Business

Know the Business — Wise plc

Bottom line. Wise is a vertically integrated cross-border payments network that earns money two ways: a per-transaction take rate on £145bn of FY25 volume, and a large, rate-sensitive interest spread on £21.5bn of customer balances it holds in trust. The headline business is the take-rate engine — Wise is deliberately cutting it (0.67% to 0.58% in FY25, 0.52% in H1 FY26) to widen its cost moat and pull volume away from banks. The market most likely overweights short-term interest income (which inflates reported PBT) and underweights Wise Platform — the API/embedded business that just signed Morgan Stanley, Standard Chartered, Nubank and Itau.

All figures in GBP (£). Wise's fiscal year ends 31 March; FY2025 = 12 months to 31 March 2025.

1. How This Business Actually Works

The engine has two pedals: cross-border revenue (volume × take rate) and Card & Other revenue (interchange + interest income on the float). They are connected — the Wise Account is the product that lets Wise hold balances, and balances are what generate interest income.

The yellow band — interest income on customer balances — is what inflates reported profit when central bank rates are high. Wise's own "underlying income" metric retains only the first 1% of yield and treats the rest as a customer benefit. That is how you should read the business too: at neutral rates, Wise is a £1.4bn-revenue, ~21% underlying PBT margin company; the rate cycle has been adding ~£280m of one-off operating profit on top.

The flywheel runs as follows:

What makes the loop unusual: each turn lowers revenue per pound moved. Take rate fell from 70bps in FY21 to 52bps in Q2 FY26. Management's bet is that the loop's drag (lower take rate) is more than offset by volume growth (23% FY25, 24% H1 FY26) and by Card & Other income from a deeper account relationship.

The cost moat is not software — it is regulatory and operational. Wise holds 70+ licences and has 6 live direct connections to domestic payment systems (UK FPS, SEPA, AUS NPP, Singapore FAST, Hungary, Philippines InstaPay; Brazil PIX and Japan Zengin in flight). A direct connection cuts payment cost by ~8x and pushes 90% of transfers through instantly. Roughly 50% of Wise transactions now ride proprietary rails; the other half still touches a partner bank. Each new direct connection is a multi-year regulator-and-engineering project — that's the moat.

2. The Playing Field

Notes: Wise figure is FY25 underlying income (£1,645m incl. all interest income). Listed peers are calendar FY2025. Revolut figures from publicly disclosed FY2024 accounts (latest available); margin and customer count approximate. Margins use operating-income-equivalent.

Three things this picture reveals:

1. Wise is the only listed peer combining 15%+ growth with 20%+ underlying margin. Remitly is growing faster but barely profitable; Western Union is the inverse. PayPal is in a different revenue league but growing 4%.
2. Revolut is the real strategic threat, not the listed peers. It bundles Wise's value proposition into a wider neobank account, charges a slightly higher take rate, and grew 72% in 2024 to ~£3bn revenue at ~37% reported operating margin (FY24 accounts). Revolut has more customers (65m+) than Wise. The fact Revolut is excluded from most sell-side comp tables is a gap.
3. Western Union's margin is the cautionary tale. A 19% operating margin on -4% revenue says "harvest mode incumbent." That margin is what awaits Wise if it stops investing — but the take rate it would need to charge to keep that margin is exactly the price umbrella that lets digital challengers eat its lunch.

The peer set also explains why Wise Platform matters. Wise can't out-distribute Western Union's 600k retail agents or PayPal's 434m accounts. It can rent out infrastructure to firms that already have distribution (Standard Chartered, Morgan Stanley, Nubank, Itau, Monzo). Platform is the only path to "trillions" volume management keeps invoking — guided from ~5% of total today to 50%+ long-term.

3. Is This Business Cyclical?

Not in the industrial sense — Wise has no inventory, no commodity input, no construction-cycle demand. The cyclicality is interest-rate-driven, with a secondary, smaller exposure to global migration and SME formation.

Read this carefully. In FY22, when central bank rates were near zero, interest income on customer balances was effectively zero. By FY24 it was £485m — bigger than the entire reported operating profit. In FY25 it grew to £594m. More than 100% of FY24 and FY25 reported operating profit is attributable to interest income on float that did not exist as a meaningful line item three years earlier.

