Mercado Libre: E-commerce Empire - [Business Breakdowns, EP.227]

Summary

• Daniel Wu’s core framing: Mercado Libre is “an amalgamation of Amazon Retail and Alipay in Latin America” — a $21B-revenue company in 2024, up from $1.4B in 2018, compounding the top line at 55% since 2019 with “barely a growth hangover” post-COVID. Matt Reustle notes the ~$120B market cap makes MELI three times the size of the eBay it once resembled. The two segments have proven counter-cyclical: the 2021 Brazil credit-card launch offset the 2022 commerce slowdown that weighed on other pandemic darlings.
• MELI’s LatAm e-commerce share is mid-to-high 20s percent — “higher than the next eleven operators combined” — spanning 80%+ in Argentina and ~40% in Brazil, against regional penetration of only ~15% vs the low-20s US and 30%+ China/UK. Wu sees nothing structurally capping penetration, implying double-digit market growth plus share gains and take-rate expansion could keep commerce revenue growing above 20% “for the foreseeable future.”
• The June cut of Brazil’s free-shipping threshold from 79 to 19 reais was read by the market as defense against Shopee — which announced it was EBITDA-positive in Brazil in Q3 last year — but Wu argues it’s offensive reinvestment, “a page out of Amazon’s playbook.” Both prior threshold cuts drove sales and logistics efficiency that absorbed the cost. Shopee opened its first Brazilian fulfillment center in September 2024; MELI expects to end 2025 with twenty-two: “the network scale and service quality is incomparable.”
• Take rate has expanded from ~17% (comparable basis, 2021) to 21% while the core seller fee barely moved from 12% to ~13% — “substantially all” of it from value-added services, “what you wanna see as a shareholder.” Two big levers remain largely unpulled: logistics is mostly unmonetized until Brazil fulfillment penetration (60%) reaches Mexico’s 70s — possibly next year — and ads sit at just 2% of GMV vs Amazon’s 7%, with “no fundamental reason” they can’t reach 3-5%.
• Credit is arguably the most important growth driver of the whole company: the portfolio grew from $480M end-2020 to $9.3B, with credit cards from zero to $4B — yet cards are structurally the low-margin product despite Brazil’s 15%/month revolving rate that “annualizes to 450% a year.” Nearly 80% of Brazilian card receivables are interest-free installments, so the paying 20% must fund losses, funding costs, and margin for the whole pool; only 51% of the Brazil card book is NIMAL-positive today and NPLs have been around 28% in recent quarters.
• Prepayments — “the original buy now, pay later” — exist because Brazil settles credit cards D+30 and consumers split purchases into up to 12 interest-free parcelados, leaving merchants waiting up to 360 days for cash. MELI estimates a ~3% net spread on anticipating those receivables; the prepayment fee is described as risk-free because the card issuer guarantees the receivables, but the reported spread above 4% is overstated because 40%+ of card receivables now sit on MELI’s balance sheet, funded through wholesale sources and deposits whose interest cost the disclosure omits.
• After 26 years Marcos Galperin steps back to executive chairman, with commerce president Ariel Sharfstein promoted to CEO — a choice over the fintech president that “suggests the long-term opportunity is still centered around the Mercado Libre marketplace.” Wu’s closing lesson: the power of reinvestment and reinvention — MELI was never “objectively cheap,” but “shareholders have certainly done very well by paying a fair price and just letting Galperin and his team execute.”

