Nigel Morris on Capital One, Nubank, and what actually changed in Latin America

The J Curve is an English-language podcast and newsletter covering Latin America's technology and venture capital ecosystem, hosted by investor Olga Maslikhova. In this episode: Nigel Morris, who co-founded Capital One inside a small Virginia bank in 1994 and grew it from roughly $1 billion to nearly $20 billion in market value, then co-founded QED Investors and backed Nubank before it existed — after David Vélez pitched him "a Capital One in Brazil" over breakfast in Mexico City. He explains why Nubank, Revolut, Klarna and Monzo are all now aiming at the same market, why banks structurally cannot innovate, why a company growing 40% with an NPS of 89 currently cannot raise growth equity, and what he thinks genuinely changed in Latin America — which is not capital, regulation, or technology.

Why did credit cards need to be reinvented in 1986?

Because there was effectively one product. Every customer paid a 19.8% APR and a $20 annual fee regardless of their risk, which meant the safest borrowers were subsidising the middle and the riskiest couldn't get access at all. Morris had spent his consulting years on insurance economics, where risk-based pricing was already unremarkable — nobody objects that a 22-year-old with a drink-driving conviction pays more for car insurance than a 50-year-old. He and Rich Fairbank took that to fifteen or so bank management teams and were told either "we already do this," which they didn't, or that the idea was sound but the two of them couldn't execute it.

"Credit cards circa 1986: any colour you like so long as it is black."

Nigel Morris

How do you turn a bank into a laboratory?

Morris abandoned clinical psychology as an undergraduate because he thought Freud and Jung were "more philosophy than psychology," and retreated into experimental method and statistics. That background, married to the annuity economics of insurance, produced the founding insight: in a business with a large enough N, every variable is testable. Red envelope or yellow. Fee on the front or the APR. Call a delinquent customer at one day, thirty, sixty or ninety. Thousands upon thousands of tests, each driven by a hypothesis.

The sharpest example is one that isn't in any Capital One case study. Two applicants, identical age, income, job tenure and FICO — one applies at midday, one at midnight. Marketing holds that data; credit doesn't. But the midnight applicant is desperate and the midday applicant is on their lunch break, and adverse selection is one of the largest drivers of credit risk. "Those two people know something that you don't know, and it's embedded in the time of day." Integrating the two datasets is what forced Capital One to invert a bank's standard functional org chart and put quantitative people at the strategic fulcrum.

What did Capital One get right in its first two years?

Three decisions. They stopped Signet Bank throwing away data tapes, which it had been doing because storage cost $5 a month. They refused to recruit through the bank's normal graduate programme, insisting on hard-science backgrounds from top schools put through six to ten case interviews — and fought the bank over paying them twice the going internal rate. And they collapsed the wall between credit and marketing, on the grounds that they are the same discipline. Underneath all three was a culture where a 23-year-old eighteen months into the job made decisions worth tens of millions, which the parent bank found frankly deranged.

Why did Nigel Morris leave Capital One at the top?

By 2004 the company had compounded earnings at 26% a year, gone from about $1 billion to nearly $20 billion in market cap, expanded from cards into installment loans, mortgages, home equity, deposits and wealth, and from the US into Canada, the UK, France, Italy, Spain and South Africa — all inside a decade. And Morris was bored. He describes the entrepreneurial drive as "a monster" he still fights to control at 67, and the later Capital One years as earnings calls and regulator meetings that didn't tax him. A board member told him to buy a red sports car and diagnosed a midlife crisis.

Worth noting because it recurs later: he had also concluded that scale is entropic. "Gravity will make us bigger and slower and more political and hierarchical with each passing day unless we all get up every day and fight back against it."

Why did fintech specialists beat generalist VCs?

