Stock Analysis · Bulls vs Bears

Nu Holdings Ltd. NU

🐂 Bulls vs Bears 🐻Report #212 Oct 2026
★ Willem owns this stock · 2.3% of the portfolio
33
moderate risk▼ −11 vs last
🐂 9vs10 🐻
Bears lead
40%
▲ Chance of the biggest tailwind the forecasts miss
30%
▼ Chance of the biggest headwind the forecasts miss
The verdict

Where it leans: the evidence leans bull on the business. Nu runs at a 19.5% efficiency ratio and 33% ROE, it is the 2 card issuer in Brazil growing volumes twice as fast as the market, and its diluted share count was flat year on year. What changed since report 1: the risk is more clearly concentrated in two places. Brazilian household credit is at record delinquency. Nu's own Q2 margin jump leaned partly on…

🎙️ Bulls vs Bears podcast — Nu Holdings Ltd.
Alex and Sam debate every bull and bear argument · 14:37
0:0014:37
⤓ Download

Written by Claude (Anthropic’s AI) from public sources · Willem owns NU · AI can make mistakes — numbers in brackets point to the sources · Not financial advice

🐂 Bull arguments 9

  1. 1A $1 monthly cost to serve against $17 of revenueStrong
  2. 2ARPAC up 37% in a year, still far below incumbentsStrong
  3. 3Card volumes growing twice the market as the 2 issuerMedium
  4. 4$45B of retail deposits at 88% of the interbank rateMedium
  5. 5Reserves at 244% of 90+ loans, built ahead of lossesMedium
  6. 6Mexico broke even in six years and is now a bankMedium
  7. 7Half of the $1B buyback done at ~$12.31; diluted shares flatMedium
  8. 816.6x forward earnings for a 33% ROE, below its own historyMedium
  9. 9US entry capped at under one point of efficiencyWeak

🐻 Bear arguments 10

  1. 1Brazilian household delinquency hit a record as Nu's 90+ roseHigh
  2. 2The Q2 margin jump leaned on Desenrola and chosen riskMedium
  3. 3A one-country stock up 27% into a close runoffMedium
  4. 4Taxes rise in 2027–28 and Nu still can't size themMedium
  5. 5CET1 fell from 13.0% to 11.9% in six monthsMedium
  6. 6Private payroll loans now get repaid before Nu's cardsMedium
  7. 7Mercado Pago and PicPay are growing credit faster than NuMedium
  8. 8A founder with 74% of votes, Monzo, and only insider sellingMedium
  9. 9Mexico deposits shrank 15% and credit quality is undisclosedMedium
  10. 10Revolving-card charges face court and political pressureLow

Bull arguments show how well they survive the devil’s advocate (strong / medium / weak); bear arguments show the damage they could do after the counter-argument (high / medium / low).

Bull case

🐂 Bull #1A $1 monthly cost to serve against $17 of revenue

  • The point: Nu's cost base is a fraction of incumbents'. That lets it price below them, pay more on deposits and still earn a 33% ROE.
  • Evidence:
    • Cost to serve is $1.0 per active customer a month, against ARPAC (average monthly revenue per active customer) of $17.1 [1].
    • The efficiency ratio (operating cost ÷ net revenue) has gone 21.4% → 21.3% → 20.3% → 19.9% → 17.6% → 19.5% over the six quarters to Q2-26 [4].
    • Q2 net revenue rose 8% q/q FX-neutral to $4,132M; opex was $806M [4].
    • Management guides "about 20%" for FY26. It said Q1's 17.6% was "not a run rate": about two-thirds of the dip was timing of real-estate and marketing spend [7].
    • Headcount is flat at about 10,400. AI agents handle more than 60% of Brazilian support chats [7].
  • Why it matters: each new product (payroll loans, insurance, the Croma "super core" card) can be profitable at prices rivals cannot match. The cost gap is also what lets Nu absorb a tax rise better than a branch bank.
  • 😈 Devil's advocate:
    • The ratio is no longer falling. Opex grew 20% q/q FX-neutral in Q2, faster than revenue's 8% [4].
    • The Q4-25 call warned of "upward pressure" over 4–6 quarters, with a return to office worth 80–100 bp (summary only) [53].
    • Cost to serve rose from $0.8 to $1.0 in a year [1].
    • Three bank-level regulators (Brazil after Porto Real, Mexico, the US) add supervision cost.
  • Verdict: Strong The gap to incumbents is several-fold, so even a ratio that drifts to 21–22% keeps the edge. The "ever-improving efficiency" part of the story has stalled, though.

🐂 Bull #2ARPAC up 37% in a year, still far below incumbents

  • The point: Nu already has the customers. Revenue now grows by selling them more, and that is cheap revenue.

  • Evidence:

    • ARPAC is $17.1 versus $12.5 a year earlier (+37%) [1].

    • Vélez cites incumbents at $40–45 but says Nu "doesn't expect to reach that level" [7].

    • Nu's own slide splits Brazil's ~$100B consumer and SME gross-profit pool into five segments [4]:

      Segment Pool Nu's position
      Mass market (144M adults) $30B Penetration bar about two-thirds
      Super core (12.4M adults, served by the new Croma card) $23B
      High income (5.3M adults) $21B
      SMEs (21M businesses) $18B
      Super-high income $8B Not served

  • Why it matters: ARPAC growth can carry revenue after Brazilian customer growth slows. Nu already has 118M Brazilian customers, about 62% of adults [1].

  • 😈 Devil's advocate:

    • About 60% of ARPAC is credit-related (Citi) [22].
    • Growing ARPAC has mostly meant lending more to the same people. That is the exposure behind Bear #1.
    • Customer saturation is real: Vélez says Nu will "run out of Brazilian customers" at some point [7].

  • Verdict: Strong. The runway is measurable, but it is bought partly with credit risk.

🐂 Bull #3Card volumes growing twice the market as the #2 issuer(new)

  • The point: in Brazil's largest consumer-credit product, Nu is taking share from the big banks every quarter.
  • Evidence:
    • In Q2-26 Nu's card volume was R$128.5bn, +23.5% y/y, a 15.2% share. That makes it #2 behind Itaú (R$206bn, +10.4%, 24.3%) [43].
    • Bradesco grew 8.8%, Banco do Brasil 5.5% and Santander 2.6% [43].
    • The market grew 11.4% (ABECS) [43].
    • The data is BTG/Card Monitor via Finsiders, so this is secondary coverage.
  • Why it matters: card spend drives interchange fees, the revolving and instalment book, and primacy. Taking share from the slowest incumbents is a cheap source of growth that does not require new risk segments.
  • 😈 Devil's advocate:
    • PicPay grew card volume 56% and C6 18.4%. Nu is not the only share-taker [43].
    • Cards are 66% of Nu's loan book (calc. from $26.0B of $39.4B) [1]. Card share gains are therefore also concentration gains.
    • The overdue >90-day card ratio rose from 6.1% to 6.3% in six months [2].
  • Verdict: Medium Strong. Hard data that the franchise is still winning, though this is not new risk-free growth.

