MELI Local-Currency Revenue Grew 81% for the Year - Net Income Still Fell 38%

A Currency Gap Wide Enough to Swallow the Whole Year, and a Recovery That Didn't Fully Show Up in the Numbers

MercadoLibre's Form 10-K for the fiscal year ended December 31, 2014 confirms, at the full-year level, the pattern every quarterly filing this coverage tracked in 2014 already showed: net revenue grew 17.8% in reported U.S. dollars, to $556.5 million from $472.6 million, while the same business grew 80.7% measured in the local currencies where its subsidiaries actually operate. That's a 62.9 percentage-point gap for the full year - narrower than Q3 2014's single-quarter record of 68.8 points, but only because Q1 and Q2's smaller currency gaps diluted the average; Venezuela alone posted 202.5% local-currency revenue growth for the year, driven by an approximately 69% Venezuelan inflation rate the filing directly cites, a product shortage pushing up seller asking prices, and a shifting mix of categories sold.

What's genuinely new this filing is that the "clean recovery" this coverage credited MercadoLibre with after Q3's return to profitability didn't fully hold up once the full year closed. GAAP net income attributable to shareholders fell 38.2% to $72.6 million (diluted EPS $1.63, down from $2.66) - almost entirely a mechanical consequence of the $49.5 million Venezuela impairment booked back in Q2 2014, which never gets restated away in the full-year GAAP figures the way the Company's own non-GAAP adjustments strip it out quarter by quarter. But even backing that one-time charge out entirely, full-year operating margin still fell - to 30.54% from 32.49%, a 1.95 percentage-point decline - because product-and-technology development expenses grew 31.1% and sales-and-marketing expenses grew 23.4%, both meaningfully faster than the 17.8% reported revenue growth that funds them. Derived Q4-standalone operating margin, calculated as the full year less the first nine months already reported, came in at 28.03% - down 10.68 percentage points from Q4 2013's 38.71%, and a sharper single-quarter decline than anything the Venezuela impairment alone can explain, since Q4 2014 carried no fresh Venezuela charge at all. A reader who took Q3's rebound at face value and stopped there would have missed that the underlying cost structure kept growing faster than the top line straight through the fourth quarter.

The one place the story is unambiguously good: free cash flow. The office-building buying spree that collapsed 2013's free cash flow to just $24.8 million - Caracas and Buenos Aires purchases that pushed 2013 capex to $117.6 million - finally ended, and 2014 capex fell back to $34.4 million even as operating cash flow itself grew 38.1% to $196.8 million. The result is $120.7 million of free cash flow, up 386.7% year-over-year, MercadoLibre's best full-year cash generation this coverage has measured.

The Prescription

MercadoLibre should tie its own cost growth to its own reported revenue growth the way it already ties revenue disclosure to local-currency and reported figures side by side. Product-and-technology development growing 31.1% and sales-and-marketing growing 23.4% against 17.8% reported revenue growth is exactly the kind of gap the Company's investor materials would flag immediately if it showed up in a competitor's numbers - yet neither expense line gets a local-currency-adjusted framing the way revenue does, even though a meaningful share of both increases (LTRP compensation tied to the Company's own rallying stock price, cloud/CDN infrastructure costs, new engineers) is genuinely controllable rather than currency-driven. Disclosing an "underlying" or currency-adjusted opex growth rate alongside the existing revenue reconciliation would let investors see whether the operating-margin erosion flagged above is a real efficiency problem or another artifact of translation math.

What it should stop doing: continuing to treat the São Paulo tax claim's "remote" risk classification as though nothing has changed, now that a full year of active judicial litigation - with a real $15.4 million court deposit sitting in escrow - has passed without any court ruling at all. This is the sixth consecutive filing this coverage has flagged the identical disconnect between an escalating judicial posture and a risk label frozen since the claim's original administrative stage. A claim this old, this large, and this procedurally advanced deserves a risk assessment that can actually move with the facts - and right now, it hasn't moved once in over a year of litigation.

