MELI Revenue Nearly Doubled in 2020 - So Why Did the Company Barely Break Even?

A Year of Record Growth That Ends at Break-Even

MercadoLibre's full-year 2020 results are a study in how a genuinely strong operating year can still produce an unremarkable bottom line. Net revenue grew 73.1% to $3,973.5 million (Commerce $2,559.8 million, +90.1%; Fintech $1,413.7 million, +48.8%), Gross Merchandise Volume^» reached $20,926.8 million (+49.5% nominal), and Total Payment Volume^» hit $49,756.8 million (+75.3%) - the fastest annual growth this coverage has measured for MercadoLibre by a wide margin, driven by the same pandemic e-commerce and Fintech tailwind flagged in the Q3 2020 post. Unique active users, a metric disclosed for the first time this filing, reached 132.5 million, up 78.6% from 74.2 million a year earlier.

And yet full-year net income was a loss of just $0.7 million - effectively break-even - on a pretax profit of $81.3 million that a 100.9% effective tax rate consumed almost entirely. That headline number, on its own, badly understates how much actually went right operationally this year, and just as badly obscures how much went wrong in a single quarter. MercadoLibre posted three consecutive profitable quarters (Q1 through Q3 2020) before a fourth-quarter relapse: Q4 2020 alone produced an operating loss of roughly $25.1 million and a net loss of $51.0 million, reversing Q3's operating profit even though Q4 is normally MercadoLibre's strongest seasonal quarter, driven by Christmas-season shopping in every country the Company operates in. Gross margin fell to 36.8% in Q4, from 45.7% a year earlier, as the Company leaned harder into first-party (1P) inventory sales at lower product margins and invested in excess logistics capacity specifically to protect delivery service levels during the holiday peak - a deliberate, disclosed choice, not an unexplained miss.

The Prescription

MercadoLibre should keep making the 1P and logistics-capacity investments that dented Q4 margin, because the volume numbers say they worked: items sold grew 110% YoY even in the seasonally hardest quarter to execute well in, and managed-network shipping penetration reached almost 80% by year-end, up from 43% just twelve months earlier. A company willing to trade a quarter of thinner gross margin for durable market-share and service-level gains during the exact season competitors are also fighting hardest for is making a real strategic bet, not stumbling into a bad quarter - the Company should say so more explicitly in its own disclosures next time, rather than letting the margin decline read as unexplained deterioration.

What it should stop doing: letting a full-year effective tax rate swing from -60.4% to +100.9% in consecutive years without giving investors a clearer normalized-tax-rate baseline to model against. Four factors drove this - Argentina's suspended tax holiday, Mexico/Colombia valuation allowances, a non-deductible FX loss on share buybacks, and higher dividend withholding tax - and all four recurred from the Q3 2020 post's discussion of the same issue. A tax rate this volatile, on a company this large, deserves a forward-looking framework from management (even a rough range) rather than requiring a reader to reverse-engineer one filing's worth of one-off adjustments each quarter.

Commerce and Fintech: A Full Year of Diverging Segment Stories

Commerce revenue grew 90.1% to $2,559.8 million for the full year, driven by the broad-based Latin American e-commerce surge already detailed in the Q3 post - Brazil, Argentina, and Mexico all posted local-currency GMV growth well above 100% in the fourth quarter alone. Fintech revenue grew a slower 48.8% to $1,413.7 million, continuing the pattern flagged last quarter where Commerce is now carrying more of the growth than Fintech, even as Fintech's own underlying metrics - off-platform TPV reaching $9.2 billion in Q4 alone (+150% FX-neutral), a Mercado Credito loan portfolio more than doubling to $479 million - show a business still expanding its addressable market independently of the marketplace.

By geography, Brazil ($2,194.0 million, 55.2% of revenue) and Argentina ($980.3 million, 24.7%) again carried the business, both posting solidly positive direct contribution margins (Brazil's direct contribution was $428.1 million; Argentina's $271.6 million). Mexico posted a negative direct contribution of $(10.8) million for the full year on revenue of $575.2 million - the one segment where MercadoLibre is still losing money at the direct-contribution level even as its top line grows fast (Mexico's own items-sold contribution "surpassed Argentina" during Q4, per the Company's investor presentation, positioning it as the second-largest geography by volume). This is worth watching: a segment that's winning on volume but still direct-contribution-negative is either early in a genuine land-grab investment phase or has a cost structure that hasn't yet found product-market fit on pricing - the filing doesn't say which, and a reader should look for this to turn positive within the next few quarters as evidence it's the former.

Beyond the Usual

A material weakness that was actually fixed this time

The material weakness in internal control over financial reporting first disclosed in the Q2 2020 10-Q, and still unremediated as of the Q3 2020 10-Q (see the prior post), was remediated during the second half of 2020: management implemented new control owners, more frequent control execution, new IT tooling for accounts-receivable-from-means-of-payment reconciliation, and additional outsourced and internal-audit staffing. As of December 31, 2020, both management and the Company's independent auditor concluded disclosure controls and internal control over financial reporting were effective. This is a plain factual update to a real prior finding, not a new one - but a reader who read the Q3 2020 post is owed the resolution, and it's a genuinely favorable one.

