Nu Holdings Ltd (NYSE: NU)

Disclaimer This report has been prepared solely for academic and educational purposes. It does not constitute investment advice, financial advice, or a recommendation to buy, sell, or hold any security. The analysis is based on publicly available information believed to be reliable, but no representation is made as to its accuracy or completeness. Any opinions, forecasts, or valuation estimates are those of the author as of the date of publication and are subject to change without notice. The author does not hold any position in the securities discussed. This report is prepared in a personal academic capacity and is not associated with, endorsed by, or representative of any investment adviser, financial institution, current employer, or other organisation. Readers should conduct their own independent research and seek professional advice before making any investment decision.

Executive Summary

• Growth Duration is Underestimated. Brazil has proven the model, but Mexico is earlier in the same curve. The market is underpricing the duration of high-ROE growth if Mexico follows even part of Brazil’s path from card-led acquisition into deposits, lending and primary banking. Colombia is not yet central to earnings, but adds long-duration optionality.

• Four Costs Moat. Nu’s advantage compounds across customer acquisition, servicing, funding and risk. Cost to serve remained around US$0.8 per active customer per month, while deposits reached US$41.9bn in FY2025, funding a US$32.7bn credit portfolio. This creates a deposit-funded lending flywheel that incumbents cannot easily replicate without dismantling branch-led cost structures.

• ARPAC Runway. Nu does not need another customer boom to compound earnings. The next leg is getting existing users to behave more like mature Brazil cohorts: account, card, deposits, personal loan, investments, insurance and SME products. Q4’25 ARPAC reached US$15, while the model assumes a pathway towards higher mature-cohort monetisation over time.

Industry Analysis

Industry Overview

Competitive Landscape

• Private valuation ** IPO valuation

Brazilian banking landscape: Unique, concentrated, profitable, but structurally disrupted

Industry specifics and brand positioning

Framework Analysis

Industry Revenue and Expense Analysis

• Net interest income (NII): spread between interest-earning assets and deposits/funding. n LATAM, high nominal rates and high consumer-credit spreads make this large, but asset quality is the binding constraint.

• Interchange and card fees: attractive at early scale, but regulated and not enough alone to justify large market caps.

• Payment and merchant fees: strongest for commerce-linked wallets and acquirers such as Mercado Pago.

• Subscription / premium fees: more relevant in Europe and affluent segments than in mass-market LATAM. However, Nubank is also expanding premium offering (“Ultravioleta”)

• Wealth, insurance and marketplace commissions: cross-sell revenue with lower capital intensity, but often slower to scale than credit.

• Branches, manual back-office processes and legacy IT are replaced by technology, cloud, fraud, customer service, marketing and credit losses. The digital model should show falling non-credit opex as a percentage of revenue as the platform scales.

• The more important debate is not opex but credit cost: unsecured lending can expand ARPAC quickly, but it also increases provisions and capital needs.

• In Q1 2026, Nu reported deposits of $42.4bn, a credit portfolio of $37.2bn, LDR of 58.3%, NIM of 21.1%, risk-adjusted NIM of 9.5%, and net income of $871m with 29% ROE. This is a useful benchmark for what successful LATAM neobank economics can look like after scale: high gross spreads, low cost-to-serve and high ROE, offset by visibly cyclical credit loss dynamics.

• Active customers (rather than registered customers);

• ARPAC;

• Cost-to-serve per active customer;

• Deposit cost as a percentage of interbank rates;

• Loan-to-deposit ratio;

• NIM and risk-adjusted NIM;

• 15-90 and 90+ NPLs;

• Provision coverage;

• Monthly purchase frequency;

• Products per customer;

• Contribution profit by cohort.

Business Segments

Pricing Strategy

1. Free or near-free entry product: no annual fee credit card, free digital account, app-first servicing, transparent pricing.

