discount rate should be 2.1%age points higher than a sober CAPM build justifies.
This gap represents the market pricing risk for the Credit Card Competition Act, the credit cycle turning, and the possibility that Variable Customer Engagement leverage stalls. It is not a fair read on the underlying business quality, but it is what the market currently believes. When the direct DCF and the reverse DCF both say the price is a modest discount to intrinsic, the analytical case for owning AXP at $326.17 gets stronger.
The reverse DCF is the second independent check on the direct DCF. When both point in the same direction (a real but thin discount to intrinsic), the case is stronger than any single number could give. It also quantifies the specific risk premium the market is charging, which becomes the yardstick for tracking whether news over the next few quarters justifies or reduces that premium.
Owner-earnings Yield versus the Graham Hurdle
A shortcut check on any equity: divide free cash flow by market capitalisation. This gives the direct return the current price offers before any growth is added. AXP generates approximately $12.1 billion of FY26 net income against a market cap of $223 billion at 25 July 2026 close, which is a yield of 5.44%.
Ben Graham’s rule is that this yield should at minimum clear the risk-free rate plus the equity risk premium (in today’s market, 4.49% plus 4.24% = 8.73%). AXP falls short of the static Graham hurdle by 329 bp on this test. However, the shortfall matters less than it looks. AXP is growing net income at 14 to 16% annually. A wide-moat compounder earns its investment-grade return through the growth, not through the starting yield. Over a five-year holding period, the growth closes the gap and then some.
Triangulation
The three methods converge. The direct DCF says Base MOS is +18.0%, real but short of the 30% hurdle. The reverse DCF says the market is charging a 210 basis point risk premium above the fundamentals, which is meaningful but not extreme. The owner-earnings yield falls short of the static Graham hurdle by 329 bp but the 14 to 16% EPS growth closes that gap comfortably over the holding period. All three point to a modestly undervalued franchise. None points to a deeply discounted entry point. The disciplined action is to hold the existing position and wait for either a further pullback or Q3 confirmation of the Platinum acceleration.
Three separate valuation methods all agree. At $326.17 the market offers a real but thin cushion below every intrinsic value estimate except Bear. Base MOS is +18.0%, still 12%age points short of the 30% wide-moat hurdle. Bull now clears the hurdle for the first time this cycle at +40.0%.
THE DCF WALKTHROUGH FOR AMERICAN EXPRESS
The context behind each step. The story behind the specific numbers. The mechanics that drive the company toward its final intrinsic value.
PILLAR 7: MARGIN OF SAFETY
Margin of safety is the cushion between what a share is worth and what the market charges for it. If intrinsic value is $500 and the market charges $400, the buyer gets a 20% margin of safety. If the market charges $520, the buyer pays a 4% premium and has no cushion at all. The size of the required cushion depends on the quality of the business: wide-moat businesses need less, narrow-moat and cyclical businesses need more.
For AXP, the required cushion is 30%, the standard wide-moat compounder threshold. This is less than the 35 to 40% required for narrow-moat businesses like AMD, because AXP’s moat is genuinely wide (per the Pillar 2 verdict) and credit metrics are best-in-class. Today the Base MOS is +18.0%, twelve percentage points short of the hurdle but improving from +14.1% on 28 June as the price re-rate closed 4%age points of the gap.
A pullback into the $278 to $285 range would give the Base case a 30% cushion (clearing the hurdle). A pullback into the $250 to $260 range would give My Case a 40% cushion. Either would be the trigger to scale the position to the full 6 to 8% portfolio weight target.
Munger Placement
A wonderful business at a fair-to-slightly-cheap price. Munger’s framework is that wonderful companies at fair prices are the prize; fair companies at wonderful prices work only with a deep discount. AXP today sits on the boundary between the wonderful-at-fair quadrant and the wait quadrant. Q2 2026 moved it slightly toward the buy side, but the cushion is still short of the required 30%. The discipline is to hold the existing starter and wait for either a further pullback or Q3 FY26 confirmation of the Platinum acceleration.
The stock market is a device for transferring money from the impatient to the patient.
