MercadoLibre's revenue roared 49% in Q1 — so why did operating profit just fall 20%?

---

There is a particular kind of magic trick that lending companies can perform, and MercadoLibre's first quarter of 2026 is a textbook performance of it. You originate a loan. The interest and fees on that loan land in revenue immediately, this quarter, in full. The losses on that loan — the borrowers who will stop paying — arrive later, gradually, over the life of the book, and are estimated today through a provision that management itself sets. When your loan book is growing 87% a year, the revenue is recognized on a book that is enormous and the losses are provisioned against a book that, only a few quarters ago, was barely half the size. Growth, in lending, mechanically flatters the present and defers the reckoning. This is not an accusation of fraud. It is a description of accrual accounting working exactly as designed. The only question that matters is whether MELI's loss estimates are conservative or optimistic, and the first quarter gave us several uncomfortable hints.

Let us be precise about what MercadoLibre reported on May 7, 2026, because the headline and the body of the income statement tell opposite stories. Net revenues and financial income reached $8,845 million, up 49% as reported and 46% on an FX-neutral basis — genuinely the strongest growth rate the company has posted since the second quarter of 2022. Gross merchandise volume hit $19.0 billion, up 42%; items sold rose 47% to 721.7 million; total payment volume climbed 50% to $87.2 billion. By every top-line measure, this is a business firing on all cylinders. And yet income from operations fell to $611 million, a decline of 20% year over year, with the operating margin compressing roughly 600 basis points to 6.9% from 12.9% a year earlier. Net income slipped to $417 million from $494 million, a 4.7% net margin. Diluted EPS of $8.23 missed consensus by more than twelve percent. A company growing revenue at its fastest pace in four years simultaneously produced its worst margin compression in years. Those two facts cannot both be the whole truth, and reconciling them is the entire thesis.

The denominator is doing the lifting

Start with the metric the bull case leans on hardest: asset quality. Management told investors the 15-to-90-day non-performing loan ratio was roughly stable, in the area of 8% on the consumer book, and the company reported provision coverage of 149% of loans more than fifteen days past due. Stable delinquency on a fast-growing book sounds reassuring. It should not be. A delinquency ratio is a fraction: bad loans in the numerator, total loans in the denominator. When the denominator grows 87% in a year — when $14.6 billion of receivables sits where roughly $7.8 billion sat twelve months ago — a flat ratio does not mean losses are flat. It means losses grew nearly as fast as the book. The newest loans, which dominate a book that nearly doubled, have barely had time to go bad. In consumer lending this is called seasoning, and a loan that was originated three months ago cannot yet be ninety days delinquent. A rapidly expanding book is structurally young, and a young book always looks healthier than it is. The stable NPL ratio is not evidence that underwriting is holding; it is exactly what you would expect to see in the months before a fast-growing book begins to age into its losses.

What the company itself admitted about the margin

The most damning evidence in the quarter did not come from a short seller. It came from MercadoLibre's own explanation of why margins fell. Management attributed the compression to two forces. First, a richer mix of the lower-spread credit card, which more than doubled year over year to roughly $6.6 billion and now accounts for somewhere in the mid-40s percent of the credit portfolio, up from the low-40s a year ago. Second — and this is the line that should give every bull pause — "higher provisions in our Brazil consumer portfolio due to longer-duration loans and a broader reach for personal loans." Read that again. The company is reaching further down the credit spectrum for personal loans in Brazil, lengthening the duration of those loans, and taking higher provisions as a result. Net interest margin after losses (NIMAL), the metric that captures what the credit business actually earns after expected defaults, fell to 17.8% from 22.7% a year earlier — a decline of nearly five full percentage points. The spread the lending machine throws off, net of the losses management is willing to acknowledge, is shrinking even as the book explodes. That is not a footnote. That is the business model changing shape in real time.

Bought growth, not organic strength

Here is the frame the 49% headline is engineered to obscure. Strip the credit book out and ask what is actually driving incremental revenue. Total payment volume grew 50%, but assets under management in Mercado Pago's investment products grew 77% to $19.9 billion, and the credit portfolio grew 87% — both materially faster than payment volume, marketplace GMV, or active users. The fastest-growing parts of MercadoLibre are not the marketplace at all; they are the financial-services balance sheet. Unique active buyers, the cleanest measure of organic commerce demand, grew 26% to 84.1 million — a strong number, but a fraction of the 87% credit growth. The revenue acceleration to its fastest pace since 2022 is being manufactured, in significant part, by lending more money to more people, faster. That is a legitimate strategy. But it is bought growth, and bought growth has a price that does not appear until the loans mature. When a company's revenue line accelerates because its loan book is accelerating, the revenue growth and the credit risk are the same phenomenon viewed from two ends. You cannot celebrate one and ignore the other.

