Four Parents, Four Very Different Fintech Bets
Every super-app or marketplace eventually spins up a financial-services arm, and the pitch is always the same: more users transacting on the core platform generates payment and credit data, that data lets the company underwrite loans nobody else can price as cheaply, and the resulting fintech unit becomes a second profit engine riding on top of the first one's user base. The interesting question isn't whether that loop exists — it's which of these four companies has actually turned it into money.
Mercado Pago is MELI's fintech segment — payments, a digital wallet, and a fast-growing consumer/merchant credit book layered on top of Latin America's largest e-commerce marketplace. Monee (renamed from SeaMoney in 2025) is SE's fintech arm, sitting alongside Shopee and Garena and increasingly functioning as a standalone digital bank across Southeast Asia. GoTo Financial is GOTO's fintech segment, built around GoPay and Bank Jago-linked lending on top of Gojek's ride-hailing and delivery base in Indonesia. Grab's Financial Services segment (GFS) is the smallest and youngest of the four, built around GXS Bank (Singapore), GX Bank (Malaysia), and a minority stake in Indonesia's Superbank, layered on top of Grab's own ride-hailing and delivery base — the direct regional rival to GoTo Financial's model.
All four tell a version of the same story — rapid volume growth, expanding user bases, credit books scaling fast — but they disclose that story very differently, and the gaps in disclosure are themselves informative. Mercado Pago is the largest of the four by revenue and the most opaque about profitability. Monee is the steadiest, most consistently profitable compounder. GoTo Financial is the smallest of the three profitable segments but has the most dramatic two-year arc, flipping from a rounding-error loss to the segment that now earns more Adjusted EBITDA» than GOTO's core marketplace business. Grab's GFS is the outlier: the only one of the four still segment-loss-making in every quarter shown, even as its loss narrows and its loan book scales the fastest of any of them.
The Prescription
Judged purely on which fintech arm has the strongest trajectory right now, GoTo Financial wins on momentum, Monee wins on quality, and Grab's GFS is the one still proving it can get to breakeven at all. GoTo Financial's Adjusted EBITDA has gone from roughly break-even to overtaking its own parent's On-Demand Services segment in two years — the sharpest inflection of the four — but it's compounding off a tiny base, in a single country, with a loan book still under $700 million and no disclosed return-on-assets to validate the credit quality behind that growth. Monee has grown revenue and Adjusted EBITDA every single quarter in this window without a single step backward, has the largest loan book of the four relative to its parent, and already produces real (if modest) operating income — the least exciting story, and the most trustworthy one.
Grab should keep doing exactly what's narrowing GFS's loss — its Q1 2026 Adjusted EBITDA loss of $17 million is the smallest of the entire two-year window even as revenue and the loan book both kept accelerating — but it should also start disclosing the dollar figures behind its credit-quality claims: management has repeated the same "second-half 2026" breakeven timeline unchanged since it was first disclosed in Q1 2023, and a reader has no way to independently verify the improving expected-credit-loss ratio management points to as the reason the loss keeps narrowing, since Grab doesn't break out the actual dollar figures behind that percentage claim (see Beyond the Usual).
Mercado Pago should stop letting Mercado Libre disclose Fintech revenue and TPV» every quarter while never publishing a segment EBITDA or operating margin line — Monee, GoTo Financial, and even loss-making Grab's GFS all manage to disclose a genuine segment EBITDA figure, and Mercado Pago's credit book is now larger than any of theirs, which makes the profitability silence more conspicuous, not less. If Mercado Pago wants credit for being MELI's second growth engine, it needs to show the segment math the way its three peers already do.
The Two-Year Scorecard, Quarter by Quarter
Same calendar quarter, compared across all four companies. Figures are each company's own disclosed fintech-segment numbers — not the consolidated group.