Key Financial Metrics

Year ended December 31, 2014 vs. year ended December 31, 2013 - consolidated, reported in USD (MercadoLibre reports natively in dollars; no FX conversion is needed)

Metric	FY2014	FY2013	YoY
Net revenues	$556.5M	$472.6M	✅ +17.8% reported (✅ +80.7% at local currency - a 62.9pp gap for the full year, versus Q3 2014's single-quarter record of 68.8pp)
Cost of net revenues	$(159.0)M	$(130.1)M	⚠️ +22.2%
Gross profit (71.4% margin)	$397.6M	$342.5M (72.5% margin)	✅ +16.1%, margin -1.0pp
Product and technology development	$(53.6)M	$(40.9)M	⚠️ +31.1% - faster than revenue growth
Sales and marketing	$(111.6)M	$(90.5)M	⚠️ +23.4% - faster than revenue growth
General and administrative	$(62.4)M	$(57.6)M	✅ +8.3% - slower than revenue growth
Impairment of long-lived assets	$(49.5)M	$0.0M	🔴 Entirely the Venezuela impairment booked in Q2 2014
Income from operations (21.7% margin)	$120.5M	$153.5M (32.5% margin)	🔴 -21.5%, margin -10.8pp (ex-impairment: $170.0M, 30.5% margin, -1.95pp)
Blended tax rate	40.3%	27.9%	🔴 +12.4pp - Venezuela's non-deductible losses plus Argentina's expired tax holiday
Effective tax rate	54.8%	30.3%	🔴 +24.5pp, same drivers, compounded by lower pre-tax income
Net income attributable to MELI	$72.6M	$117.5M	🔴 -38.2%
Diluted EPS	$1.63	$2.66	🔴 -38.7%
Free cash flow» (Company's own reconciliation)	$120.7M	$24.8M	✅ +386.7% - capex collapsed as the office-building purchases behind 2013's cash crunch finally closed
Capital expenditures (property & equipment, cash basis)	$34.4M	$113.8M	✅ -69.8%
Cash, ST & LT investments (period end)	$577.2M	$262.6M	✅ +119.8%
Loans payable and other financial liabilities (period end)	$283.8M	$15.9M	⚠️ Almost entirely the $330M convertible notes, carried at $283.8M net as the debt discount amortizes

Every year-over-year decline in this table traces to a single event: the $49.5 million non-cash Venezuela impairment and its associated FX re-measurement loss, both booked in Q2 2014 and never reversed for GAAP purposes even after Q3's clean quarter. Strip that one item out and operating income still only reaches $170.0 million - still 1.95 percentage points of margin worse than 2013, since product-and-technology and sales-and-marketing spending both grew faster than revenue for the full year, not just in the quarter the impairment landed.

Key Operational Metrics

Year ended December 31, 2014 vs. year ended December 31, 2013

Metric	FY2014	FY2013	YoY
Gross merchandise volume» (GMV, excludes motor vehicles/vessels/aircraft/real estate)	$7,081.9M	$7,305.3M	🔴 -3.1% - the first full-year GMV decline this coverage has measured
Total payment volume» (TPV)	$3,523.2M	$2,497.7M	✅ +41.1%
Total payment transactions	46.3M	31.5M	✅ +47.0%
Confirmed registered users (cumulative, period end)	120.9M	99.5M	✅ +21.5%
New confirmed registered users (in period)	21.5M	18.0M	✅ +19.4%
Items sold	101.3M	83.0M	✅ +22.1%
Take rate» (net revenues / GMV)	7.86%	6.47%	✅ +1.39pp for the full year (versus Q3 alone's +2.24pp single-quarter jump, diluted by weaker Q1/Q2 take-rate readings)
TPV as % of GMV	49.7%	34.2%	✅ +15.5pp - MercadoPago penetration accelerating across the full year

Full-year GMV falling 3.1% while items sold grew 22.1% - a 25.2-point gap - confirms at the annual level what Q3 alone showed at 32.7 points: a large and growing share of what's transacted on the platform is priced in currencies (chiefly the Bolivar and the Argentine peso) that lost dollar value faster than transaction volume itself grew, so the dollar-denominated GMV figure understates real platform activity more with every passing quarter of 2014.

Five Country Segments - Argentina Was the Only Segment to Improve Full-Year Margin; the Venezuela Impairment Wiped Out What Would Otherwise Have Been a Strong Year There

MercadoLibre reports the same five geographic segments adopted in Q3 2010 - Brazil, Argentina, Mexico, Venezuela, and Other Countries.