The free-shipping subsidy and the disappearing operating leverage

Credit is only half the margin story. MercadoLibre also leaned hard into free shipping, marketing, and credit-card issuance costs during the quarter, with management explicitly framing the spending as long-term value creation over near-term margin. This is the language every growth company uses, and sometimes it is true. But notice what it implies: at 49% revenue growth — the fastest in four years — MercadoLibre could not generate operating leverage. It generated operating de-leverage. A business with a genuine moat and pricing power should see margins expand when revenue grows this fast, because fixed costs spread across a larger base. Instead, the costs grew faster than the revenue they were meant to win. Free shipping is a subsidy to the consumer; it is the cost of buying GMV that might otherwise go to Amazon's Brazilian operation, to Shopee's Sea-backed push, or to Shein and Temu's cross-border flood. When the marketplace must subsidize shipping to defend share, and the fintech arm must subsidize spreads to grow the book, you are looking at a company spending heavily on two fronts at once to keep the growth narrative alive — and reporting a 6.9% operating margin as the receipt.

Cyclical risk priced as a secular certainty

The defining bet embedded in MercadoLibre's valuation — trading in the low-to-mid 40s on trailing earnings and roughly the low-40s on a forward basis as of mid-June 2026, with a market capitalization in the low-$80-billion range — is that the company is a secular compounder whose every dollar of reinvestment will compound. But a credit business is, by its nature, cyclical. It lends into Brazilian and Argentine consumers whose ability to repay depends on local interest rates, inflation, employment, and currency. Argentina's financial system is already showing rising delinquency across the sector; MercadoLibre says its own Argentine 15-90 day NPL actually fell sequentially while spreads widened, which the bulls will cite as proof of superior underwriting. Maybe. Or maybe it is proof that a book growing this fast in a country emerging from hyperinflation simply has not seasoned yet. Brazil, where the company is taking "higher provisions" and "broader reach," is the larger and more telling tell. A secular compounder does not need to broaden its credit reach to keep growing. A company chasing a growth number does. The market is paying a secular multiple for what is, at the margin, an increasingly cyclical consumer-credit lender wearing an e-commerce costume.

The FX trap underneath everything

Every figure above is reported in U.S. dollars, but almost none of MercadoLibre's economics happen in dollars. Revenue, loans, and losses are denominated in Brazilian reais, Argentine pesos, and Mexican pesos, then translated. The 49% reported growth was 46% FX-neutral — a gap that, this quarter, happened to be modest. It is not always modest, and it can cut the other way violently. Argentina remains a serial source of currency devaluation and inflation accounting; a sharp peso move can vaporize dollar-reported revenue and, worse, can impair the real value of a peso-denominated loan book while the dollar-reported provisions lag. The company carries genuine exposure to Brazilian and Argentine policy rates: when the central bank in Brasília or Buenos Aires moves, MercadoLibre's funding costs and its borrowers' repayment capacity both move, often in opposite directions for the company. A lender this leveraged to two of the most macro-volatile economies in the hemisphere, priced for the serenity of a secular compounder, is carrying a tail risk the multiple does not contemplate.

Quality of earnings: where did the cash go

The cleanest way to see the bought-growth problem is to ask where the money goes. A marketplace that compounds on capital-light network effects should throw off cash as it grows. A consumer lender that grows its book 87% must fund that book — every incremental loan is cash out the door today against revenue recognized over time. This is why explosive credit growth and reported profit can move in opposite directions: the income statement books the spread, but the balance sheet swallows the principal. MELI's net income of $417 million against a credit book that grew by roughly $6.8 billion year over year tells you the lending machine is a voracious consumer of capital, not yet a generator of it. The earnings are real in an accounting sense, but their quality is hostage to provisioning assumptions on a book that has not been tested through a full Brazilian or Argentine credit cycle. Investors paying 40-plus times earnings are paying for the spread today and trusting management's loss estimates for the bill tomorrow.

The moat question nobody is asking

MercadoLibre's marketplace moat — logistics density, brand trust, the Mercado Envios shipping network — is genuine and is the part of the company most worth owning. But the growth at the margin is coming from fintech, and fintech moats in Latin America are thinning fast. Nubank now serves over a hundred million customers across the region and is a direct competitor in credit cards and lending. Brazil's instant-payment rail, Pix, commoditized the payments business that Mercado Pago once dominated, turning a profit center into table stakes. The "PayPal of LatAm" framing flatters a business whose payment economics are under structural pressure and whose lending economics are, by the company's own admission, compressing. When the moated part of the business (commerce) grows 26-42% and the un-moated, capital-hungry, cyclically exposed part (credit) grows 87% and drives the revenue headline, the quality of the franchise is quietly deteriorating even as the top line accelerates. The market is extrapolating the moat onto the part of the company that does not have one.

