Flywire Corp FLYW

Key data

The business

Flywire moves money across borders for payments that are large, infrequent and a nightmare to reconcile. Its first and still largest business is tuition: a student in Mumbai owes a university in Boston or Melbourne, and Flywire collects the local-currency payment, handles the FX, and lands the exact dollar figure in the school's account with the student ID attached so the bursar knows who paid. It plugs into the local rails a student already uses and sits inside the university's billing page, and that two-sided integration, awkward to rip out on either end, is the moat. On top of education it has bolted travel, healthcare and B2B supplier payments, and the Sertifi acquisition carried it deeper into hospitality. Net revenue, what it keeps after network and bank costs, runs about 57% gross margin. The payer never sees a "Flywire" bill; they see their own school's checkout and a receipt in rupees.

The numbers

The quarters are seasonal and lumpy, not smooth. Q3 is the fall-enrollment spike, the year's biggest quarter, and the off-season quarters swing to small losses.

What matters under the seasonality is the year-over-year line: Q2 2026 revenue grew 27% and Q1 grew 41%. Yet Q2 still printed a loss and missed a small estimated profit, because stock compensation runs near a tenth of revenue and swallows the thin operating margin.

Revenue roughly tripled from 2021 to 2025, about 33% a year, and operating income crossed from a $21M loss in 2023 to an $11M profit in 2025. This is a business that just reached the far side of breakeven and is now testing whether scale drops to the bottom line. Two things complicate the read. $9.4M of first-half 2026 revenue was pure FX translation as the dollar weakened, so organic growth sits a few points below the reported 27%. And trailing free cash flow is puffed up by client-fund timing, so the FY2025 figure below is the honest cash number. What this memo believes the tape does not fully price: the headline rate overstates the underlying engine once FX and acquisitions come out, and the print that settles it is organic, constant-currency growth over the next two quarters.

Management

Founder Michael Massaro still runs it, and the capital record is the tell. In the first half of 2026 the company bought back 4.1M shares at an average near $12.79, well below today's $18.55, which is disciplined repurchasing rather than the price-insensitive kind. Against that, the stock compensation shown above is doing much of the diluting the buyback is undoing, so the net share count barely moves. Insider selling looks heavy on paper, $23.5M across a year, but the largest seller is Voss Capital, an activist fund that also bought $2.6M in December, and Massaro's own $2.1M September sale carries no disclosed 10b5-1 status. With near-zero debt against the cash above, the balance sheet is not the worry.

How it fails or surprises you

Student-visa concentration. Education is the biggest vertical, and Canada-plus-Australia student flows were about 15% of revenue at the last disclosure. Both governments capped foreign enrollment, and that was the watch-item coming in; it held rather than broke, because group revenue still grew 27% as other verticals absorbed the drag. A second tightening, or a US move, is the clearest path to growth halving.

The FX-and-loss gap. The fact explained least well is that a 27%-growth quarter still lost money, and that $9.4M of first-half growth was only the dollar falling. If the dollar reverses, reported growth and the sliver of profit compress together. The print that proves the bull wrong: constant-currency growth under 20% with operating margin back below zero.

Cross-sell re-rates it (right tail). If travel, healthcare, B2B and the Sertifi hospitality book keep compounding and take rate holds, the visa fear capping the multiple proves overblown, and a 26-times-forward name with a 33% growth history re-rates hard. The market pays for the risk today, not the diversification.

Closing thoughts

This is mostly a distribution the market already prices, not a hidden mispricing. The 72-times-trailing multiple looks absurd only because earnings just crossed zero; on forward earnings and on cash, 26 times and about 26 times EBITDA is an ordinary price for a mid-20s grower with a genuine integration moat and net cash. The edge, if there is one, is small and sits in which way the FX-flattered growth resolves. The left tail is a real visa shock that turns 27% growth into single digits while stock comp keeps GAAP profit near zero; the right tail is the non-education engine proving the concentration fear stale. The fatter tail is probably the upside, but only modestly, and only for a holder patient enough to sit through lumpy, loss-printing off-season quarters.

The bet is still that Flywire keeps handling the tuition that foreign students wire to universities in another country, and keeps loading more schools, plus hospital bills, travel invoices and supplier payments, onto the same pipes, faster than the visa crackdowns and a normalizing dollar can pull growth down. What breaks it is one pair of numbers read together, constant-currency organic growth and operating margin; if both fall at once, the diversification story is failing in real time and the multiple has nowhere to hide.