This is why Wise discloses an "underlying" view that retains only the first 1% of yield as Wise's. At a structural Bank of England base rate of ~3%, the other ~£280m flows back to customers via Wise Assets ('Interest' product) or is set aside as a customer benefit. The market debate is what level of rates is durable. The H1 FY26 average yield on customer balances was 3.95% — meaningfully off the FY24 peak but still well above the FY22 trough.

Beyond rates, the cycle exposures are mild:

• Migration / remittance flows — global remittance volumes are surprisingly counter-cyclical (workers send more home in downturns). 280m+ migrants form a structural demand floor.
• SMB volumes — Wise Business is more cyclical (B2B trade slows in recession). ~27% of cross-border volume is from SMBs (£38.8bn of £145.2bn FY25).
• FX volatility — Wise hedges, so volatility is largely a wash; very wide spreads can briefly help cost-of-sales.

What is not cyclical: customer count and balances. Even in March 2020 (COVID crash) Wise grew customers and balances. Stickiness comes from local account details (UK sort code, US routing number, EU IBAN) that customers don't want to re-paper.

4. The Metrics That Actually Matter

Stop staring at revenue growth and reported PBT. These are the metrics that drive value:

Note: Platform partner counts are illustrative — Wise discloses notable partners but not a full count. Scorecard from FY25 annual report and analyst presentation.

The pattern matters: customers > volume > holdings > income. Each rung loses some growth to the deliberate take-rate cut. The day customer growth slows below 15% is the day the model starts breaking, because there's no longer enough volume growth to fund the price cuts.

Two metrics most analysts get wrong:

• Take rate. Lower is not bearish — it is the strategy. Fixate on whether take-rate cuts are matched by volume growth (FY25: -9bps, +23% volume; H1 FY26: -10bps, +24%). The day cuts stop generating volume growth, the bear thesis is real.
• Customer holdings. Reported as a balance-sheet curiosity but it is the single largest determinant of interest income and the leading indicator of whether the Wise Account is becoming a primary account. £21.5bn at FY25 is an average of ~£1,400 per customer — still small vs a primary current account, so room to grow before saturation.

5. What I'd Tell a Young Analyst

Watch what the market is mispricing, not what it is pricing.

What's likely overestimated:

• Reported earnings durability. ~£280m of FY25 PBT is the rate-cycle bonus. Strip it out and the underlying business trades at a multiple closer to 30x than the headline 20x. Every 100bps of base-rate cuts trims roughly £200m of pre-tax income.
• Take rate as a "yield" metric. It is not yield — it is the price Wise charges customers to send money. Wise wants it lower. Modeling a stable take rate is a structural error.
• TAM in developed-market consumer remittance. Wise already has ~5% of global consumer cross-border volume; in the UK, US, EU it is materially higher. Easy growth from here is in business and platform, not consumer.

What's likely underestimated:

• Wise Platform. Three FY25 wins (Morgan Stanley, Standard Chartered, Itau) collectively touch over 100m end-customers. Platform is guided from ~5% of total volume today to "more than 50%" long-term. If true, this is a B2B infrastructure compounder, not a consumer fintech.
• The cost moat. ~£3bn cumulative infra spend, 70+ licences, 6 direct connections, 850+ engineers and ~33% of staff in financial-crime work. None of that shows up in P/E. Replicating it would take a competitor 5+ years and £2bn — and they would still need to acquire the customer base.
• The customer-balance flywheel. As the Wise Account becomes a primary account (already 50% multi-product adoption), balances per customer should grind from £1,400 toward bank-account levels (£3-5k). That is non-linear interest-income upside even at lower rates.

What would actually change the thesis:

• Take-rate cut without volume response (i.e. the price advantage stops creating share gain). H1 FY26 needs to keep the pattern; one bad period is noise, two is a problem.
• Material loss of a Platform partner or a major regulator restricting embedded payments models.
• The CEO/founder situation. Käärmann's ~49% voting control and his FCA review (£350k tax fine in 2024, ongoing) are governance overhangs the dual-listing US move will spotlight. A forced CEO departure is not priced.
• A US dual-listing that fails to attract incremental US investor demand — neutral case is fine; downside is "failed listing" optics.

One number to remember: at FY25 underlying income of £1,362m and 16% guided growth, Wise should clear £2bn of underlying income by FY27. At a 13-16% underlying PBT margin, that is £260-320m of PBT before any rate-cycle bonus. Anything you pay for Wise above ~£10bn market cap is a bet on either (a) Platform succeeding at scale, or (b) interest rates staying above 3% indefinitely. Pick which one you are underwriting.