Segment	FY2014 Revenue	YoY Revenue	Direct contribution margin, FY2014	Direct contribution margin, FY2013	Margin change
Brazil	$273.6M	✅ +32.6%	42.11%	42.14%	⚠️ -0.03pp (essentially flat)
Argentina	$150.7M	✅ +23.4%	46.06%	45.55%	✅ +0.51pp
Mexico	$37.7M	✅ +14.7%	36.11%	43.50%	🔴 -7.39pp
Venezuela	$58.0M	🔴 -31.4%	-13.88%	65.03%	🔴 -78.91pp (entirely the impairment; ex-impairment margin ≈71.4%, ✅ +6.4pp)
Other Countries	$36.5M	✅ +37.0%	44.81%	53.72%	🔴 -8.91pp
Total	$556.5M	✅ +17.8%	37.12%	47.86%	🔴 -10.74pp (ex-impairment ≈46.00%, ⚠️ -1.86pp)

Argentina was the only segment to post a genuine full-year margin improvement, continuing the rebound from Q1 2014's tax-credit write-off dip even as its own tax holiday expired mid-year. Brazil, the largest segment at $273.6 million and 49.2% of total revenue, held margin essentially flat despite 32.6% revenue growth - a healthy scale story on its own, but one that didn't produce the operating leverage a reader might expect from growth that fast. Mexico and Other Countries were the year's real laggards: Mexico's margin fell 7.39 percentage points as take-rate-driven marketplace revenue growth in that segment came with a currency devaluation drag, and Other Countries posted the steepest non-Venezuela margin decline (-8.91pp) even as its revenue grew fastest of any segment at 37.0%. Venezuela's headline margin (-13.88%) looks catastrophic only because the entire $49.5 million impairment sits inside that one segment's direct contribution; strip it out and Venezuela's underlying margin actually improved to roughly 71.4% from 65.0%, consistent with the segment's pattern in every quarterly filing since the impairment - the segment's operations, as distinct from the one-time asset write-down, had a genuinely strong year.

Beyond the Usual

The São Paulo tax claim remains active litigation with zero reserve - the sixth consecutive filing this coverage has flagged the same disconnect

As of December 31, 2014, the lawsuit before the 9th Treasury Court of the City of São Paulo - contesting taxes and fines assessed against MercadoLibre's Brazilian subsidiary for the 2007-2010 period, backed by an approximately R$41 million ($15.4 million at the December 31, 2014 exchange rate) court deposit - remains exactly where it was a year ago: the lower court's ruling was still pending as of the filing date, more than a year after the lawsuit was originally filed.

This is the sixth consecutive filing this coverage has flagged the same pattern: a lawsuit in active judicial litigation, backed by a real eight-figure court deposit, that management's legal counsel continues to classify as "remote" risk with zero reserve accrued - a classification that has not changed once since the claim first exhausted its administrative appeals in October 2013. A claim old enough and large enough to have survived a full year of active court proceedings without a ruling deserves closer scrutiny than a risk label frozen in time.

A subsequent event discloses Venezuela's currency regime was overhauled again in February 2015, at a rate roughly 3.4 times weaker than the one used throughout 2014

In February 2015 - after this fiscal year closed but before the 10-K was filed - the Venezuelan government unified the SICAD 1 and SICAD 2 exchange systems into a single "SICAD" mechanism, with an initial published rate of 12 BsF per U.S. dollar, and separately created a new "Sistema Marginal de Divisas" (SIMADI) market, whose first published rate on February 12, 2015 was 170.04 BsF per U.S. dollar. MercadoLibre used the SICAD 2 rate of roughly 50 BsF per U.S. dollar to re-measure its entire Venezuelan balance sheet and results throughout 2014, including the very numbers reported in this filing. The Company states it was still assessing the new regulations as of the filing date and has not yet determined what accounting effect, if any, they will have on 2015 results.

A currency re-measurement rate roughly 3.4 times weaker than the one this filing's own Venezuela segment figures are built on was already public knowledge by the time this 10-K was signed - and the Company disclosed it only as an unresolved subsequent event, with no estimated impact given. The same mechanical playbook that produced [Q2 2014's first-ever operating loss](/analysis/meli/2014-06/#a-venezuela-accounting-entry-not-a-business-problem) - a forced shift to a weaker official rate, followed by a re-measurement loss and a long-lived-asset impairment - is the obvious risk sitting in Q1 2015, and this filing gives no indication of its likely size.

