AXP Stock ($335.39) Trades at 17.75x Forward as Card Fees Grow 15% and Delinquencies Fall to 1.2% — Upside to $385

American Express closed Monday at $335.39, up $9.22 or 2.83% from Friday's $326.17 — recovering roughly two-thirds of the ground it lost when second-quarter results landed on July 24.

The reaction to that print was the story. Amex beat on earnings, missed slightly on revenue, raised full-year revenue guidance, and the stock fell 5.58% in premarket before closing down 4.3%. That is an unusual response to a quarter in which card member spending grew at its fastest pace in three years.

The recovery since has been partial. At $335.39, the shares sit roughly 13.4% below the 52-week high of $387.49 and about 16.3% above the 52-week low of $288.34 — squarely in the middle of the range, and below the 200-day simple moving average. The all-time closing high of $381.78 was set on December 11, 2025.

The twelve-month performance context matters. AXP was down 1.4% year-to-date entering the print, against an S&P 500 that has returned north of 20% over the trailing year. Market capitalization sits near $226 billion to $229 billion across 675 million to 682 million shares.

The valuation has compressed alongside that underperformance. The stock trades at 20.71 times trailing earnings on TTM EPS of $16.02, and 17.75 times forward. Trailing twelve-month revenue reached $79.54 billion with a 34.01% return on equity. Beta sits at 1.05.

The dividend is $3.80 annually for a 1.09% to 1.15% yield on a payout ratio near 21%. The most recent quarterly declaration was $0.95, with an ex-dividend date of July 2 and payment on August 10.

Context from the broader payments complex is favourable. Visa closed at $364.81, up 0.63%, and reports fiscal third-quarter results after tonight's bell. Mastercard sits at $558.75, up 1.28%, and reports Thursday. PayPal reported this morning with total payment volume up 10% to $486.4 billion and raised full-year guidance.

Volume growth across the payments network business is not in question. What is in question for Amex specifically is whether a company deliberately choosing to reinvest an earnings beat rather than bank it deserves a premium multiple.

That decision is the reason the stock fell on a beat, and it is the entire investment debate.

Q2 Delivered $4.53 on $19.64 Billion, Beating EPS and Missing Revenue

The numbers themselves were strong across nearly every operational line.

Diluted earnings per share came in at $4.53 against $4.08 a year earlier, an 11% increase, topping consensus estimates that ranged from $4.40 to $4.45. Total revenues net of interest expense reached $19.637 billion versus $17.856 billion, up 10% and 10% on an FX-adjusted basis. Consensus sat between $19.70 billion and $19.90 billion depending on the compiler, making the revenue miss anywhere from $60 million to $260 million.

Pretax income rose 15% to $4.071 billion, a 20.7% pretax margin. Net income came in at $3.110 billion against $2.885 billion, up only 8% — the gap between 15% pretax growth and 8% net growth reflects prior-year tax discrete items rather than any operational deterioration.

Billed business reached $455.8 billion on an FX-adjusted basis against $416.8 billion, up 9%.

The first-half figures reinforce the trajectory. Revenue for the six months totalled $38.544 billion, up 11%, with EPS at $8.81 versus $7.71, a 14% increase. Billed business across the half reached $883.8 billion, up 10%.

Average diluted shares outstanding fell to 679 million from 699 million, a 3% reduction that contributes roughly three points to the EPS growth rate independent of operations.

The revenue composition tells the strategic story. Discount revenue — the largest line — grew 9% to $10.163 billion, or 8% FX-adjusted, driven by higher billed business. Net card fees surged 15% to $2.862 billion from $2.48 billion, the fastest-growing line in the business. Net interest income advanced 11% to $4.649 billion on balance growth and net yield expansion. Service fees and other revenue rose 7% to $1.963 billion.

One offsetting item on net interest income: the exit of a small business cobrand held-for-sale portfolio partially reduced the growth rate. Card balances and other loans increased 8%, in line with billed business, with total balances up 9% FX-adjusted.

Management raised full-year 2026 revenue growth guidance to a firm 10%, up from a 9% to 10% range. Full-year EPS guidance was left unchanged at $17.30 to $17.90.

