Verdict: REMOVE | LR 0.7 (bearish)

Revenue growing 39%. Earnings growing 4.5%. Management says they won't fix this. At 42x trailing earnings with no margin catalyst in the next 15 weeks, this is an expensive name to hold through Q1 earnings.

The regression tells the same story from the other direction: trailing idiosyncratic alpha is -31.8% annualized. That's not macro. That's not LatAm FX. That's MELI-specific value destruction — margin compression, credit provisioning, competitive subsidies — all showing up in the residual after removing market and momentum factors. Orthogonal Sharpe: -0.91.

Factor Decomposition

iev regress MELI (n=250, daily)

r_MELI = -31.8% + 0.62 x SPY + 0.23 x MTUM + e

Variance decomposition:
  SPY     12.8%
  MTUM     5.4%
  Idio    81.8%    (above 75% target)

a = -31.8%   s_idio = 35.1%   R^2 = 18.2%

A note on beta: yfinance reports univariate beta of 1.53 to SPX. The multivariate regression separates momentum from market — the "true" market beta after controlling for MTUM is 0.62. The univariate number inflates market exposure by absorbing the momentum loading. This is the Oakmark Select problem: univariate beta conflates factor exposures.

The 82% idiosyncratic variance is actually bad news for the bull case. It means the negative alpha is company-specific, not macro noise you can hope reverses with a LatAm rotation. MELI's fundamentals are doing this to MELI.

Fundamental Snapshot

Source: 10-K filed Feb 25, 2026; Q4 2025 earnings call Feb 24, 2026

Revenue-earnings disconnect: $8.1B of incremental revenue (+39%) produced only $77M of incremental net income (+4%). Operating leverage is running in reverse.

Credit book:

From the 10-K (line 4696-4703): "The provision for doubtful accounts increased by $1,233 million mainly due to the increase in originations growth at 61% (mainly related to the credit card, consumer and merchant products)."

Provisions are growing faster than originations (66% vs 61%). Write-offs doubled (+103%) while the portfolio grew 90%. These are small signals of marginal quality deterioration masked by rapid volume growth.

The Bear Case (Primary Sources)

1. Management explicitly rejects short-term margin optimization

From the Q4 2025 call, CFO Martin de Los Santos:

"Not hesitate invest order capture opportunities as past, short-term margin pressure, not optimize short-term margin. Manage business long term long-term perspective."

This is not new language. The same message appeared in Q3 2025, Q1 2025, Q4 2024, and Q3 2024. Five consecutive quarters of "we will not optimize for short-term margin." When management tells you the same thing five times under securities law, believe them.

The 5-6 percentage points of deliberate margin compression come from four investments: lowered free shipping thresholds in Brazil, credit card expansion (not yet NIMAL positive on average), 1P e-commerce (not standalone profitable), and cross-border trade scaling. None of these reach inflection within 15 weeks.

2. Credit card book is not profitable and growing fast

From the Q4 2025 call:

"Credit point, not NIMAL positive on average, half portfolio already NIMAL positive in Brazil... So on average not profitable yet."

MELI issued ≈3M new credit cards in Q4 alone, accelerating from 2M in Q3 and 1.5M in Q2. The older Brazil cohorts (2021+) are NIMAL positive. The newer cohorts — which are now more than half the book — are not. Each quarter of 3M new cards adds more unprofitable portfolio weight that takes 2+ years to season.

Consumer credit NIMAL is "30s to high 40s" percent. Merchant credit NIMAL is high. Credit cards are the drag — and credit cards are 39% of the credit portfolio (up from 25% a year prior), still growing as a share. Mix shift toward the lowest-margin product continues.

3. Competition disclosure is unusually candid

From the 10-K (lines 1758-1761):

"In 2025, several new global and regional entrants, including rapidly expanding Asian e-commerce platforms, gained significant market share in Latin America through low-price strategies, direct-from-manufacturer supply chains and cross-border logistics models."

This is Tier 1 evidence — 10-K risk factors, signed by the CEO, reviewed by auditors, executives face criminal liability for material misrepresentation. When a company writes that competitors "gained significant market share" in its own filing, the competitive pressure is real enough that the legal team insisted on disclosure.

MELI's response: lowering the free shipping threshold in Brazil, which drove the 35% GMV growth but compressed gross margins (46.1% to 44.5%). This is a defensive subsidy — buying volume with margin.

