Between AstroPay, PayRetailers and EBANX who’s actually eating the smallest portion of a…

Heard that FX landmine story from Curitiba twice this year already and it still stings. The “real BRL” line is a two-tier game: AstroPay front-ends your PIX traffic, then sweeps the credit to BRL in a second hop. Where most players assume one transparent FX step, you actually get stacked twice—once on entry (credit card FX into AstroPay USD) and again when AstroPay converts USD back to BRL for the merchant payout. In the last settlement batch I ran, the second hop widened by an average 1.6 pp vs. the commercial mid, and that’s before they debit the promised 0.8 % CB. In practice, the total erosion lands between 2.4 % and 3.0 % of deposit volume, which is half a basis point higher than PayRetailers’ single FX layer and a full basis point above EBANX’s blended quote when the latam liquidity pool is calm. Add the rolling reserve sweepstakes: AstroPay holds 5 % for 90 days on any PIX ticket unless you keep six clean quarters behind you. That reserve alone wiped another 0.3 pp off NGR in Q1. On a 2 M USD deposit book that’s 6 k USD parked for three months—cash that could sit on your operating account earning 3 % but is instead tied up earning zip. So yes, Rob_WL is spot-on; the “hidden” FX isn’t hidden so much as deliberately two-layered. If your margin stack can’t stomach that second conversion plus the reserve drag, you’re better off letting EBANX ride the float themselves and eat the 1.1 % blended spread upfront—because at least you see the cost in one line item and can price it into the NGR forecast.

Ever seen a vendor’s cost sheet that looks cleaner than a freshly laundered soccer kit—only to find the stains still there once you put it through the wash? AstroPay’s "real BRL" pitch? That’s the same trick: they sell you a two-for-one FX discount when it’s actually a double dip, and the second conversion is where the knife comes out. I’ve got a client in Fortaleza running 1.8 M USD of PIX volume last quarter. Their monthly AstroPay settlement sheet looked tidy—until we layered in the intercompany FX sheets. The second hop on PIX exits hit an average 1.9 pp above the mid versus PayRetailers’ 0.4 pp and EBANX’s 0.2 pp in the same period. And SoftAndReadyAndScaling’s math tracks—AstroPay’s rolling reserve on new PIX traffic starts at 5 % until you hit six consecutive clean quarters, so every new joint brought into the pool is immediately a cash-flow drag. The client had to book a 9 k USD top-up to cover reserves in Q2 while still eating the CB line at 3.8 %. Question I keep asking: why does the industry let AstroPay book the "real BRL" claim when the actual FX math lands two layers deep? If you can’t control the second hop, you’re not buying PIX—you’re buying a FX tourniquet disguised as a payment rail.

Wait… what’s this "intercompany FX sheets" thing Steve just mentioned? Are they the breakdowns of how AstroPay shuffles money between their own accounts to make that second hop? Or is it something operators are supposed to request separately to see where the extra bleed is coming from?

so OffshoreForeverLoyal are you still running funds through a single mid-office ledger or have you let the finance guys loosen up the Excel spreadsheet handcuffs? the intercompany FX sheets are nothing spooky—imagine AstroPay in Curacao has a BRL sub-account because they can’t get a local license, so when a player in São Paulo pays via PIX the money lands in the sub-account at the local Brazilian bank. now AstroPay sweeps that BRL off their books into their master USD pool using their own internal rate instead of the Visa MasterCard commercial mid, then they pay you out of the same USD pool and quote you “real BRL” while the invoice you signed months ago never mentioned the second hop. it’s like ordering a taxi from the airport in Lisbon and getting charged twice: once in euros and again when the driver converts your euros to escudos inside the cab. the sheets just track those two conversions and show you how far the second hop drifts from the mid you thought you locked in.

Stepped into a local cashier in São Paulo last week to withdraw BRL for petty expenses and watched the teller key in two FX rates on the screen before hitting confirm—first the Visa mid, then a second rate that clipped 0.8 pp off the top. That’s the exact play AstroPay runs behind the curtain: the player thinks they’re clearing a PIX invoice at mid-market, but once AstroPay squirrels the BRL into their own USD pool via the second hop, the margin they shave isn’t disclosed anywhere on the operator contract—only buried in the final settlement line when you run the numbers at month-end.

You seeing two conversions in the AstroPay waterfall isn't paranoia—it's physics. I ran a soft launch in Florianópolis last month to test PIX vs. cards, and the FX bleed on exits alone cost us 2.7 % of deposit volume against EBANX's single-layer spread. That’s not peanuts when your NGR on Brazilian traffic sits at 38 %. You ever try explaining to the CFO that the "real BRL" promise just swallowed two-thirds of their projected margin?

Look, the thing that truly bites is the mismatch between what they tell you upfront and the real bleed that shows up on the settlement statement. I had a joint in Rio running 1.2 M USD monthly PIX deposits through AstroPay last quarter—clean traffic, six-figure FTDs every month—and the final settlement sheet told a different story. The FX on the exit conversions wasn’t 0.8 %, it was 1.9 % above the commercial mid, and that spread widened by 30-40 bps when BRL volatility spiked after Copom minutes. What killed us wasn’t the CB line (they stayed at 0.8 %); it was the hidden FX layer that only appears as “currency adjustment” in the invoice grid, not as a separate line. The CFO’s margin deck still forecasts 0.8 % FX cost on Brazilian traffic; the CFO’s margin deck needs a serious rewrite if they want to keep paying those exit conversions out of operating cash.

Found the FX sheet from my São Paulo client last month—AstroPay processed 950k USD in PIX deposits, and the exit conversion landed me at 2.1 pp over mid on real exits. That’s not 0.8 % they promised—it’s more than double. And the kicker? The settlement line called it “FX adjustment” with no breakdown, just like TurnkeyHQ said. Next quarter I’m forcing finance to run two FX layers through their model before we even sign a new MID. Lock-in math is free until you audit it.

So the AstroPay FX sheet is basically where your margin goes to die and nobody tells you the name of the coffin until the CFO screams. 😬 Between the PIX pickup at commercial mid, the intercompany vault to USD that smashes you with 1.9-2.1 pp on exit, and the hidden “currency adjustment” row that isn’t a line item but a knife twisting slower every Copom minute, the bleed hits far north of the 0.8 % CB fee they love to quote. EBANX still looks like a straight pipe—single-layer spread, no second hop, invoice hits your USD ledger without magic tricks—so their exit cost stays close to the Visa mid minus a thin strip of margin. PayRetailers sits somewhere in-between, but last time I saw their real exit spread it came out at roughly 0.45 % above mid after accounting for the local acquiring fee baked in. Either way, if you model 2.7 % bleed on every USD that exits via AstroPay and your Brazilian NGR is already 38 %, you’re effectively donating 7 % of your operating cash to FX theatre—just to keep the lights on in Florianópolis. Guess the open question is: why does Curacao even allow a licensed money-mover to pretend the mid is something other than a spot price?