2. High-yield deposits where the market is early or acquisition is strategy: Mexico and Colombia have used promotional/competitive deposit yields to pull balances into the ecosystem. Mexico’s Cajitas (“money boxes”) show 6.5% liquid yield, higher rates for frozen balances; Cajitas Turbo (boost) shows 13% promotional rate as of May 20266; Mexico’s central bank rate (BoM rate) is currently 6.5%. Colombia’s Cajitas reported up to 11.25% AER, starting from 9 April 2026.7 Colombia’s policy rate set by Central Bank of Colombia is 11.25%.

1. Credit pricing is not cheap, in absolute terms: Mexico’s credit card pages shows a zero annual fee; but a high average CAT (“Costo Anual Total”, total annual costs) of 145.8% and weighted annual fixed rate of 92.5%8, valid for the stated offer period. Colombia’s May 2026 credit card tariff sheet shows zero management fee and zero insurance; but a 28.17% AER credit interest rate – the proposition is no nuisance fees and transparent control, not low APR.

2. Risk-based limit expansion: Nu gives small initial limits; then observes behaviour; then increases exposure. This allows them to price risk while reducing upfront selection risk. According to Nu release, NuFormer, its proprietary foundation model set is already used for credit card decisioning in Brazil and Mexico.

1. Incoming Pricing: Reducing friction, winning the first relationship

1. Lock-in pricing: turn product usage into behavioural stickiness

• Deposit-side Lock-in: the structure is usually Liquid savings → goal-based pockets/Caixinhas/Cajitas → frozen/term deposits → higher-yield conditional products. That means Nu can segment customers by liquidity preference. Customers who only need daily liquidity receive a lower rate; customers willing to freeze balances receive a slightly higher rate; customers who satisfy engagement conditions can access promotional or Turbo rates. In Mexico, Nu explicitly separates 24/7 Cajitas from frozen balances and Cajita Turbo, with higher rates for less-liquid or more conditional balances. In Colombia, CDT Nu plays the same lock-in role: custoomers can open fixed-term deposits from Cajitas, with a minimum deposit of COP 50k and 90/120-day terms. The point is not only funding; it makes Nu part of the customer’s savings goals, rather than a spending card.

• Lending-side lock-in: On the lending side, lock-in comes from limit progression. NU starts with controlled exposure, observes behaviour, then expands limits or offers adjacent credit products. In Brazil, its card page explicitly points customers towards monitoring their limit and using saved money/investments to increase credit access through secured-limit structures. This is more of a change in customer psychology: the customer is encouraged to keep deposits, payments and credit behaviour inside Nu, given the improved access to credit over time.

1. Portfolio pricing: monetise through spread, risk and mix

NU’s funding cost is not simply “pay the highest rate”. In FY2025, total deposits were US$41.9bn, up 29% YoY, while cost of funding was 87% of interbank rates. That means NU was still funding below the broad interbank benchmark despite offering attractive visible deposit products.

Brazil illustrates the mature version: Nubank’s disclosed card interest range includes 2.75%–19.99% per month for revolving credit and 0.99%–19.75% per month for instalment of the bill. Mexico shows the expansion-market version: no annual card fee, but an average CAT of 145.8%; personal loans show an average CAT of 88.2% and weighted fixed annual rate of 84.8%. Colombia is more regulated/earlier-stage, but its May 2026 card tariff sheet still shows zero handling fee and zero insurance, alongside a 28.17% E.A. credit interest rate in the simulation.

NU’s portfolio strategy is to use a large, relatively low-cost deposit base to fund credit card receivables and loans. The company gives up traditional banking fees at the front end, but recovers economics through card interest, loan interest, interchange, late fees and balance-sheet spread.

Market and Competitor Analysis

SWOT Analysis

Investment Thesis

• Brazil has validated the playbook. NU reached 131m global customers in FY2025, with Brazil at 113m customers and 62% adult penetration, showing that the model has already moved from customer acquisition into monetisation.

• Mexico sits earlier in the same curve. Mexico reached around 15% adult penetration and became the leading issuer of new credit cards, but remains far behind Brazil in product depth, credit penetration and ARPAC maturity.