A currency-comparability note before the tables below. Three of these four companies — MELI, SE, and GRAB — report their fintech segment natively in US dollars. GOTO reports in Indonesian Rupiah. Every table below still converts GoTo Financial's Rupiah figures to USD (using Bank Indonesia's quarter-end reference rate, the same method every other GoTo post on this site already uses) so a reader can see the absolute-dollar comparison at a glance — but a USD-converted growth rate for GoTo Financial is not a clean read of the underlying business, because it bundles two different things together: how much the business actually grew, and how much the Rupiah moved against the dollar over the same stretch, which have nothing to do with each other. Over this post's roughly two-year window, the Rupiah went from IDR 15,138 to IDR 17,908 per US dollar (Bank Indonesia's quarter-end reference rate, Q3 2024 to Q2 2026) — a genuine 18.3% depreciation in the currency itself, unrelated to anything GoTo Financial's business did. Because of this, every table below shows GoTo Financial's own Rupiah (local-currency) YoY growth rate as the primary, comparable figure, with the USD-converted absolute figure alongside it as a secondary, supporting number for scale comparison only — not the reverse. For MELI, SE, and GRAB, local currency and USD are the same currency, so this split doesn't apply to them; their disclosed YoY growth rates are already directly comparable to GoTo Financial's Rupiah-denominated growth rate, and it's that Rupiah figure — not the USD-converted one — that should be read against theirs.
On top of that, the tables below now also add a genuine constant-currency ("FX-neutral") USD growth rate for GoTo Financial, computed by converting both the current quarter's and the year-ago quarter's Rupiah figure at the same exchange rate — the current quarter's own rate — instead of letting each quarter float on its own historical rate. A worked example is in "GoTo Financial's Growth, Isolated From the Rupiah" below. The short version of what it shows: GoTo Financial's Rupiah (local-currency) YoY growth is already, by definition, FX-neutral — currency never enters a same-currency comparison — so the newly added constant-currency USD figure converges almost exactly on the Rupiah growth rate already shown, not on the raw USD-equivalent growth rate a naive dollar conversion would imply. It's the raw USD-equivalent growth rate that's the distorted one, since it's built from two different historical exchange rates, one per quarter, and therefore blends real business growth with the Rupiah's depreciation. That match between the Rupiah growth rate and the constant-currency figure is itself the finding: it confirms GoTo Financial's growth story is real and not a currency illusion.
Top-line volume and revenue
| Quarter | Mercado Pago Revenue | Mercado Pago TPV | Monee Revenue | GoTo Financial Revenue (Rp, YoY% local ccy) | GoTo Financial Revenue (USD equiv.) | GoTo Financial Revenue FX-neutral YoY% (USD, constant FX) | GoTo Financial Core GTV» | GRAB GFS Revenue | GRAB GFS Loans Disbursed |
|---|---|---|---|---|---|---|---|---|---|
| Q3 2024 | $2,173M (+20.8% YoY) | $14,650M (+124.1% FX-neutral) | $615.7M (+38.0% YoY) | Rp1.0T (+128% YoY) | ~$68M | n/a (no prior-year USD figure in this table) | ~Rp64.6T (+82% YoY) | $64M (+34% YoY) | $567M (+38% YoY) |
| Q4 2024 | $2,505M (+28.6% YoY) | $17,080M (+101.4% FX-neutral) | $733.3M (+55.2% YoY) | Rp1.2T (+95% YoY) | ~$74M | n/a (no prior-year USD figure in this table) | ~Rp71.6T (+71% YoY) | $74M (+38% YoY) | $639M |
| Q1 2025 | $2,632M (+43.3% YoY) | $17,986M (+112.8% FX-neutral) | $787.1M (+57.6% YoY) | Rp1.2T (+90% YoY) | ~$72M | n/a (no prior-year USD figure in this table) | ~Rp76.1T (+57% YoY) | $75M (+36% YoY) | $630M (+30% YoY) |
| Q2 2025 | $2,951M (+40.3% YoY) | $20,237M (+83.1% FX-neutral) | $882.8M (+70.0% YoY) | Rp1,356B (+76% YoY) | ~$84M | n/a (no prior-year USD figure in this table) | ~Rp82.2T (+46% YoY) | $84M (+41% YoY) | not disclosed this quarter |
| Q3 2025 | $3,235M (+48.9% YoY) | $23,512M (+71.7% FX-neutral) | $989.9M (+60.8% YoY) | Rp1.5T (+55% YoY) | ~$90M | ~+55% (matches local ccy; raw USD-equivalent YoY was only +32.4%) | ~Rp95.3T (+48% YoY) | $90M (+39% YoY) | $886M (+56% YoY) |
| Q4 2025 | ~$3,781M (implied, +51%; MELI does not disclose a standalone Q4 figure — full-year Fintech revenue was $12,599M, +46.2%) | (not separately captured this quarter) | $1,131.9M (+54.3% YoY) | Rp1.7T (+45% YoY) | ~$101M | ~+45% (matches local ccy; raw USD-equivalent YoY was only +36.5%) | ~Rp116.3T (+62% YoY) | $99M (+34% YoY) | $979M (+53% YoY) |
