Cole Page
Credit Analyst · cole.page.2027@anderson.ucla.edu · colehpage.com
dLocal 4Q24: The Take Rate and the Merchant Question
Written the week DLO fell from $12.60 to $8.87 on a short report and a soft print. The take rate call was right and compression ran further than I expected. The conclusion I drew from it, that growth would stall without new merchants, was wrong.
I first published this on Substack on 08-Mar-25. The wording here has been rewritten since, but the analysis, figures and conclusions are as I wrote them. I added the mark at the end in September 2026. On the publish date I held DLO at 2.76% of my portfolio.
The Setup
DLO has had a rough two weeks. It dropped 8% to $12.60 on 20-Feb-25 on a short report from Hollenden Square Research, recovered for a little while, and then fell 30% after hours following the 4Q24 call on 27-Feb-25, which took the stock to $8.87 on 4-Mar-25 and back to where it was in August 2024.
The short report deserves little of the credit. Hollenden's piece ran to 500+ pages alleging undisclosed related parties, money laundering and accounting fraud, and compared dLocal to Wirecard and BCCI, and I read all of it. Most of the claims refer to events from more than three years ago, largely before the IPO, and they were addressed in the 2022 Muddy Waters round. Where you'd want evidence, the report asks leading questions instead ("Do these financials look like those of a legitimate payments company, or rather a potential money laundering or other dubious scheme?"). It also misreads the normal hiring of a fast-growing fintech as a governance problem, and its Argentina bond claims rest on speculation. Muddy Waters had real concerns in 2022 and produced real change, including Pedro Arnt coming over from MercadoLibre as co-CEO in January 2023, and this one is piggybacking on that work.
The earnings reaction is the part worth taking seriously, and I think it was also an overreaction, but the quarter had three trends in it that matter more than the move in the price does.
The Quarter
| 4Q24 | Value | YoY | QoQ |
|---|---|---|---|
| Total Payment Volume | $7.7B | +51% | +18% |
| Revenue | $204M | +9% | +10% |
| Gross profit | $84M | +20% | +7% |
| Net income | $30M | +4% | +11% |
| Adjusted EBITDA | $57M | +16% | +9% |
TPV grew 51% while revenue grew 9%, and the rest of this piece is me working through that gap. Cross-border grew faster than local-to-local for once, which is a good sign, since it means existing merchants are using the platform to expand internationally instead of just processing in one market.
Where the Take Rate Went
Gross profit over TPV fell to 1.1% from 1.4% a year ago. Management named three causes for that on the call, and all three of them check out.
Higher volumes, lower rates. Large merchants negotiate their rates down as they scale, the same as in any bulk arrangement, and dLocal serves Tier 0 and Tier 1 global merchants who process enormous volume and get priced accordingly. TPV grew 51%, but a larger share of it came from the merchants paying the least per transaction. That isn't bad news by itself, because winning the largest global merchants is what you want, and a lower blended rate is the price you pay for it.
Payouts growing faster than pay-ins. These are two different businesses inside the same company. With a pay-in, a customer pays a merchant and dLocal provides localization, fraud and compliance around the transaction. With a payout, a merchant pays a worker or a supplier (say Uber paying a driver in Mexico), and dLocal is moving money with far less service on top. Payouts grew 68% YoY against 44% for pay-ins, so the mix shifted toward the lower-value half and revenue per dollar processed fell.
Emerging market currency depreciation. dLocal reports in USD, so when a local currency falls, the same local volume converts into less revenue.
Nigeria
The Naira is the clearest case: the Central Bank abandoned its peg in June 2023, and the currency lost 25% in a single day and roughly 40% in total. A second devaluation in February 2024 took another 44.5% month on month. Letting it float freely without intervention opened a gap between the official and parallel rates, which was compounded by a USD shortage tied to weak oil production, and by inflation at multi-decade highs.
Nigerian revenue fell 90% YoY in the quarter and dragged Africa & Asia down to -9%, and Nigeria went from 13% of total revenue at the end of 2023 to 1.78% at the end of 2024. Egypt covered most of the gap, which is the argument for operating across many markets rather than a few. The volume didn't disappear, only its value in dollars did, and that's the same point as the take rate: TPV alone doesn't tell you what a quarter earned.