That last decision is where the stock reaction originated.

Raising Revenue and Holding EPS Is a Deliberate Choice, and It Cost the Stock 4.3%

The market wanted a guidance raise on both lines and got one on revenue only.

Management stated the rationale directly: the company raised revenue growth guidance to 10% and plans to reinvest that outperformance in growth initiatives rather than letting it flow through to earnings. Some analysts entering the print had modelled a 3% EPS guidance raise on the back of first-half momentum.

The reinvestment is going into customer acquisition, marketing, technology, new product features, and acquisitions. That is a management team explicitly prioritising long-term momentum over near-term profit optics, and it is defensible — but it removes the earnings upside investors had priced.

The expense arithmetic makes the trade-off visible. Consolidated expenses rose 12% year-over-year to $14.5 billion, faster than the 10% revenue growth. The drivers were variable customer engagement costs tied to the US Platinum Card refresh and higher card-member benefit utilization.

That is the bear case in one line: expense growth outpacing revenue growth, with the cost of maintaining premium engagement rising rapidly.

The bull rebuttal is that the expense growth is largely variable and demand-driven. Card member benefit utilization rising means cardholders are actually using the value propositions they pay annual fees for — which is what drives retention and justifies the next fee increase. Marketing spend directed at acquiring 3.0 million new cards in a quarter where 75% came on fee-paying products is customer acquisition cost, not overhead.

Whether that distinction holds depends entirely on the return. Amex is spending today for card fee revenue that amortises over the following twelve months and beyond. If the retention and spend profile of the 2026 cohort matches prior vintages, the reinvestment is accretive within a year. If premium competition compresses those economics, it is not.

For the first half, revenue grew 11% and EPS grew 14% — the reinvestment has not yet broken the operating leverage.

Management indicated it expects spending momentum to continue through the second half, with the caveat that portfolio transfers will create a modest headwind to reported billing growth. That is a technical drag on a reported metric rather than a demand signal, and it is the kind of detail that gets misread in headline numbers.

Card Member Spending Grew 9% — the Fastest in Three Years

The demand signal underneath the quarter is unambiguous and it is the strongest data point Amex has produced in years.

Card member spending grew 9% on an FX-adjusted basis in the second quarter — matching the first quarter and marking the highest rate in three years. Total spending on the company's own measure rose 9.4%. US Consumer Services billed business accelerated to 11%.

The category detail explains where it came from. Airline travel, travel and entertainment, and luxury retail all performed strongly. Luxury retail spending had risen 18% in the first quarter, the most recent period with a confirmed category breakdown, and management indicated in June that second-quarter spending was outpacing the first across key segments.

That composition matters more than the headline rate. Travel, entertainment, and luxury retail are the most discretionary categories in any consumer wallet and the first to be cut when households retrench. They accelerated.

The read-through to the broader consumer picture is genuinely useful and genuinely narrow. Amex serves the premium segment — affluent consumers and businesses whose spending behaviour diverges from the aggregate. The company's results say the high-end consumer remains highly engaged. They say nothing about the consumer who reported to a national confidence survey that jobs are hard to get, a share that reached 22.5% in June, the highest since January 2021.

That bifurcation is the honest reading. Coca-Cola reported unit case volume up 5% with growth in every segment this morning while a rival flagged tighter household budgets. UPS beat by raising revenue guidance while shedding volume. The consumer is not one thing.

For Amex specifically, the premium concentration is a feature. Affluent cardholders spend more per transaction, default less, and pay annual fees. Their spending is more volatile in a genuine recession and more resilient in a slowdown — and 2026 has been a slowdown with an energy shock, not a recession.

Balance growth kept pace with spending at 9% FX-adjusted, which indicates the spending is not being funded by an unusual drawdown of credit. That distinction is what separates healthy volume from a late-cycle credit build.

Net Card Fees Rose 15% and Will Exit the Year in the High Teens

The single most important line in this business is the one growing fastest, and its trajectory is already committed.

Net card fees reached $2.862 billion, up 15% — some measures put it at 15.4% — from $2.48 billion a year earlier. That is a record level and the company's fastest-growing revenue line. Net card fees have now grown at a double-digit rate for 32 consecutive quarters, an eight-year streak without a single quarter of single-digit growth.