| Q1 2026 | $3,977M (+51.1% YoY) | (not separately captured this quarter) | $1,241.8M (+57.8% YoY) | Rp1.9T (+58% YoY) | ~$112M | ~+58% (matches local ccy; raw USD-equivalent YoY was only +55.6%, the closest the two ever get in this window) | ~Rp131T (+72% YoY) | $107M (+43% YoY) | $1,050M (+67% YoY) |
| Q2 2026 | MELI has not yet reported Q2 2026 as of this post | — | SE has not yet reported Q2 2026 as of this post | Rp2,070B (+53% YoY, from Q2 2025's Rp1,356B; 36.6% of GOTO's group revenue) | ~$115.6M | +53% (see worked example below; raw USD-equivalent YoY was only +37.6%) | ~Rp157T (+91% YoY) | GRAB has not yet reported Q2 2026 as of this post | not yet reported |
MELI, SE, and GRAB's latest reported quarter on file is Q1 2026; GoTo has already reported Q2 2026, one quarter ahead of the other three. Where a quarter isn't yet available for a company, it's marked rather than estimated. MELI's Q4 2025 Fintech revenue is a derived arithmetic residual (full-year minus the three disclosed quarters), not a number MELI itself published as a standalone Q4 line — flagged as such rather than presented as a direct disclosure. GoTo Financial's Q2 2026 YoY growth rate is likewise an implied figure (calculated from its own two disclosed Rupiah revenue figures), since GoTo's Q2 2026 materials stated the absolute Rupiah amount and its share of group revenue but not a standalone YoY percentage for the segment.
GoTo Financial's new FX-neutral column is a constant-currency USD growth rate: both the current quarter's and the year-ago quarter's Rupiah revenue converted at the same rate (the current quarter's own Bank Indonesia reference rate). For Q3 2024 through Q2 2025, that calculation needs a year-ago USD figure this table doesn't carry (it would require Q3/Q4 2023 and Q1/Q2 2024 data, outside this post's window), so it's marked n/a rather than guessed at. From Q3 2025 onward, where a prior-year USD-equivalent figure is already in the table, the constant-currency rate lands within a rounding error of the local-currency (Rupiah) growth rate shown in the first GoTo Financial column — as it mathematically must, since holding the exchange rate constant cancels the currency out of the ratio entirely. It's the plain, uncorrected USD-equivalent-to-USD-equivalent comparison (each quarter's own historical rate) that runs meaningfully lower every time, because that comparison is still absorbing the Rupiah's depreciation. Core GTV isn't given a matching FX-neutral column: this table never converts it to USD in the first place, so its stated growth rate is already Rupiah-native and already FX-neutral by construction.
Users and credit book size
| Quarter | Mercado Pago MAU» | Monee Active Credit Users | GoTo Financial MTU» | GRAB GFS Users |
|---|---|---|---|---|
| Q3 2024 | 56.2M (+35.4% YoY) | ~60% YoY growth (no absolute disclosed) | not disclosed | not disclosed as a unified figure |
| Q4 2024 | 61.2M (+33.6% YoY) | >26M (+60% YoY) | 20.2M (+35% YoY) | not disclosed as a unified figure |
| Q1 2025 | 64.3M (+31.2% YoY) | >28M | 20.6M (+30% YoY) | not disclosed as a unified figure |
| Q2 2025 | 67.6M (+30.0% YoY) | >30M (+45% YoY) | 22.4M (+29% YoY) | not disclosed as a unified figure |
| Q3 2025 | 72.2M (+29% YoY) | 34M (+45% YoY) | not refreshed this quarter | not disclosed as a unified figure |
| Q4 2025 | 78M (+27.9% YoY) | >37M (+40%+ YoY) | 26.2M (+30% YoY) | not disclosed as a unified figure |
| Q1 2026 | 83M (+29.7% YoY) | >38M (+35%+ YoY) | 27.5M (+33% YoY) | not disclosed as a unified figure |
| Q2 2026 | not yet reported | not yet reported | 28.8M (+29% YoY) | GRAB has not yet reported Q2 2026 as of this post |
Grab never discloses a single "GFS users" or "GFS borrowers" figure the way MELI's MAU or GoTo's MTU rows do — GFS spans multiple products (GXS Bank, GX Bank, GrabFin lending, the Superbank stake) with no combined user metric. The closest disclosed proxies are individual product user counts, reported inconsistently quarter to quarter: GX Bank Malaysia reached 892,000 deposit customers in Q3 2024 and crossed 1 million by Q4 2024; Indonesia's Superbank (an equity-method associate, not consolidated) reached 2 million users by October 2024, 2.8 million by December 2024, 3.3 million by Q1 2025, 4.0 million by Q2 2025, and crossed 6 million by Q1 2026. These are useful directional signals but aren't comparable to a segment-wide MAU/MTU figure, so the table above marks the row "not disclosed as a unified figure" rather than substituting one of these partial counts for it.
