American Express

American Express is the boring crown jewel in the Berkshire Hathaway portfolio. Buffett bought his first AXP stake in 1964 (the 'Salad Oil Scandal' crisis) and has not sold a share since 1998. In Q1/2026, Berkshire still holds 151.6M shares — 22% of the entire company — worth $47.9B. That is the third-largest Berkshire position after Apple and BAC.

The 2026 story is the Millennial and Gen-Z premium-card adoption. Card members under 35 grew +21% YoY in Q1/2026 and now account for 38% of all new platinum acquisitions. The Platinum refresh in October 2025 (with $1,400 in new annual benefits including Apple TV+, Walmart+, and SoulCycle credits) lifted the annual fee to $895 — and applications still grew 14% post-launch.

The other half of the story is the spending data. AmEx members billed $1.84T in 2025 (record), with international T&E (travel and entertainment) recovering to 118% of 2019 levels. Q1/2026 EPS came in at $3.66 (vs. $3.41 consensus), and management raised full-year guidance to $15.20-$15.50 — a 14% growth midpoint.

Berkshire's AXP position is the textbook never-trim. Buffett wrote in the 2023 annual letter that AXP is one of the four 'forever' holdings (alongside Apple, Coca-Cola, and Occidental Petroleum). Berkshire's cost basis sits around $1.3B against a current value of $47.9B — a 37× return over 30 years, plus dividends.

Notable recent activity: Capital Group increased its stake by 9% in Q1/2026 to 64M shares. Vanguard added 4M shares net. The only meaningful seller among the top-20 holders was Wellington Management, which reduced by 12M shares — but most analysts read this as a routine rebalancing within their broad-market funds, not a directional call.

Insider activity (Form 4): CEO Stephen Squeri sold 188,000 shares in March 2026 at an average of $321 (routine 10b5-1 plan). CFO Christophe Le Caillec sold 45,000 shares. No insider buying in the past 12 months — but that is typical for a mature financial with established quarterly grant cycles.

American Express trades at a forward P/E of 15.7× and a P/B of 6.5× as of May 2026. Comparable franchise-quality financials — Visa, Mastercard, and S&P Global — trade at 28×, 32×, and 31× respectively. The valuation gap reflects credit-risk premium that AmEx genuinely carries, but the discount is wider than fundamentals justify if the credit cycle stays benign through 2027. Wall Street analyst targets range from $285 (Wells Fargo, credit-cycle concern) to $450 (Bank of America, premium-cohort growth bull), median $362 vs. current $316 — implying 14% upside before dividends. The dividend yield of 1.2% is modest, but the $7.5B 2026 buyback authorization (3.5% buyback yield) makes total capital return competitive.

1. July 18, 2026: Q2/2026 earnings — domestic T&E billings is the make-or-break KPI; <5% growth pressures the EPS thesis
2. September 2026: Investor Day — management expected to reiterate the 10%+ EPS-CAGR-through-2028 framework
3. November 2026: Federal Reserve stress test results — AXP must show <2.5% write-off projection in the severely adverse scenario to maintain buyback pace

American Express is the closest thing to a 'set-and-forget' Berkshire-style compounder I keep in my own portfolio. The combination of Buffett's 22% stake, the platinum-card moat that genuinely refreshes itself with each generation, and a 15.7× forward P/E in a market trading at 22× makes the risk-reward attractive. What you need to accept: there will be a credit-cycle pullback at some point in the next 3 years, and the stock will likely give back 25-30% during that window. If you can hold through that without panic-selling, AXP is one of the few stocks where the 'never trim' Berkshire playbook genuinely applies. My add-trigger is a forward P/E under 13 — implying a share price near $260.