Market News 9 min read

Nubank Stock Analysis: What the Latest NU Results Mean for Investors

Nu Holdings posted record Q3 net income of $1.09bn on a 32% ROE, while 90-plus day NPLs rose to 7.3%. NU shares closed 11.75% higher on Friday.

A sturdy glass pitcher being steadily filled with clear water, while a small crack near its base lets a thin trickle escape onto the tabletop

Nu Holdings reported record quarterly net income of $1.09bn for the third quarter of 2026 on 12 November, up 31% year on year on an FX-neutral basis, on a return on equity of 32%. Gross revenue reached a record $6.3bn, up 37%, gross profit rose 35% to $2.6bn, and the customer base closed the quarter at 143 million after 3.7m net additions. The shares closed on Friday 13 November at $13.22, up 11.75% from $11.83, on 152.4m shares against an 81.4m average.

What the latest results mean for investors is a tension rather than a verdict, and it resolves into one question: what is a 32% return on equity worth when credit quality is the swing factor? On one side, Nubank delivered a 19.3% efficiency ratio that almost no incumbent bank in the Americas can approach, and tightened full-year efficiency guidance to 19%-19.5%. On the other, the 90-plus day non-performing loan ratio rose to 7.3% from 6.9% in the second quarter, coverage of those loans fell to 234% from 244%, and a one-off charge on receivables acquired from payment institutions placed under liquidation by Brazil's central bank pulled the reported risk-adjusted net interest margin down to 12.1%. Both sides are set out below.

What actually happened

The quarter was, on the growth lines, close to a clean sweep:

  • Net income: $1,091.6m, a record, up 31% year on year FX-neutral.
  • Return on equity: 32%.
  • Gross revenue: a record $6.3bn, up 37% year on year.
  • Gross profit: $2.6bn, up 35%.
  • Customers: 143 million, after 3.7m net additions in the quarter.
  • Efficiency ratio: 19.3%, improved from 20.3% a year earlier, with full-year guidance tightened to 19%-19.5%.
  • Credit portfolio: expanded to $41.6bn, up about 33%.

The credit and margin lines are where the quarter gets interesting, and they require the same care with basis that any bank's disclosures do. Nu reports delinquency on more than one portfolio definition:

  • 90-plus day NPL ratio: 7.3%, up from 6.9% in the second quarter, measured on the total portfolio.
  • 15-90 day NPL ratio: 4.9%, against 4.8% in the second quarter, also on the total portfolio.
  • On the interest-earning portfolio basis, the 15-90 day measure improved to 8.0% from 8.2%. That figure is not comparable with the 4.9% above and the two should never be quoted side by side.
  • Coverage of 90-plus day NPLs: 234%, down from 244%.

On margin, the reported risk-adjusted net interest margin was 12.1%, down from 12.4% in the second quarter. Excluding non-recurring items it rose to 12.7%, from 11.9% in the second quarter. The difference between those two sentences is the entire story of the quarter's margin: reported NIM fell while underlying NIM improved by roughly 80 basis points, and the gap is a charge on receivables Nu had acquired from payment institutions that Brazil's central bank subsequently placed under extrajudicial liquidation.

Management explained the rise in the 90-plus day ratio partly as a function of portfolio mix: the interest-earning share of the book is growing, and interest-earning balances carry structurally higher delinquency than transactional balances. That is a credible explanation and it is also what you would expect a management team to say. The way to test it is coverage and the early-stage bucket, which is why the 15-90 day ratio at 4.9% against 4.8% matters more than its small movement suggests — early-stage delinquency barely moved, which is not what a deteriorating credit cycle normally looks like.

Why the market reacted the way it did

An 11.75% gain on nearly twice average volume, in a quarter where reported earnings per share came in fractionally below consensus, needs explaining. The answer is that the market looked through the reported lines to the underlying ones.

First, the efficiency ratio. At 19.3%, Nu is operating a 143 million-customer bank at a cost-to-income level that traditional Latin American incumbents, carrying branch networks and legacy technology, cannot structurally reach. Tightening full-year guidance to 19%-19.5% told the market that this is a durable operating characteristic and not a quarter of underspending. For a bank, the efficiency ratio is the closest thing there is to a moat expressed as a number.

Second, the quality of the margin miss. A reported risk-adjusted NIM of 12.1% looks like deterioration until you see that the underlying figure rose to 12.7% from 11.9%. The charge that caused the gap arose from Brazilian regulatory plumbing — the central bank liquidating payment institutions whose receivables Nu had bought — rather than from Nu's own underwriting. Markets routinely pay for earnings they believe will recur and discount those they believe will not, and this quarter handed investors a clean example of each.

