People
The People Running Nu Holdings
Governance grade: B. Founder-controlled, deeply aligned, and stacked with top-tier independent directors — but the dual-class structure gives David Vélez 74% of the vote on far less than 20% of the economics, and the controlled-company exemption blunts the protections minority holders would otherwise rely on. Trust is high on capability and alignment; structural power asymmetry is the offset.
Governance Grade: B
Skin-in-Game (1-10)
Founder Voting Power
Independent Directors
The People Running This Company
Nine people decide what Nu does. Two of them — David Vélez and Cristina Junqueira — founded the company together in 2013, still run it, and between them control about 84% of the votes. Everyone else is a hired executive or an independent director.
The team passes the threshold question — does it look like a real bank? Yes. Lago (CFO) and Fragelli (CRO) bring orthodox global banking pedigrees (Credit Suisse, HSBC). The hires of Roberto Campos Neto (former Brazilian central bank president, architect of Pix) and Eric Young (Google/Snap/Amazon engineering) in 2025 directly address Nu's two scaling risks: regulatory navigation as it expands across LatAm and pursues a US charter, and platform engineering as it crosses 130M customers. Campos Neto's hire is the most consequential governance event of the past year — it is rare for a sitting G20 central banker to join a private bank within a year of leaving office, and it materially strengthens Nu's standing with regulators in Mexico, Colombia, and the US.
The Lahrech departure (May 2025) and Vélez resuming direct command is the one watch-item. Founders who reclaim operational control after a President exits sometimes signal either decisiveness or stretched bandwidth — Nu's Q1-Q4'25 execution suggests the former, but the hiring cadence (CTO, CDO, CPO, Vice-Chairman, plus a new CPO in May 2026) shows succession is being built around Vélez rather than under him.
What They Get Paid
Nu discloses aggregate, not individual, compensation — it is a Cayman-domiciled foreign private issuer and is not required to. The headline number is $91.3M in aggregate FY25 compensation for the board and key management, of which the overwhelming majority is share-based.
Aggregate Board + KMP Comp (FY25, $M)
% of FY25 Net Income
RSUs Outstanding (M)
Options Outstanding (M)
The 84% equity weighting is unusually high for a regulated bank and is driven by the Contingent Share Award (CSA) granted to Vélez in 2021: a one-time long-term grant worth ~1% of Nu's share capital, but only earned if the share price hits specific thresholds (with the first threshold around $18.69 — roughly 3x the IPO price). Vélez has publicly committed via the Giving Pledge to donate the entirety of any CSA proceeds to a family philanthropy focused on LatAm youth — so the program functions more as a control-and-retention device than personal wealth creation.
Is pay sensible? $91.3M in aggregate is roughly 0.6% of FY25 revenue ($15.8B) and 3.1% of net income ($2.9B). For a $63B-market-cap bank with 33% ROE and 10,000 employees, this is in line with — and arguably below — what a large US or European bank pays its top tier. The Compensation Committee is chaired by Jacqueline Reses (ex-Square Financial Services, CEO of Lead Bank) and includes Doug Leone (Sequoia) and Luis Moreno (former IDB president) — three people who have seen elite executive comp at scale and are not pushovers. The clawback policy was adopted in October 2023 and is current with SEC rules. What you do not get with a Cayman FPI: individual NEO disclosure, a true Pay-vs-Performance table, or a binding say-on-pay vote. That is a structural protection minority shareholders simply do not have.
Are They Aligned?
This is the strongest section of the file. By every meaningful test of alignment — founder economic stake, co-founder stake, equity-weighted comp, restraint on dilution, absence of related-party self-dealing — Nu is at the top of its peer group. The asymmetry is in control, not in alignment.
Vélez's Class B shares carry 20 votes each versus 1 for Class A. He owns about 18.6% of the company economically but controls 74.4% of the votes. That gap — 56 percentage points — is the single biggest governance risk in the name. It means:
- Strategic decisions, board composition, and amendments to the M&A all run through Vélez alone.
- Nu invokes the NYSE controlled-company exemption, electing not to be required to have a majority-independent board or a fully independent comp committee (Nu chooses to anyway, but it could stop).
- The Shareholder's Agreement gives Vélez personal consent rights over major actions (M&A, dividends, share issuances over 20% of equity, related-party transactions, etc.) so long as he holds 10%+ of votes. He is not just the largest holder — he is a one-man second board.
Insider activity. Insider selling is minimal and routine. The one discretionary sale in the past year was Cristina Junqueira selling 300,000 shares at $14.81 on March 23, 2026 — a $4.4M transaction that reduced her direct holdings by 11.48% but is trivial against her ~$1.7B total stake (she still controls ~131M shares through trusts and LLCs). The CRO had a tax-withholding disposition in April 2026 — these are mandatory, not discretionary. There has been no insider open-market buying disclosed, but founder-controlled companies rarely show it.
Dilution. Nu's equity plans authorize up to 5% of fully diluted shares per year. At ~4.86B shares outstanding, that is a notional ceiling of ~243M shares — meaningful. Actual outstanding RSUs + options sit at 74.9M (21.8M options + 53.1M RSUs), about 1.5% of shares. So the live overhang is small; the cap is high. SBC ran roughly $400M in FY24-25, real but absorbed by 33% ROE.