The driver is the US Platinum Card refresh. Amex raised the annual fee and enhanced the value proposition, and the revenue recognition works with a lag: higher fees are recognised gradually as existing customers are repriced at renewal, with each fee amortised over twelve months.

That mechanic makes the forward path unusually visible. Management guided card fee growth to accelerate in the third quarter and exit the year in the high teens. Those fees are already contracted — the repricing has occurred, and the revenue recognition follows automatically as renewal dates arrive.

This is the closest thing to a subscription revenue stream in the payments industry, and the market has not been paying a subscription multiple for it.

The quality of that revenue is the point. Card fees are recurring, high-margin, and disconnected from transaction volume. A cardholder who spends less in a recession still pays the annual fee. Discount revenue is cyclical. Net interest income is rate-sensitive. Card fees are neither.

At $2.862 billion quarterly, card fees are running near $11.5 billion annualised — roughly 14.5% of trailing revenue. Growing that line in the high teens while the total business grows 10% means the revenue mix is shifting toward the most durable component every quarter.

The offsetting cost is exactly the expense line that spooked the market. Higher fees require enhanced benefits, and enhanced benefits get utilised. The Platinum refresh generated both the 15% fee growth and a substantial portion of the 12% expense growth. They are the same decision viewed from two sides of the income statement.

The judgement to make is whether high-teens fee growth exiting 2026 justifies the engagement cost incurred in 2026. Thirty-two consecutive quarters suggests management has repeatedly gotten that calculation right.

Provisions Fell 23% and Delinquencies Dropped to 1.2%

The credit picture is the quietest strength in the quarter and the most durable.

Consolidated provisions for credit losses came in at $1.1 billion, down from $1.4 billion a year earlier — a 23% reduction driven primarily by a reserve release in the current period against a reserve build in the prior one. That swing alone is worth a meaningful share of the pretax income growth.

The underlying metrics support the release rather than merely enabling it. The net write-off rate on consumer and small business principal held at 2.0%, flat year-over-year. The delinquency rate declined to 1.2%, improving from 1.3% a year earlier.

For context on how favourable that is: a 1.2% delinquency rate and a 2.0% write-off rate are exceptional numbers in card lending at any point in a cycle, and they are being achieved while balances grow 9%. Portfolios usually deteriorate as they grow, because new vintages season into losses. Amex's are improving.

Management attributes this directly to the premium strategy. Investing in the value propositions of premium products attracts higher-credit-quality customers, and those customers default less. The company frames improving credit performance as a consequence of the product strategy rather than as a separate achievement, which is analytically correct.

Guidance is for credit metrics to remain generally stable throughout the year. That is deliberately unexciting language, and stability at these levels is the desired outcome.

The forward risk is worth stating. A reserve release is a one-time earnings contribution that cannot repeat indefinitely. If the second half brings a build rather than a release, the year-over-year provision comparison inverts and becomes a headwind to reported earnings — even with unchanged underlying credit quality. That is a mechanical drag investors should model rather than a deterioration to fear.

The balance sheet carries $59.05 billion in total debt against $45.24 billion in cash and equivalents. Debt-to-equity screens at 642%, which is a normal figure for a lender and a misleading one when compared against non-financial companies. Amex holds A-level credit ratings and passed Federal Reserve stress testing.

The financial strength score of 3 out of 10 assigned by one quantitative framework reflects that leverage screen rather than any distress.

Three Million New Cards, and 75% Came on Fee-Paying Products

The acquisition data is where the strategy becomes measurable, and the mix is the metric that matters.

American Express acquired 3.0 million proprietary new cards in the second quarter. US Consumer Services contributed 1.3 million, Commercial Services 0.8 million, and International Card Services 0.9 million.

Seventy-five percent of new accounts in the quarter came on fee-paying products — the highest proportion since the company increased its focus on premium offerings. Year-to-date, more than 70% of new accounts have been fee-based.

Sixty-five percent of global consumer new accounts came from Millennial and Gen-Z customers. In the first quarter, Gen-Z spending grew 38%.

