Nu Holdings: A False Narrative, A Catalyst, And A Valuation The Market Won't Sustain

Two analysts downgraded Nu Holdings this spring and the stock sold off into the $13 range. Bank of America slashed its price target from $16 to $10. Susquehanna cut its target from $18 to $13. Both cited rising credit provisions and margin pressure following Q1 results. The stock fell nearly 37% from its 52-week high.

The disconnect is that the business didn't slow down. The downgrade pile-on doesn't match the growth rate, and Q2 earnings on August 13th should force the market to choose between the narrative and the math.

Credit provisioning is seasonal, not structural.

The bear case centers on one number: the 15-to-90-day non-performing loan ratio (loans 15 to 90 days past due, a measure of near-term credit stress) rose to 5.0% in Q1 2026, up 89 basis points from Q4. Credit loss allowances — the reserve set aside for expected loan defaults — jumped 33% quarter-over-quarter to $1.79 billion. Risk-adjusted net interest margin (the spread between lending income and funding costs after provisioning for losses) compressed 100 basis points to 9.5%.

On its own, that looks bad. In context, it's first-quarter seasonality in Brazilian consumer credit. The 5.0% NPL figure was only marginally higher than the same metric a year prior. The older 90+ day delinquency ratio — the measure that actually tracks whether loans are going bad permanently — declined 10 basis points to 6.5%, below the 7.0% peak reached in Q3 2025.

Seasonal provisioning compresses margins in one quarter. It doesn't rewrite the business model. The credit portfolio expanded 40% year-over-year to $37.2 billion. Older delinquencies are improving. The provision spike is a temporary hit to earnings, not a signal of underwriting failure.

The real variable is revenue per customer, and it's climbing.

Q1 2026 revenue surpassed $5 billion for the first time. Net income grew 41% year-over-year to $871 million. Customer count hit 135 million. Average monthly revenue per active customer — the single best measure of how much value each user generates — reached $16, up 23% from the prior year. The efficiency ratio (operating expenses as a percentage of revenue, where lower is better) improved to a record 17.6% from 19.9% in the prior quarter.

The market saw provisioning headwinds and sold the growth story. That's a false narrative when the per-customer monetization curve is accelerating.

Mexico just flipped from cost center to profit center.

Nu crossed 15 million customers in Mexico and reached break-even profitability. That's the second-largest market for a digital bank operating in Latin America, and it just stopped losing money. Brazil, where Nu holds 115 million customers, captures roughly 7% of a country addressable profit pool exceeding $100 billion in annual gross profit. Colombia is approaching 5 million customers. The U.S. expansion is being pursued with a measured approach — management said the maximum investment drag would stay below 100 basis points on the efficiency ratio through 2027.

Mexico break-even changes the trajectory of international expansion from cost sink to optionality. That's the kind of inflection point that doesn't show up on a quarterly margin chart but alters the long-term earnings profile.

The $1 billion buyback signals management conviction.

In June, Nu's board authorized a $1 billion share repurchase program over 12 months, announced while the stock was sliding. With $13.9 billion in cash, a zero net debt position, and $1.2 billion in trailing operating cash flow, the company has the firepower to execute. A buyback of this size is management's way of saying the dislocation is a pricing error, not a business error.

Management doesn't buy back shares when they think credit losses are about to blow out. They buy them when the stock is temporarily out of sync with earnings power.

Q2 earnings is the catalyst.

Consensus expects $5.39 billion in revenue for Q2 2026, roughly 46% year-over-year growth, with EPS around $0.20 versus $0.14 in Q2 2025. Revenue per customer is the metric to watch — if it holds above $16 or pushes higher, it confirms monetization is scaling alongside the customer base. The efficiency ratio should improve sequentially if first-quarter credit provisioning normalizes.

If Nu delivers in-line or above expectations, the disconnect between the beaten-down stock and the accelerating fundamentals forces a repricing. That's the setup.

The valuation number that closes the argument.

Nu's PEG ratio (the forward price-to-earnings multiple divided by the revenue growth rate, where below 1.0 means the multiple is cheaper than the growth rate) sits at 0.45. In plain English: the stock's forward PE is less than half its revenue growth rate. The enterprise value-to-sales multiple is 3.0x on a business growing revenue at 45% year-over-year. The stock trades at 28 times forward earnings when the underlying growth rate exceeds that multiple.

At that PEG, the market has already priced in a growth slowdown that hasn't happened.

The break condition.

The thesis breaks if credit losses turn structural — a hard economic downturn in Brazil driving 90+ day delinquencies back above 7%, sustained margin compression pushing the efficiency ratio above 25%, or a material miss on Q2 revenue that signals the customer monetization curve is flattening. If those happen, the forward math changes.

Absent that, a PEG of 0.45 on a $5+ billion-per-quarter revenue base with 135 million customers and international optionality is a valuation the market won't sustain once Q2 results arrive on August 13th.