Aug 13, 2026 earnings call — Nu's AI-driven growth and Brazil product push
Nubank reported nearly $5.9 billion in gross revenue and its first quarter with more than $1 billion in net income.
Revenue $5.9B (est $5.4B) · EPS $0.22 (est $0.20)
Key quotes
"About a year ago, we introduced the NuFormer, our foundation model for financial behavior."
"In July, we launched Croma, a subscription-based tier for our super core customers, a segment with an even larger profit pool than high income and one where we already have significant penetration."
"I'm proud to announce that in the past quarter, for the first time, we generated more than $1 billion in net income."
Analyst Q&A 10
Desenrola effect on credit costs
Q (Daer Labarta): Analyst asked how Desenrola affected provisions and risk-adjusted margin, and whether it impacted the bottom line.
A: Management said Desenrola raised cost of credit by about 5% and that it also affected risk-adjusted net interest margin, but was a minority of the impact versus seasonality and strong underlying credit.
"Yes, I did mention that Desenrola had an impact of about 5% on our cost of credit." — Rob Livingston (CFO)
AI and risk-adjusted margin drivers
Q (Jorge Kuri · Morgan Stanley): Analyst asked which KPIs show AI-driven credit improvements and whether AI makes the 12% risk-adjusted NIM a floor.
A: Management said their AI models are more powerful and tracked the same way as historical models, with back and forward testing. They clarified they did not call 12% a floor but said they expect margins to be in that ballpark and that model-enabled intentional risk expansion is paying off.
"I would correct one thing you said, though, where you said that 12% is a floor." — Rob Livingston (CFO)
OpenAI board role and Nubank benefit
Q (Eduardo Rosman · BTG Pactual): Analyst asked how David's OpenAI board role helps OpenAI and how it might help Nubank.
A: David said AI is the most important technology transformation and is already changing Nubank across credit and customer-facing areas. Being close to OpenAI gives useful insight and he hopes to contribute to how OpenAI executes, while learning to apply AI at Nubank.
"I think effectively is we discussed very openly that -- internally and externally that we think artificial intelligence is the most important technology transformation in our history." — David Velez-Osomo
Government renegotiation portfolio and payroll product
Q (Pedro Leduc · Itau BBA): Two questions: whether the government renegotiation accounts appear in Stage 3 and their loss coverage; and an update on private payroll lending and rollout pace.
A: Management said the expected credit loss impact was small, less than $10 million, and the portfolio shows up in Stage 3. On private payroll, they are accelerating month-over-month, getting more comfortable, and moving cautiously to avoid early high-risk customers while positioning to win over 18–24 months.
"There was a small impact on expected credit losses, as I mentioned in my prepared remarks, but it wasn't material, less than $10 million." — Rob Livingston (CFO)
How to grow net income above $1 billion
Q (Yuri Fernandes · JPMorgan): Analyst asked how Nubank can keep growing net income — through ARPAC, new products, or Mexico expansion.
A: Management said growth will come from raising ARPAC in Brazil (we're at 17 and cohorts reach mid-20s), plus big opportunities in Mexico and Colombia. They also highlighted Croma for higher segments and a large SME opportunity with over 6.5 million SMEs to monetize.
"From that perspective, there's going to be a lot of the opportunity in Brazil is to continue increasing ARPAC." — David Velez-Osomo
Headcount and AI productivity
Q (Geoffrey Elliott · Autonomous): Analyst asked about small headcount decline and future hiring plans, and how AI affects workforce productivity.
A: Management said headcount is roughly flat after rehiring attrition from return-to-office, and they don't expect big changes in total employees. They expect a large productivity increase from AI and that current headcount will be "2, 3, 4, 5x more productive" over the next few years.
"Looking forward, we are seeing a huge amount of productivity increase with AI, and we're very excited about the potential that, that creates." — David Velez-Osomo
AI price optimization for loans and deposits
Q (Daniel Vaz · BofA): Analyst asked where AI-driven price optimization has biggest impact: credit card financing repricing or deposits/funding cost.
A: Management said optimization is often about balancing product quality and cost, not just lowering cost. For deposits, they can optimize funding cost but may choose to reinvest gains into customer experience, since they remain focused on growth and product quality in early markets like Mexico and Colombia.
"a lot of this optimization is not necessarily about minimizing cost, but it could also be about optimizing quality." — David Velez-Osomo
Why risk-adjusted NIM beat expectations
Q (Mario Pierry · BofA): Analyst asked why risk-adjusted NIM jumped to 12.4% versus prior guidance around 10.5–10.8%, and about NPL seasonality versus slide detail.
A: Management said about one-third of the upside versus Q1 expectations came from Desenrola, and two-thirds came from stronger-than-expected credit performance and increased earning balances driven by earlier growth. They also explained Page 20 is credit-card-only and uses a rolling average, so it smooths seasonality versus the company-wide chart.
"about 1/3 of that benefit relative to what we were expecting is coming from Desenrola." — Rob Livingston (CFO)
Source of NPL deterioration
Q (Daniel Vaz · Safra): Analyst pressed whether NPL deterioration is coming mainly from unsecured personal loans or mix shifts.
A: Management said the increase is driven more by mix shift in where they are lending rather than deterioration within unsecured lending. They noted disaggregated views show steady or improving performance across income segments, but portfolio mix changes are raising the overall 90-plus trend.
"So it's -- I wouldn't quite characterize it as that as much as I would characterize it as a mix shift that we are seeing." — Rob Livingston (CFO)
Building US credit models and timing
Q (Craig Maurer · FT Partners): Analyst asked how existing models translate to the U.S. and how long to rebuild confidence there.
A: Management said the NuFormer platform and credit expertise translate quickly, but building U.S.-specific data richness and foundational testing will take time. They estimated between 12 and 30 months to mature U.S. models before expanding with similar confidence, and they will start by testing and learning.
"the actual data richness and building the experience of foundational testing and having the models in place that are specifically tooled for the U.S. market will take somewhere between 12 and 30 months, depending on the degree of maturation of those curves." — Rob Livingston (CFO)