That demographic skew is the most valuable data in the release. Card relationships are long-duration assets — a premium cardholder acquired at 28 generates fee and interchange revenue for decades. Amex acquiring two-thirds of its new consumer accounts from cohorts with rising income trajectories is building a revenue annuity, and the cost of that acquisition sits in the current period's expense line while the revenue arrives across the next twenty years.

The accounting mismatch between those two facts is exactly why the stock fell on a beat.

The commercial side added 0.8 million cards, supported by product launches including a business card developed with a professional association for solo practitioners and small firms, offering tiered cash back and issued through a partner bank on the Amex network. That structure — Amex network, third-party issuer — is how the company extends reach without underwriting every account itself.

International Card Services contributed 0.9 million, roughly 30% of the total, against a business that generates a smaller share of revenue. International remains the largest structural growth opportunity, with Amex operating in roughly 130 countries and holding meaningfully lower share outside the US.

Partnerships continue to broaden the value proposition: a membership rewards partnership with a global hotel group announced July 22, an existing relationship with a major sports merchandise platform, and Membership Rewards points now redeemable directly within Apple Pay's checkout experience.

The company also ranked first in the 2026 US credit card mobile app and online satisfaction studies.

The Closed Loop Is Why Amex Is Not a Network Stock

The structural distinction between Amex and its peers is frequently collapsed, and it changes how the business should be valued.

American Express operates a closed-loop network for the cards it issues. It is simultaneously the issuer, the network, and in many cases the acquirer — which means it captures substantially more of the economic profit from a single card payment than either a pure network like Visa or a pure issuer like a bank.

The consequence appears in the financials. Trailing twelve-month revenue of $79.54 billion against Visa's roughly $43 billion looks like Amex is the larger business, but the comparison is meaningless: Amex's revenue includes interest income and card fees that Visa never touches, while Visa's revenue is nearly pure network toll. Visa's net margin runs above 51%. Amex's runs 14.11%.

Different businesses, different margin structures, different multiples. Visa trades at 25.5 times forward earnings. Amex trades at 17.75 times.

The closed loop delivers two advantages beyond economics. It gives Amex transaction-level data on both the cardholder and the merchant side, which feeds fraud detection, targeted offers, and the analytics that justify premium annual fees. And it means Amex controls the entire customer relationship rather than sharing it with an issuing bank.

The corresponding disadvantage is scope. Amex operates more narrowly than large issuers, offering fewer deposit and lending products, and it carries direct credit risk that Visa and Mastercard do not. When credit deteriorates, Amex absorbs it. When credit improves — as it has, with delinquencies at 1.2% — Amex captures the benefit through provision releases.

That credit exposure is precisely why Amex trades at a discount to the networks and why the discount is partly justified. It is also why a quarter with a 23% provision reduction and improving delinquencies deserved a better reception than it received.

Competitive pressure is intensifying. Card networks and issuers are all repositioning for AI-agent commerce, with Amex explicitly building its tradition of trust, security, and service into agent-initiated transactions. The company has also been acquiring — a restaurant reservation platform, an expense management build-out, and a payments infrastructure acquisition agreed in April.

Whether any of that generates incremental revenue is a 2027 question.

17.75 Times Forward for 34% Return on Equity

The valuation case rests on a comparison the market has been reluctant to make.

Amex trades at 20.71 times trailing earnings on TTM EPS of $16.02, and 17.75 times forward. Against full-year 2026 EPS guidance of $17.30 to $17.90, the multiple at $335.39 works out to 18.7 to 19.4 times — the midpoint of guidance implies roughly 19 times.

Return on equity runs 34.01%. Gross margin sits at 61.03%, net margin at 14.11%.

A business generating a 34% return on equity while growing revenue 10% and earnings 14% at 19 times forward earnings is not obviously expensive. The five-year median trailing P/E has been 18.58, meaning the stock currently sits at a modest premium to its own history — and that premium has compressed substantially as the shares have gone nowhere for twelve months.