• Colombia remains early-stage optionality. Colombia exceeded 4m customers in FY2025, but monetisation is still nascent and should be treated as long-duration growth rather than near-term earnings support.

• Market is treating NU as a mature Brazil neobank. The investment case is that Brazil is no longer the whole story; the next phase is the repeatability of the Brazil monetisation curve across Mexico and, later, Colombia.

• Cost to acquire is structurally lower. NU’s digital-first model allows customer acquisition at a fraction of branch-based incumbents, supporting profitable scaling even in lower-income or thin-file segments.

• Cost to serve is structurally lower. FY2025 cost to serve remained around US$0.8 per active customer per month, while efficiency improved materially as revenue scaled.

• Cost of funding is improving through deposits. NU ended FY2025 with US$41.9bn of deposits, up 29% YoY, giving it a self-funded base for credit expansion rather than relying primarily on wholesale funding.

• Cost of risk is managed through data, not avoided. NU is still exposed to consumer credit risk, but the underwriting advantage comes from transaction, payment, deposit and app-behaviour data accumulated across a large active base. NU’s advantage is not just that it is digital. The moat is that all four cost lines improve together: acquisition, servicing, funding and risk. A branch-based bank may compete on app design or deposit pricing, but it cannot easily replicate NU’s lower structural cost base without undermining its existing distribution model. NU’s low-friction acquisition model drives customer growth; customer growth creates more data; data improves underwriting and cross-sell; cross-sell increases ARPAC; higher ARPAC then supports further reinvestment in product and risk infrastructure.

• Customer growth is no longer the only driver. NU already has 131m customers; the next earnings leg depends on converting existing users into higher-value relationships.

• ARPAC expands as cohorts mature. The product path is account, card, deposits, personal loan, investments, insurance and SME services. Time increases product penetration and revenue per active customer.

• Credit is the main ARPAC engine. Credit cards and loans remain the largest monetisation drivers, with FY2025 interest income from credit cards and loans forming the core of revenue generation.

• Operating leverage magnifies ARPAC gains. Because incremental servicing cost is low, each dollar of ARPAC expansion has high flow-through potential if credit losses remain controlled. The market is too focused on headline customer growth. NU does not need another 100m customers to compound earnings; it needs existing customers to behave more like mature Brazil cohorts. The customer journey is gradual: first account or card, then deposit balance, then credit limit expansion, then personal loan, then investments, insurance or SME products. This is why ARPAC is more important than raw customer count at this stage.

Valuation

Financial Summary

• Revenue modelling: A bank-style revenue build has been adopted, using Net Interest Income plus non-interest income net of transactional expense. Forecast growth is driven by customer growth, monthly active customers, ARPAC expansion, loan book growth and deposit-funded balance sheet deployment. Headline revenue is not used as the main valuation revenue line because Nu’s economics are better captured through NII, fee income, provisions and efficiency ratio.

• Customer growth: Total customers are projected from 131.0m in FY2025A to 177.5m in FY2028E. Brazil growth is modelled to decelerate as penetration matures, from 113.0m customers in FY2025A to 138.4m in FY2028E. Mexico remains the main growth engine, increasing from 14.0m to 29.6m, while Colombia rises from 4.0m to 9.5m over the same period.
• Brazil: Growth slows from high-penetration expansion to monetisation, with forecast customer growth stepping down from 8.0% in FY2026E to 6.0% by FY2028E.
• Mexico: Customer growth remains structurally higher, forecast at 30.0% in FY2026E, 28.3% in FY2027E and 26.7% in FY2028E, reflecting earlier-stage penetration.
• Colombia: Growth is modelled from a low base, with customers increasing to 9.5m by FY2028E, but Colombia is treated as optionality rather than near-term core profit driver.

