Did Greg Abel Already Make His First Big Mistake as Berkshire Hathaway CEO?

Berkshire Hathaway continues to treat American Express as a core long-term holding under new CEO Greg Abel, despite fully exiting Visa and Mastercard and AXP underperforming those peers year-to-date. Berkshire's effective AXP ownership has risen to roughly 24%, worth $49.2 billion, solely through American Express buybacks; AXP has reduced its share count by nearly 59% over 31 years. American Express forecasts 10% revenue growth in 2026 and trades at a 16.3x forward P/E versus about 25x for Visa and Mastercard, although higher rewards and advertising spending has compressed margins.

Analysis

The relevant valuation gap is not a simple mispricing versus V/MA: AXP retains consumer-credit, funding, and rewards-liability exposure that open-loop networks largely outsource. The incremental earnings sensitivity to a softer U.S. affluent consumer is therefore materially higher at AXP, while its marketing/rewards investment can depress near-term operating leverage before any retention benefit is visible. A lower multiple is warranted unless AXP demonstrates that billed-business growth and cardmember acquisition translate into stable spend per account, lower churn, and disciplined credit losses.

Berkshire's passive ownership increase creates a second-order technical issue: AXP's effective free float is shrinking, which can support the stock during buybacks but raises downside air-pocket risk if capital returns slow or Berkshire ever becomes a seller. This is not an independently verifiable new fundamental catalyst; portfolio-manager commentary should not change AXP earnings power. Near term, the more important catalyst is the next earnings release's revenue-growth guide, rewards expense trajectory, net write-off rate, and reserve build; over 6-18 months, normalization in revolving balances and funding costs determines whether AXP deserves to close any of its multiple discount.

Consensus may be too focused on AXP's headline P/E discount and too little on the fact that buybacks are most accretive only if credit costs remain contained and shares are repurchased below normalized intrinsic value. Conversely, the market may underappreciate AXP's premium-card franchise resilience if affluent spend remains insulated from lower-income consumer stress. The cleanest expression is relative, not an outright bet on a modestly positive retail article.

Key Decisions for Investors

• Maintain a 1-3 month watch, not a fresh outright AXP long, until quarterly disclosures show cardmember-loan growth without a sequential deterioration in net write-offs/reserves and rewards expense does not outpace revenue. A guidance raise or sustained double-digit billed-business growth would justify upgrading.
• For a neutral financials book, consider long AXP / short V in equal beta-adjusted dollars only after AXP underperforms V by a further 5-7% without a corresponding rise in AXP credit losses. Target 8-12% relative upside over 6-12 months; stop if AXP's credit-loss/reserve metrics worsen materially or V reaccelerates cross-border volume growth.
• Avoid treating BRK.A/BRK.B as a direct AXP catalyst trade: concentration support is already observable and the position is too small relative to Berkshire's operating businesses and cash allocation decisions. Reassess only if Berkshire reports an actual AXP purchase or sale rather than ownership changes caused by issuer repurchases.
• Set an alert for AXP reducing buybacks while maintaining elevated rewards/marketing spend. That combination would weaken EPS support and could trigger multiple compression toward bank-like valuations; it falsifies the capital-return component of the bull case.