Attributed to Warren Buffett
The required cushion for AXP is 30% below intrinsic value (standard wide-moat threshold). Today the Base MOS is +18.0%, still 12%age points short of the hurdle. Bull now clears at +40.0% for the first time this cycle. Discipline is to hold, refuse to add, wait.
RISK MITIGATION AND WATCHLIST
Five thesis-killers. Each is high severity and high likelihood over a three-to-five-year holding period. If any fires it is a reason to trim or exit, not to hold and average down. The point of a written watchlist is to force pre-commitment to the exit rule before the news arrives.
1 · Credit Card Competition Act enacted
The Credit Card Competition Act (bills S.1838 in the Senate and H.R.3881 in the House) would require AXP and Visa to enable alternative network routing for merchant transactions. If enacted, the effective merchant discount rate on the closed-loop network falls from 2.3% toward the open-loop level of 1.5 to 1.8%. This breaks the Base case revenue projection in Pillar 6 by roughly 15% and drops Base intrinsic to the mid-$300s. The single largest tail risk in the whole document.
Monitor. Congress.gov bill tracker S.1838 and H.R.3881. AXP quarterly 10-Q Risk Factors section. National Retail Federation policy statements. First signal would be committee-vote scheduling in either chamber.
2 · Credit cycle with net charge-off rate sustained above 3.5%
AXP carries $229.5 billion of Card Member loans and other loans as of Q2 2026 (per Pillar 4). The net write-off rate at Q2 was 2.0% principal-only, flat year-on-year. If U.S. unemployment rises above 5.5% the NCO rate has historically climbed to 3.5 to 4.0%. That would trigger reserve builds and directly compress the pretax margin trajectory in Pillar 6.
Monitor. AXP monthly Credit Statistics disclosure. AXP 10-Q Card Member Receivables and Loans section. FRED unemployment rate. Federal Reserve G.19 consumer credit release.
3 · Premium card share lost to JPM Sapphire Reserve and Capital One Venture X
JPMorgan relaunched Sapphire Reserve at $795 annual fee (matching Platinum) with Sapphire Lounge expansion. Capital One post-Discover invests aggressively in Venture X. AXP Platinum and Gold annual fees generate roughly 15% of total revenues (per Pillar 1). A 20% share loss over three years would break the USCS revenue projection in Pillar 4 and the moat-widening trajectory in Pillar 2.
Monitor. AXP quarterly earnings supplements USCS billed business growth. JPM Consumer and Community Banking segment disclosures. Capital One Domestic Card segment disclosures. Any AXP Platinum retention comment on the earnings call.
4 · Pretax margin sustained below 18% for a full fiscal year
The Base case requires pretax margin to ramp from 19.1% in FY25 to 22.0% in FY30 (per Pillar 6). Failure of the Variable Customer Engagement leverage algorithm invalidates the core margin thesis. Q2 2026 printed 20.7% (materially ahead of schedule), but a slip below 18% for a full fiscal year would drag Base intrinsic to the low $300s.
Monitor. AXP 10-Q Operating Expenses by line item, in particular Card Member Rewards and Marketing. AXP earnings call remarks on Variable Customer Engagement. FactSet consensus margin trajectory.
5 · International FX or geopolitical shock to ICS revenue
ICS revenue at $13.0 billion is 18% of total (per Pillar 1). The Pillar 6 model assumes ICS compounds at 14 to 18% through FY30 driven by the Swisscard consolidation and premium adoption abroad. A material USD-strengthening cycle or a geopolitical break in European travel spending would compress this line and drag the Bull scenario in particular.
Monitor. Bloomberg DXY dollar index. AXP International Card Services segment table. AXP travel and entertainment billed business quarterly.
Five specific developments would break the thesis. Each is pre-committed as a sell trigger, so the exit decision is not made in the heat of news. Each cross-references the pillar where its assumption lives.
CORE GRAPHICS →
6 key graphics for American Express (AXP): visual trends for the actuals and the projected figures for total revenue, net income, revenue growth %, and operating margin.