🐂 Bull #4$45B of retail deposits at 88% of the interbank rate

  • The point: Nu funds itself with cheap retail deposits and lends out only part of them, so loan growth needs no wholesale money.
  • Evidence:
    • Deposits were $45.3B (+18% y/y FX-neutral): Brazil $36.4B, Mexico $5.7B, Colombia $3.3B [4].
    • Cost of deposits was 88% of the interbank rate, against 91% a year earlier [4].
    • Total borrowings are only $4.68B: financial bills $2.81B and a $1.87B margin-loan facility [2].
    • There are no financial covenants, and Nu Holdings guarantees none of the borrowings [2].
    • Q1 deck: available funding was $39.1B against a $22.8B net credit portfolio (1.7x) [5].
  • Why it matters: shifting deposits from government bonds (~13.75% Selic) into cards and loans that yield far more is a built-in margin lever. In a funding scare Nu does not depend on capital markets.
  • 😈 Devil's advocate:
    • 98.7% of deposits are contractually due within 12 months [2]. These are savings balances that move if a rival pays more.
    • Float is 34% of gross profit [2]. It shrinks with every Selic cut, and five cuts have taken the Selic from 15% to 13.75% [37].
    • $2.83B of borrowings mature within 12 months [2].
  • Verdict: Medium Strong. A real structural edge; part of today's profit is a high-rate gift that fades.

🐂 Bull #5Reserves at 244% of 90+ loans, built ahead of losses

  • The point: Nu provisions early. If losses rise, a cushion is already on the balance sheet, and its data-driven underwriting (NuFormer) lets it tighten faster than rivals.
  • Evidence:
    • Total coverage over 90+ NPL was 244% in Q2-26 and 252% in Q1-26. It has stayed in a 224–252% band since Q4-23 [4].
    • The allowance rose from $6,142M to $6,642M in Q2 [4].
    • Allowance build as a share of new 15+ NPL formation was 113% in Q2 and above 100% in every quarter since Q4-23 [4].
    • Nu says card 90+ delinquency improved in every income band since July 2025 while peer segments deteriorated, and that primary-bank customers default at about half the average [1].
    • NuFormer now uses 15 data sources and runs in production on three portfolios [4,7].
  • Why it matters: the reserves would absorb a moderate turn without a big earnings hole.
  • 😈 Devil's advocate:
    • Coverage measures the reserve against today's 90+ stock. If that stock grows, the ratio holds only if Nu keeps building.
    • Stage 3 loans (credit-impaired) jumped from 6.2% to 8.3% of the loan book in six months. Stage 3 coverage slipped from 67.2% to 66.2% [2].
    • The income-band claim is company-only data.
    • NuFormer has not been through a full Brazilian consumer bust.
  • Verdict: Medium Provisioning discipline is consistent and visible, but it buffers losses rather than preventing them.

🐂 Bull #6Mexico broke even in six years and is now a bank

  • The point: Mexico is less banked than Brazil, and a full bank licence lets Nu take salary deposits and lend more.
  • Evidence:

    • Nu had 16.0M customers by July 2026, 16.5% of adults [4].

    • Mexico broke even in Q1-26 [4,5]. Vélez: "six years in Mexico compared with 8 years in Brazil" [7].

    • The full bank launched on 2026-08-06 [15,2].

    • Mexico at Q2-26 compared with Brazil at the same penetration (Q2-20) [4]:

      Metric Mexico Q2-26 Brazil Q2-20
      ARPAC $12.3 $5.6
      Interest-earning portfolio $1.3B $0.6B
      Cost to serve $1.2 $1.7

    • Mexico revenue was $314.6M in Q2, +79% y/y [2].

    • Mexico's capital ratio is 14.9% against a 10.5% minimum [2].

    • Management's base case is a business "60%, 70% of Brazil" [7].

  • Why it matters: a second profit pool would cut Brazil dependence (Bear #3).
  • 😈 Devil's advocate:
    • Deposits fell from $6.7B to $5.7B y/y in dollars, "by design" [4].
    • The loan-to-deposit ratio is still only 35% [1].
    • Mexico is 7.2% of revenue [2].
    • Nu does not publish Mexico NPLs. See Bear #9.
  • Verdict: Medium The trajectory is good and the licence is done, but profit is tiny and credit data is opaque.

🐂 Bull #7Half of the $1B buyback done at ~$12.31; diluted shares flat

  • The point: Nu now generates more capital than it needs and returns some of it. Stock pay is no longer diluting holders.
  • Evidence:
    • The $1.0B program was authorised 2026-06-04, valid to 2027-06-03 [8].
    • By 06-30 Nu had bought 40.66M shares for $500.4M, an average of about $12.31 (calc.). $499.6M remains [2].
    • Weighted diluted shares were 4,904.8M in Q2-26 against 4,899.8M a year earlier (+0.1%) [2].
    • Equity rose $1.93B in H1 to $13.25B, after the buyback [2].
  • Why it matters: with the stock at $16.12 [51], the shares bought so far are already ~31% above cost. A second program at Investor Day (12-08) is plausible.
  • 😈 Devil's advocate:
    • SBC was $134.4M in Q2 (+33% y/y), about 12.7% of net income [2]. RSUs and options total 87.0M (~1.8%) [2].
    • The repurchased shares sit in treasury and are not cancelled [2].
    • Capital leaving Brazil needs Central Bank approval: $130M of excess was "under approval by BCB" in Q1 [5].
    • Brazilian CET1 is falling (Bear #5).
    • No purchases after 06-30 have been disclosed.
  • Verdict: Medium Shareholder-friendly and well-timed, but limited by Brazilian capital rules.

🐂 Bull #816.6x forward earnings for a 33% ROE, below its own history(new)

  • The point: after a year of downgrades the multiple has compressed. Even after the October rally, the stock is not priced for perfection.
  • Evidence:

    • At $16.12 (2026-10-09) NU trades at 16.6x forward earnings and 5.9x book, on a 31.6% ROE [17].

    • Peer multiples [17]:

      Company Forward P/E P/B ROE
      Itaú 10.9x 2.6x 21.5%
      MercadoLibre 44.8x 12.2x 27.5%
      SoFi 21.7x 1.8x 7.1%

    • Trailing P/E was 38x in 2023 and about 33x in late 2025, against about 22x now [17,50].

    • Consensus EPS is $0.89 for FY26 (9 analysts) and $1.13 for FY27 [18].

    • Short interest is about 3.6% of float, 1.7 days to cover [20].

  • Why it matters: at ~14x FY27 consensus, the stock does not need multiple expansion. EPS growth alone can carry it if credit holds.
  • 😈 Devil's advocate:
    • "Cheap versus its own history" means little when the history was a growth-bubble multiple.
    • On P/B Nu is 2.3x Itaú. That premium exists only if a ~30% ROE is durable, and Citi argues the sustainable level is 25% [22].
    • Itaú BBA's FY27 earnings estimate is 15% below consensus [21].
  • Verdict: Medium Not expensive for the returns, but the valuation is only as good as the credit cycle.

🐂 Bull #9US entry capped at under one point of efficiency

  • The point: the US launch is a call option. Management caps its cost, so the downside is small.
  • Evidence:
    • Launched 2026-09-10 via partner Lead Bank: 3.50% APY savings, a 1.5% cashback card, and "Nu Global" USDC/EURC accounts [11].
    • The OCC gave preliminary conditional approval for Nubank N.A. on 2026-01-29. FDIC and Fed approvals are still pending [6].
    • The spend cap is under 100 bp of the efficiency ratio in 2026 and 2027 [5,7].
    • Citi's widely cited "$21B by 2030" is a loan-book size, not a business value. It assumes 2% share in California, Texas and Florida, and Citi doubted Nu could match Brazilian profitability [30]. Report #1 described it loosely.
  • Why it matters: if it works it adds a new market. If it fails, the cost is capped.
  • 😈 Devil's advocate:
    • The US has no expensive incumbent oligopoly to undercut.
    • Nu has no US credit data and needs 12–30 months to build models [7].
    • No traction numbers have been published.
  • Verdict: Weak as a value driver before 2028. It matters mainly if the cap is breached.