Deep dive

1. A dot-com survivor that refused to burn capital

• Wu’s one-liner to level-set: MELI is “an amalgamation of Amazon Retail and Alipay in Latin America” — the region’s largest e-commerce platform plus the Mercado Pago wallet and Mercado Crédito credit business, with major operations in Brazil, Mexico, and Argentina and a presence in fifteen other countries. Reustle admits his mental model was “stuck in the past” as the eBay of LatAm.
• The founding is “classic Silicon Valley”: a 1999 Buenos Aires garage, with Marcos Galperin and co-founders Hernan Kazah and Stelio Kulter. Galperin was studying for his Stanford MBA when he developed the idea, and the founders later returned to Argentina to launch MELI. They copied eBay’s auction model alongside ~80 LatAm e-commerce startups — many faster-growing and better-funded.
• Why MELI is the one still discussed: the founders focused “on building organic growth engines for the long term instead of chasing a big IPO by burning through capital on user acquisition with negative economics,” pacing growth to what infrastructure could sustain. That let them survive the dot-com bust and ride two decades of tailwinds. By the 2007 IPO, fixed-price was already the main format; through the 2010s, when the company reported the metric, Wu believes fixed-price sales were 95% or more of GMV.

2. First mover meant building LatAm’s missing infrastructure

• The tailwinds: LatAm internet users exploded from ~18M to 122M+ between 2000 and 2007 — yet penetration was still only ~20% vs 70%+ in the US — while a rising middle class faced limited physical retail options.
• The frictions MELI had to solve itself: cultural skepticism of online transactions, low credit-card penetration, a lack of secure digital payments infrastructure, patchy logistics, and diverse regulations. Hence Mercado Pago in 2003 (playing PayPal’s trust-and-escrow-like role for eBay) and Mercado Envíos in 2013, letting sellers print labels and use MELI’s negotiated carriers.
• On Reustle’s innovation question, Wu is candid: the founders themselves “would admit that they weren’t particularly innovative.” The homegrown exception is MELI Delivery Day, which stages a buyer’s slow-shipping packages to one fixed weekday, raising route density enough to fund free shipping under 79 reais. The verdict: MELI won “by bringing existing innovations into the LatAm market at scale and then just executing better” — including against Amazon itself.

3. Two engines that hedge each other

• Revenue hit nearly $21B in 2024, from $1.4B in 2018, with commerce just under 60% of the mix. The flywheel: marketplace consumption drives payments; credit and wallet adoption drive consumption back on the marketplace.
• Reustle’s pushback on the claimed counter-cyclicality — “It’s not logical to me why that would actually happen” — draws the idiosyncratic answer: the Brazil credit-card launch in 2021 ramped exactly as commerce growth tapered, so fintech carried the top line through 2022 and MELI dodged the post-COVID hangover that hit Amazon.
• Geographic split: Brazil 52% of LTM revenue, Mexico and Argentina 22% each, rest of LatAm just 4%. Argentina runs 65% fintech mix with mid-40s contribution margins; Brazil is less than half that; Mexico sits in the high teens.

4. Shopee is the fight that matters — and MELI is punching first

• The playing field: LatAm e-commerce penetration ~15% vs the low-20s US and 30%+ in the UK and China. MELI’s share is mid-to-high 20s regionally — over 80% in Argentina, ~40% in Brazil, “higher than the next eleven operators combined” — in a market where physical retailers’ long-tail online operations have been the main share donors.
• The Shopee arc: entered Brazil in December 2019 as cross-border, actually exceeded MELI’s monthly active users sometime in 2022, hibernated through the rate-hike cycle when Sea retreated to Southeast Asia, then re-emerged — announcing Shopee Brazil EBITDA-positive in Q3 last year and growing while profitable.
• MELI’s response — merchant fee cuts in May, then the free-shipping threshold slashed from 79 to 19 reais in June — was “perceived as a defensive move” against Shopee’s low-end threat. Wu’s reframe: it’s reinvestment in engagement and frequency, and both prior threshold cuts drove sales and logistics efficiency that offset the cost. “That’s generally pretty hard to argue against.” One Shopee fulfillment center vs MELI’s expected twenty-two by the end of 2025: “incomparable.”
• On comping profitability: Brazil’s contribution margin bounces around the high teens to 20%, mostly e-commerce today — call it mid-teens with corporate costs allocated.

