Business

Figures converted from GBP at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

Know the Business

Wise is a low-cost cross-border payments network funded by an unusually large customer-deposit float, dressed as a fintech. Personal money transfers and card spend are the volume engine; the $27.8bn in customer balances quietly carries roughly half of reported profit through interest income — which means rate cuts hit harder than the headline 16% income growth implies. The market is right that the unit economics are best-in-class, but it consistently under-prices how much of the recent profit beat was a transient rate-cycle gift, and over-prices the operating leverage from here as management deliberately reinvests every basis point of efficiency back into lower prices.

1. How This Business Actually Works

Wise runs two engines bolted to one customer relationship, and the mix matters more than the headline.

The first engine is the cross-border take. A customer sends money in one currency and receives it in another; Wise charges roughly half a percent of volume (53bps in Q4 FY2025, down from 67bps a year prior) and clears that transfer through one of eight direct connections to domestic payment systems instead of correspondent banking. About 65% of transfers settle in under twenty seconds. This engine generated $1,088m of revenue on $188bn of volume in FY2025 — the take rate is falling on purpose, and the volume is growing 23%.

The second engine is the deposit float. Customers hold $27.8bn in balances inside their Wise Account so they can spend, get paid, or sit on multi-currency cash. Wise invests this in cash, money-market funds and short bonds, and earns the prevailing rate. By design, only the first 100 bps of yield drops to Wise's "underlying" income ($195m in FY2025); everything above that is shared with customers via the Assets feature or paid out as benefits — but a meaningful chunk still lands in reported P&L as net interest after benefits paid. Total reported interest income was $769m in FY2025 against $209m of customer benefits paid, a $560m net contribution that swings with central-bank policy rates.

The cost base is the interesting part. Cost of sales (banking partner fees, payment scheme fees, card-issuance costs, credit losses) ran at 25% of revenue in FY2025; gross margin on a full-income basis is 80%. The variable cost of moving the next dollar is shrinking faster than the price Wise charges for it, which is why management can cut take rates 14bps and still expand gross margin two percentage points in the same year. Below the gross line, half of opex is people — about 6,500 employees, with roughly a third dedicated to financial-crime and compliance work that is regulatory table stakes in 70+ jurisdictions.

The business model is best understood as "Costco for cross-border money": every efficiency gain is given back to customers as lower prices, which drives volume, which drives further unit-cost improvement. That flywheel only works if scale economics keep showing up — and so far they have.

2. The Playing Field

No single peer is a clean comp; together they triangulate where Wise actually sits.

Read the chart top-right to bottom-left. Adyen sits where Wise wants to live in five years — same gross margin profile, lower growth, much higher absolute margin because it stopped reinvesting price. Wise has the better growth-and-margin combination today, but only because management is choosing to bank some of the float income; strip out interest above the 1% yield and operating margin drops closer to 20%. PayPal is the cautionary tale of what happens to a payments network when growth stalls and the moat becomes brand recognition rather than infrastructure. Western Union is the whole reason Wise exists — a 174-year-old incumbent shrinking at a 3% CAGR, trading at 1.2x EV/EBITDA because the market has stopped pretending it is anything other than a melting ice cube. Remitly is the closest pure-play in personal remittance and is barely profitable; that contrast tells you Wise's float economics, not its take rate, are doing the heavy lifting at the bottom line. dLocal is the emerging-markets specialist — narrower geographic moat, similar growth, much thinner gross margin. Payoneer is closest to the Wise Business segment alone.

What "good" looks like in this industry: 75%+ gross margin, double-digit volume growth, take rate trending down (because the alternative is margin compression from churn), and a deposit franchise that gives you optionality on rates. Wise is the only name in this set that hits all four.

3. Is This Business Cyclical?

The cross-border franchise is structurally non-cyclical; the earnings are increasingly tied to interest rates.

Volumes don't blink. People sending money home, paying overseas tuition, or running cross-border payroll do this through cycles; remittance flows held up through COVID and through the 2022-2023 inflation shock. Wise's cross-border volume grew 23% in FY2025 and 24% in H1 FY2026 against a backdrop where consumer-discretionary names were guiding down. The Business segment (24% of underlying income) is more SME-cyclical, but it is also the fastest-growing slice — volumes up 35% in H1 FY2026 — so the cyclical drag is not visible yet.