Bear case

🐻 Bear #1Brazilian household delinquency hit a record as Nu's 90+ rose

  • The point: Nu's book is 92% unsecured (cards and personal loans) to mass-market Brazilians, and system-wide household credit is deteriorating fast.
  • Evidence:

    • Brazilian household delinquency, BCB data released 2026-09-29 [35,36]:

      Series Jun Jul Aug
      Household free-credit delinquency 7.37% 7.78% 8.01%
      Total household delinquency 5.42% 5.80% 6.03%
      • The 8.01% free-credit figure is a record since the series began in March 2011.
      • Card delinquency is 10%. Revolving card interest is 444.9% a year [36].
      • Correction to report #1: the 4.58% → 5.04% series quoted there was total system delinquency (SGS 21082), not household.

    • Nu's own credit ratios (90+ NPL = loans more than 90 days late; 15–90 = early delinquency) [1,4,5]:

      Quarter 90+ NPL 15–90 NPL
      Q2-25 6.5% 4.5%
      Q1-26 6.5% 5.0%
      Q2-26 6.9% 4.8%
      • Q2-26 is the highest 90+ reading since Q3-24 (7.0%) [5].

    • Loan Stage 3 rose from 6.2% to 8.3% in six months [2].

    • The CFO conceded that "the general trend is upwards, and that's being driven by the mix" [7].

    • Secured loans are only $3.1B, 8% of the $39.4B book [2].

    • BCB says about half of the system rise comes from an accounting change (CMN Res. 4,966) [36].

  • Why it matters:
    • Cost of credit ($1.69B in Q2, managerial) is Nu's largest cost line [1].
    • On a $48B average 2027 book (estimate), every 1 point of extra cost of risk is about $480M pre-tax, or ~$0.06 of EPS (6% of FY27 consensus).
  • 🛡️ Counter-argument:
    • The 15–90 ratio, the leading indicator, fell q/q from 5.0% to 4.8% [1].
    • Coverage is 244% [4].
    • Half the system rise is accounting [36].
    • Nu's card 90+ improved by income band while peers' deteriorated (company data) [1].
    • The Selic is falling, which eases debt service. Household debt service was 28.7% of income in July [35].
  • Severity: High This is still the variable that moves earnings most, and the macro data moved the wrong way since report #1.

🐻 Bear #2The Q2 margin jump leaned on Desenrola and chosen risk(sharpened)

  • The point: the leap in risk-adjusted net interest margin (NIM after credit losses) from 9.5% to 12.4% beat management's own target early. Part of the beat will not repeat, and part was bought with extra risk.
  • Evidence:
    • The Q1 guide was to get back to H2-25 levels (10.5–10.8%) by end-2026. Q2 came in at 12.4% [7].
    • The CFO said about one-third of the beat versus expectations came from Desenrola. That is a government debt-renegotiation program; it "just about 5%" lowered cost of credit [7].
    • More than 4/5 of Desenrola's effect landed in Q2, with a little more in Q3 [7].
    • Of the $500M Q2 allowance build, $170M was booked for "intentional risk expansions", meaning deliberate loosening of underwriting [4].
    • Those expansions added +24 bp to the 15–90 NPL, partly offsetting −37 bp of seasonality [4].
    • On whether 12% is a floor, the CFO said: "I didn't say that it was a floor" [7].
    • The Q1 miss (net income $871M vs ~$918–936M consensus) shows how fast the margin moves [29].
  • Why it matters: consensus EPS of $1.13 for FY27 (+27%) assumes the margin holds "in the same region" [18,7].
    • Itaú BBA (09-16) sees FX-neutral earnings growth of only 14% in 2027 [21].
  • 🛡️ Counter-argument:
    • Two-thirds of the beat was underlying performance and balance growth [7].
    • Risk expansion is priced: the margin is measured after losses.
    • Lower rates cut funding cost as well as float income.
  • Severity: Medium High. Q3 (~11-12) shows how much was one-off.

🐻 Bear #3A one-country stock up 27% into a close runoff

  • The point: about 91% of revenue is Brazilian. The stock trades as a leveraged bet on Brazil, and it has just priced in a market-friendly election result.
  • Evidence:

    • Brazil was $3,990M of $4,381M Q2 segment revenue (91.1%) [2].

    • First round, 2026-10-04: Flávio Bolsonaro 47.03%, Lula 45.16%. The runoff is 2026-10-25 [32].

    • Runoff polls since then [31,32]:

      Poll Flávio Lula
      Datafolha (10-06/07) 49% 45%
      AtlasIntel 51.1% 45.7%
      Ipespe 49% 44%
      Vox Brasil 42.7% 44.2%

    • Before the first round, Datafolha had Lula ahead 48–45 [31].

    • NU rose from $12.66 (09-30) to $16.12 (10-09), +27% [27,51].

    • USD/BRL fell from 5.22 to 4.99 [35]. The Ibovespa rose 11.2% on the week [33].

  • Why it matters: a Lula win would likely reverse part of the BRL move and the rally. Every 1% move in BRL changes dollar EPS by about 1%.
  • 🛡️ Counter-argument:
    • Most polls show Flávio ahead, and the first round beat the polls.
    • Funding is in local currency, so the FX hit is translation only.
    • Nu grew through both Lula and Bolsonaro governments.
  • Severity: Medium High. The binary outcome on 10-25 is now one-sided in the price.

🐻 Bear #4Taxes rise in 2027–28 and Nu still can't size them

  • The point: Brazil is closing the tax gap between fintechs and banks. Two legislated increases arrive in 2027–28, and Nu gives no number for either.
  • Evidence:

    • CSLL (a federal profit tax) under LC 224 [2,38]:

      Entity type 2026–27 From 2028
      Payment institutions 12% 15%
      Credit companies (SCFI) 17.5% 20%
      Banks 20% 20%

    • Nu's 2026 combined Brazilian statutory rate is 42.5% [2]. Which Nu entity books how much profit is not disclosed. After Porto Real, the bracket is not stated.

    • IBS/CBS (the new consumption taxes) apply to financial services from 2027-01-01, at about 10.85% initially and rising to 12.5% by 2033 [2,39]. Nu says the impact is "currently under assessment" [2]. No analyst asked about tax on the Q2 call [7].

    • Revenue-based taxes in the managerial accounts rose from about 2.9% of revenue (Q1/Q2-25) to 4.1% (Q1/Q2-26) [3]. The filings give no explanation.

    • The reported (IFRS) tax rate was 14.2% in Q2, against 34.9% on the managerial basis [1,3]:

      • $671M of deferred tax benefit offset $846M of current tax [2].
      • The deferred tax asset reached $3.65B, 28% of equity [2].
  • Why it matters: each point of extra tax rate on ~$8.5B of FY27 pre-tax profit (estimate) is ~$85M, or ~$0.017 a share.
  • 🛡️ Counter-argument:
    • Banks already pay more, so a level playing field removes a political target.
    • IBS/CBS replaces PIS/COFINS/ISS that Nu already pays, so the net change could be small.
    • Flávio campaigns on revising the tax reform [34].
  • Severity: Medium The direction is certain; the size is unknown until Investor Day at the earliest.

🐻 Bear #5CET1 fell from 13.0% to 11.9% in six months(new)

  • The point: fast loan growth is using up regulatory capital in Brazil. That limits buybacks and could force slower growth.
  • Evidence:

    • Brazil prudential conglomerate, end-2025 to Q2-26 [2]:

      Measure Dec-25 Q2-26
      CET1 13.0% 11.9%
      Total capital ratio 16.6% 15.7%
      Excess over minimum $1.89B $1.85B
      Risk-weighted assets $31.1B $35.7B (+15%)

    • The Q1 capital stack was $9.4B, including $3.7B of holding-company cash [5].

