American Express: A Beat-and-Raise Into an 18% Drawdown — and a New Word for the Risk
Record revenue, raised EPS guidance, an OCC penalty, 5% of the workforce restructured, and a strategy pivot away from the big banks — all in the same 60 minutes
AXP · Earnings Call · 2026-10-23
Nothing broke — except the multiple
American Express delivered the kind of quarter it has been delivering for years: record revenue of $20.1 billion, up 9%; GAAP EPS of $5.42, up 31%; adjusted EPS of $4.68, up 13%; card fees accelerating to 17% FX-adjusted on roughly $3 billion for the quarter; 3.1 million new cards acquired with 76% on fee-paying products. Management reaffirmed 10% revenue growth and raised full-year EPS guidance to $18.30–$18.60 reported. And the stock sits 18.6% below its December 2025 peak, down 8% over the last 90 trading days alone. That gap is the story. This is a company whose operating engine is running better than its share price, and the call itself contains three separate explanations for why the market has stopped paying up for it: an accounting-clean quarter stuffed with discrete items, a regulator's consent order, and a genuine strategic acknowledgement that two of its pillars — the premium card value proposition and the network partnership model — need to be re-engineered.The quarter beneath the quarter
Strip out the $1.11 per share gain on the Global Business Travel Group sale, the $0.37 restructuring charge, the $0.51 legal reserve (which includes the unreserved portion of the OCC civil money penalty), a $0.29 tax benefit and a $0.24 CECL model benefit, and you get $4.68 — a clean, mid-teens-growth quarter, exactly as advertised. The effective revenue base of $21.6 billion is up 8% year over year and 3% sequentially, and net interest income of $4.6 billion grew 11%, continuing its habit of out-running balance growth. Credit remains the quiet flex. Provisions were $985 million including a $211 million release; the delinquency rate has sat at 1.2% or 1.3% for fifteen straight quarters. Christophe Le Caillec attributed the 2.6% allowance coverage — below the level at CECL adoption — to the "premiumization of the portfolio." That is a real, measurable shift, and it is why provision expense is running at roughly a third of its 2020 quarterly peak. What the market sees instead is the cost side. The efficiency ratio at 67.1% is 2.3 points worse year over year, the direct arithmetic consequence of spending the GBTG gain on restructuring and stepping up marketing and technology.“Rewards maxing” enters the lexicon
The most revealing moment came from an analyst, not management. Sanjay Sakhrani and later Erika Najarian raised a concern that did not exist in this company's keyword history until this quarter: rewards maxing — the fear that AI agents will let cardholders systematically arbitrage the value proposition to extract more in benefits than they pay in fees. Steve Squeri's answer was defensive in structure and revealing in substance. On the economics: “we really do not have a breakage model. So we do not benefit when people do not use our rewards.” — Steve Squeri, Chairman and Chief Executive Officer · 2026-10-23 On the mechanics: threshold-based controls like the $75,000 spend gate on Platinum passes. And on the technology, a line that positions Amex against the banks that are simply blocking bots: “We're just not going to let assistants run rogue through our systems.” — Steve Squeri, Chairman and Chief Executive Officer · 2026-10-23 The company will allow agents — but only through approved API or MCP channels, and it does not want AI AI agents "pinging Resy every four seconds." That is a materially different posture from the one struck six months ago. In April, Squeri described agentic commerce as “we're warming up at a bull pen” — Stephen Squeri, Chairman and Chief Executive Officer · 2026-04-23 — pre-season, unquantifiable. In October he spent a third of his prepared remarks laying out five concrete AI domains (value propositions, acquisition, engagement, payments, service) and named a developer kit (ACE), an Intent Intelligence framework, and Agent Purchase Protection. The keyword agentic commerce has moved from a philosophical aside to an operating roadmap in two quarters.The network quietly abandons the big banks
Buried in the Q&A, prompted by a question about the MasterCard settlement and honor-all-cards, Squeri conceded something that reverses years of effort:The new network strategy is Synchrony, Bread Financial, NFL, Lowe's — smaller and mid-size issuers reaching segments Amex cannot reach organically. This is a real pivot, and notably it is the first quarter the keyword big banks has surfaced at all, arriving as a negative. Two quarters ago, management was still framing Capital One's Brex deal and Chase's refresh as the competitive reference point; now the framing is that the co brand portfolio exits (Amazon, Lowe's) cost about 60 basis points of billed business and roughly a point of revenue growth — a headwind they are choosing to run into.we are not looking to partner with the bigger banks. I think we made a lot of efforts to attempt to partner with some of the bigger banks. That did not really pay off for us.