5. Take-rate expansion from services, not merchant squeeze

• The marketplace served 100M+ unique buyers on $51B of 2024 GMV — buyers have more than doubled since pre-COVID yet remain under 40% of the adult population in the three core markets. First-party is just over 6% of GMV in the latest quarter (vs Amazon’s probably 30-50%), used to sharpen pricing in electronics and push into grocery where 3P economics currently don’t work.
• The reported 3P take rate reached 21% in H1 vs a comparable ~17% in 2021, while the core final value fee moved only from ~12% to ~13%. So “substantially all of the take rate expansion has come from value-added services… rather than from squeezing merchants on core fees — and that’s what you wanna see as a shareholder,” mirroring Amazon layering FBA and ads atop stable base fees.
• Logistics runs three tiers — Full (FBA analog), cross-docking, Flex — and stays capital-light because MELI leases rather than owns, expanding “as-needed” rather than ahead of demand like post-COVID Amazon. It’s also largely unmonetized: management is waiting for Brazil fulfillment penetration (now 60%, up from 50%) to reach Mexico’s 70s before “hitting that monetization button,” potentially next year. Meanwhile the Brazil capacity-doubling announced September last year has dented margins, since new warehouses apparently take a few years to reach peak utilization.
• Ads are just 2% of GMV (from 50bps in 2019) vs Amazon’s 7% — a full-stack self-serve console only arrived in early 2023 — but “there’s no fundamental reason why ad penetration can’t get up to three or four or five percent,” a meaningful earnings driver given ad margins. MELI+ relaunched late 2023 at $2/mo ($5 with Disney+), though Wu believes there is still a 19-reais free-shipping minimum. Wu’s estimates for the ~$12B commerce segment are ~$6B 3P fees, ~$2B 1P sales, ~$3B net shipping, and ~$1B ads — summing to the 24% blended take rate.

6. Brazil’s D+30 world makes prepayments the original BNPL

• Fintech TPV was $197B in 2024 excluding free P2P: $55B on-marketplace, $88B off-marketplace acquiring — the main growth driver for payment processing since it opens a TAM far beyond marketplace GMV — and $54B of wallet volume. Merchant discount rates run 1-3% debit, 4-5% credit, 0-1% for Pix, skewing high because Pago serves micro and small merchants; rates are structurally declining as MELI deliberately pushes up the merchant-size spectrum, though higher processing volumes and revenue dollars should more than offset that pressure.
• The Brazilian quirk: card settlement is D+30 vs the international D+1/D+2, and consumers love parcelados — 3 to 12 interest-free installments — so a 12-month installment sale pays the merchant its final slug 360 days after the transaction, “which is pretty crazy if you think about it.” Anticipation of receivables solves it: “a win-win for all the parties” — merchant gets discounted upfront cash, consumer splits interest-free, and the acquirer earns “a fat, risk-free fee” because the card issuer guarantees the receivables in the event of default.
• MELI can either hold the receivables on its balance sheet or sell them to a bank; either route carries funding costs. Wu’s forensic catch: the disclosed net spread widened past 4%, which “on the face of it looks pretty attractive” — but MELI now holds 40%+ of card receivables on balance sheet (vs ~11% in 2019), funded increasingly through wholesale sources and deposits. The reported prepayment revenue includes the cost of discounting receivables to a bank but not the interest cost of wholesale funding, so Wu estimates the true net spread remains ~3%, above the acquiring MDR.

