Plutux
MercadoLibre's credit-card bet in Brazil lives or dies on whether losses scale slower than originations insight cover
Industry NewsMELI · NU · PAGS8 min read

MercadoLibre's credit-card bet in Brazil lives or dies on whether losses scale slower than originations

Reuters reports that Mercado Pago is expanding Brazil credit cards even as household stress rises—yet the reported delinquency snapshots are high enough to force heavy provisioning. In its own filings, MercadoLibre shows credit losses climbing in line with faster credit growth, making credit quality the key swing factor for near-term fintech earnings and the medium-term credit-cycle narrative.

Published Sep 3, 2026Updated Sep 3, 2026

Credit cards issued (Q2)

2.6M

Second quarter 2026, reported by Reuters

Delinquency (15–90 days)

4.6%

Q2 2026, reported by Reuters

Delinquency (90+ days, total loan book)

7.0%

Q2 2026, reported by Reuters

Delinquency (90+ days, loans more broadly)

18.7%

Metric referenced by Reuters alongside the 90+ discussion

Mercado Pago credit expansion amid Brazil’s household-debt stress

The playbook: grow card originations, then prove the underwriting still holds when consumers get overextended

Brazil’s consumer-debt stress is increasingly concentrated in unsecured products like credit cards and other revolving balances. Reuters’ latest read is that Mercado Pago is not pulling back—instead it is expanding its credit-card book while pointing to card delinquency levels it says remain manageable.

The investor lens is MercadoLibre: Mercado Pago’s credit cards are the highest-leverage path from “payment engagement” to “credit monetization,” but they are also the fastest path from a credit-cycle downturn into mark-to-model losses and provisioning.

Verified event details (primary sources)

Reuters says cards are still growing—and delinquency is already flashing yellow

Credit cards issued (Q2)

2.6M

Second quarter 2026, reported by Reuters

Delinquency (15–90 days)

4.6%

Q2 2026, reported by Reuters

Delinquency (90+ days, total loan book)

7.0%

Q2 2026, reported by Reuters

Delinquency (90+ days, loans more broadly)

18.7%

Metric referenced by Reuters alongside the 90+ discussion

The market’s problem is not that delinquency exists—it’s that growth plus delinquency can force higher provisions before lenders see defaults. Reuters’ reported levels mean MercadoLibre must convert originations growth into spread without letting provision rates outrun revenue growth.

What the market fears vs. what the filings show

In MercadoLibre’s filings, the credit stack is rising—allowance and exposure measures move together with growth

A key question for investors is whether delinquency is staying stable (good underwriting, disciplined growth) or whether it is migrating into higher-loss buckets (worse underwriting, credit-cycle compression). MercadoLibre’s SEC reporting gives two useful checks: (1) how much card credit is outstanding, and (2) how quickly expected credit losses (allowance) are building.

In its Q2 2026 10-Q, the company provides a direct credit-card receivables breakout and an allowance and exposure view.

Credit-card receivables, allowance, and off-balance-sheet exposure in [MercadoLibre](meli)’s reported loan disclosures
PeriodCredit cards loans receivable (gross, $m)Allowance for doubtful accounts (related, $m)Net credit card loans receivable (net, $m)Off-balance-sheet unused agreed loan commitment (credit risk exposure, $m)
Jun 30, 2026$7,676$1,440$6,236$14,047
Dec 31, 2025$5,656$1,107$4,549$9,001

Two things jump out.

First, card credit receivables rose meaningfully from year-end to mid-year. Second, the company’s allowance and unused-agreed commitments increased at the same time—exactly what you would expect if originations are scaling and credit losses are being modeled through CECL.

The crucial nuance is that this is not “charge-offs vs. charge-offs” yet; it’s CECL allowance plus exposure. That still matters: allowance grows when models see higher expected lifetime losses, even before defaults peak.

Quantifying the credit-cycle pressure inside fintech economics

Credit revenues climbed with originations—then loss provisioning escalated in the same direction

MercadoLibre discloses how fintech revenues move, including credit revenues tied to credit-card transactions. In the Q2 2026 10-Q, the company states that fintech growth was supported by higher originations driving credit revenues.

Separately, the company’s 2025 10-K provides a clear mechanism for how sensitive its allowance is to changes in probability of default, and it quantifies how provisioning increased with originations growth.