Credit book size and deposits
| Quarter | Mercado Pago Credit Portfolio | Monee Loan Principal Outstanding | GoTo Financial Loan Book | GRAB GFS Loan Portfolio | GRAB GFS Customer Deposits |
|---|---|---|---|---|---|
| Q3 2024 | $6,016M (+76.8% YoY) | $4.6B (+73.2% YoY) | ~$284M (Rp4.3T) | $498M net (+81% YoY) | $1.1B (GXS + GX Bank) |
| Q4 2024 | $6,573M (+74.0% YoY) | $5.1B (+63.9% YoY) | ~$322M (Rp5.2T) | $536M net (+64% YoY) | $1.2B |
| Q1 2025 | $7,780M (+75% YoY) | $5.8B (+76.5% YoY) | ~$344M (Rp5.7T) | $566M net (+56% YoY) | $1.4B |
| Q2 2025 | $9,347M (+90.6% YoY) | $6.9B (+94.0% YoY) | ~$407M (Rp6.6T) | $708M net (+78% YoY) | $1,543M |
| Q3 2025 | $11,022M (+83.2% YoY) | $7.9B (+69.8% YoY) | ~$429–$456M (Rp7.16–7.6T) | $821M net / $903M gross (+65% YoY net) | $1,311M (-15% QoQ) |
| Q4 2025 | $12,508M gross / $9,365M net | $9.2B (+80.4% YoY) | ~$523M (Rp8.777T) | $1,180M net / $1,278M gross (+100%+ YoY) | $1.6B |
| Q1 2026 | $14,555M gross / $10,735M net | $9.9B (+71.3% YoY) | ~$583M (Rp9.9T; redefinition to include merchant loans) | $1,438M gross (Gross Loan Portfolio, now the headline metric; more than doubled YoY from $625M) | $1,630M |
| Q2 2026 | not yet reported | not yet reported | ~$614M (Rp11.0T) | not yet reported | not yet reported |
Grab's own headline lending metric changed mid-window: through 2025 it emphasized the net loan portfolio (after expected-credit-loss provisions), disclosed the gross figure alongside it starting Q3 2025, then made the Gross Loan Portfolio the sole headline metric from Q1 2026 onward. GFS's customer deposits combine only GXS Bank (Singapore) and GX Bank (Malaysia) — the two digital banks Grab consolidates — and exclude Superbank, which Grab holds as a minority equity stake rather than a consolidated subsidiary.
Bottom line: segment profitability and returns
| Quarter | Mercado Pago Segment EBITDA | Monee Adjusted EBITDA | GoTo Financial Adjusted EBITDA (Rp, local) | GoTo Financial Adjusted EBITDA (USD equiv.) | GoTo Financial Adjusted EBITDA FX-neutral YoY% (USD, constant FX) | GoTo Financial Real Operating Result | GRAB GFS Adjusted EBITDA |
|---|---|---|---|---|---|---|---|
| Q3 2024 | not disclosed | +$187.9M (+13.4% YoY) | -Rp65B (roughly breakeven) | ~-$4.3M | n/a (base too small/negative for a meaningful %) | n/a | -$26M |
| Q4 2024 | not disclosed | +$211.0M (+42.1% YoY) | +Rp14B (first positive quarter) | ~+$0.9M | n/a (base too small/negative for a meaningful %) | -Rp197B | -$27M |
| Q1 2025 | not disclosed | +$241.4M (+62.4% YoY) | +Rp47B | ~+$2.8M | n/a (no YoY% disclosed for this quarter) | n/a | -$30M |
| Q2 2025 | not disclosed | +$255.3M (+55.0% YoY) | +Rp88B | ~+$5.4M | n/a (no YoY% disclosed for this quarter) | -Rp168B | -$26M |
| Q3 2025 | not disclosed | +$258.3M (+37.5% YoY) | +Rp136B (4th straight profitable quarter) | ~+$8.2M | n/a (no YoY% disclosed for this quarter) | -Rp131B | -$28M |
| Q4 2025 | not disclosed | +$263.1M (+24.7% YoY); FY Adjusted EBITDA $1,018.1M | +Rp226B (record quarter); FY Rp497B, GOTO Financial's first profitable full year | ~+$13.5M | n/a (no YoY% disclosed for this quarter) | -Rp50B | -$25M (FY2025: -$110M, +5% wider YoY) |
| Q1 2026 | not disclosed | +$275.2M (+14.0% YoY); operating income $265.6M | +Rp364B (+674% YoY, 6th straight profitable quarter) | ~+$21.4M | +674% (matches local ccy; raw USD-equivalent YoY was only +664%) | +Rp130B (first-ever positive real operating result) | -$17M (-45% narrower YoY, best improvement of this window) |
| Q2 2026 | not yet reported | not yet reported | +Rp481B (+447% YoY) — first time overtaking On-Demand Services' segment Adjusted EBITDA | ~+$26.9M | +447% (matches local ccy; raw USD-equivalent YoY was only +398%, the widest FX drag of any comparison in this post) | +Rp272B (second straight positive quarter) | GRAB has not yet reported Q2 2026 as of this post |