Regional
LatAm revenue was $152.9M, up 16% YoY and 75% of the total, and the growth came from Argentina and Colombia rather than the usual two. Brazil declined because of regulatory changes affecting the largest merchant there, who moved to processing their own payments to stay compliant. It's worth noting how that ended, though, because dLocal kept the merchant and moved them onto other products. The revenue was lost to regulation, which isn't a verdict on the platform, and the retention is the encouraging part.
Mexico slowed because of the same Tier 0 mix effect I described above.
Existing vs New Merchants
This is my largest concern, and it's the reason I haven't sized up.
| 4Q24 revenue | Amount | YoY |
|---|---|---|
| Existing merchants | $198.3M | +13% |
| New merchants | $6.1M | -48% |
New merchant revenue was $11.8M in 4Q23 and it's $6.1M now. The trend isn't new, and it has gotten worse over the four years since the IPO. Net revenue retention of 106% says existing merchants keep spending more, which is good, but that number has been falling steadily and is stabilizing as merchants exhaust the products available to them.
Upselling existing clients isn't a strong long-term strategy on its own. I wouldn't expect substantial increases in NRR from here and wouldn't rely on it for growth, and if dLocal keeps failing to bring in new merchants, revenue growth slows.
I see three reasons this is happening, and all of them carry into 2025:
Competition. Stripe, Adyen and MercadoLibre are pushing into dLocal's most important markets, and so are local processors in Africa with sharper regional products. The cross-border payments market is forecast to reach $320tn by 2032, so everyone is paying attention.
Enterprise focus. Chasing one large merchant beats chasing a hundred small ones on cost to acquire, but it concentrates the risk, since losing one Tier 0 merchant (or repricing one) moves the whole number.
Go-to-market execution. I don't want to assume anything without evidence, but a 48% decline in new merchant revenue in a market this competitive raises real questions about sales, pricing and onboarding. New leadership has had barely a year, and I want to see structural change here before I stop worrying about it.
2025 Guidance
| Metric | Guided YoY |
|---|---|
| TPV | 35–45% |
| Revenue | 25–35% |
| Gross profit | 20–25% |
| Adjusted EBITDA | 20–30% |
Sorted by whether management controls them, the three risks look like this:
- Currency volatility. This one is environmental, and hedging and renegotiating with processors help at the margin.
- Take rate compression. It's largely structural, and partly the price of winning larger merchants.
- New merchant growth. This is the one dLocal actually controls, and it's the one where the execution hasn't been visible.
I'm cautiously bullish and will likely add while the price is here. The business is growing, its compliance posture in hard markets is a real edge, and the reaction to this quarter had more to do with the tape than with the print. Still, the third risk is the thesis, and if 2025 passes without traction in new merchants, growth slows regardless of how much volume runs through the platform.
How It Played Out
Marked as of 03-Sep-2026.
| 4Q24, as written | 2Q26 | |
|---|---|---|
| TPV | $7.7B, +51% | $17.7B, +92% |
| Revenue | $204M, +9% | $399.7M, +56% |
| Gross profit | $84M, +20% | $127.2M, +29% |
| Gross profit / TPV | 1.1% | 0.72% |
| Net revenue retention | 106% | 153% |
| DLO | $8.87 (04-Mar-25) | $15.78, +78% |
Two of the three calls held up. The selloff was an overreaction, and the stock is up 78% from the low. Take rate compression continued and ran further than I expected, with gross profit over TPV now at 0.72%, down from 1.07% a year ago and from the 1.1% I was writing about, and every reason I gave for it is still operating.
The growth conclusion was wrong. I had the fact right, in that new merchant revenue was small and shrinking, and I drew the wrong inference from it. I said NRR at 106% was decaying and shouldn't be relied on for growth, and it's now 153% and has been above 140% for five straight quarters, with TPV retention at 188%. Existing merchants adding countries, payment methods and products turned out to be the entire engine, and it was a far better one than I gave it credit for. Management raised 2026 guidance to TPV growth of 60 to 70% and gross profit growth of 25 to 30%.
What I got wrong was treating net revenue retention as a maturing metric with a ceiling, when the product surface underneath it was still expanding. A merchant count of 760 across 60+ markets doesn't need to grow for revenue to compound, as long as each merchant keeps buying more. My bear case assumed revenue couldn't compound without new logos, and it did.