A full-year effective tax rate of 100.9% - the tax bill essentially equaled the entire pretax profit

FY2020's 100.9% effective tax rate (versus -60.4% in FY2019 and 44.1% in FY2018) means income tax expense of $82.0 million came within $0.7 million of consuming the Company's entire $81.3 million pretax profit. The drivers - Argentina's suspended knowledge-based-economy tax holiday, Mexico/Colombia deferred-tax valuation allowances, non-deductible FX losses on share repurchases, and higher dividend withholding tax - are the same ones flagged for Q3 alone; seeing them compound across all four quarters means a reader should treat FY2020's reported net income as an especially poor proxy for underlying earnings power, even more so than any single quarter in isolation.

Two new Brazilian tax assessments, filed against MercadoPago.com Representações Ltda. and Ebazar.com.br Ltda. in October and November 2020, claim a combined $27.7 million ($15.2 million and $12.5 million respectively) in additional income tax on the grounds that intercompany technology-service payments made to MercadoLibre's own Argentine and Uruguayan affiliates weren't adequately substantiated as deductible expenses. The Company filed defenses in December 2020 arguing the contracts and project-level documentation support the deductions, and management's counsel views the position as "more likely than not" to succeed - a genuinely new footnote-level item this filing (not a continuation of the long-running São Paulo tax and IR-withholding disputes, which remain open and unreserved for a seventh consecutive year with no material change), worth tracking into 2021 given the dollar amount is larger than several of MercadoLibre's other disclosed contingencies combined.

MercadoLibre's own Long Term Retention Plan (LTRP) - a cash-settled, stock-price-indexed retention program running ten overlapping annual tranches (2011 through 2020) - carried a total contractual obligation fair value of $307.2 million at year-end, up from $273.2 million at the end of Q3 2020, tracking the stock's own Q4 rally. It's a genuinely unusual compensation structure: employees are paid in cash, not equity, but the payment amount moves with the stock price anyway, meaning the Company effectively carries synthetic equity-compensation exposure on its balance sheet (an accrued liability of $136.8 million within Salaries and social security payable) without ever issuing a share.

The buyer protection program's maximum potential exposure grew again, to $2,535.0 million at year-end from $1,365.8 million a year earlier - an 85.6% increase roughly tracking GMV growth - while the actual provision recorded against it rose to $8.4 million from $3.8 million, still just 0.3% of the disclosed maximum, consistent with the historical-loss-rate justification management gives for the low reserve.

As a subsequent event, MercadoLibre issued its first-ever labeled bond: $400 million of 2.375% Sustainability Notes due 2026 and $700 million of 3.125% Notes due 2031, both closed January 14, 2021, with the Sustainability Notes' proceeds earmarked for environmental or social "Eligible Projects." Part of the combined $1.1 billion raise funded a genuinely expensive housekeeping move: repurchasing $440 million of principal on the 2.00% Convertible Senior Notes due 2028 for $1,865.1 million in total cash (principal, accrued interest, and a conversion premium) - a reflection of just how far the stock has moved above the Notes' conversion price. The same footnote discloses that the 2028 Notes' own contingent-conversion price threshold was actually triggered during Q4 2020, meaning holders gained the right to convert the remaining Notes into common stock between January 1 and March 31, 2021 - the buyback is best read as pre-empting further dilution from that trigger, not as an isolated capital-structure move.

Target Valuation Range

DCF fair-value range: roughly $20.3 billion (bear/reversion case) to $63.2 billion (bull case) enterprise value, against an actual $81.0 billion enterprise value - overvalued even under the bull case. MercadoLibre's stock rallied 54.8% during the same quarter the Company posted a net loss, pushing the price meaningfully ahead of what even a bull-case discounted cash flow can support at today's discount rate.

The stock closed 2020 at $1,675.22, up 472.1% over the trailing two years (from $292.85 at the end of 2018) and up 54.8% in the fourth quarter alone - a rally that happened during the same three months the Company posted an operating and net loss, underscoring how disconnected the stock's move was from this specific quarter's GAAP results (the market was evidently looking through Q4's margin investment to the underlying volume growth). MercadoLibre has never split its common stock, so this is the nominal price actually quoted on NASDAQ, not a split-adjusted figure.

No preferred stock remains at year-end (the last 50,000 shares of Series A Perpetual Preferred converted to common in September and November 2020).

This is a very different starting point from Q3's roughly $51.8 billion EV, reflecting how much of the stock's move outran the business's own growth in a single quarter. FY2020's actual free cash flow of $935.5 million (operating cash flow of $1,182.6 million less $247.0 million of capex - up 197.7% YoY, and this quarter's filing finally gives a clean full-year figure rather than the nine-month proxy used last quarter) is used for all three DCF scenarios below at the same 11% discount rate.

The base case is materially short, versus Q3's ~84% on the same methodology applied to a smaller EV. The bull case still falls short of today's price too, a reversal from Q3 where the bull case cleared EV. The bear/reversion case now implies the stock is worth roughly a quarter of its current price if growth normalizes toward pre-pandemic trend.

The single-stage reverse DCF now implies the market is pricing in a perpetual growth rate of roughly 9.6%, up from Q3's already-demanding 8.6% - a rate this coverage has not seen a Latin American e-commerce/Fintech business sustain over a full economic cycle. None of this changes the read that MercadoLibre had a genuinely strong operating year; it changes the read on whether today's price already assumes that strength continues indefinitely, and by this analysis, it assumes more of it than even an optimistic scenario delivers.

Coverage Table