7. Credit: 450% APYs, and cards are still the low-margin product

• Credit grew from 33% to 42% of fintech revenue between 2021 and 2024, driven by a portfolio up from $480M end-2020 to $9.3B, with cards zero to $4B — while gross interest margin after losses fell from an annualized 50% to 28%.
• The episode’s best exchange: Brazil’s revolving rate is 15% a month — Reustle guesses 30-40% annualized; “It actually annualizes to 450% a year.” The resolution of the margin paradox: ~80% of Brazilian card receivables are interest-free installments (vs two-thirds interest-earning in the US), so the paying 20% must cover losses, funding, opex, and margin for the entire pool — blending to a ~40% gross yield vs 80%+ on personal loans. The entire 2023 Brazil card cohort only turned NIMAL-positive this year; 51% of the Brazil book is positive, Mexico is still negative, and Argentina’s credit-card portfolio is expected to be NIMAL-negative for its first few years after launch.
• On blowup risk — which Wu concedes is “fair to be concerned about” in an EM — NPLs have been around 28% in recent quarters: “a quarter of the credit book is essentially past due at any given point in time.” Comfort comes from behavior: when conditions deteriorated in mid-2022, MELI immediately cut originations and tightened underwriting; “none of the leaders of the credit team have a KPI where you’ve got to grow the credit book.” Proprietary underwriting built on marketplace transaction data (FICO or VantageScore do not apply to much of the underbanked), plus the ability to sweep merchant cash, reduces risk versus a standalone Brazilian issuer.

8. Three bites at one GMV dollar; margins as a tale of two halves

• The consolidated picture: 55% revenue CAGR since 2019, 43% since the 2021 COVID peak. The remarkable feature: “three opportunities to earn revenue on the same dollar of marketplace GMV” — the marketplace fee, the prepayment fee on card transactions, and interchange plus net interest if it’s a Pago card.
• Margin history tells the reinvestment story: mid-30s operating margins through 2013, falling to 21% amid hyperinflationary issues in Venezuela, MELI’s highest-margin country, then to 5% when free shipping launched in 2017 and negative in 2018-19, recovering to mid-teens by 2023 before the current reinvestment phase — with pressure expected through next year. Crucially, free shipping launched while unit growth was still healthy; without it, MELI “would have a smaller share of a smaller e-commerce pie today.”
• Below the line: adjusted FCF (stripping fintech distortions) was $1.3B in 2024 and $1.4B in 2023, with approximately 60-80% net-income conversion after additional non-cash adjustments. Capital allocation is almost entirely organic — just over $1B of buybacks and ~$200M of M&A since the 2007 IPO — and management “doesn’t really manage the business through margins,” sacrificing short-term profit to extend the runway.

9. Risks, succession, and the lesson

• Wu’s risk ranking: Shopee specifically, not international competition generally — Amazon never caught on outside Mexico, and Temu’s only weapon is “a very, very, very low price.” Temu saw a spike in Mexican downloads and MAUs that came back quickly once advertising and user acquisition stopped. Pix has digitized 90%+ of personal consumption in Brazil, eating debit share — but might be a long-run card tailwind as the underbanked graduate into credit. Regulatory rumblings (aligning D+30, capping revolving rates) haven’t gained traction; the Argentine banks’ antitrust suit is “pretty hard to argue against… MELI being a monopoly in Argentina,” but “it’s not really clear to us who is being hurt,” and Milei disdains excessive regulation anyway. FX distorts everything — 20% USD growth vs 80% FX-neutral tells you little, so units sold is the cleanest underlying metric.
• On succession: after 26 years Galperin becomes executive chairman, with commerce president Ariel Sharfstein CEO starting next year — and Wu reads the choice of commerce over the more experienced fintech president as a signal “the long-term opportunity is still centered around the Mercado Libre marketplace.” Culture is Galperin’s “professional sports team” — cliché post-Reed Hastings, Wu admits, though Galperin nearly went pro in rugby — with long-tenured executives and departures typically followed by internal promotions.
• The outlook and closing lesson: commerce plausibly compounds above 20%, payments decelerates (Mexico, where cash usage for consumer purchases remains in the 40% range, is the remaining growth market), and margins expand non-linearly. The transferable insight is “the power of reinvestment and reinvention” — and for investors, that MELI was never “objectively cheap,” yet “shareholders have certainly done very well by paying a fair price and just letting Galperin and his team execute against the long-term vision.”