American Express (AXP): 34% ROE, $15.1B FCF, and a 20× P/E below Visa and Mastercard (2026)

The pick

American Express Company (NYSE: AXP) is this week's candidate. It is a global payments and lending company: it issues cards, acquires merchant transactions, runs a payments network, and earns interest and fees from cardmember loans. Its operating segments are U.S. Consumer Services, Commercial Services, International Card Services, Global Merchant and Network Services, and Corporate and Eliminations. 1

The valuation snapshot uses the July 31, 2026 closing price of $336.25. At that price, American Express had a $227.07 billion market capitalization, a 20.41x trailing P/E, and a 52-week range of $288.34 to $387.49. The same data page showed a $341.70 premarket quote on August 3, so the multiples will move with the stock. 2

The case is easy to state: AXP combines a premium card franchise with a fast-growing payments-and-lending base, and it trades at roughly two-thirds of Visa's and Mastercard's trailing P/E. The caveat is equally concrete: its current P/E is about 9% above its own five-year average, while credit losses, antitrust litigation, and a $700 million acquisition create real monitoring work.

Scope note: This is a verified qualified candidate from the current U.S.-listed-company screen and the cited peer checks. The public data pages did not expose a reproducible complete-universe export using identical ROE, FCF, and valuation definitions, so this article does not claim that AXP was the only U.S. qualifier.

Screen check

The ROE figures below use the labeled annual ROE series from StockAnalysis, which calculates annual ROE using average shareholder equity. TTM means the twelve months ended June 30, 2026. FCF is the same mechanical series throughout: operating cash flow less capital expenditures. For a financial company, that FCF definition is less economically clean than it is for an industrial company because customer funding and lending cash flows run through the cash-flow statement. The screen still applies the same definition consistently. 3 4

Hard gate	Evidence	Result
ROE above 15% in each of the last 3 fiscal years	FY2023 31.74%; FY2024 34.74%; FY2025 33.99%; TTM 34.38%	Pass
Positive free cash flow	FY2023 $16.996B; FY2024 $12.139B; FY2025 $16.003B; TTM $15.052B	Pass
Reasonable valuation	Current trailing P/E 20.41x; forward P/E 18.11x; PEG 1.42x; FCF yield 6.63%	Pass, but not a deep discount

For this edition, I define a reasonable valuation as three conditions: trailing P/E no more than 10% above the average of the five completed fiscal-year P/Es, trailing P/E below the selected peer median, and positive FCF yield. AXP's five completed P/Es were 16.33x, 15.00x, 16.71x, 21.18x, and 24.05x, averaging 18.65x. The current 20.41x is 9.4% above that average, just inside the screen's boundary. The peer comparison is much more favorable, as shown below.

The high ROE also needs interpretation. American Express buys back shares and carries a lending balance sheet, so ROE is not the same as an unlevered return on an industrial operating business. The screen records the labeled ratio; the balance-sheet and credit sections explain what supports it.

What American Express actually sells

American Express sits in the middle of the transaction rather than merely licensing a card logo. It can acquire a merchant, issue a card, operate the network, price rewards, and use the resulting spending and repayment data to manage the relationship. The company describes its brand as a global payments and premium lifestyle platform built around trust, security, service, and Membership value, with relationships with millions of merchants. 5

That integrated model creates a different moat from Visa or Mastercard. Those networks are larger transaction rails, but American Express controls more of the customer proposition and takes direct credit exposure. Its advantage is the loop between premium cardmembers, differentiated rewards and service, merchant acceptance, and data-driven offers. The trade-off is that a downturn can hurt spending and credit at the same time.

Dining is one current example of the strategy. In June, AXP proposed buying TheFork from Tripadvisor for $700 million in cash, subject to customary adjustments, labor consultation, and regulatory approvals. The deal is expected to close before the end of 2026 and would add to the company's Resy and Tock dining assets, taking the estimated bookable-venue network to about 75,000. That is a product and engagement expansion, not a reason to assume immediate earnings accretion. 6

The sources checked do not establish a single reliable market-share figure, so I am not calling AXP the market-share leader. The more defensible claim is narrower: it has a differentiated closed-loop model, a premium brand, direct merchant and cardmember relationships, and a growing dining ecosystem. Visa and Mastercard remain the appropriate scale and valuation anchors, not interchangeable business models.