Third, the compounding arithmetic. 3.7m net new customers and a credit book up about 33% on a 32% return on equity is a business reinvesting at a rate that makes the current period's earnings a weak guide to its earning power. That is the case the bulls were buying on Friday.

The Openbook read

Profitability is the strongest factor and the least contestable. A 32% return on equity with a 19.3% efficiency ratio is top-decile for any bank of any size anywhere, and it is being achieved while the company spends heavily on growth in Mexico and Colombia. The nuance a factor score should capture is that this ROE is being earned on a thin equity base relative to a fast-growing loan book, which mechanically flatters the ratio compared with an over-capitalised incumbent. High, genuine, and partly a function of leverage on a growing balance sheet.

Growth is almost as strong. Revenue up 37%, gross profit up 35%, the credit portfolio up about 33% and 3.7m net customer additions in a single quarter is compounding at a rate that places Nu among the fastest-growing large financial institutions in the world. The relevant caveat is that growth in financial services can be bought by loosening underwriting, and the only way to tell the difference is to watch credit metrics in the quarters that follow the growth. Which brings us to the factor that matters most here.

Solvency — read as asset quality and capital adequacy together — is the factor carrying the risk in this name, and it has genuinely weakened this quarter. A 90-plus day NPL ratio at 7.3% from 6.9%, with coverage down to 234% from 244%, is deterioration on both the numerator and the cushion against it. Neither number is alarming in isolation: 234% coverage is substantial, and Brazilian consumer credit has always run at delinquency levels that would be extraordinary in developed markets. But the direction of both is the same, and the payment-institution liquidation charge — which cost 63 basis points of risk-adjusted margin and, with other non-recurring items, about $53m of net income — is a reminder that Nu's balance sheet carries exposures beyond its own lending decisions.

Momentum is strong on this print — an 11.75% close-based gain on 1.87 times average volume — though a single session following results tells you about positioning rather than trend.

Reward/risk is the factor that frames the rest, and the honest framing is that provisions are the swing variable. The reward case is a 32% ROE compounding a 143 million-customer base with a cost structure incumbents cannot match, in markets where banking penetration is still rising. The risk case is that at a 7.3% delinquency ratio with coverage falling, a modest change in the loss assumption moves net income by a large proportion — because Nu's earnings are the residual after provisions on a $41.6bn book. A 100 basis point deterioration in expected losses on that portfolio is a large number set against $1.09bn of quarterly net income. That sensitivity, not the growth rate, is what a reader should carry away. Our screener is the place to compare that profile against regional peers.

The distinction worth holding onto: Nu's growth engine and its credit engine should be scored separately. The first is exceptional and improving. The second is good and deteriorating at the margin. The share price reflects the first more than the second.

The read-across

For Brazilian incumbents — Itaú Unibanco, Banco Bradesco and Banco do Brasil — a 19.3% efficiency ratio at a competitor with 143 million customers is a structural problem rather than a quarterly one. Nu's cost advantage is not cyclical and cannot be matched by cost-cutting programmes within a branch-based operating model. The read-across is on deposit pricing and on fee income: every quarter Nu adds millions of customers is a quarter in which incumbents defend share with price.

The second read-across is on credit, and it runs the other way. If Nu's 90-plus day ratio is rising because of portfolio mix, that is Nu-specific. If it is rising because the Brazilian consumer is weakening, it will appear in the incumbents' books too, with a lag — they lend to overlapping customers in the same economy. The incumbents' next reporting round is therefore a direct test of which explanation is right, and it is the cheapest test available to an investor in Nu.

The third is regulatory. The central bank liquidating payment institutions and leaving receivable holders with a charge is a reminder that Brazilian financial regulation lands on the profit and loss account with little notice. Any listed name that buys receivables from, or lends against, the Brazilian payments sector carries some version of that exposure.

What to watch next

  • The fourth-quarter result, due around February. The single most important number will be whether the 90-plus day ratio stabilises or continues past 7.3%, and whether coverage holds near 234%.
  • The full-year efficiency ratio against the 19%-19.5% guide. Delivery would confirm the cost advantage as structural; a miss would suggest the Mexico and Colombia build is more expensive than modelled.
  • Risk-adjusted NIM on a clean basis. Management said it expects to carry the underlying strength into the fourth quarter. The test is whether the 12.7% ex-items figure holds without new non-recurring items appearing.
  • The interest-earning share of the portfolio. This is the mechanism management used to explain rising delinquency, so the mix disclosure is what makes that explanation checkable.
  • Brazilian incumbent results for independent evidence on whether consumer credit is deteriorating system-wide.
  • Further central bank action on payment institutions, which would indicate whether this quarter's charge was isolated or the first of several.
Was this useful?
Sign in to vote
3 of 4
Compare NU, ITUB, BBD