Greg Abel Just Sold Both Visa and Mastercard Stock but Kept This Top Buffett Favorite

American Express was retained in Berkshire Hathaway’s portfolio while smaller positions in Visa and Mastercard were sold, underscoring continued confidence in AXP’s business model. The article highlights AmEx’s fee-based membership model, affluent customer base, and closed-loop system, and notes that its total return has beaten Visa, Mastercard, and the S&P 500 over the past five years. This is primarily a bullish stock-picking commentary rather than new fundamental news, so near-term market impact should be limited.

Analysis

AXP looks less like a generic card issuer and more like a premium consumer-finance compounder with unusually sticky unit economics. The market is implicitly paying up for the fact that its moat is behavioral, not just contractual: affluent cardholders tend to self-select into higher-fee ecosystems and then resist switching because the reward stack, status signaling, and expense-account acceptance all reinforce one another. That creates a far cleaner path to sustained fee growth than the lower-friction, lower-take-rate networks.

The second-order issue is that Berkshire’s trimming of V and MA may be read as a relative signal on where payments economics are going next. If portfolio concentration is the lens, AXP is the one most exposed to an upside scenario where premium spend remains resilient and credit losses stay contained; V/MA are the cleaner monetization vehicles in a world of volume growth, but their upside is more dependent on merchant pricing power and global payment mix, which is less differentiated. In other words, the market may keep rewarding AXP until credit normalization or higher charge-off expectations force investors to separate “brand moat” from balance-sheet leverage.

The main risk is not competition, it is cycle duration. AXP’s operating leverage can reverse faster than the market expects if affluent spending shifts from discretionary travel/dining into smaller-ticket essentials, because fee growth then has to outrun a slower revolving book and higher provisioning. Over a 3-6 month horizon the stock can keep working on sentiment and buy-side “quality at a premium” flows; over 12-18 months the key catalyst is whether consumer delinquencies broaden from subprime to prime credit, which would compress the multiple even if revenue growth holds up.

Contrarian view: the consensus is probably underestimating how much of AXP’s outperformance is already a quality-premium trade, not a fresh fundamental rerating. That means incremental upside is likely more limited than the recent relative chart suggests, while the left tail is meaningful if unemployment ticks up or travel spend normalizes after a multi-year air pocket. The better expression may be to own AXP on pullbacks rather than chase it here, especially versus the network names that benefit more from secular cash-to-card conversion with less direct credit exposure.

Key Decisions for Investors

• Buy AXP on a 5-7% pullback over the next 1-3 months; target a 10-12% rebound with a tighter stop if delinquency commentary worsens at the next earnings update.
• Relative-value pair: long AXP / short V or MA for 3-6 months to express the view that premium consumer-credit economics remain underappreciated versus pure network monetization; risk is a sudden de-rating of the entire payments complex.
• If already long AXP, finance upside with a covered call 6-9 months out; the stock likely remains range-bound to mildly higher unless underwriting remains pristine, so premium capture improves risk-adjusted return.
• Avoid chasing V and MA into strength until there is evidence that network growth reaccelerates outside of pure take-rate expansion; they are cleaner businesses, but the article’s signal suggests relative sentiment is weaker near term.
• Use earnings season as the catalyst window: if AXP reports stable spend but higher loss reserves, reduce exposure immediately—the stock’s biggest vulnerability is not revenue miss, it is a multiple reset on forward credit costs.