Fintech revenue driver and allowance sensitivity that link originations to credit costs for [MercadoLibre](meli)
DisclosureWhat it saysWhere it appears
Credit revenues$4.290B for six months ended Jun 30, 2026 (vs $2.472B for six months ended Jun 30, 2025); $2.278B for three months ended Jun 30, 2026 (vs $1.326B for three months ended Jun 30, 2025).MercadoLibre Q2 2026 10-Q
Credit-loss allowance sensitivityHypothetical +10% probability of default increases allowance for loans receivable and unused agreed commitments on the credit cards portfolio by approximately $99M.MercadoLibre FY2025 10-K
Provisioning (allowance movement driver)Provision for doubtful accounts rose to $3,091M in 2025 vs $1,858M in 2024, attributed mainly to higher originations growth at 61% (mainly related to credit card, consumer and merchant products).MercadoLibre FY2025 10-K
In a household-debt storm, growth without a parallel risk-engine improvement can compress margins quickly. Based on disclosed credit-driven revenue increases and the allowance framework, MercadoLibre is effectively betting that losses scale sublinearly vs. originations.

Supply-chain awareness (credit-risk supply chain + cash flow realities)

The credit chain has three choke points: funding cost, underwriting signals, and loss recognition timing

  • Funding cost pressures spread as Brazil’s rates rise; fintech models need stable net interest income to prevent credit growth from turning into lower profit per account.
  • Underwriting signals determine how much delinquency migrates from 15–90 days into 90+ and then into CECL and write-offs.
  • Loss-recognition timing determines whether earnings lag the credit-cycle downturn; CECL allowance can rise before defaults peak, making results look worse sooner.

Unlike a traditional bank, MercadoLibre runs a vertically integrated engagement funnel: marketplace + wallet + credit. The positive case is that it can monitor consumer behavior at high frequency and tighten underwriting faster than competitors.

The negative case is that rapid expansion into a worse consumer cohort raises expected losses, and CECL moves early—so profits can deteriorate even if cash recoveries later turn out better.

Horizons: what to watch next in the tape and in the model

Short-term: delinquency bucket migration and allowance growth speed. Long-term: whether the business can “outgrow” the credit cycle

Near-term, the sequencing matters. Reuters’ reported delinquency snapshot shows the risk is already present. The next read should be whether the company’s credit-card receivables continue growing faster than allowance, and whether past-due totals rise in step with exposure.

Over 1–3 years, investors should focus on whether MercadoLibre can maintain credit growth while keeping expected loss rates and write-offs contained, even as Brazil’s household stress changes composition.

The thesis split is simple: if MercadoLibre keeps credit growth while stabilizing expected-loss ratios, the market can justify premium multiples; if allowance accelerates faster than credit revenue, the stock is likely to de-rate.

Synthesis for investors

What MELI is really telling the market: “We can scale credit without letting it become a provisioning story”

Reuters frames the moment as Mercado Pago being “unfazed” by consumer-debt worries while expanding. MercadoLibre’s SEC disclosures show the mechanics: credit-card receivables and off-balance-sheet unused commitments are rising, and allowance/provisioning is built using CECL models that react to expected default and loss-given-default patterns.

So the live question for investors is not whether delinquency exists—it does. The live question is whether CECL allowance as a percentage of the expanding credit book can flatten, even if delinquency remains elevated.

Listed public equity exposure to the LatAm credit-cycle test

MMercadoLibre, Inc.MELI--
--Vol --
-
Mixed
  • Reuters’ reported card delinquency makes allowance sensitivity a first-order risk for near-quarter earnings.
  • In the Q2 2026 10-Q, credit-card receivables and off-balance-sheet commitments rose together, implying CECL costs can scale fast.
  • In FY2025 results, provision for doubtful accounts rose on 61% higher originations growth, so the model must show loss ratios don’t deteriorate further.
NNu Holdings Ltd Class ANU--
--Vol --
-
Watch
  • Nu’s consumer-credit mix makes it a key comp for whether LatAm fintech can maintain credit quality during household stress.
  • Watch for whether Nu’s delinquency and provision trends diverge from MercadoLibre’s as rates stay high.
PPagSeguro Digital Ltd - Class APAGS--
--Vol --
-
Watch
  • If Brazil consumer risk spills into digital credit, PagSeguro’s credit-related economics could face similar CECL upward pressure to MercadoLibre.
SStoneCo Ltd - Class ASTNE--
--Vol --
-
Watch
  • Brazil’s unsecured-credit stress tests fintech funding resilience; StoneCo is a proxy for whether payments-led models can keep credit losses controlled.
BBanco do Brasil S.A.BBAS3.KS--
--Vol --
-
Mixed
  • If consumer credit quality worsens broadly, large banks face higher provisions and slower asset growth, which can raise credit spreads and pressure consumer demand.

Plutux is not an investment adviser. Market data and AI-generated analysis are for information and education only, not investment advice. Disclaimer

© Plutux Technology Limited 2026