FINAL THESIS
Current quality
A high-quality, wide-moat business at a discount to value. American Express owns a closed-loop premium network with an affluent, high-spend customer base, producing recurring discount revenue, sticky card fees, and a high return on equity. The central lever is spend: if premium-customer growth and spending compound and credit stays disciplined, revenue rises from about $72.2B toward $106B by 2030 with operating margins widening. The one caveat is credit risk on the lending book through the cycle.
What to focus on next quarter
Q3 FY26 reports in October 2026. The central read is billed-business (spend) growth, card-member acquisition, and the credit trend, net write-offs and delinquencies. Watch rewards cost as a share of revenue; a persistent rise there is what erodes the margin the thesis depends on.
Holding period and sell discipline
Three to five years for the thesis to compound through premium-spend growth, international scale, and per-share buybacks. The position is sold on a broken thesis, not on a falling price.
Closing
As of 25 July 2026 (Q2 FY26), at $326.17, American Express is a high-quality business at a discount. Against My Case intrinsic of $426.50 this offers a positive margin of safety, at 23.5 percent, a 30.8 percent upside to the market. The discipline here is to own, sized to the one real risk: that a credit downturn pressures the lending book.
In an open-loop network like Visa or Mastercard, the card issuer, the network, and the merchant's bank are separate companies splitting a thin fee. In a closed loop, one company is all three at once. It keeps the full economics of each transaction and, equally importantly, sees all the data, which sharpens underwriting, rewards, and merchant offers.
Every time a cardholder spends, the merchant pays American Express a small percentage of the purchase. This discount revenue is the largest and highest-quality line: it scales with spending, carries no credit risk, and rises with an affluent customer base that spends more.
When customers carry a balance, American Express earns interest on it, funded largely by customer deposits. Net interest income is the spread between the two. It is higher-return than discount revenue but brings credit risk, because some borrowers do not repay, which is why the company holds reserves.
A moat is what keeps competitors from eroding a company's returns. The wider and more durable it is, the longer high returns persist. Moats come from switching costs, network effects, scale, and brand, and the best businesses have several at once.
Compared with cost of equity, it is the cleanest test of value creation for a bank. Above cost of equity = value creation. Below cost of equity = value destruction.
AXP creates value at approximately 2× the rate of the next-best major bank. This is the mathematical signature of a wide-moat franchise, and it justifies the 30% MOS hurdle rather than the 35-40% applied to narrow-moat businesses.
Rising reserve coverage signals deteriorating credit outlook. Falling reserve coverage signals improving credit outlook or reserve release.
AXP’s falling reserve coverage reflects an improving forward credit outlook. AXP’s super-prime cohort continues to perform well.
It is the discount rate used in a DCF, and the hurdle a company must clear on ROTCE (for banks) or ROIC (for industrials) to create value.
Every dollar AXP reinvests must earn at least 8.40% to break even. AXP’s ROTCE at 37.8% clears this hurdle by 29%age points, which is the mathematical signature of a wide-moat compounder.
MOS is the discipline that makes value investing survive. Buffett: "When you build a bridge to hold 30,000 pound trucks, only 10,000 pound trucks drive over."
AXP required cushion is 30% (wide-moat standard). Current Base MOS of +18% is short of the hurdle by 12%age points but improving from +14% on 28 June. Bull clears the hurdle for the first time this cycle.
A DCF values the enterprise, the sum of all future cash flows to every provider of capital. To reach value per share, add cash, subtract debt to get equity value, and divide by shares outstanding.
Net debt shows what a company would owe if it used all its cash to pay down debt. For a lender, customer deposits fund the loan book and are treated separately, so the net debt here reflects only the corporate funding outside deposits.
A buyback returns cash to owners by shrinking the share count, which lifts earnings and value per remaining share, and creates value when done below intrinsic value. It is a core lever of American Express's capital return alongside the dividend.
When a business has high fixed costs, each additional dollar of revenue drops more to operating income once those costs are covered, so margins rise with scale.
Instead of forecasting to a value, a reverse DCF starts from the current price and solves for the assumptions that justify it, revealing what the market already believes so it can be judged.