Remitly Global, Inc. RELY

Key data

The business

Remitly moves money across borders for people who send part of a paycheck home. A nurse in Texas sends $300 to her mother in Manila, Remitly debits her card, and within minutes the cash is ready for pickup or has landed in a bank account or mobile wallet on the other side. It reaches about 150 receiving countries and earns two ways: a transfer fee, and a spread on the exchange rate the customer barely notices, which is where much of the margin lives.

The moat is habit and trust. Sending money home is a recurring, emotionally loaded errand, and once someone has watched a few transfers arrive correctly they stop shopping around. That reluctance to switch, plus a payout network stitched across banks, wallets and cash agents in scores of countries, is what a new entrant cannot copy quickly. The weakness is that Remitly does not set the corridor price. Western Union, Wise, and a shifting cast of app rivals do, and the fee is one tap away from comparison.

Growth has come from adding customers and pushing them into new corridors, not from raising prices. Revenue rose from $654M in 2022 to $1.64B in 2025, compounding about 36% a year, while the losses that funded that land-grab finally reversed.

The numbers

The story is operating leverage arriving all at once: revenue climbs steadily and the cost of running the business has stopped climbing with it.

Read the last two rows against the first three. June-quarter revenue grew 20% year over year, a step down from the mid-30s, but net income went from single-digit millions to a figure unthinkable in the loss years. Most of that jump is not operating: $141.3M of the $205.9M came from releasing a valuation allowance against U.S. deferred tax assets, an accounting gain, not cash from customers. Strip it out and pretax income was $64.6M, still the cleanest quarter the company has printed. The real inflection was Q4 2025, when the business first earned meaningfully after tax.

Free cash flow tells the same story without the tax noise: negative $63M in 2023, positive $177M in 2024, $296M in 2025. That is the number that makes this a business rather than a story.

On the compounding math: at about $165M of normalized pretax earnings and rising, against a $5.5B market value, you pay roughly 33 times that, and over 40 times once a normal tax rate applies. The reported 19 times is a mirage. The honest figure asks the business to nearly double earnings to grow into the price. The variant this memo holds is that revenue is decelerating faster than the multiple assumes, and the September quarter's growth line settles it.

Management

CEO Sebastian J. Gunningham leads the company as of September 2026. Co-founder Matthew Oppenheimer served as CEO through at least April 2026 and drew $296K in total 2024 compensation, under half a percent of that year's net income, with no new stock award. The record with shareholder money is mixed. Remitly bought back $51M of stock in 2025 at an average near $19, well below today's $26, a good use of cash. Against that, diluted shares grew about 30% in three years as stock compensation, running near 12% of revenue, did the diluting, so the buyback is a bucket bailing a boat still taking on water. Insiders have been sellers only: 53 sales and no purchases over the past year totaling $38.5M, led by co-founder Josh Hug unloading $17M in mid-July near the highs. This run's data does not flag which sales were pre-scheduled 10b5-1 plans versus discretionary, so read the timing, not the intent, but no insider has bought at these prices.

How it fails or surprises you

Take-rate compression (downside). Remitly cannot set corridor prices, and the exchange-rate spread that carries the margin narrows first. Watch gross margin, 60% today: a two-point slip over two quarters with revenue still growing would mean price, not cost, is moving, and the whole earnings inflection thins out.

Growth decelerating into a growth multiple (downside). This is the fact the read explains least well. Growth has fallen from the mid-30s to 20% while the stock trades over 40 times normalized earnings. If the September quarter steps down toward the high teens, the multiple and the growth rate collide and the stock re-rates down regardless of profit.

The margin flywheel outruns expectations (right tail). The leverage that flipped losses to $296M of free cash flow in two years may not be done. If revenue holds near 20% while operating costs grow mid-single digits, as they did last quarter, pretax margin could roughly double, and a profitable, growing remittance leader is a scarce thing the market pays up for. The print that reveals it: operating expenses growing slower than revenue for a third straight quarter.

Closing thoughts

A specific print settles most of this, and it is close. Nothing turns on macro or a distant catalyst. It turns on whether the September quarter shows growth stabilizing near 20% with costs flat, in which case the operating-leverage story is real and the rich multiple is defensible, or shows growth sliding toward the high teens, in which case a cooling top line meets a growth price and the reader waits a quarter before touching it. Since this name was last looked at in late August no new quarter has printed and the shares have edged toward their 52-week high, so the tax-flattered headline still stands unrepeated and the clean $64.6M pretax quarter is the number to hold in mind. The left tail is not bankruptcy: almost no debt and $296M of annual free cash flow. The left tail is paying a 40-times growth price for what has quietly become a value-priced compounder growing 20%, not 35%.

The bet is still that immigrants keep leaving the cash-wire counters for the app and keep coming back once the money arrives right. What breaks it is price, not demand: if the exchange-rate spread compresses even as customers grow, revenue and gross margin fall together. The one pair of numbers that tells you first is revenue growth and gross margin in the September quarter. If both hold, the price is earned. If margin slips while growth cools, it is not.