Argentina's tax-holiday replacement is still unapproved more than a year after the enabling decree, and Q4 2014 ran with zero benefit exactly as guided

The Industry Secretary's resolution approving MercadoLibre's application for Argentina's new software-development tax law - submitted in May 2014, more than 17 months after the enabling decree was first issued in August 2011 - was still pending as of this filing's date. As guided since Q3 2014, the fourth quarter ran with no tax-holiday benefit at all, contributing to the full-year blended and effective tax rates rising to 40.3% and 54.8% (from 27.9% and 30.3%) - though the larger share of that jump is the non-deductible Venezuela impairment and FX losses rather than Argentina alone.

Argentina represented roughly 27.1% of FY2014 consolidated net revenues. A tax-law application that has now been pending regulatory approval for the better part of a year, with no indication in this filing of when - or whether - it will be granted, is a real and recurring drag on the Company's effective tax rate that shows no sign of resolving.

Full-year operating margin fell even after excluding the Venezuela impairment entirely, and the decline accelerated into the fourth quarter rather than fading

Ex-impairment operating margin fell 1.95 percentage points for the full year (30.54% versus 32.49%), and the derived Q4-standalone reading - 28.03% operating margin, down 10.68 percentage points from Q4 2013's 38.71% - shows the compression getting worse, not better, in the one quarter that carried no fresh one-time charge at all. Product-and-technology development spending grew 31.1% for the year and sales-and-marketing spending grew 23.4%, both comfortably ahead of 17.8% reported revenue growth; the filing attributes part of the increase to LTRP compensation cost that moves mechanically with MercadoLibre's own rallying stock price, and part to genuine headcount and cloud-infrastructure investment.

Q3 2014's [full return to GAAP profitability](/analysis/meli/2014-09/#a-clean-recovery-quarter-wrapped-around-a-currency-story-that-keeps-widening) read, at the time, like the Venezuela-driven loss was a one-off interruption to an otherwise healthy cost structure. The full year shows a different picture: operating margin kept eroding even without any Venezuela charge at all, and the fourth quarter - the one period genuinely free of one-time items - posted the steepest single-quarter margin decline of the year.

The State of São Paulo fraud claim's risk classification was quietly reclassified from "remote" to "reasonably possible" this year, following an ambiguous appellate ruling

This is a separate, older case from the 9th Treasury Court tax claim above: a state prosecutor's 2007 claim seeking to hold MercadoLibre's Brazilian subsidiary jointly liable for seller fraud and buyer damages. On May 23, 2014, the State Court of Appeals issued a ruling the Company itself describes as unclear, but which "could be understood" to hold the subsidiary jointly and severally liable - a genuine escalation from the claim's prior years of favorable procedural posture. MercadoLibre's appeal to the Brazilian Federal Superior Court of Justice remained pending as of the filing date, and management's own risk assessment now reads "reasonably possible" rather than "remote."

A risk reclassification driven by an appellate court's own ambiguous language is a genuinely different situation from a claim sitting unchanged in "remote" purgatory - this is a case where the facts on the ground actually moved, and the Company's disclosure moved with them. Worth tracking alongside the São Paulo tax claim above, since both cases now carry real uncertainty, for different reasons.

The Rio Grande do Sul marketplace-monitoring claim first flagged in Q1 2014 resolved favorably this year: on October 16, 2014, the State Court of Rio Grande do Sul ruled the interlocutory appeal in MercadoLibre's favor, definitively revoking the injunction that had ordered the Brazilian subsidiary to monitor and prevent unlawful listings - though the underlying case-merits ruling by the trial court remained pending as of the filing date. On September 12, 2014, MercadoLibre exercised its call option to acquire the remaining 40% of Autopark LLC (the Mexican real-estate and vehicle classifieds business AP Clasificados) for $4.0 million, paid December 10, 2014, ending the put/call structure this coverage has tracked since the original 2011 acquisition and bringing the business to full ownership. The Chile/Mexico real-estate classifieds acquisition's fully-closed purchase price allocation shows total goodwill of $20,783,292 - a small further upward revision from the $20.3 million interim figure disclosed in Q3 2014. This filing also discloses, for the first time in this coverage, a standalone $7.0 million purchase-obligations line item (minimum commitments for advertising, technology equipment, and other goods and services) layered on top of $14.9 million of non-cancelable operating lease commitments; the Company continues to disclose no off-balance-sheet arrangements or variable interest entities. In December 2014, MercadoLibre acquired Business Vision S.A., an Argentine software-development company, for $4.8 million - the same bolt-on engineering-talent playbook it used for a similar Córdoba, Argentina acquisition in 2013.