• Activity and ARPAC: Monthly active customers grow from 109.3m in FY2025A to 150.7m in FY2028E, with activity rate improving from 83.4% to 84.9%. Monthly ARPAC is forecast to increase from US$13.3 in FY2025A to US$20.2 in FY2028E, driven by product penetration, credit usage and cohort ageing.

• Net Interest Income: NII is projected from US$8.9bn in FY2025A to US$18.9bn in FY2028E. The main driver is growth in credit card receivables and customer loans, partly offset by deposit funding costs. The model assumes NIM stabilises and gradually improves from 16.2% in FY2025A to 17.9% in FY2028E, rather than assuming a straight-line margin expansion.

• Credit book: Net loans grow from US$27.7bn in FY2025A to US$55.6bn in FY2028E. Credit card receivables and customer loans remain the key monetisation assets, while secured lending provides a lower-risk growth channel. Nu’s FY2025 filing separately disclosed loans to customers of US$10.9bn, including US$2.7bn of secured loans, confirming the shift towards a broader lending mix.

• Provisions: Provision expense is modelled to rise from US$4.2bn in FY2025A to US$8.6bn in FY2028E, broadly in line with credit book growth. This avoids overstating profitability by assuming that incremental lending scales without corresponding credit cost. The model therefore treats underwriting advantage as a margin/risk selection benefit, not as a removal of credit-cycle risk.

• Operating expenses: Non-interest expense increases from US$2.8bn in FY2025A to US$4.9bn in FY2028E, but grows slower than revenue. Cost to serve is held broadly stable at US$0.8 per active customer per month, allowing the efficiency ratio to improve from 25.4% to 21.2% by FY2028E. This is the main operating leverage assumption.

• Deposits and funding: Deposits are projected to grow from US$41.9bn in FY2025A to US$75.5bn in FY2028E. The model assumes Nu remains primarily deposit-funded, with loan/deposit ratio increasing from 66.0% to 73.6%. This reflects balance sheet utilisation, not a wholesale funding-led growth model.

Valuation Parameters

Relative Valuation

Sensitivity Analysis

Investment Risk & Catalysts

Risks

Catalysts

Disclosure & Important Notes

1. Brazil’s Pix set for next leap with launch of recurring payments, Reuters, 4 June 2024, Accessed on 10 April 2026. https://www.reuters.com/world/americas/brazils-pix-set-next-leap-with-launch-recurring-payments-2025-06-04/↩︎

2. Competition in Mobile Payment Services – Note by Mexico, OECD, 18 June 2025, Accessed on 10 April 2026. https://one.oecd.org/document/DAF/COMP/WD(2025)12/en/pdf↩︎

3. Brazil’s Nubank partners with convenience store Oxxo to expand in Mexico, Reuters, 13 January 2025, Accessed on 10 April 2026. https://www.reuters.com/world/americas/brazils-nubank-partners-with-convenience-store-oxxo-expand-mexico-2025-01-13/↩︎

4. 5-Bank Asset Concentration for Brazil, FRED(Federal Reserve Bank of St. Louis), Accessed on 16 April 2026, https://fred.stlouisfed.org/series/DDOI06BRA156NWDB↩︎

5. Nu Holdings Ltd. Reports First Quarter 2026 Financial Results, Nu Holdings, 14 May 2026.↩︎

6. Cajitas Nu te ayudan a crecer tu dinero hasta 13% anual, Nu Mexico, Accessed on 23 May 2026, https://nu.com.mx/cuenta/rendimientos/↩︎

7. Nu Colombia mejora sus cuentas de ahorro: nueva tasa y cómo acceder al beneficio, Enter.Co, 9 April 2026, Accessed on 23 May 2026. https://www.enter.co/colombia/nu-colombia-mejora-sus-cuentas-de-ahorro-nueva-tasa-y-como-acceder-al-beneficio/↩︎

8. La tarjeta de credito sin anualidad ni cobros sorpresa - Respaldada por Mastercard ®, Nu Mexico, Accessed on 23 May 2026, https://nu.com.mx/credito/↩︎