The members have never spent more. The market has stopped counting the spend and started counting the perks.
Revenue has grown double digits four quarters running, billed business hit +9% in Q2 and ROE is 36% — yet AXP trades at ~18.5× trailing earnings, near its March low, because revenue guidance rose while EPS guidance held and expenses grew 12%. Four lenses became six, three scenarios, four horizons. Next catalyst: Q3 results on Oct 23.
Gray = dated closing anchors (Sep 26 ’25 to Oct 6 ’26; the Dec 12 high, Jan 12 and Mar 20 low are intraday; the window is 12 months because sourced daily history beyond that was unavailable). Colored paths = scenario midpoints, probabilities bear 30% · base 45% · bull 25% — tilted to the bear side because Q3 is 17 days away, expense growth is running ahead of revenue and the rate-cap overhang is unresolved. Street 12-month consensus ≈ $372 (range $329–$450). Non-zero price axis.
Re-weight the scenarios
Those probabilities are a judgment call — make them yours. Drag the bear and bull weights (base takes the remainder); the blended target, the dotted line and the prob-weighted row of each scenario card update live.
Six analyst lenses, six answers
Synthesized perspectives, each with its own 12-month target. Panel range $216–$383, mean $322 — below the $372 Street consensus.
The Compounder
Four straight double-digit revenue quarters, billed business +9% (best FX-adjusted pace in three years) and net card fees +15% for a 32nd double-digit quarter. Younger members are the fastest-growing cohort. A 36% ROE earns a premium multiple.
The Cash Counter
TTM free cash flow of $15.1B is a ~7.3% yield; $7.5B of TTM buybacks plus $2.4B of dividends retire 2.5% of shares a year. At 15× 2027E EPS the stock is the cheapest it has been in over a year while CET1 sits at 10.4%.
The Short Thesis
Card Member services cost is +50% and card adds slowed to 1.3M a quarter. Revenue guidance rose but EPS guidance did not, which says growth is being bought. A 10% rate-cap push and stablecoin or agentic bypass cap the multiple.
The Closed-Loop Case
Amex is issuer, network and merchant acquirer at once, so it keeps the data and the merchant relationship. 76,800 employees run ~$456B a quarter of billed business, and the affluent base has kept delinquency near 1.2% for three years.
The Tape Reader
Price sits below the 50-day ($326) and 200-day ($332) averages with RSI near 36, 4.7% above the March low. Estimate cuts and target trims are clustering ahead of Q3, which is typical of an oversold base, not yet a reversal.
The Cycle Watcher
A K-shaped consumer favors Amex’s affluent base, but higher inflation, firmer gas prices and tighter Fed messaging pressure rate-sensitive financials. Regulatory risk (rate cap, interchange) is a standing discount.
Wall Street 12-month price targets
Sorted low to high. Consensus ≈ $372 (+22%); the dashed line is today’s $304.55. Holds are clustered at $320–$345 after this week’s cuts.
Targets from firm notes dated Sep 23–Oct 6, 2026 (TipRanks/TheFly via StockAnalysis); *JPMorgan’s $400 Overweight dates from mid-July. Rating mix Sep ’26: 14 Buy/Strong Buy, 14 Hold, 0 Sell. Consensus conflict: $372.33 on the Oct 6 statistics page vs $377.91 on the Oct 5 forecast page; the later figure is used.
Where the ledger leads
Scenario midpoints (mid-year) vs today’s $304.55, each = calendar-year EPS × exit multiple. Illustrative frameworks, not forecasts.
1 Year
Mid-20272 Years
Mid-20283 Years
Mid-20295 Years
Mid-2031▸ Bull case — show the assumptions & math
▸ Base case — show the assumptions & math
▸ Bear case — show the assumptions & math
Revenue, capex, free cash flow & debt ($B)
For a card issuer, debt funds the loan book rather than signaling leverage, so read free cash flow against capex.
*2026E: revenue and FCF are consensus estimates ($79.5B, $11.9B); capex ($3.4B) and total debt ($59.1B) are trailing to Jun 30, 2026. Revenue is company-basis net of interest expense ($60.5B/$65.9B/$72.2B); aggregator series run ~$5B lower. Capex has roughly doubled since 2023 ($1.6B to $3.4B TTM). TTM FCF is $15.1B, so the 2026E estimate looks conservative or reflects timing.