Adjusted EBITDA FX-neutral figures are only computable for the two quarters GoTo itself gave a local-currency YoY% for (Q1 and Q2 2026) — the earlier quarters either lack a disclosed prior-year comparison or moved through a near-zero/negative base where a percentage isn't meaningful. Notably, the Q1 2026 EBITDA gap between constant-currency (+674%) and raw USD-equivalent (+664%) is small, while Q2 2026's gap (+447% vs. +398%, roughly 49 percentage points) is the largest FX distortion measured anywhere in this post — a reminder that the size of the FX drag on a USD-converted growth rate isn't constant, it scales with how much the Rupiah moved in that specific stretch.
Mercado Pago's segment EBITDA row reads "not disclosed" for every quarter in this window because MELI genuinely never publishes it — the closest MELI discloses is NIMAL» (net interest margin after losses), a spread metric on the credit book, not a profitability-per-dollar-of-revenue figure. NIMAL has run from the high-teens to high-20s across this window (17.8%–28.2%, with the low point in the most recent quarter, Q1 2026), which is healthy on its own terms, but it answers a different question than "how much did Fintech actually earn," and it isn't comparable to Monee's, GoTo Financial's, or GRAB's own disclosed EBITDA rows above.
GRAB's GFS is the only segment of the four that is loss-making in every single quarter shown. But read as a trend rather than a level, it's also the segment improving fastest in percentage terms: the loss narrowed from $30 million (its widest point, Q1 2025) to $17 million by Q1 2026 — a 45% improvement in a single year — while revenue kept accelerating and the loan book more than doubled. Whether that trend holds through Grab's own guided second-half-2026 breakeven target is the open question this segment carries into its next few quarters, discussed further below.
No company in this comparison discloses a genuine segment return-on-assets» figure. MELI and SE never publish a segment asset base for their fintech arms, so a real ROA can't be computed for Mercado Pago or Monee from disclosed numbers, and Grab doesn't either. GoTo is the one company where a rough approximation is at least possible, because it discloses both a segment Adjusted EBITDA and a loan book figure each quarter — and the same rough approximation can now be built for GRAB's GFS too, since it discloses the same two ingredients. Adjusted EBITDA ÷ period-end loan book is a single-quarter, non-annualized, EBITDA-based ratio, not a real net-income-based ROA, and it's shown below only as a directional signal, not a genuine profitability-on-assets metric comparable to a bank's disclosed ROA.
| Quarter | GoTo Financial Adjusted EBITDA ÷ period-end loan book | GRAB GFS Adjusted EBITDA ÷ period-end loan book |
|---|---|---|
| Q3 2024 | negative | -5.2% |
| Q4 2024 | ~0.27% | -5.0% |
| Q1 2025 | ~0.82% | -5.3% |
| Q2 2025 | ~1.33% | -3.7% |
| Q3 2025 | ~1.9% | -3.1% |
| Q4 2025 | ~2.57% | -2.0% |
| Q1 2026 | ~3.68% | -1.2% |
| Q2 2026 | ~4.37% | not yet reported |
That climb is real and directionally meaningful for both companies, even though they're moving in opposite directions on an absolute basis: GoTo Financial's Adjusted EBITDA is growing faster than its loan book, not just alongside it, while GRAB's GFS loss is shrinking faster than its loan book is growing — both readable as "this segment's profitability relative to its lending assets is improving quarter over quarter," just from very different starting points. Neither should be read as a stated, audited ROA figure. (GRAB's ratio uses the net loan portfolio for Q3 2024–Q2 2025, the only figure disclosed at the time, and the gross Loan Portfolio from Q3 2025 onward, once Grab began disclosing it — a methodology break worth noting, though it doesn't change the direction of the trend.)
