Greg Abel's Berkshire Hathaway Collected $479 Million in Dividends Last Year from an American Express Stake That Cost $1.3 Billion. You Can Put the Same Principles to Work for Yourself.

American Express reported second-quarter EPS growth of 11% year over year, revenue growth of 10%, and a 9% increase in card-member spending. Berkshire Hathaway, which paid $1.3 billion for its stake through 1995, owned 22% of AmEx worth $56 billion at year-end 2025 and received $479 million in dividends that year. AmEx’s annual dividend rose to $3.54 per share from $2.08 in 2022; its forward P/E of 15.3 was below its five-year average of 16.8.

Analysis

The durable bull case is not Berkshire’s historical yield-on-cost; that is irrelevant to a new buyer. For AXP, the investable question is whether premium-card fees and spending growth can outpace rewards, customer-acquisition costs, and credit losses. Its integrated card-and-network model may support customer retention, but it also leaves earnings exposed to affluent discretionary spending and the economics of competing for cardmembers against Visa, Mastercard, and large banks. Buybacks can amplify EPS growth, yet should not be mistaken for stronger underlying demand.

The valuation discount to AXP’s own history could reflect a reasonable market haircut for credit normalization or tougher competition, not a mispricing by itself. The article’s positive operating indicators are encouraging, but without delinquency, charge-off, reserve, and rewards-cost trends, they do not establish that incremental spending is converting into durable profit growth. Berkshire’s large holding may signal patient ownership, but it is not a near-term catalyst; its concentration also makes the stake’s mark-to-market value sensitive to AXP’s multiple.

Near term, price action is likely to depend more on guidance and credit data than on the dividend narrative. Over 1–3 months, watch cardmember spending, net card fees, and loss trends. Over 6–18 months, a consumer slowdown or more expensive rewards competition could pressure returns even if headline revenue continues growing. The thesis weakens materially if credit losses rise faster than revenue or management cuts earnings guidance; stable credit metrics and sustained fee/spending growth would support it.

Key Decisions for Investors

• AXP: consider a modest starter long rather than buying on the dividend story alone; add only if upcoming results confirm spending and fee growth without deterioration in delinquencies, charge-offs, or reserve coverage. The own-history valuation discount offers some cushion, not a floor.
• Before adding, verify net card-fee growth, credit-loss trends, rewards and acquisition costs, and full-year guidance. If losses accelerate or guidance is reduced, stand aside or exit; if operating growth persists with stable credit, retain the position through the next reporting cycle.
• BRK.A: treat the AXP stake as evidence of long-term conviction, not a fresh catalyst. Avoid extrapolating Berkshire’s yield-on-cost to prospective returns or using the stake’s disclosed value as a short-term valuation anchor.
• No options trade is warranted from this article alone: the key missing inputs are AXP’s current credit-quality trajectory and the market’s expectations embedded in the shares.