Nu Holdings Expands Lending: Can Credit Quality Hold Up?

Nu Holdings' Q2 2026 credit portfolio rose 37% year over year to $39.4 billion, driving a 9% sequential increase in net interest income to $3.7 billion and a 180bps NIM expansion to 22.9%. The faster-growing, higher-risk lending mix lifted returns and reduced cost of credit 9% sequentially to $1.7 billion, but 90-plus-day delinquencies increased 35bps to 6.9%. Management cites 244% reserve coverage of 90-plus-day NPLs and materially lower delinquencies among primary-bank customers as mitigants, while investors will need to monitor whether credit quality holds as risk appetite expands.

Analysis

NU is optimizing reported profitability by moving down the credit curve, but the key underwriting question is vintage seasoning rather than the current reserve ratio. A high-yield unsecured book can show improving early arrears and lower provisions simultaneously during rapid origination; the relevant test arrives over the next 2-4 quarters as newer cohorts migrate into late-stage delinquency. If loss emergence exceeds pricing assumptions, incremental provisions will absorb a disproportionate share of net interest income and challenge the premium multiple.

ITUB is the cleaner relative beneficiary: its lower-risk balance-sheet mix gives it capacity to take prime borrowers if NU tightens underwriting, while its earnings are less exposed to a Brazilian consumer-credit downturn. MELI is differentiated by merchant, payments, and commerce data, but its credit business is also cyclically exposed; a consumer stress event would pressure both lenders, albeit MELI has a larger non-credit earnings cushion. The market may be underpricing correlation between Brazilian household liquidity, FX volatility, and fintech credit losses.

Near term, no standalone directional catalyst is evident absent monthly Brazilian delinquency data or NU’s next cohort disclosure. Over 1-3 months, monitor whether late-stage arrears stabilize despite continued unsecured growth; a sustained deterioration would force 2026-27 earnings-estimate cuts and multiple compression. Over 6-18 months, the structural upside case requires proof that primary-account adoption—not merely higher-risk acquisition—is driving loan growth; otherwise growth becomes increasingly capital- and reserve-intensive.

Key Decisions for Investors

• Initiate a 3-6 month relative-value position: long ITUB / short NU in equal dollar exposure. ITUB offers defensiveness if Brazilian consumer credit weakens, while NU has greater sensitivity to adverse vintage development; reassess if NU reports two consecutive quarters of stable-to-lower late-stage delinquency while maintaining unsecured growth.
• Do not add directional NU long exposure until the next earnings release confirms that provision expense and late-stage arrears remain contained as recent unsecured cohorts season. A material upward revision to expected-loss guidance or a sequential rise in early arrears is the short trigger.
• For existing NU longs, use a 6-9 month put spread around the next two earnings dates rather than selling outright; the asymmetric risk is an abrupt reserve-build narrative, while stronger underwriting evidence could preserve upside.
• Maintain MELI as the preferred growth-fintech exposure versus NU only if credit remains a minority contributor to incremental operating profit. Monitor credit-loss expense, receivable growth, and Brazil consumer indicators; reduce the preference if lending becomes the principal source of MELI earnings acceleration.