Coverage Table

Theme	FY2014	FY2013	YoY	Why it matters
Net revenue, reported (USD)	$556.5M	$472.6M	✅ +17.8%	Reported growth decelerated slightly from FY2013's 26.5%
Net revenue, local currency	n/a	n/a	✅ +80.7%	62.9pp gap vs. reported growth for the full year
Net income attributable to MELI	$72.6M	$117.5M	🔴 -38.2%	Entirely the Q2 2014 Venezuela impairment; ex-impairment operating margin still fell 1.95pp
Free cash flow	$120.7M	$24.8M	✅ +386.7%	Office-building capex binge behind 2013's cash crunch finally ended
São Paulo tax claim	Same posture as Q3 2014, lower court ruling still pending	n/a	🔴	Sixth consecutive filing flagged; active litigation, still "remote," zero reserve
Venezuela currency regime	New SICAD/SIMADI system announced Feb. 2015 at ~3.4x weaker rate	n/a	🔴	Subsequent event; no estimated 2015 accounting impact disclosed yet

Target Valuation Range

DCF fair-value range: roughly $2,682 million (base case) to $5,372 million (bull case) in enterprise value - against the current $5,343.8 million enterprise value, the base case covers about 50.2% and the bull case just clears the actual price at ~100.5%, for the first time since Q2 2014. The stock's rally kept outrunning even an updated bull case through most of 2014, but the year-end numbers - a free-cash-flow figure nearly five times larger than 2013's, on a much lighter capex base - finally gave the bull case room to clear the actual price again. The base case still covers only about half of enterprise value, so this remains a stock priced for a scenario meaningfully better than the base case, not a name trading at a discount to its own numbers.

Market cap rose to $5,637.2 million even after the stock pulled back from a November 2014 high of $140.94 to close the year at $127.67, up 17.5% from Q3 2014's ~$4,797.3 million (MercadoLibre has never split its common stock since its August 2007 IPO, so no split adjustment applies).

Market cap → enterprise value	FY2014
Share price (period-end, Dec 31 2014)	$127.67
Shares outstanding	44,154,572
Market capitalization	$5,637.2 million
Total liabilities (loans payable, mostly convertible notes)	$283.8 million
Less: cash and equivalents	$577.2 million
Enterprise value	$5,343.8 million

Peer-multiple sanity check	Q3 2014 (TTM)	FY2014	Change
Net revenue	~$529.8 million	$556.5 million	-
Net income (attrib. to MELI)	~$79.2 million	$72.6 million	⚠️ down
Free cash flow	~$64.2 million	$120.7 million	up
EV/Sales	8.52x	9.60x	up
P/E	~60.6x	77.7x	up sharply
EV/FCF	~70.3x	44.3x	down (improvement)

The P/E jump is driven entirely by the Venezuela impairment dragging GAAP earnings down even as the stock rallied; EV/FCF improved because free cash flow (unaffected by the non-cash impairment) grew far faster than enterprise value. Which multiple a reader trusts more depends on whether the impairment is treated as a real economic cost or a one-time non-cash item.

Scenario	Key assumption	Implied EV	% of current EV
Current (period-end close)	Actual Dec 31, 2014 close	$5,343.8 million	100%
Base	Revenue growth 20%→10% over 5 years; FCF margin 21.7%→28% (verified year-end margin vs. Q3's 12% interim estimate); 13% discount, 4% terminal growth	$2,682 million	50.2% (up sharply from Q3's ~39%)
Bull	Revenue growth 24%→11% over 5 years; FCF margin 26%→34%; 11% discount, 5% terminal growth	$5,372 million	100.5% (clears EV, reversing Q3's first-ever failure)

The bull case clears the actual price again this quarter, though only just, and only because free cash flow's full-year performance came in well ahead of what Q3's interim estimate assumed.

Reverse DCF: holding the 13% discount rate and solving for the perpetual FCF growth rate that would justify today's $5,343.8 million enterprise value on the $120.7 million trailing FCF base gives roughly 10.5% growth, forever - down from Q3 2014's ~11.4%, continuing the multi-quarter easing this coverage has tracked since Q1 2014's ~12.4% high.

Whether the Venezuela subsequent event flagged above produces a fresh 2015 impairment on the scale of Q2 2014's, and whether the operating-margin erosion that continued into Q4 despite no new one-time charges is a genuine trend or a one-quarter blip, are the two clearest swing factors for whether this valuation gap keeps closing or reopens once 2015's numbers start landing.

---