Revenue, earnings, and cash flow

American Express has compounded revenue across the last five completed fiscal years, although net income did not rise every year. Dollar figures below are in billions and use the StockAnalysis financial-statement series; the latest quarter is cross-checked against the company's SEC earnings filing.

Fiscal year	Revenue	YoY revenue	Net income	YoY net income	Net margin	FCF
2021	$34.630B	—	$8.060B	—	23.27%	$13.095B
2022	$42.967B	24.1%	$7.514B	-6.8%	17.49%	$19.224B
2023	$49.086B	14.2%	$8.374B	11.4%	17.06%	$16.996B
2024	$50.406B	2.7%	$10.129B	21.0%	20.09%	$12.139B
2025	$54.865B	8.8%	$10.833B	6.9%	19.74%	$16.003B
TTM to Jun. 30, 2026	$57.601B	—	$11.445B	—	19.87%	$15.052B

The pattern is better than a simple growth label suggests. Revenue grew 8.8% in FY2025, but net income grew only 6.9%; FCF recovered from its FY2024 trough but remained below the FY2022 peak. The cash-generation gate is therefore a quality check, not a growth claim.

American Express's operating margin is not a clean cross-company measure because interest income and funding costs are part of the core business. I use net margin and pretax margin instead: FY2025 net margin was 19.74% and pretax margin was 25.14%, versus 20.09% and 25.58% in FY2024. The TTM page shows a 19.87% net margin and 20.82% pretax margin. 4

The latest quarter was stronger on demand. Q2 2026 revenue net of interest expense was $19.637B, up 10% year over year; net income was $3.110B, up 8%; diluted EPS was $4.53, up 11%; and billed business rose 9% to $455.8B. The company raised full-year 2026 revenue-growth guidance to 10% and set adjusted EPS guidance at $17.30 to $17.90. 7

Credit costs were not absent from that result. Q2 provision for credit losses was $1.084B, down from $1.405B because of a reserve release, while card net write-offs increased 8% to $1.207B. The headline earnings beat is more durable if spending growth remains near 9% without a material deterioration in those loss numbers.

Valuation against history and peers

Company	Trailing P/E	Forward P/E	P/B	EV/EBITDA	PEG	FCF yield
American Express (AXP)	20.41x	18.11x	6.62x	n/a	1.42x	6.63%
Visa (V)	31.17x	25.32x	19.40x	24.07x	1.88x	3.13%
Mastercard (MA)	31.52x	26.94x	89.60x	23.77x	1.62x	3.33%

The two-peer medians are 31.35x trailing P/E, 26.13x forward P/E, 54.50x P/B, 23.92x EV/EBITDA, 1.75x PEG, and 3.23% FCF yield. AXP is cheaper on every comparable measure. The P/B median is distorted by Mastercard's very small post-buyback book equity, so it should not drive the conclusion. 8 9

EV/EBITDA is not reported for AXP because EBITDA is not a meaningful operating measure for a card issuer; the same page also shows interest-coverage and debt/EBITDA as not available. That is a limitation of the metric, not a hidden valuation pass. For AXP, P/E, P/FCF, PEG, credit performance, and capital ratios are more useful.

The valuation is reasonable rather than obviously cheap. The stock's 20.41x trailing P/E is near the upper edge of its own five-year range and 9.4% above the five-year average. The investment case needs the company's 10% revenue-growth guidance and mid-teens EPS growth to persist; if growth falls toward the low single digits, the historical premium becomes harder to defend.

Balance sheet and credit health

Measure	Latest reading
Cash and equivalents	$45.75B
Total debt	$59.09B
Long-term debt	$57.017B
Debt/equity	1.72x
Current ratio	1.55x
Interest coverage	n/a
Customer deposits	$156.973B
CET1 capital ratio	10.4%
Q2 principal-only card net write-off rate	2.0%
30+ days past due, consumer and small business	1.2%

Cash fell 22% from $57.937B at December 31, 2025, while customer deposits rose 5% to $156.973B and long-term debt fell 2% to $57.017B. The company says it maintains a $6.0B committed syndicated bank facility that matures on September 24, 2028. The key bank-style capital number is CET1, not an industrial company's interest-coverage ratio: Q2 CET1 was 10.4%, within management's stated 10% to 11% target range. 10

The Q2 credit picture was still controlled: principal-only consumer and small-business net write-offs were 2.0%, unchanged year over year, and 30+ day delinquency was 1.2%. My monitoring trigger—not company guidance—is a principal-only write-off rate above 2.5% for two consecutive quarters, especially if CET1 moves below the 10% target floor. That would tell an investor that the current earnings and valuation framework needs to be rebuilt.