    • Colombia's capital ratio fell from 16.9% to 15.3% [2].

  • Why it matters: if RWA keeps growing ~30% a year, Brazilian capital needs to be topped up from holding cash or retained earnings. That competes with buybacks and M&A.
  • 🛡️ Counter-argument:
    • Group equity/assets is 16% (6.2x leverage), very low for a bank [2].
    • $1.06B of quarterly profit adds capital fast.
    • The Porto Real licence brings "no additional capital or liquidity requirements" [13].
  • Severity: Medium Low. A constraint on capital returns, not a solvency issue.

🐻 Bear #6Private payroll loans now get repaid before Nu's cards

  • The point: Brazil's 2025 private-sector payroll loan ("Crédito do Trabalhador") is deducted from salary first. Nu's unsecured loans are therefore effectively behind it in the queue, and Nu is small in that product.
  • Evidence:
    • Citi's June downgrade (to Neutral, target cut from $18 to $13) [22,23]:
      • Payroll loans "effectively subordinate" cards.
      • About 9% of Nu's credit customers already have such loans elsewhere.
      • Citi cut its sustainable ROE to 25% from 30%.
    • Private-payroll delinquency reached 10.6% in August, from 5.2% a year earlier (secondary, citing BCB) [36].
    • PicPay holds R$7.2bn of private payroll, a 6.4% BCB market share [41]. That implies a ~R$112bn market (calc.).
    • Nu's private-payroll balance and share are not disclosed. Vélez expects to be "one of the leading players… over the next 18, 24 months" [7].
  • Why it matters: if borrowers' salaries are pre-empted, card and personal-loan losses rise without any change in Nu's own underwriting.
  • 🛡️ Counter-argument:
    • Nu's slowness avoided the worst early defaults. Early adopters were "very high risk" [7].
    • Payroll lending is still small relative to card balances.
    • If Nu becomes a lead payroll lender, the subordination works in its favour (see tailwinds).
  • Severity: Medium

🐻 Bear #7Mercado Pago and PicPay are growing credit faster than Nu

  • The point: Nu's next growth areas are contested by rivals with marketplace data or aggressive pricing.
  • Evidence:
    • Mercado Pago's gross loans (all countries) rose from $12.5B to $16.4B in six months (+31%) [42].
    • PicPay's credit book doubled to R$31.9bn (+99% y/y), and its card volume grew 56% [41,43]. PicPay (PICS) is also a holding in this portfolio.
    • Nu's book grew 37% y/y FX-neutral [1].
  • Why it matters: competition could cap Nu's pricing in cards and payroll, and rivals' aggressive lending raises system leverage on Nu's customers.
  • 🛡️ Counter-argument:
    • Rivals are buying growth with worse credit [41,42]:
      • Mercado Pago: 18.7% of its gross book is 91+ days past due (calc.). Its operating margin fell to 6.7%.
      • PicPay: 90+ NPL of 9.8% against Nu's 6.9%.
    • Mercado Pago still has no Brazilian bank licence [42].
  • Severity: Medium Real competition, but so far it is costing rivals more than Nu.

🐻 Bear #8A founder with 74% of votes, Monzo, and only insider selling

  • The point: outside shareholders cannot check management, and 2026 brought three surprises.
  • Evidence:
    • Vélez holds about 74.4% of the votes [6].
    • The 2026 surprises:
      • The CFO change (06-01). BofA cut to Underperform with a $10 target [9,24].
      • Reports of a £8–10B Monzo approach. NU fell 8–10% on 09-28 before Nu said it is "not pursuing a transaction" [10,27].
      • Vélez joined OpenAI's boards on 2026-07-21 [16].
    • Form 4s from 03-25 to 10-07 show no open-market purchases by any insider [14]:
      • Junqueira sold 465k shares ($6.9M), partly under a 10b5-1 plan adopted 03-27.
      • CRO Fragelli sold 221,707 shares at $15.81 on 08-14.
      • Director Sands sold 21,000 shares at $12.24.
      • Vélez's only filings are tax-withholding events.
  • Why it matters: a large stock-funded deal could dilute holders by high single digits. A CEO with an OpenAI board seat may have less bandwidth.
  • 🛡️ Counter-argument:
    • The Monzo denial came within days, which shows pushback works.
    • The sales are small against holdings.
    • The buyback (Bull #7) is real capital returned.
  • Severity: Medium Low.

🐻 Bear #9Mexico deposits shrank 15% and credit quality is undisclosed(new)

  • The point: the "second engine" is opaque. The only Mexico credit figures available are secondary and poor.
  • Evidence:
    • Mexico deposits went $6.7B → $6.1B → $6.3B → $5.9B → $5.7B over the five quarters to Q2-26 [4].
    • Mexico's interest-earning portfolio is $1.3B [4].
    • XP data via a Brazilian outlet (secondary; not in Nu's filings) [48]:
      • Mexico Q2 net income about $7M.
      • Cost of risk 26.3%.
      • NPL 11.3%.
      • That source's Q1 profit figure ($14M) conflicts with Nu's own "break-even" slide, so treat it with caution.
  • Why it matters: if Mexican NPLs are really ~11%, scaling the loan book through the new bank could cost more than the market assumes.
  • 🛡️ Counter-argument:
    • Mexico ARPAC is twice Brazil's at the same stage [4].
    • High cost of risk sits alongside a ~33% NIM (same secondary source) [48].
    • Mexico is 7% of revenue, so a miss is small at group level.
  • Severity: Medium Low.

🐻 Bear #10Revolving-card charges face court and political pressure

  • The point: Brazil's card rates are a constant political target, and Nu is the second-largest card issuer.
  • Evidence:
    • The Federal District's public civil action is in the 7th Public Treasury Court [45,46]:
      • It seeks suspension of charges above 100% of original debt and R$10m of collective damages.
      • The trigger case was a debt that grew from R$12.3k to R$788k.
      • The court gave Nu 15 days to respond in July. No injunction ruling has been found since.
    • The government asked the Central Bank and Finance Ministry in February to study cuts to revolving rates [34].
    • Vélez warned that a pricing cap "is going to hurt more the players that were too fast" [56].
    • The 20-F lists possible-loss civil and labour claims of only about $4.4M [6].
  • Why it matters: the money at stake is the precedent, not the R$10m. A national rule on revolving or interest-free instalments would hit Nu's largest product.
  • 🛡️ Counter-argument:
    • The 100% cap on revolving charges has been in force since 2024 [6].
    • A right-leaning Congress and a likely Flávio win lower the odds of new caps.
  • Severity: Low Medium.

Tailwinds NOT in the forecasts

Consensus: FY27 EPS $1.13 and revenue ~$28.2B [18]. The revenue figure comes from a page with a visible data error, so treat it as approximate. Diluted shares are 4.905B [2]. FY27 pre-tax profit is about $8.5B (estimate, from $1.13 × 4.905B ÷ 0.65).