EPS path underpinning the targets ($)
Each scenario price is an EPS estimate times an exit multiple. This is the ladder under the base case.
2024 ($13.35) is derived from FY25 $15.38 and its +15.2% growth; 2026E–2027E ($17.65, $20.16) are consensus (company guide $17.30–$17.90 for 2026); 2028E–2031E are this desk’s base extension of +13% fading to +10%. Base 2031E ≈ $31.2 × 17× ≈ $530.
The business is still growing
Q2 FY26 year-over-year — read against a stock 21% below its high.
Every line is positive and net card fees (clay) lead. The cost mirror is Card Member services expense at +50%: that is the bear’s whole point, and the number Q3 must reconcile. Billed business is +9% FX-adjusted.
Bull vs. Bear
The valuation argument compresses into one disagreement: are the perks an investment or a giveaway?
▲ THE BULL CASE
- Growth is accelerating. Q2 revenue +10% (fourth straight double-digit quarter), billed business +9%, six-month EPS +14%.
- Fees are the engine. Net card fees $2.86B, +15%, the fastest-growing revenue line and a 32nd straight double-digit quarter.
- Returns are elite. ROE 36.4% with CET1 at 10.4%, and $2.9B returned to holders in Q2.
- Credit is clean. Delinquency 1.2%, write-offs below 2019 levels, and a $191M reserve release in Q2.
- The multiple already reset. P/E fell from ~25 at end-2025 to ~18.5; 15× 2027E EPS is below its recent range.
- Street still sees upside. Consensus ≈ $372 (+22%), 14 Buy/Strong Buy vs 14 Hold and no Sells.
▼ THE BEAR CASE
- Costs are eating the upside. Q2 expenses +12% to $14.5B; marketing guided ~+10% in H2; EPS guide held at $17.30–$17.90.
- Platinum perks are expensive. Card Member services expense +50% as benefits expand and usage rises.
- Card adds are slowing. US consumer adds fell to ~1.3M a quarter from ~1.5M.
- Rate-cap overhang. A 10% credit-card rate cap floated in Jan ’26 cut AXP ~4% in a day; it needs legislation but is not dead.
- Write-offs were flattered. June’s 1.4% consumer rate included a ~0.3pt asset-sale benefit; July–Aug ran 1.7%.
- Momentum is broken. −21% from the high, below both moving averages, with a string of target cuts into Oct 23.
Risk map — likelihood × impact
The hot cell is likely and high-impact: costlier growth. Regulatory and macro risks sit one row down in “possible.”
- FX & one-off noise
- Slower card adds
- Costlier growth
- Rivals & credit drift
- Rate-cap law
- Consumer downturn
- Network bypass
Costlier growth
Perk and acquisition spend grow faster than fees, so revenue beats keep failing to reach EPS.
Rate-cap law
A statutory cap on card APRs would cut net interest income of ~$24B a year.
Consumer downturn
A hit to affluent spend lifts write-offs from a 2.0% base and slows billed business.
Slower card adds
Fewer new Platinum and Gold accounts weakens the fee flywheel.
Rivals & credit drift
Chase and Capital One premium cards pressure acquisition; normalized write-offs (~1.7% consumer) reset the credit story.
FX & one-off noise
The pending GBT stake sale and co-brand portfolio sales move reported lines without changing the core.
Network bypass
Stablecoin or agentic-commerce rails reroute payments around closed-loop economics.
The jargon, decoded
Hover the dotted terms in the metrics, or scan the working definitions.
- Billed business
- Total spend on Amex cards before rewards — the headline demand gauge. $455.8B in Q2.
- Net card fees
- Annual fees net of the cost of card benefits credited to members. $2.86B in Q2.
- VCE ratio
- Variable customer engagement costs (rewards, services, cash-back) as a share of revenue. 44.6% in Q2.
- CET1
- Common equity Tier 1: the regulatory capital cushion. 10.4% in Q2.
- Net write-off rate
- Share of card balances written off as uncollectable. 2.0% (principal only) in Q2.
- P/B and ROE
- For a lender, EV/EBITDA is not meaningful because debt funds the loan book; use price/book (6.0×) and ROE (~34–36%).
- Exit multiple
- The P/E assumed at the end date. EPS × multiple gives each scenario price.
- Prob-weighted
- Each scenario price × its probability, summed into one expected value.