The CEO who was told he’d never run American Express has made Amex cool again—and is beating JPMorgan, Visa, and the S&P 500

American Express says the Platinum Card refresh lifted U.S. new account acquisitions to double in the three weeks after the September fee increase to $895, while retention stayed high. The company also reported FY2023 revenue 40% above 2019 levels, revenue growth averaging 11% annually since fiscal 2022, and EPS growth of 16% on controlled expense growth. The article portrays CEO Stephen Squeri’s premium-focused strategy as a major success, though he faces rising competition, overcrowding at lounges, and macro/cycle risks.

Analysis

AXP’s edge is no longer just brand; it is pricing power plus customer mix engineering. The key second-order effect is that premium cards are turning into a quasi-subscription business with high annual fee elasticity but low churn when benefits map to daily behavior rather than sporadic travel. That makes AXP structurally better insulated than networks whose value proposition is more points-driven, because the incremental spend is being pulled into dining, streaming, wellness, and lifestyle merchants where utilization is frequent and habit-forming.

The bigger competitive takeaway is that AXP is forcing rivals into a costly response function: either copy the benefit stack and compress economics, or let affluent cohorts drift toward a more complete lifestyle bundle. That is a hidden negative for C/COST-style disintermediation economics in co-branding and for JPM/GS/other premium issuers that may need to raise acquisition spend to defend share. The payment networks (V/MA) are not direct losers on volume, but if premium issuers are successfully monetizing exclusivity through merchant-funded perks, more economics migrate from network rails to issuer-led ecosystems.

Risk is mostly second-derivative and medium-term. The main bull case breaks if labor-market softness or AI-linked white-collar job losses start impairing the very cohorts AXP is upselling, because high-fee products are most vulnerable when consumers reassess whether they are truly “using” the bundle. Another risk is lounge and benefit saturation: once the premium experience becomes crowded or commoditized, the incremental willingness to pay can fall faster than headline retention suggests. That would likely show up over 2-4 quarters in slower acquisition, not immediate churn.

Contrarian view: the market may be underestimating how cyclical the premium consumer really is. AXP’s recent outperformance is partly a multiple expansion story built on confidence in sustained fee growth; if macro data softens, the stock can de-rate even while fundamentals remain intact. The more attractive setup may be relative rather than absolute: AXP is still the cleanest expression of affluent-consumer resilience, but the best risk/reward likely comes from pairing it against more macro-sensitive financials with weaker brand pricing power.

Key Decisions for Investors

• Go long AXP on a 3-6 month horizon into any macro-led pullback; use weakness below recent highs as entry, targeting continued fee-led EPS compounding with downside buffered by premium customer stickiness.
• Pair trade: long AXP / short C for 6-12 months. AXP has stronger pricing power and brand monetization; C has more legacy exposure to lower-quality consumer mix and weaker proof of durable premium attachment.
• Pair trade: long AXP / short JPM or GS for 3-6 months if market re-rates financials on credit worries. AXP’s revenue model is more insulated from capital markets volatility and more levered to affluent consumer spend than balance-sheet-heavy peers.
• Sell/underweight V and MA versus AXP on a 6-12 month view. If issuers keep winning economics through proprietary benefits, the value capture shifts toward issuers, while the networks remain lower-growth toll collectors.
• For option exposure, buy AXP 6-9 month call spreads after any post-earnings consolidation. The setup favors moderate upside with defined risk, but avoid chasing outright calls until credit trends and unemployment data confirm the next leg.