Note on FX and this ratio: because both the numerator (Adjusted EBITDA) and denominator (loan book) for GoTo Financial are converted to USD using the same quarter's own exchange rate, the currency cancels out of the ratio automatically — this is a same-quarter snapshot, not a year-over-year growth comparison, so it carries none of the FX distortion the growth-rate tables above do, and no constant-currency adjustment is needed or applicable here. The same logic applies to the loan-book USD levels shown earlier in the "Credit book size and deposits" table: a point-in-time size comparison (GoTo Financial's ~$614 million loan book against GRAB's ~$1,438 million) is a spot conversion, not a growth rate, so the constant-currency method used elsewhere in this post doesn't change or apply to it — the comparison would only be FX-distorted if it were expressed as a growth rate across quarters using different rates, which it isn't.
GoTo Financial's Growth, Isolated From the Rupiah
Take Q2 2026 revenue as the worked example, applying the same FX-neutral method used for Adjusted EBITDA above. GoTo Financial reported Rp2,070 billion in Q2 2026 revenue, up from Rp1,356 billion in Q2 2025 — a Rupiah (local-currency) growth rate of +53%. The USD-equivalent figures shown in the table above are ~$115.6M for Q2 2026 and ~$84M for Q2 2025, computed using each quarter's own Bank Indonesia reference rate at the time. Dividing those two dollar figures directly gives a raw USD-equivalent growth rate of only +37.6% — roughly 15 percentage points below the Rupiah growth rate, because that comparison is quietly absorbing the Rupiah's depreciation between the two quarters on top of the real revenue growth.
To isolate the business growth from the currency move, hold the exchange rate constant: convert both quarters' Rupiah figures using Q2 2026's own rate (~17,908 IDR/USD) instead of each quarter's own historical rate.
- Q2 2026 revenue at Q2 2026's rate: Rp2,070B ÷ 17,908 ≈ $115.6M (same as the disclosed figure, since it's already Q2 2026's own rate)
- Q2 2025 revenue restated at Q2 2026's rate: Rp1,356B ÷ 17,908 ≈ $75.7M (lower than the ~$84M actually disclosed for Q2 2025, because Q2 2025's own rate was stronger for the Rupiah than Q2 2026's)
- FX-neutral YoY% = $115.6M ÷ $75.7M − 1 ≈ +53%
That FX-neutral figure matches the Rupiah growth rate exactly, not approximately — and it has to, algebraically: once the same exchange rate appears in both the numerator and the denominator of the growth ratio, it cancels out, leaving only the underlying Rupiah-denominated growth. This is the real three-way distinction worth keeping straight: (a) the Rupiah (local-currency) YoY% is already FX-neutral by construction — currency never enters a same-currency comparison; (b) the raw USD-equivalent YoY% (each quarter converted at its own historical rate) is the one figure genuinely distorted by FX, since it's built from two different exchange rates; and (c) the FX-neutral-recomputed USD YoY% converges on (a), not on (b) — proving the Rupiah growth story is real business performance, not a currency artifact. The same method is applied silently to every other GoTo Financial YoY figure in the tables above (results in the "FX-neutral YoY%" columns), without re-showing the arithmetic each time.
Mercado Pago: The Biggest Number, The Least Transparent Bottom Line
Mercado Pago is, by a wide margin, the largest fintech operation of the four — its Q1 2026 revenue of $3,977 million is roughly 3.2x Monee's, over 35x GoTo Financial's USD-converted figure, and more than 37x GRAB's GFS for the same period. Its credit portfolio has also grown the fastest in absolute dollar terms, from $6,016 million gross in Q3 2024 to $14,555 million gross by Q1 2026, more than doubling in six quarters. TPV growth has been consistently strong throughout, running above 70% FX-neutral in every quarter shown except the two most recent, where MELI stopped breaking the figure out with the same granularity.