Related-party transactions. The 20-F discloses only ordinary credit cards and personal loans to executives and directors, on the same terms offered to other customers, and a de minimis "Others" line of $1,795 in FY24. No promoter loans, no acquisitions from affiliates, no real-estate side deals. This is exceptionally clean for a controlled company.
Capital allocation. Nu does not pay a dividend and has done no buybacks — every dollar of $2.9B in FY25 net income is being reinvested in customer growth, credit book expansion, and the US launch. Given 33% ROE, this is the right call. A buyback or special dividend would actually be a yellow flag at this stage.
Skin-in-the-game score: 9/10. Founder and co-founder hold ~$13B in stock between them, executive comp is overwhelmingly equity, the CEO's largest grant only vests on share-price thresholds, and pledged philanthropy on those proceeds removes the personal-enrichment angle. One point off for the dual-class structure, which converts excellent alignment into excellent control — minority holders ride with him, but cannot challenge him.
Board Quality
Seven of nine directors are formally independent. More importantly, the substantive quality of the board is unusually high for an emerging-markets bank. Each independent director brings either operating experience at a globally-scaled platform, deep banking/audit expertise, or regulatory standing.
Board — Fintech Strategy Expertise (1-10)
Audit & Risk Committee — chaired by Rogério Calderón Peres, a former PwC Brazil audit partner and CFO at Bunge Brasil, Unibanco/Itaú Unibanco, and HSBC Brasil. He is one of the strongest bank-audit-chairs in LatAm capital markets. Joined by Anita Sands (ServiceNow board) and Thuan Pham (former Uber/Coupang CTO) — a balanced mix of financial-reporting and tech-risk expertise. This is a real committee, not a rubber stamp.
Compensation & People Committee — chaired by Jacqueline Reses, ex-Square Financial Services and currently CEO of Lead Bank. Includes Doug Leone (Sequoia Global MP 2012-22), Luis Moreno (15-year IDB president), and David Marcus (ex-PayPal President, ex-Meta payments, now Lightspark CEO). On paper this committee could push back on any pay or equity proposal — whether it actually does is unverifiable given the lack of individual disclosure.
Where the board is thin: real cybersecurity-specialist depth (Sands is the closest); a director with hands-on consumer-credit underwriting/loss-given-default experience separate from management; and an independent voice from Mexico or Colombia (the two growth markets). Marcus's crypto/Bitcoin/stablecoin focus at Lightspark is a watch-item if Nu accelerates NuCrypto.
The structural caveat. Substantive independence ≠ structural independence. Vélez retains the right under the Shareholder's Agreement to nominate up to 5 directors (a majority) for as long as he owns 40%+ of votes. He currently nominates fewer than that. But if a real disagreement arose between the independent directors and Vélez over a strategic decision, Vélez wins — he could replace them. The current board is excellent because Vélez chooses excellent people. There is no governance mechanism preventing him from choosing differently.
The Verdict
Overall Governance Grade: B
Letter grade: B.
The strongest positives. (1) Founder economic alignment is among the best in any large public bank globally — Vélez and Junqueira together hold ~$13B in stock and pay themselves overwhelmingly in equity. (2) The board is substantively excellent: a former Brazil central bank president, a former IDB president, Sequoia's former Global MP, an ex-PwC bank-audit partner as audit chair, and a current digital-bank CEO as comp chair. (3) Related-party transactions are clean — no self-dealing, no promoter loans, no affiliate acquisitions. (4) The clawback policy is current and untriggered. (5) The CEO's largest equity grant only vests if the stock crosses performance thresholds, with proceeds pledged to philanthropy.
The real concerns. (1) The dual-class structure gives Vélez 74.4% of votes on 18.6% of economics — a 56-point gap. The Shareholder's Agreement converts this into one-man consent rights over M&A, dividends, large issuances, and related-party transactions. (2) Nu uses the NYSE controlled-company exemption — minority holders have weaker protection than at a typical NYSE issuer. (3) Cayman Islands incorporation removes US-style derivative-suit and proxy-rule protections; no individual executive comp disclosure, no say-on-pay. (4) The ESOP cap is 5% of fully-diluted shares — high relative to global peers. (5) Vélez resumed direct command in 2025 after the President exited; succession depth below him is being built but not yet proven.
What would upgrade this to an A. A binding sunset on the Class B super-voting structure (most peers have a 7-10 year automatic sunset; Nu has none). Or, voluntary individual executive comp disclosure. Either move would close the alignment-vs-control gap.
What would downgrade this to a C. A material related-party transaction with Vélez's family office or Rua California Ltd.; replacement of any current independent committee chair with a Vélez nominee under contested circumstances; or a debt-funded special dividend / leveraged buyback that disproportionately benefits the controlling shareholder.
The bet on Nu is, in governance terms, a bet on David Vélez's judgment and stewardship. The infrastructure around him — board, audit chair, CRO, CFO, the Campos Neto hire — is among the best available, and his own economic interests are vastly more aligned with the public float than the typical founder-CEO. The structural risk is that none of the protections are binding against him. For a fundamentally sound, profitable, founder-driven bank growing at 40%+, that is an acceptable trade — but it is a trade, not a free option.