  • BRL stays below 5.00 after a Flávio Bolsonaro win — 40% chance through FY27

    • Impact on revenue: +4.6% / about +$1.3B vs consensus FY2027
    • Impact on EPS: +$0.052 per share / +4.6% vs consensus FY2027
    • How we got the number (estimate):
      • Consensus is assumed built at roughly the Aug–Sep average of 5.18 BRL/USD [35]; the consensus FX assumption was not found.
      • 5.18 ÷ 4.95 − 1 ≈ 4.6% translation uplift; $1.13 × 4.6% ≈ +$0.052.
    • Expected value: 0.40 × $0.052 ≈ +$0.021
    • What it is: a market-friendly government keeps the real strong and lets the Selic fall faster. That lifts dollar earnings and lowers borrowers' debt-service load.
    • Why it's not priced in: analysts update FX slowly. The stock has priced the election odds, but most estimates predate the 10-05 move.
    • Trigger to watch: runoff 10-25; USD/BRL on 10-26; Copom 11-04.
    • Why this probability: polls give Flávio roughly a 60–65% chance, and the real also has to hold its gains for a year.

  • Nu becomes a top private-payroll lender by 2027 — 30% chance by end-2027

    • Impact on revenue: +$0.84B / +3.0% vs consensus FY2027
    • Impact on EPS: +$0.057 per share / +5.0% vs consensus FY2027
    • How we got the number (estimate): $2B of extra payroll book × (~42% yield − ~12% funding − ~8% losses) × (1 − 35% tax) ÷ 4.905B ≈ +$0.057. The revenue figure is $2B × 42%.
    • Expected value: 0.30 × $0.057 ≈ +$0.017
    • What it is: with a R$112bn market [41], a 10% share is about R$18bn ($3.6B). Our estimate assumes about $2B of it is above consensus. Payroll also moves Nu to the front of the repayment queue (the mirror of Bear #6).
    • Why it's not priced in: Nu discloses no balance, so analysts cannot model it.
    • Trigger to watch: a payroll balance at Q3 or Investor Day; Nu's share in BCB data.
    • Why this probability: Vélez targets leadership within 18–24 months, but PicPay already holds 6.4% and delinquency is 10.6%.

  • AI pushes the efficiency ratio to ~18% in FY27 — 30% chance

    • Impact on revenue: none (cost effect)
    • Impact on EPS: +$0.053 per share / +4.7% vs consensus FY2027
    • How we got the number (estimate): 2 points × ~$20B FY27 net revenue (Q2 $4.13B × 4 × ~1.2 growth) = $400M × 0.65 ÷ 4.905B ≈ +$0.053.
    • Expected value: 0.30 × $0.053 ≈ +$0.016
    • What it is: flat headcount, AI agents doing >60% of support, and NuFormer underwriting.
    • Why it's not priced in: management guides "about 20%" and warns of upward pressure.
    • Trigger to watch: an efficiency target at Investor Day; Q3/Q4 ratio.
    • Why this probability: Q1 showed 17.6% is possible, but Q2 opex rose 20% q/q.

  • Mexico salary accounts lift loan-to-deposit from 35% to 50% — 35% chance by end-2027

    • Impact on revenue: +$0.38B / +1.4% vs consensus FY2027
    • Impact on EPS: +$0.015 per share / +1.3% vs consensus FY2027
    • How we got the number (estimate): 15 points × $5.7B deposits = $0.86B extra loans × ~13% net spread over TIIE after losses × 0.65 ÷ 4.905B ≈ +$0.015. Revenue is $0.86B × ~45% gross yield.
    • Expected value: 0.35 × $0.015 ≈ +$0.005
    • What it is: payroll direct deposits make lending cheaper and safer.
    • Why it's not priced in: Nu gives no Mexico targets; Investor Day may.
    • Trigger to watch: Mexico deposits turning up in Q3; the payroll-deposit count.
    • Why this probability: the licence is in hand, but deposits are still shrinking "by design".

  • Second buyback of $1B at Investor Day — 50% chance by mid-2027

    • Impact on revenue: none
    • Impact on EPS: +$0.007 per share / +0.6% vs consensus FY2027
    • How we got the number (estimate): $1B ÷ $16 = 62.5M shares (+1.3%) ≈ +$0.015, less ~$40M of lost interest on holding cash (−$0.008).
    • Expected value: 0.50 × $0.007 ≈ +$0.004
    • What it is: a renewal once the current program completes.
    • Why it's not priced in: EPS consensus assumes no share-count shrinkage.
    • Trigger to watch: Q3 6-K buyback disclosure; Investor Day 12-08.
    • Why this probability: half the program went in four weeks, but CET1 is falling (Bear #5).

  • US business scales — 10% chance of more than 5M accounts by 2029

    • Impact on revenue / EPS: not quantifiable. There are no traction numbers, and costs are capped through 2027, so there is no FY27 upside.
    • Expected value: n/a
    • What it is: the call option in Bull #9.
    • Why it's not priced in: no consensus model includes US revenue.
    • Trigger to watch: US account numbers at Investor Day; FDIC/Fed approval.
    • Why this probability: a saturated market with no Nu data edge.

Headwinds NOT in the forecasts

  • Brazilian credit turn lifts Nu's cost of risk 1.5 points in 2027 — 30% chance

    • Impact on revenue: −$0.8B / −3% vs consensus FY2027 (estimate: tighter underwriting slows loan growth)
    • Impact on EPS: −$0.095 per share / −8.4% vs consensus FY2027
    • How we got the number (estimate): 1.5% × ~$48B average 2027 book = $720M × 0.65 ÷ 4.905B ≈ −$0.095.
    • Expected value: 0.30 × −$0.095 ≈ −$0.029
    • What it is: a record household delinquency, payroll pre-emption and Nu's deliberate risk expansion combine.
    • Why it's not priced in: consensus FY27 EPS of +27% assumes the margin stays near 12% [18,7].
    • Trigger to watch: the Q3 15–90 NPL rising y/y above 4.8%; BCB household free-credit delinquency above 8%; Stage 3 above 9%.
    • Why this probability: delinquency is at a record, but half of it is accounting, rates are falling, and Nu's leading indicator improved q/q.

  • IBS/CBS and CSLL add 2 points of tax drag by FY27–28 — 45% chance

    • Impact on revenue: none on reported revenue (the tax is netted below revenue)
    • Impact on EPS: −$0.050 per share / −4.4% vs consensus FY2027
    • How we got the number (estimate): a 1-point rise in revenue-based tax ($28B × 1% × 0.65 = $182M) plus a 1-point rise in the income-tax rate ($8.5B × 1% = $85M) = $267M ÷ 4.905B ≈ −$0.054. Rounded to −$0.050 because part of the effect only arrives in 2028.
    • Expected value: 0.45 × −$0.050 ≈ −$0.023
    • What it is: a new consumption tax on financial services from 2027, and the CSLL step-up in 2028.
    • Why it's not priced in: Nu says "under assessment", and no analyst asked about it on the call.
    • Trigger to watch: 2027 tax guidance at Investor Day; the Q4-26 managerial tax rate.
    • Why this probability: the laws are passed. The net effect after replacing PIS/COFINS/ISS is genuinely unknown.

  • Lula wins the runoff — 37% chance on 2026-10-25

    • Impact on revenue: −$1.3B / −4.5% vs consensus FY2027
    • Impact on EPS: −$0.051 per share / −4.5% vs consensus FY2027
    • How we got the number (estimate): BRL reverts from 4.99 to ~5.22, the pre-first-round level: −4.5% translation × $1.13.
    • Expected value: 0.37 × −$0.051 ≈ −$0.019
    • What it is: a reversal of the October rally, plus a higher Brazilian risk premium and possible credit-policy activism.
    • Why it's not priced in: the stock rose 27% on the first-round result.
    • Trigger to watch: the final Datafolha and Quaest polls; the 10-25 result.
    • Why this probability: five of six post-first-round polls show Flávio ahead by 4–6 points, but the 2022 runoff was decided by under 2 points.