The mix shift toward the worst kind of growth

There is a quieter story buried in the portfolio composition that deserves its own examination, because it reveals the direction of travel rather than just the snapshot. The credit-card sub-portfolio more than doubled, growing 104% year over year to roughly $6.6 billion, with 2.7 million new cards issued in the quarter alone, and it now represents the largest single slice of the book. Management is candid that this is the lower-spread product — credit cards earn less per dollar lent than the merchant and consumer loans that built the franchise. So the fastest-growing piece of the fastest-growing engine is also the least profitable piece, which is precisely why NIMAL fell almost five points. The company is, in effect, sprinting toward thinner economics. A card book is also a different animal from merchant financing: merchant loans are secured against a seller's verifiable cash flows on the very platform that disburses them, while credit cards are unsecured consumer debt extended to millions of newly banked Latin Americans with thin or nonexistent credit histories. Doubling the unsecured, lower-spread, harder-to-collect portion of the book in a single year, into economies with rising sector-wide delinquency, is the highest-risk, lowest-reward growth a lender can choose — and it is exactly the growth MercadoLibre chose. The mix is shifting toward the segment most likely to disappoint when the cycle turns.

Priced for perfection, exposed to the ordinary

Strip away the narrative and look at the asymmetry a buyer accepts at today's price. At roughly 43 times trailing earnings and a market capitalization in the low-$80-billion range, MercadoLibre is priced for the continuation of near-flawless execution: marketplace share gains, fintech monetization, benign credit, and stable currencies, all at once, for years. That is a great deal of perfection to underwrite in a hemisphere whose defining economic feature is its lack of perfection. The danger of a priced-for-perfection asset is not that the bear case is certain — it is that the bull case is already fully paid for, so the upside from things merely going well is thin while the downside from anything going ordinary is severe. A single quarter of credit normalization — provisions catching up to a seasoned book, a peso or real devaluation impairing dollar revenue, a Brazilian rate cycle squeezing borrowers — would not require a catastrophe to compress the multiple. It would only require the ordinary volatility that Latin America delivers as a matter of routine. The first quarter of 2026 was, by the company's own framing, a good quarter, and operating profit still fell 20%. That is the asymmetry: when the good quarters already hurt the bottom line, the bad ones do not bear contemplating at this valuation.

What the bulls genuinely get right

It would be dishonest to short this story without conceding how strong the bull case truly is, because in MercadoLibre's case it is strong on the merits. This is not a fad or a fraud; it is one of the genuinely great franchises in emerging-market technology. The marketplace numbers are extraordinary by any standard: 84.1 million unique active buyers, up 26%; 721.7 million items sold, up 47%; same- or next-day delivery on 199 million shipments. That logistics network is a real, compounding, capital-intensive moat that took fifteen years and billions of dollars to build, and it is nearly impossible to replicate. The fintech franchise, whatever the credit risk, is attached to 82.9 million Mercado Pago monthly active users and $19.9 billion in assets under management growing 77% — a deposit-and-investment flywheel that most digital banks would envy. The margin compression is, plausibly, a deliberate and defensible choice: management is electing to invest free shipping and credit growth ahead of profit precisely because the secular runway in under-penetrated Latin-American e-commerce and finance is enormous, and capturing share now is worth more than optimizing margin today. The 15-90 day NPL really did hold roughly flat, and the company really does have a fifteen-year track record of underwriting through Argentine and Brazilian volatility that few Western lenders could survive. Provision coverage at 149% of past-due loans is genuinely conservative on its face. If the credit book seasons gracefully and the marketplace continues to compound, the spending that crushed this quarter's margin will look, in hindsight, like exactly the right call — and the bears will look like people who could not see past one noisy quarter. None of the forensic concerns above are proof of impairment. They are reasons to demand that the multiple price the risk, which today it does not.

The kicker

So weigh the two stories one final time. The bull holds a fast-growing, moated, fifteen-year compounder electing to invest ahead of profit in a vast under-penetrated market — a defensible, even admirable, posture. The bear holds a consumer-credit lender whose revenue acceleration is being manufactured by an 87% surge in loans to Brazilian and Argentine borrowers, whose own management admits to broader reach and higher provisions, whose net lending spread after losses fell five points, and whose operating profit fell 20% in the very quarter the headline growth looked best — all priced at forty-plus times earnings as if the credit risk did not exist. Both stories are built from the same true numbers. The difference between them is entirely a question of whether MercadoLibre's loss estimates, set today against a book that has nearly doubled in twelve months, prove conservative or optimistic when those loans finally age. That answer does not exist yet; it lives in quarters not yet reported. And that is the whole trade.

The revenue is recognized today, the losses are provisioned today, and the truth about the difference between them will not arrive until the loan book that grew 87% in a single year finally grows old enough to tell us which one management got wrong.