The gap is what MELI doesn't say. There's no Fintech segment EBITDA, no operating margin, no segment-level income statement — only revenue, TPV, MAUs, and the credit-book size, plus NIMAL as the sole profitability proxy. NIMAL held in a healthy 22–28% band through this window, but a spread metric doesn't tell a reader whether Mercado Pago as a segment is actually profitable once its opex, marketing, and provisioning are allocated against it, the way Monee's, GoTo Financial's, and even loss-making GRAB's own disclosed EBITDA figures do for their businesses. Given that Mercado Pago's credit book is now the largest of the four, that silence carries more weight than it would for a smaller operation.
Monee: The Quietest Compounder
Monee is the one segment in this comparison that has grown Adjusted EBITDA every single quarter across the full two-year window, with no reversals — from $187.9 million in Q3 2024 to $275.2 million in Q1 2026, alongside revenue growing from $615.7 million to $1,241.8 million over the same stretch. Its loan book, at $9.9 billion by Q1 2026, is also the largest of the four in absolute terms after Mercado Pago's, and the only one of the four where the company discloses a genuine operating-income line ($265.6 million in Q1 2026), not just an adjusted figure — a real accounting profit, not only an EBITDA add-back.
The one place Monee discloses less than its peers is volume: SE never publishes a GTV or TPV figure for Monee, only loan principal outstanding, active credit-user counts, and revenue/EBITDA. That's a genuine disclosure gap relative to Mercado Pago's TPV, GoTo Financial's Core GTV, or GRAB's disclosed loan-disbursal figures — but it's a narrower gap than MELI's missing profitability line, since a reader can still judge Monee's actual earnings quality quarter to quarter, just not its raw transaction volume.
GoTo Financial: The Sharpest Turnaround, Off the Smallest Profitable Base
GoTo Financial's story is the most dramatic of the three profitable segments, purely on the shape of the curve. It went from a roughly break-even -Rp65 billion in Q3 2024 to +Rp481 billion of Adjusted EBITDA in Q2 2026 (its own Rupiah-denominated growth, not a currency-inflated one — see the currency-comparability note above and the worked FX-neutral example below) — and in that same quarter, its Adjusted EBITDA overtook GOTO's own On-Demand Services segment for the first time, meaning the fintech arm is now the more profitable half of the parent company on this metric. Core GTV nearly doubled year-on-year in Q2 2026 (+91% in Rupiah terms), and the loan book crossed $614 million, up from $284 million two years earlier. Recomputing that same Q2 2026 Adjusted EBITDA growth at a constant exchange rate (holding FX fixed instead of letting each quarter float on its own historical rate) lands at +447% — matching the Rupiah figure exactly, not the +398% a raw USD-to-USD comparison implies — which is the clearest evidence in this whole comparison that GoTo Financial's turnaround is a real operating story, not partly a byproduct of the Rupiah's depreciation working the other direction on the way in.
But scale still matters for how much confidence that trajectory deserves. GoTo Financial's Q2 2026 revenue of roughly $106 million is a fraction of Mercado Pago's or Monee's, its loan book (~$614 million) is smaller than GRAB's GFS loan book despite GoTo Financial already being profitable and GFS not, and it operates in a single country rather than across a region. The approximate EBITDA-to-loan-book ratio climbing from negative to ~4.4% over two years is the single most eye-catching number in this entire comparison — but it's calculated from a much smaller, more concentrated credit book than either MELI's or SE's, and GoTo, like the other three, discloses no real segment ROA to validate that the underlying credit quality is holding up as the book scales this fast.
GRAB's GFS: Still Losing Money, But Closing the Gap Fastest
Grab's Financial Services segment is the odd one out in this comparison — the only one of the four still segment-loss-making in every quarter shown, and the youngest of the four as a genuine profit-and-loss story. But it's also the segment moving fastest toward the other three's territory. Its Q1 2026 revenue grew 43% year-on-year to $107 million, the fastest revenue growth rate of any of the four companies that quarter, while its Adjusted EBITDA loss narrowed 45% year-on-year to just $17 million — a smaller absolute loss than GoTo Financial's own worst real (unadjusted) operating quarter in this window (Q4 2024's -Rp197 billion, roughly -$12 million at that quarter's rate) despite GFS running a Gross Loan Portfolio more than twice the size of GoTo Financial's.