  • A large stock-funded acquisition — 15% chance within 36 months

    • Impact on revenue: not quantifiable (it depends on the target)
    • Impact on EPS: −$0.087 per share / −7.7% from dilution before any target profit
    • How we got the number (estimate): $6B of stock at $16 = 375M new shares = +7.7% on 4.905B. The target's profit is not quantifiable.
    • Expected value: 0.15 × −$0.087 ≈ −$0.013 (gross of target earnings)
    • What it is: a Monzo-style deal, reported at £8–10B.
    • Why it's not priced in: Nu says its capital-allocation framework is "unchanged" [10].
    • Trigger to watch: press reports; Investor Day M&A comments.
    • Why this probability: the appetite was revealed, then withdrawn after a 10% share-price drop.

  • A national cap on revolving rates or interest-free instalments — 12% chance by end-2027

    • Impact on revenue / EPS: not quantifiable. Nu does not disclose the revolving share of its card balances.
    • Expected value: n/a
    • What it is: the DF case becomes a precedent, or a populist post-election rule.
    • Why it's not priced in: no draft rule exists.
    • Trigger to watch: a DF injunction ruling; any CMN agenda item on revolving credit.
    • Why this probability: the 2024 cap is already in force, and a right-leaning Congress lowers the odds.

Investment Risk Score: 33/100 (moderate)

# Factor Weight Measured value Score Why
1 Liquidity & funding runway 15% Profitable (Q2 net income $1,061M). Cash $13.55B + securities ~$19.2B + compulsory deposits $9.15B vs deposits $45.3B (98.7% due <12 months). Borrowings $4.68B ($2.83B due in 12 months), no covenants. No holdco debt found. Holdco cash $3.7B (Q1). 25 Self-funding. The residual risk is deposit flight, a bank-specific risk that "net cash" anchors understate.
2 Leverage 10% Equity/assets 16% (6.2x). CET1 11.9% (13.0% at Dec-25). Total capital 15.7%. Excess capital $1.85B. Net debt/EBITDA not meaningful for a bank. 25 Lightly levered for a bank, but capital headroom is shrinking.
3 Dilution (last 12 months + overhang) 10% Diluted weighted shares +0.1% y/y (4,899.8M → 4,904.8M). RSUs and options 87.0M (~1.8%). SBC $134M a quarter. Buyback 40.7M shares. 25 Flat after the buyback; the overhang is small.
4 Profitability 10% ROE 33%. Net income / managerial revenue 18.1% (Q2), 16.4% (Q1). LTM net income $3.61B. 10 Highly profitable. FCF is not meaningful for a bank, so net margin and ROE are used.
5 Revenue quality & concentration 10% No customer ≥10% of revenue. 91.1% of revenue from Brazil. 61% of managerial revenue is credit income, 25% float. Cards 66% of the book. 35 No customer risk, but one country and one product (cards) dominate.
6 Execution & track record 10% NIM target beaten early; Mexico break-even and bank launch delivered; efficiency in line. EPS vs consensus: about 3 beats and 2 misses in 7 quarters. No material weakness; KPMG unchanged. Surprise CFO change; Monzo episode. 35 Mostly hits its own targets. Quarterly earnings are less predictable than the guidance record suggests.
7 Competitive position & disruption 10% #2 card issuer (15.2% share, +23.5% vs market +11.4%). Cost to serve $1. Mercado Pago loans +31% in 6 months; PicPay doubling. 30 A strong, still-widening position; rivals are buying growth with worse credit.
8 Cyclicality & commodity/macro sensitivity 10% 92% unsecured book. Household free-credit delinquency at a record 8.0%. Nu 90+ 6.9%, rising. Float 34% of gross profit tied to the Selic. ~1% EPS per 1% move in BRL. 62 Highly sensitive to the Brazilian consumer cycle, rates and the currency.
9 Regulatory, legal & geopolitical 5% Binary runoff on 10-25. Unquantified 2027 IBS/CBS and 2028 CSLL increases. Revolving-card lawsuit. Porto Real approval pending. US Section 301 action names Pix. 60 Material, with several pending decisions, but none existential.
10 Valuation & expectations 10% Forward P/E 16.6x vs Itaú 10.9x; P/B 5.9x. Trailing P/E ~22x vs 33–38x in 2023–25. Short interest 3.6% of float. Consensus assumes a ~12% margin holds. 45 Rich against Brazilian banks and fair against its own growth. The election rally has added expectation.

Overall: weighted average = 0.15×25 + 0.10×25 + 0.10×25 + 0.10×10 + 0.10×35 + 0.10×35 + 0.10×30 + 0.10×62 + 0.05×60 + 0.10×45 = 3.75 + 2.5 + 2.5 + 1.0 + 3.5 + 3.5 + 3.0 + 6.2 + 3.0 + 4.5 = 33.45 → 33.

Change vs last: −11 (44 → 33).

  • Method (most of it): the old method put 25% on business and competition and 15% on concentration. The 10-factor method gives separate weight to liquidity, dilution and profitability, where Nu scores 10–25.
  • Substance, risk up:
    • Record household delinquency.
    • The $170M risk-expansion build.
    • CET1 down 110 bp.
    • The stock rallied into the runoff.
  • Substance, risk down:
    • Flat diluted share count.
    • Measured card-share gains.
    • Rivals' credit visibly worse than Nu's.

Ranking check: at 33, NU sits between MNTN (32) and AVGO (34). It is as financially strong as either, but its single-country unsecured-credit cyclicality offsets their concentration and competition risks. It is clearly safer than VIST (41), the other one-country Latin American holding, which also carries commodity risk.

Conclusion

  • What you are really betting on: the Brazilian mass-market consumer staying solvent through 2027. Nu's cost and funding edge (Bulls #1, #3, #4) is proven. Whether it turns into +27% EPS growth depends almost entirely on credit losses and on the real.
  • The 2–3 things that will decide the outcome:
    1. Whether Q3 and Q4 risk-adjusted margin holds near 12% once Desenrola is gone and the risk expansion seasons.
    2. The 10-25 runoff and the BRL/Selic path after it.
    3. Whether Investor Day (12-08) quantifies the 2027 tax hit and sets credible Mexico and payroll targets.
  • Signposts:
    • Bull confirmed:
      • Q3 15–90 NPL at or below 4.5% with a margin ≥ 11.5%.
      • Mexico deposits turn up after the bank launch.
      • A payroll balance disclosed at more than $1B.
      • A second buyback announced while CET1 holds above 11.5%.
    • Bull broken:
      • Q3 90+ NPL above 7.2% or Stage 3 loans above 9%.
      • A 2027 tax guide implying a managerial rate above 37%.
      • A national revolving-rate rule or a DF injunction.
      • A stock-funded deal above $5B.
  • New since the last report:
    • Nu's own Q2 bridge shows $170M of the $500M reserve build was deliberate risk expansion. The record 8.0% household free-credit delinquency replaces report #1's mislabelled 5.04%.
    • Brazil CET1 fell 110 bp in six months as RWA grew 15%. Diluted share count was flat, and buyback shares were bought at ~$12.31.
    • Card volumes grew +23.5% vs the market's +11.4%, at a 15.2% share. Rivals' credit is visibly worse (Mercado Pago 18.7% 91+ past due, PicPay 90+ at 9.8%). The stock has rallied 27% on a runoff Flávio leads in five of six polls.