4. Argentina is a volatile swing factor, not a stable growth driver

Argentina is 13.6% of consolidated revenue ($3,928M, +62.9% YoY) — the fastest-growing segment. But the effective tax rate spiked from 21.4% to 29.7%, driven by Argentina inflation adjustment mechanics. Under Milei's stabilization, lower inflation means smaller tax deductions, pushing the rate toward statutory. Meanwhile, MELI announced $3.4B of Argentina investment for 2026 (30% above 2025). The stock dropped 6.6% on that announcement (March 14).

Argentina giveth the revenue growth and taketh the earnings through tax mechanics. For a 15-week window, this is unhedgeable noise.

Bull Case (Steelman)

The long-term thesis is real:

• Revenue compounding: 28 consecutive quarters above 30% growth. At $28.9B, MELI is still growing like a company one-third its size.
• Fintech TAM: 78M MAUs on Mercado Pago, leading NPS in four countries. Less than 20% of Mexicans hold credit cards. The digital bank thesis has structural runway.
• Logistics moat: Mercado Envios handles the majority of shipments. 14 new fulfillment centers planned for Brazil in 2026. This is hard to replicate.
• Advertising: +67% FX-neutral growth, high-margin revenue stream still small as % of GMV.
• Credit quality: Credit card NPL fell to an all-time low of 4.4% in Q4. The scoring models are improving. NIMAL on older cohorts is positive and rising.
• Investment grade: Moody's upgraded to Baa3 stable. The credit market is endorsing the balance sheet.

But this is a 12-24 month thesis, not a 15-week thesis. Management has explicitly and repeatedly said margins are not recovering near-term. The next catalyst within our window — Q1 2026 earnings on May 7 — is unlikely to show margin improvement given the investment posture. The logistics fee adjustment in Brazil (March 2026) adds an estimated +3% to EPS per Itau BBA. That's a tailwind, but small relative to 5-6pp of investment-driven margin compression.

The bull case requires patience that a 15-week filtration window does not afford.

Consensus Map

25 Buy / 1 Hold / 0 Sell. Average PT $2,684 (65% above current). Lowest PT $2,100 (JPMorgan, downgraded March 12).

The Street is overwhelmingly bullish on the long-term thesis and hasn't yet fully adjusted to the margin compression reality. JPMorgan's whiplash — upgrade to Overweight with $2,800 PT on Feb 12, downgrade to Neutral with $2,100 PT on March 12 — illustrates the instability. The stated reason for the upgrade ("more benign competitive environment" as Shopee raised take rates) was reversed four weeks later ("Shopee willing to sacrifice margins, 15% downside to consensus EBIT").

Forward P/E: 21.8x. This implies consensus expects earnings to roughly double from $39.40 to ≈$75 NTM. With operating margins compressing and management explicitly not optimizing for near-term profitability, that expectation carries significant downside risk. If earnings come in at $60 instead of $75 — which is what JPMorgan's "15% downside to consensus EBIT" suggests — the stock needs to re-rate to justify current price.

Counterparty analysis: At $83B market cap with 25 buy ratings, the informed market owns this thesis. The bull case is not differentiated information — it's consensus. Any edge in MELI would need to be contrarian, and the contrarian view is bearish.

Valuation

MELI trades at a 35% trailing P/E premium to QQQ despite 40pp of underperformance over the past year. The premium is justified only if you believe earnings growth will re-accelerate to match revenue growth. Management is telling you it won't — not yet.

At 52-week lows ($1,631 vs $2,645 high, 1% of range), the stock could be at capitulation or at fair value. The RSI of 32.4 is near oversold. But "oversold" is a technical condition, not a fundamental floor. A stock with compressing margins, accelerating provisions, and management explicitly investing through earnings can stay "oversold" for quarters.

Filtration Decision

P(MELI underperforms QQQ over 15 weeks): 70%

What would change this: MELI pre-announces positive Q1 results (P < 10%). Shopee withdraws from Brazil (P < 5%). Credit card NIMAL turns positive on average (not before Q1 report). Strong EM rotation with BRL appreciation >5% (P ≈15%).

Where I might be wrong: The 70% probability could be too high given RSI oversold conditions. A sharp technical bounce is possible regardless of fundamentals. But the fundamental overhang — 42x P/E with 4.5% EPS growth and management explicitly not optimizing margins — limits the magnitude and duration of any bounce. The Q1 earnings report on May 7 would need to surprise positively on margins, and management has given you no reason to expect that.

Redistribute 0.47% to survivors.