That loan book is where GFS actually leads the comparison: its $1,438 million Gross Loan Portfolio at Q1 2026 is more than double GoTo Financial's own Q1 2026 loan book of ~$583 million, and roughly a seventh the size of Monee's $9.9 billion, having grown from $498 million just two years earlier — the fastest percentage growth of any lending book in this comparison. That scale is split across three different products (GXS Bank Singapore, GX Bank Malaysia, and a growing minority stake in Indonesia's Superbank, now above 6 million customers), which makes it harder for a reader to judge underwriting consistency across the segment as a whole, since Grab doesn't disclose a combined delinquency or provisioning figure across all three. Management has guided to segment breakeven "no later than the second half of 2026" since first disclosing that target in Q1 2023 — the loss trend through Q1 2026 is consistent with that target being reachable, but it's also a timeline management has now had over three years to hit, worth watching rather than taking on faith (see The Prescription above and Beyond the Usual below).
Beyond the Usual
Mercado Pago is the only one of the four with no disclosed segment profitability line at all
Mercado Libre discloses Fintech-segment revenue, TPV, MAUs, and credit-portfolio size every quarter, but never a segment EBITDA, operating margin, or income statement — the only profitability proxy offered is NIMAL, a credit-spread metric, not a full segment P&L. Monee, GoTo Financial, and even loss-making Grab's GFS all manage to disclose a genuine segment EBITDA figure each quarter despite being smaller, and in GoTo Financial's and Grab's cases, newer to (or still short of) profitability. Given that Mercado Pago's credit book is now the largest of the four, this is worth flagging as an ongoing disclosure gap rather than a one-off omission — a reader following this segment quarter to quarter has no way to verify whether its rapid TPV and credit growth are translating into proportionate earnings.
Three of the four fintech arms are now profitable; Grab's GFS is the one still not there
Monee has been Adjusted-EBITDA-positive every quarter this blog has tracked; GoTo Financial crossed into positive Adjusted EBITDA in Q4 2024 and into real (unadjusted) operating profit in Q1 2026; Mercado Pago's segment profitability can't be verified either way given its disclosure gap above. Grab's GFS is the one segment in this comparison still posting a loss every quarter, though the loss has narrowed from $30 million to $17 million over the trailing year even as revenue and the loan book both accelerated. This is a genuine, structural difference between the four businesses, not a rounding artifact — GFS is younger as a multi-market digital-banking operation (GXS Bank and GX Bank both launched within this blog's tracked window) than GoTo Financial's single-market GoPay/Bank Jago lending model, which may partly explain the gap, but it's worth tracking whether GFS actually reaches its own guided second-half-2026 breakeven target.
None of the four companies discloses a real segment return on assets
No blockquote needed here beyond what's already covered above in the scorecard section — worth restating plainly, though, since it's easy to assume a "credit business" automatically comes with a disclosed ROA the way a bank's does. It doesn't, for any of these four, and the approximations built for GoTo Financial and GRAB's GFS in this post (Adjusted EBITDA ÷ period-end loan book) are this site's own construction, not a number any of the four companies publishes.
Grab's improving credit-quality claim for GFS isn't broken out in dollar terms
Grab's own management has attributed GFS's narrowing loss partly to improving expected-credit-loss[»](/glossary/#ecl-expected-credit-loss) provisioning as a percentage of the loan book, even as that loan book more than doubled year-on-year through Q1 2026 — but the underlying dollar figures behind that percentage claim aren't disclosed, so a reader can't independently verify whether credit quality is actually holding up or whether the provisioning percentage is simply lagging a fast-growing denominator. This is the same category of disclosure gap flagged for Mercado Pago's NIMAL above, just smaller in scale given GFS's size relative to Mercado Pago's credit book.
Mercado Libre's Q3 2024 through Q1 2026 earnings releases, shareholder letters, and 10-Q/10-K filings (Fintech segment disclosures); Sea Limited's Q3 2024 through Q1 2026 earnings releases and 20-F/6-K filings (SeaMoney/Monee segment disclosures); GoTo Gojek Tokopedia's Q3 2024 through Q2 2026 earnings press releases, results presentations, and interim/annual consolidated financial statements (GoTo Financial segment note), with IDR figures converted to USD at Bank Indonesia's quarter-end reference rate; Grab Holdings' Q3 2024 through Q1 2026 earnings releases, results presentations, and SEC 6-K/20-F filings (Financial Services segment disclosures). Internal comparisons also draw on this site's already-published quarterly posts for each of the four tickers.