Open questions & data gaps

  • Q3-26 results (~2026-11-12; date from aggregators, not confirmed by Nu): the decisive data point for Bears #1 and #2. Not available yet.
  • Mexico NPL and profit: only a secondary XP-based source (NPL 11.3%), which conflicts with Nu's Q1 figures. Nu's deck gives no Mexico credit metrics. This matters for Bull #6 and Bear #9.
  • Tax:
    • The managerial revenue-based tax rate rose from ~2.9% to ~4.1% of revenue. The cause was not found.
    • Not found: the profit split by entity (payment institution / SCFI / bank) and the CSLL bracket after Porto Real.
    • This matters for the size of Bear #4.
  • Private payroll: Nu's balance and market share are not disclosed. The Morgan Stanley note was not found. The two private-payroll delinquency series (8.6% June vs 10.6% August) come from different sources and may measure different things.
  • Consensus: the FY28 consensus and the FX, NPL and tax assumptions behind FY26–27 were not found. The FY27 revenue consensus comes from a page with a visible data error.
  • Other gaps:
    • The buyback executed after 06-30 is not disclosed until Q3.
    • The US FDIC/Fed approval status is unknown.
    • BCB approval of Porto Real has not been found, ahead of the ~late-November Joint Resolution 17 deadline.
    • The Q3-25 and Q4-25 call promises are available only in summary.
  • The EPS beat/miss record: aggregator figures conflict for Q2-25 and Q4-25. The "3 beats, 2 misses" summary is approximate.
Sources
  1. Nu Holdings Ltd. Reports Second Quarter 2026 Financial Results (6-K) — Nu Holdings / SEC — 2026-08-13 — https://www.sec.gov/Archives/edgar/data/1691493/000129281426004222/nupr2q26_6k.htm
  2. Unaudited Interim Condensed Consolidated Financial Statements, 30 June 2026 (6-K) — Nu Holdings / SEC — 2026-08-13 — https://www.sec.gov/Archives/edgar/data/1691493/000129281426004225/nufs2q26_6k.htm
  3. Managerial P&L Reconciliation Report Q2'26, with KPMG limited assurance (6-K) — Nu Holdings / SEC — 2026-08-13 — https://www.sec.gov/Archives/edgar/data/1691493/000129281426004228/nu20260813_6k1.htm
  4. Q2'26 Earnings Presentation (6-K, slides 7–9, 14, 18–19, 22–23 read) — Nu Holdings / SEC — 2026-08-13 — https://www.sec.gov/Archives/edgar/data/1691493/000129281426004226/nu20260813_6k.htm
  5. Q1 2026 Earnings Presentation (6-K) — Nu Holdings / SEC — 2026-05-14 — https://www.sec.gov/Archives/edgar/data/1691493/000129281426003055/nu20260514_6k.htm
  6. Form 20-F for FY2025 — Nu Holdings / SEC — 2026-04-08 — https://www.sec.gov/Archives/edgar/data/1691493/000129281426002166/nuform20f_2025.htm
  7. Nu (NU) Q2 2026 Earnings Call Transcript (third-party transcript) — The Motley Fool — 2026-08-20 (call 2026-08-13) — https://www.fool.com/earnings/call-transcripts/2026/08/20/nu-nu-q2-2026-earnings-call-transcript/
  8. Nu Holdings Announces US$1.0 Billion Share Repurchase Program (6-K) — Nu Holdings / SEC — 2026-06-04 — https://www.sec.gov/Archives/edgar/data/1691493/000129281426003326/nu20260604_6k.htm
  9. Nubank appoints Rob Livingston as CFO (6-K) — Nu Holdings / SEC — 2026-06-01 — https://www.sec.gov/Archives/edgar/data/1691493/000129281426003267/nu20260601_6k.htm
  10. Statement Regarding Media Reports (Monzo) (6-K) — Nu Holdings / SEC — 2026-09-30 — https://www.sec.gov/Archives/edgar/data/1691493/000129281426004752/nu20260930_6k.htm
  11. Notice to the Market: US launch and Nu Global (6-K) — Nu Holdings / SEC — 2026-09-10 — https://www.sec.gov/Archives/edgar/data/1691493/000129281426004518/nu20260909_6k.htm
  12. Save the Date: Investor Day, 8 December 2026 (6-K) — Nu Holdings / SEC — 2026-09-02 — https://www.sec.gov/Archives/edgar/data/1691493/000129281426004456/nu20260831_6k.htm
  13. Form 6-K: agreement to acquire Banco Porto Real de Investimentos — Nu Holdings / SEC — 2026-07-20 — https://www.sec.gov/Archives/edgar/data/1691493/000129281426003814/nu20260720_6k.htm
  14. Form 4 filings, 28 filings 2026-03-25 to 2026-10-07 (e.g. accession 0000950103-26-015323) — SEC EDGAR — 2026 — https://www.sec.gov/Archives/edgar/data/1691493/000095010326015323/ownership.xml
  15. Nu México receives authorization to begin operations as a bank — Nu newsroom — 2026-07-10 — https://nu.com/en/newsroom/company/nu-mexico-receives-authorization-to-begin-operations-as-a-bank
  16. David Vélez and Robin Vince join the boards of the OpenAI Foundation and OpenAI Group PBC — OpenAI Foundation — 2026-07-21 — https://openaifoundation.org/news/david-velez-robin-vince-join-openai-boards
  17. NU statistics and peer statistics pages (ITUB, INTR, PAGS, MELI, PICS, SOFI) — StockAnalysis — 2026-10-09 — https://stockanalysis.com/stocks/nu/statistics/
  18. NU analyst forecast — StockAnalysis (S&P Global data) — 2026-10-06 — https://stockanalysis.com/stocks/nu/forecast/
  19. NU analyst forecast and ratings history — MarketBeat — 2026-10-09 — https://www.marketbeat.com/stocks/NYSE/NU/forecast/
  20. NU short interest (FINRA data) — CurvedTrading — through 2026-09-30 — https://curvedtrading.com/stocks/nu/short-interest/
  21. Nu Holdings: Itaú BBA stuft die Aktie ab und senkt das Kursziel auf 18 US-Dollar — Goldesel — 2026-09-16 — https://goldesel.de/aktien/news/nu-holdings-itau-bba-stuft-die-aktie-ab-und-senkt-das-kursziel-auf-18-us-dollar
  22. Citi rebaja acción de Nubank, pero Wall Street todavía ve potencial alcista de casi 40% — Bloomberg Línea — 2026-06-16 — https://www.bloomberglinea.com/mercados/citi-rebaja-accion-de-nubank-pero-wall-street-todavia-ve-potencial-alcista-de-casi-40/
  23. O risco que levou o Citi a rebaixar o Nubank — Forbes Brasil — 2026-06-17 — https://forbes.com.br/forbes-money/2026/06/o-risco-que-levou-o-citi-a-rebaixar-o-nubank/
  24. Nu Holdings Aktie: Bank of America senkt auf Underperform — Börse Global — 2026-06 — https://www.boerse-global.de/nu-holdings-aktie-bank-of-america-senkt-auf-underperform/793154
  25. Nu Holdings keeps imploding: Goldman reiterated its belief that NU shares are going to double — 24/7 Wall St — 2026-09-29 — https://247wallst.com/investing/2026/09/29/nu-holdings-keeps-imploding-goldman-sachs-rieterated-its-belief-that-nu-shares-are-going-to-double/