Quantitative frameworks disagree on where fair value sits. One assigns a value of $323.90 against the current price, implying the stock is roughly 3.4% overvalued at $335.39. Another puts fair value at $440.45, recently trimmed by about $3, implying substantial undervaluation. That $117 spread reflects genuine disagreement over how much credit to give the premium strategy.

The overall quality scores are strong: an 80 out of 100 composite with growth ranked 9 out of 10, offset by the financial strength score of 3 that reflects a lender's balance sheet rather than distress.

Capital return provides mechanical support. Average diluted shares fell 3% year-over-year to 679 million. At that pace, buybacks contribute roughly three percentage points to EPS growth annually. Management indicated it will reinvest some upside into growth initiatives rather than focusing solely on repurchases — which caps how much the buyback can offset the reinvestment decision.

The dividend of $3.80 on a 21% payout ratio yields 1.09% to 1.15%. That is well below the 4.2% top quartile for US dividend payers and the 1.8% peer average. Amex is not owned for yield, and the low payout ratio is deliberate: it preserves capital for buybacks and growth investment.

Insider activity has been one-directional. Ten open-market trades over six months, all sales, totalling roughly $2.4 million over the past three months.

Analysts Trimmed Targets After Cutting Them Higher Three Weeks Earlier

The sell-side sequence around this print is a useful illustration of how quickly sentiment moves without the fundamentals changing.

In the first week of July, ahead of results, three desks raised targets substantially: one to $386 from $340, another to $364 from $322, and a third to $380 from $345. Those were 13% to 14% increases in the space of days, reflecting confidence in the first-half momentum.

After the July 24 print, the same desks trimmed. One cut to $382 from $385. Another moved to $370 from $380. A third reduced to $385 from $391. Those are small adjustments — one to three percent — that reflect modelling the unchanged EPS guidance rather than any change in view.

The ratings distribution held. Buy ratings were reaffirmed across multiple houses following the results, with one desk maintaining Hold and another reaffirming Hold. Across 30 covering analysts, the average rating is Buy with a twelve-month price target of $374.94, implying roughly 11.8% upside from $335.39. A narrower panel of 13 analysts carries a median target of $385, implying 14.8%.

One independent research view upgraded to Buy ahead of the print with a $422 target, citing sequential acceleration in billed business, plummeting delinquencies, achievable guidance, and a mid-30s return on equity justifying a higher multiple. That thesis was largely validated by the results — spending accelerated, delinquencies fell, guidance was raised on revenue — and the stock fell anyway.

Another view rated the stock Hold at roughly 19 times earnings, describing it as a sector premium with no urgency to sell but preferring pullbacks toward $300 before adding.

The spread between $370 and $422 across professional estimates on the same quarter tells you the disagreement is about multiple, not about the business. Everyone models similar numbers. They disagree on what those numbers are worth.

Next earnings are estimated for October 23, 2026.

What Visa and Mastercard Will Tell Us This Week

The read-across from the rest of the payments complex arrives inside 72 hours and provides the cleanest test of whether Amex's spending data generalises.

Visa reports fiscal third-quarter results after tonight's close, with consensus at $3.22 per share on $11.35 billion of revenue, up 8.4% and 11.6% year-over-year. Total payment volume consensus implies 8.8% growth. Mastercard follows Thursday.

PayPal already reported this morning: total payment volume up 10% to $486.4 billion, revenue up 5% to $8.7 billion, transactions up 8% to 6.8 billion, with raised full-year transaction margin and EPS guidance.

Three payment companies growing volume between 8% and 10% in the same quarter is a consistent signal. Amex's 9% FX-adjusted spend growth sits directly inside that band, which argues the number reflects broad payment volume trends rather than anything specific to the premium cohort.

The divergence to watch is cross-border. Visa's international transaction revenue carries consensus at just 7.9% growth, the softest line in its model, reflecting concern about Middle East travel disruption through the quarter. Amex reported airline travel and travel and entertainment as strong performers. If Visa's cross-border comes in soft while Amex's travel spending was strong, the explanation is that premium travellers kept flying while the broader corridor thinned — which would be a genuine competitive data point in Amex's favour.

