Phase Space AI

Trade Construction

MercadoLibre [MELI]

MercadoLibre [MELI] — Trade Construction & Risk

The memo issues NO position verdict. Criteria are scored; the book decides. This section sizes and constructs the expression if the book allocates.

Vehicle: EQUITY

Equity is the default and needs no argument (MEMO_BRIEF, "Vehicle"). No LEAP is proposed because the required input — measured option IV minus trailing 252-day realised vol — was not retrieved inside the time box. Trailing 252-day realised vol is 39.8% (screen record, vol_252d_pct). Without a measured IV the vol spread is unquantified, so per the standing rule the vehicle defaults to equity. Recording this as an explicit non-computation rather than asserting a reason the document cannot support.

Entry / exit

Spot $1,833.15
12-month target (3-yr p25 EV/EBIT 36.82x on NTM EBIT) $2,288, +24.8%
Low case (multiple flat at today's 14th percentile) $2,031, +10.8%
Downside case (NIMAL continues falling — named cause) $1,370, −25.2%
Reward : risk vs the named downside 1.0 : 1.0

That reward:risk is not attractive on its own. What carries the name is the implied-path margin (+28.2pp) and its robustness across the entire 18-cell exit-multiple × FCF surface, not the 12-month skew. A book that ranks on valuation margin should look at this differently from one that ranks on 12-month upside, and this memo is explicit about which is which.

Invalidation — three falsifiable tests, each with a print date

  1. NIMAL below 15% on any quarterly print. It has gone 22.7 → 17.8 in four quarters. Below 15% the credit book is not earning its cost of capital after losses and the terminal-margin bridge fails at its load-bearing assumption.
  2. Provision for doubtful accounts above 15% of revenue for two consecutive quarters (Q1'26: 14.1%). That would mean the CECL front-loading is not front-loading — it is loss experience.
  3. GMV growth below 20% with the food-delivery redefinition still in the base. Volume is the leg of the thesis that does not depend on any accounting judgement; if it goes, nothing else matters.

Sizing constraint

size_bucket: large ($92.9bn market cap), 39.8% realised vol, ADV not a constraint. Evidence grade B — per valuation.md rule 5, uncertainty reduces position size and never the operating assumption. The two named evidence gaps (unquantified GMV redefinition; contested FCF definition) argue for the lower half of whatever band the book assigns to a B-grade State-C name.

Correlation note for the book: MELI, RBA and BKNG were underwritten together as "network-effect marketplaces". They are not one exposure. MELI is LatAm consumer credit and FX; RBA is North American used-equipment and insurance-salvage supply; BKNG is European travel demand. The shared label is a research convenience, not a factor.