The cycle bites through the float. Wise's reported PBT is a function of central-bank policy more than most fintech investors realise. In H1 FY2026, interest income above the first 1% yield fell 11% year-over-year purely because UK and US base rates came down; this dragged reported PBT 13% lower despite underlying income growing 13% and volume growing 24%. The asymmetry is the point: a 100bp move in policy rates flows roughly $207-259m into or out of pre-tax profit on the current balance base, against a reported PBT of $731m in FY2025. Customer balances themselves are pro-cyclical — they grow with consumer wealth and SME confidence — so a rate cut combined with an SME slowdown would compound the hit.

This is the single most important thing to internalise: when management says they target a 13-16% underlying PBT margin, they are guiding to a normalised view that already strips out the rate windfall. Anyone modelling FY2024-FY2025 reported earnings as the new baseline is going to be wrong as the rate cycle continues to mean-revert through FY2026-FY2027.

4. The Metrics That Actually Matter

Five metrics tell you whether this business is working; everything else is noise.

Cross-border take rate is the price signal. It must trend down — 67 → 58 → 53 bps in three years is a feature, not a bug. If take rate ever stops falling without volume re-accelerating, the moat is gone, because it means scale economies are no longer feeding back to the customer. Watch the Q4 number, not the annual, since that is where the latest pricing actions show through.

Active customer growth and volume per customer together explain underlying revenue. FY2025 added 2.7m net customers and Personal volume-per-customer hit $4,143 (up 7%). The Costco analogue: customer count is membership growth; VPC is basket size. Card-only customers (~20% of personal active) have $650-$1,300 quarterly VPC and are the lower-engagement cohort; the high-volume cohort responds aggressively to price cuts.

Customer holdings ($27.8bn, +33%) is the deposit franchise. This is the hidden value driver: every $1.3bn of additional balances is worth roughly $10-13m of after-cost interest at the first-1% yield alone, before the above-1% sharing. Holdings have grown 5.7x in four years. If Wise gets the OCC US national trust bank charter it has applied for, this becomes a meaningfully bigger franchise.

Underlying gross profit margin (75%) measures whether unit economics are improving. Up two points in FY2025 against a 14bp take-rate cut tells you variable costs fell faster than price.

Instant transfer share (65%, from 49% three years ago) is the only durable product moat that is hard for incumbents to copy. Each new domestic-rail integration (PIX, Zengin, InstaPay added in FY2025-2026) ratchets this number and widens the gap with bank wires that still settle in days.

Conventional metrics like revenue growth and EPS are misleading here — the first conflates float income with operating performance, and the second is being pushed around by a 11% share-count reduction (1.42bn → 1.27bn) from buybacks and the dual-class collapse.

5. What I'd Tell a Young Analyst

Three things to anchor on, in order.

Decompose every "income" number into volume, price, and float. Wise reports an "underlying income" number and a "reported income" number, and the gap between them is interest income above the first 1% yield — a pure rate-cycle line item. Build your model from cross-border volume × take rate + card revenue + customer balances × yield, separately. If you can't tell a portfolio manager which of those four lines beat in a given quarter, you don't yet understand the business.

The market is mispricing the rate cycle in both directions. Bears focus on the H1 FY2026 PBT decline and miss that volume and customer growth accelerated. Bulls extrapolate the FY2024 reported margin and ignore that the entire above-1% interest stack will halve if base rates normalise to 2-3%. The right framing is: what is the steady-state PBT margin on underlying income at a 2.5% policy rate? Management's 13-16% target is that number, and the stock currently trades as though 18-20% is the new normal.

Watch three things and ignore most of the rest. First, the take-rate exit run-rate each quarter — if it stops falling, the flywheel is broken. Second, customer-balance growth — anything below 25% YoY would signal the deposit franchise is maturing earlier than expected. Third, Wise Platform partner wins (Morgan Stanley and Standard Chartered are the recent signals; Nubank and Itaú already shipped). Platform is the optional second act that turns Wise from a consumer fintech into infrastructure that competes with SWIFT — it is currently in "Other revenue" and not large enough to model precisely, but it is the line that would make this stock worth materially more than 30x earnings if it inflects.

The thesis-killer is not regulation, not competition, not credit losses; it is take-rate stagnation. If pricing power flips from "we choose to cut" to "we are forced to hold," the entire scale-economy-shared narrative collapses and the multiple compresses to PayPal levels. That single line in the quarterly KPI table is worth more than every other slide in the deck.