  26. Nu Holdings spikes 12% as Brazil election upset drives rally — 24/7 Wall St — 2026-10-05 — https://247wallst.com/investing/2026/10/05/nu-holdings-spikes-12-as-brazil-election-upset-drives-rally-in-brazilian-assets-robinhood-edges-higher-sofi-stays-put/
  27. Nubank says it is not pursuing Monzo takeover after shares plunge — City AM — 2026-10-01 — https://www.cityam.com/?p=2462260
  28. Nubank (NYSE: NU) ainda vale a pena? BTG vê alta de 65% — Suno — 2026-05-13 — https://www.suno.com.br/noticias/nubank-nu-ainda-vale-a-pena-btg-ve-potencial-de-alta-mt/amp/
  29. Nubank bate o consenso em receita mas fica abaixo no lucro — NeoFeed — 2026-05-14 — https://neofeed.com.br/negocios/nubank-bate-o-consenso-em-receita-mas-fica-abaixo-no-lucro-mexico-chega-ao-breakeven/en/
  30. Citi: Nubank nos EUA pode chegar a US$21 bilhões em 2030 — Let's Money — 2026-04 — https://www.letsmoney.com.br/noticias/nubank-eua-citi-21-bilhoes-2030-junqueira/
  31. Flávio Bolsonaro slightly ahead of Lula in Brazil runoff, Datafolha poll shows — Reuters via The Star — 2026-10-09 — https://www.thestar.com.my/news/world/2026/10/09/flavio-bolsonaro-slightly-ahead-of-lula-in-brazil-runoff-datafolha-poll-shows
  32. Opinion polling for the 2026 Brazilian presidential election — Wikipedia (compilation; secondary) — retrieved 2026-10-12 — https://en.wikipedia.org/wiki/Opinion_polling_for_the_2026_Brazilian_presidential_election
  33. Ibovespa recorde, dólar abaixo de R$5 — Forbes Brasil — 2026-10 — https://forbes.com.br/forbes-money/2026/10/ibovespa-recorde-dolar-abaixo-5-reais-pesquisas-eleitorais/
  34. Fintechs chegam à eleição com agenda concreta e pouco espaço nos planos de governo — Finsiders Brasil — 2026-10-02 (updated 10-05) — https://finsidersbrasil.com.br/reportagem-exclusiva-fintechs/fintechs-chegam-a-eleicao-com-agenda-concreta-e-pouco-espaco-nos-planos-de-governo/
  35. SGS time series (1, 432, 21082, 21084, 21112, 29034, 29037) — Banco Central do Brasil — retrieved 2026-10-12 — https://api.bcb.gov.br/dados/serie/bcdata.sgs.21082/dados
  36. Inadimplência das famílias bate recorde e chega a 8% em agosto — Gazeta do Povo — 2026-09-29 — https://www.gazetadopovo.com.br/economia/inadimplencia-familias-bate-recorde-chega-8-agosto/
  37. Copom communications list (Selic 13.75%, 2026-09-16) — Banco Central do Brasil — 2026-09-16 — https://www.bcb.gov.br/api/servico/sitebcb/copom/comunicados
  38. Client Alert: Lei Complementar nº 224/2025 — Demarest — 2026-01 — https://www.demarest.com.br/wp-content/uploads/2026/01/Client-Alert-LEI-COMPLEMENTAR-No-224-2025-vf2.pdf
  39. Alíquotas de CBS e IBS sobre serviços financeiros serão de 10,85% em 2027 — Sefin-RO tax-reform portal — 2025-09-16 — https://reformatributaria.sefin.ro.gov.br/2025/09/16/aliquotas-de-cbs-e-ibs-sobre-servicos-financeiros-serao-de-1085-em-2027-define-relator/
  40. Resolução Conjunta nº 17 (naming rules) — BCB/CMN — 2025-11-28 — https://static.poder360.com.br/2025/11/resolucao-nomenclatura-bank-28nov2025.pdf
  41. PicPay 2Q26 results release — PicPay — 2026-08-24 — https://static.poder360.com.br/uploads/2026/08/2t26-picpay-release-24-8-2026.pdf
  42. MercadoLibre Form 10-Q for the quarter ended 30 June 2026 — SEC EDGAR — 2026-08-06 — https://www.sec.gov/Archives/edgar/data/1099590/000109959026000023/meli-20260630.htm
  43. Nubank cresce o dobro do mercado de cartões (BTG / Card Monitor) — Finsiders Brasil — 2026 (2Q26 data) — https://finsidersbrasil.com.br/pagamentos/cartoes/nubank-cresce-o-dobro-do-mercado-de-cartoes-e-os-bancoes-assistem-a-festa-pela-janela/
  44. SFG Research: Crédito Consignado — Moody's Local Brasil — 2026-08-24 — https://moodyslocal.com.br/wp-content/uploads/2026/08/SFG_Research_Credito-Consignado_24.08.2026.pdf
  45. Justiça dá 15 dias para Nubank explicar cobrança de juros em ação do GDF — Terra Brasil Notícias — 2026-07-12 — https://terrabrasilnoticias.com/2026/07/justica-da-15-dias-para-nubank-explicar-cobranca-de-juros-em-acao-do-gdf/
  46. Justiça vai analisar teto para cobrança de dívidas — Correio Braziliense (CB Poder) — 2026-07 — https://blogs.correiobraziliense.com.br/cbpoder/justica-vai-analisar-teto-para-cobranca-de-dividas/
  47. USTR announces final Section 301 tariff action on Brazil — Wiley — 2026-07-16 — https://www.wiley.law/alert-USTR-Announces-Final-Section-301-Tariff-Action-on-Brazil
  48. Nubank lucra US$7 milhões no México com provisões em alta (XP data; secondary) — Eu Quero Investir — 2026-08 — https://euqueroinvestir.com/acoes/nubank-lucra-us-7-milhoes-no-mexico-com-provisoes-em-alta
  49. Inter&Co posts record Q2 2026 results — Inter&Co — 2026-08-05 — https://www.blog-us.inter.co/interco-posts-record-q2-2026-results-with-us81m-net-income-and-rule-of-50-achievement/
  50. NU P/E ratio history — Public.com — 2026 — https://public.com/stocks/nu/pe-ratio
  51. NU quote (close 2026-10-09 $16.12) — Yahoo Finance — 2026-10-09 — https://finance.yahoo.com/quote/NU/
  52. NU Q3 2026 earnings date listing (not confirmed by company) — Earnings Whispers — 2026-10 — https://beta.earningswhispers.com/go/w/NU
  53. Nu Holdings Q4 2025 earnings call (AI summary only) — Investing.com — 2026-02-26 — https://www.investing.com/news/transcripts/earnings-call-transcript-nu-holdings-q4-2025-revenue-beats-expectations-93CH-4
  54. Nu Holdings surpasses Q2 earnings and revenue estimates (EPS history; aggregators conflict) — FXStreet / Zacks — 2026-08-14 — https://www.fxstreet.com/news/nu-holdings-ltd-nu-surpasses-q2-earnings-and-revenue-estimates-202608140603
  55. C6 Bank vê inadimplência subir — Economic News Brasil — 2026-08-19 — https://economicnewsbrasil.com.br/2026/08/19/c6-bank-inadimplencia-lucro-credito/
  56. Nu (NU) Q1 2026 Earnings Call Transcript — The Motley Fool — 2026-05-15 — https://www.fool.com/earnings/call-transcripts/2026/05/15/nu-nu-q1-2026-earnings-call-transcript/
All NU Bulls vs Bears reports
Disclaimer. AI-generated research and podcast (Claude by Anthropic, voices by Google Gemini) for educational purposes — not financial advice and not a recommendation to buy or sell. Willem (MR.INVEST) owns this stock. Check the company’s own filings before making any decision.
← All Bulls vs Bears reports