This Buffett Stock Has Crushed the Market During the Past 5 Years. Here's Why It's Still a Buy Today.

American Express reported robust 2025 fourth-quarter results with revenue up 10% year-over-year and EPS rising 16%; card fees climbed 17% y/y and the company saw no drop in renewal rates despite higher annual fees. Management highlights heavy reinvestment—$6.3 billion in marketing in 2025 (a 75% increase since 2019) and rollout of a cloud-based data analytics platform—cited as drivers of improved credit quality, retention and relationship expansion; Gen‑Z spending grew 36% y/y (now 6% of mix), indicating a long runway. The stock has substantially outperformed the S&P 500 over five years (+201% vs +91%), and Berkshire Hathaway holds a ~22.1% stake, underscoring institutional conviction.

Analysis

Market structure: AXP’s Q4 showing (revenue +10% YoY, EPS +16%, card fees +17%, $6.3bn marketing spend) reinforces a fee- and relationship-driven revenue mix that benefits issuers with affluent customer niches and high take-rates (AXP, BRK.B as a large holder). Expect gradual share gains in premium card spend vs interchange/volume players; Gen‑Z spending +36% YoY (6% of total today) implies multi‑year LTV upside if retention holds. Higher marketing and tech spending signal demand-side strength for premium credit but raise fixed-cost leverage if macro weakens.

Risk assessment: Key tail risks are regulatory rollback of merchant fees or stricter consumer credit rules, a sharp cyclical pullback in travel/entertainment spending reducing revolver balances, and a major data breach undermining trust—each could compress EPS by >10% within 12 months. Short-term (days/weeks) price moves will track macro/consumption prints and Berkshire positioning headlines; medium-term (3–12 months) depends on marketing ROAS and vintage credit losses; long-term (2–5 years) hinges on Gen‑Z cohort retention and tech personalization payback. Hidden dependency: co‑brand and acceptance network economics (merchant mix) and third‑party fintech partnerships could flip unit economics quickly.

Trade implications: Base case is constructive — establish a modest core long in AXP (2–4% portfolio) funded by reducing passive exposure to broad financials where fee capture is weaker (eg, trim large-cap bank ETFs by 1–2%). Pair trade: long AXP vs short V or MA (size short 50–75% of AXP notional) to isolate premium-fee outperformance. Use options to define risk: buy 12–18 month AXP call spreads (e.g., buy 2027 Jan 110C / sell 2027 Jan 150C) or sell covered calls if long and satisfied with +15–25% upside in 12 months.

Contrarian angles: Consensus may overrate marketing spend as purely positive; if incremental customer acquisition cost (I CAC) exceeds lifetime value, EPS could disappoint—set a watch: if marketing/investment spend as % of revenue rises >200bps sequentially without >150bps revenue growth, reassess. The Berkshire stake is both endorsement and concentration risk—any reduction by BRK.B would materially pressure AXP (stock could drop >10% intraday). Historical parallel: premium issuers outperformed after 2009–2019 cyclical rebound but were vulnerable in 2020 downturn; stress‑test positions for a 20% revenue shock.

Key Decisions for Investors

• Establish a 2–3% long position in AXP (ticker AXP) within 5 trading days, target 12‑month total return +20%, set a stop-loss at -12% from entry and plan to trim to 1% if position gains +25%.
• Implement a pair trade: go long AXP (notional $1) and short Visa (V) at 0.65 notional to hedge network-volume risk; expect alpha capture if AXP outperforms by >8% over 6–12 months.
• Buy a 12–18 month AXP call spread to cap cost: buy Jan 2027 110C and sell Jan 2027 150C (or nearest strikes) with max loss limited to premium paid; this expresses bullish view while limiting downside.
• Reduce exposure to passive large-cap financial ETFs by 1–2% and redeploy into fee-heavy consumer finance names (AXP, BRK.B) because of durable take-rates; reverse if consumer credit delinquencies rise >50bps quarter-over-quarter.
• Monitor three triggers over next 90 days before increasing size: (1) marketing ROAS disclosed or implied (seek >1.8x incremental revenue per dollar vs 2019 baseline), (2) next-quarter credit loss/vintage deterioration <+50bps, (3) any regulatory proposals on interchange/merchant fees — cut/add only if thresholds breached.