Crude's collapse this week is a forward tailwind for all of them. Brent has fallen roughly 10% across three sessions to $87.05 as the US-Iran pause held, which lowers the cost of long-haul flying and supports travel volumes into the fourth quarter.

The macro overlay is a Federal Reserve decision Wednesday at 2 p.m. Eastern, with the target range at 3.50% to 3.75% and implied hike odds near 35.8%. For Amex specifically, higher rates cut both ways: net interest income benefits from yield expansion, while a tighter consumer credit environment eventually pressures the spending that drives discount revenue.

Net interest income grew 11% last quarter on balance growth and net yield expansion. That line is the direct beneficiary of a hawkish Fed.

The Bear Case Deserves a Fair Hearing

Three objections to owning this stock have genuine substance, and dismissing them would be dishonest.

The first is expense growth. Total costs rose 12% against 10% revenue growth, and the gap is structural rather than one-time. Maintaining premium engagement — lounge access, credits, benefit utilization — costs more every year as more cardholders actually use what they pay for. Amex is in a benefits arms race with competing premium products, and arms races compress margins.

The second is the reserve release. A 23% reduction in provisions contributed materially to pretax income growth of 15%. Strip out the release and the underlying earnings growth is meaningfully lower. Credit metrics at 1.2% delinquency and 2.0% write-offs are near cycle-best levels, which means the direction of travel from here is toward normalisation rather than further improvement. The provision line becomes a headwind in 2027 arithmetic even if credit stays perfectly healthy.

The third is concentration. Amex's fortunes are tied to the premium consumer specifically. That cohort has proved resilient through an energy shock, a weakening labour market, and a confidence reading below the recession threshold since February 2025. Whether it stays resilient through an actual downturn is untested at this scale — the current premium card base is far larger and younger than it was in 2008.

Against those, the counterweights are equally concrete. Card fees have compounded at double digits for 32 consecutive quarters and are guided to accelerate. Seventy-five percent of new accounts are fee-paying, which improves the mix every quarter. Two-thirds of new consumer accounts come from Millennial and Gen-Z customers, extending the revenue duration. Credit is improving because the customer base is improving, not because underwriting loosened.

And the stock is down 1.4% year-to-date while earnings compound at 14%.

The valuation question resolves to a single judgement: is a closed-loop payments business with 34% return on equity, a subscription-like fee stream growing in the high teens, and improving credit worth more or less than 19 times earnings? The market currently says less. Every professional target says more.

Forecast: $360 on Momentum, $300 on a Credit Turn, $375 as the Consensus

The setup into the second half is defined by what management has already committed to rather than by anything unknown.

The bull path is mechanical. Card fee growth accelerates in the third quarter and exits the year in the high teens — that revenue is contracted and amortising. Spending momentum continues at 9%, credit metrics stay generally stable, and the reinvestment cycle begins converting into the 2027 revenue base. Under that outcome, the stock reclaims the $360 area first, then approaches the consensus target of $374.94, roughly 11.8% above spot. Clearing $381.78 would take out the all-time closing high, with the 52-week high at $387.49 above it. That path requires the third-quarter print on October 23 to show operating leverage returning — revenue growth at or above expense growth.

The base case has the shares range-bound between $320 and $360 through the third quarter, with the middle of the 52-week band and the 200-day average acting as the pivot. The stock has already recovered two-thirds of its post-earnings decline, which suggests buyers viewed $326 as value rather than as the start of a leg down.

The bear path runs through credit and expenses together. A second-half reserve build rather than a release, combined with expense growth staying two points above revenue growth, would compress the reported earnings trajectory and put full-year EPS at the bottom of the $17.30 to $17.90 range. That takes the stock toward $300, where one professional view has said it would prefer to buy, and toward the 52-week low at $288.34 in a harder scenario.

What would confirm the bull case: third-quarter card fee growth printing above 17%, delinquencies holding at or below 1.2%, or an EPS guidance raise at the October print. What would confirm the bear case: a provision build, expense growth exceeding 13%, or any deceleration in billed business below 8% FX-adjusted.

Amex told the market it would spend its beat rather than bank it. The stock priced that as a disappointment. Whether it was a disappointment or an investment gets answered in October.