Risks and red flags

Credit normalization. The provision decline was helped by a reserve release, while card net write-offs still increased. Watch the 2.0% write-off rate, the 1.2% 30+ day delinquency rate, and loan growth each quarter. A sustained move above the 2.5% monitoring trigger would pressure earnings and capital at the same time. 10
Merchant litigation and regulation. In two 2024 cases, merchants allege that AXP's anti-steering and non-discrimination provisions violate antitrust law; courts rejected motions to compel arbitration and appeals are pending. In the older Moskowitz case, a jury awarded $12.5M on one Illinois consumer-law claim, and the parties reached a settlement subject to court approval. AXP estimates a possible-loss range of $0 to $250M above accruals for disclosed proceedings, while warning that a ruling that materially increases merchant steering could affect results of operations. Monitor the First Circuit appeals and court approval of the settlement. 11
TheFork execution. The proposed $700M cash acquisition has labor-consultation and regulatory conditions and is expected to close before year-end 2026. Watch the closing date, final consideration, and whether management gives a measurable return or expense target; the announced 75,000-venue network figure is strategic context, not an earnings forecast. 6
Funding and capital. Debt/equity is 1.72x and cash was down 22% from year-end, even though deposits increased and CET1 stayed at 10.4%. The next hard date in the disclosed funding structure is the $6B facility's September 2028 maturity. A CET1 reading below 10% or a material rise in funding costs would change the risk/reward more than a small P/E move. 10
Insider and short-interest signals. The current snapshot shows insider ownership of 0.12% and short interest of 2.20% of float. The public data reviewed here does not provide a clean, recent open-market insider-buying cluster; compensation awards should not be treated as bullish trading evidence. Recheck Form 4 filings before acting. 2
Management changes. No material executive appointment or departure was identified in the current 2025 10-K and Q2 2026 filing excerpts. That is an absence of a disclosed change, not proof that no personnel movement occurred.

Catalysts and what to monitor next

Q3 2026 earnings — October 23, 2026, 8:30 a.m. ET. This is the next scheduled company earnings call in American Express's published calendar. The test is whether billed-business growth stays near the 9% Q2 rate while write-offs remain controlled. 12
TheFork closing before year-end. A successful close would extend the Membership Model into dining; delays or a higher final price would make the $700M capital allocation harder to assess. 6
Product refresh and spending growth. Q2 billed business grew 9%, net card fees grew 15%, and net interest income grew 11%. AXP's proposed and ongoing premium-product refreshes need to keep acquisition and retention strong without pushing credit losses higher. 7
Shareholder returns. The current data page shows an annualized dividend of $3.80, a 1.13% yield, four years of dividend growth in the displayed series, and a 2.49% buyback yield. The latest analyst snapshot is Buy from 30 analysts with a $374.54 average target, ranging from $315 to $450. That target is an opinion, not an underwriting assumption. 2 13

Verdict

American Express clears the channel's three hard gates: ROE stayed above 15% in each of FY2023–FY2025, FCF remained positive in every listed period, and the stock's 20.41x trailing P/E is below the 31.35x median of Visa and Mastercard and within 10% of its own five-year average. The valuation is not a bargain on its own history; the reason it qualifies is the combination of a peer discount, a 6.63% FCF yield, and 10% company revenue-growth guidance.

The thesis stays intact while billed-business growth remains close to high single digits, principal-only write-offs stay near 2.0%, and CET1 remains within the 10% to 11% target range. Revisit the case after the October 23 earnings call, the next credit-loss disclosure, and the TheFork closing update. AXP is a candidate for further research, not a substitute for checking those conditions against an investor's own risk tolerance.

This is a quantitative screening article, not financial advice. Valuation, price, ownership, and operating figures are snapshots and should be rechecked before trading.