QED began accidentally, as ex-Capital One "refugees" turned up with ideas and Morris, Frank Rotman and Caribou Honig applied Capital One frameworks to them — risk-based pricing, positive selection, NPV per customer. Within a few years they had found Nubank, Remitly, Credit Karma, Flywire and AvidXchange. Morris calls the durable advantage the Ghostbuster effect: any investor hears from a CEO who just beat their February numbers, but the question is who the CEO calls when they are in a pickle and don't know what to do. He argues generalists genuinely add value on culture, fundraising, IPO timing and recruiting — but that fintech is a stack of specialist verticals (fraud, credit, KYC, asset-liability management, treasury, overlapping regulators) where having lived inside the trade-offs is not substitutable.

"Any fool can lend money. The challenge is getting people to pay you back."

Nigel Morris

What does "playing the full 90 minutes" mean?

It's the football metaphor QED uses for post-investment behaviour, and it comes from a specific observation: at the first or second board meeting after a round there is often an "oh crap" moment where the numbers are nowhere near the plan. QED's instinct as operators was to lean in; they watched generalist co-investors pull back, so the partner the founder thought they were getting is replaced at the next board meeting by someone two years out of an MBA. Morris frames this as an underwriting insight, not a service promise: "Behavioural scoring is much more powerful than credit scoring. The way a company behaves is more important than the a priori diligence data."

He is honest about the cost. It is far easier to turn a 5× into a 6× than a 0× into a 1×, and his operator reflex is to spend time on the problem child rather than pour petrol on the fire — a bias he says he still has to stop himself acting on.

How did Nigel Morris meet David Vélez before Nubank existed?

Over breakfast in a Mexico City hotel courtyard, when Vélez was an associate at General Atlantic — before Stanford, before Sequoia, and before QED had properly started. They were there to look at Compartamos. Vélez said he thought there was an opportunity to build a Capital One in Brazil, and asked whether Morris would help. Morris's reply, by his own account: he had never been to Brazil and the only thing he knew about it was that it always seemed to beat England at football.

Vélez then came to Northern Virginia, where Morris and Rotman spent serious time laying out, in Morris's phrase, "in a molecular way," how Capital One actually worked. At some point Vélez gave them warrants for the help. Morris's assessment of those warrants: "gold dust." What he says he noticed was not brilliance but listening — Vélez may be the smartest person in the room and you would never know it — and then decisiveness. And Vélez didn't copy the model, he moved it: mobile-first origination and servicing rather than call centres, producing a cost to serve that Morris says is de minimis against Wells Fargo or JP Morgan.

Why are Nubank, Revolut, Klarna and Monzo all aiming at the United States?

Morris frames it as a reversal. After the financial crisis, global banks retreated to their home markets — HSBC, Barclays, Citi, Chase, and Capital One itself, which exited South Africa, Italy, France and Spain. What is happening now is the opposite: fintechs built outside the US, having won their home geographies, moving on the largest market in the world at exactly the moment the regulatory window opens. He notes he'd have said a year ago that the twenty-odd companies applying for OCC charters would never get one, and now believes they will.

"The biggest prize is not Sweden or Germany if you're Klarna. It's not the UK if you're Monzo or Revolut. And it's not Brazil if you're Nubank. It's the US."

Nigel Morris

The open question is whether comparative advantage survives contact. The US is many multiples of Brazil's TAM and also the most competitive financial market on earth — Amex, Capital One and JPMorgan at the top, Capital One again in near-prime. His cautionary case is N26, which he suspects moved before it had truly conquered home.

Second product or second country?

A generalisable answer buried in the Nubank discussion, and one of the most useful things in the episode for any founder. Successful companies reach a fork: add a second product in the home market, or take the existing product to a new one. Morris says the answer is almost always the second product — and that companies get intoxicated by the new geography, run it as a science experiment, lose money and shut it down.

Why can't a good company raise growth equity in 2026 without an AI story?

Morris's description of the current market is unusually blunt, and it is the passage most worth sending to anyone raising right now. You can have unit economics that work, a net promoter score starting with an 8, a grown-up management team, borderline profitability and 40% growth — and if your first sentence isn't stablecoins or native AI, it will be really hard.

"This market is intoxicated by AI startups. I believe in the end it's about fundamentals, and in the long run markets are efficient. Maybe I'm being delusional — I've been accused of it in the past."

Nigel Morris

What did the 2021 valuations actually cost founders?

More than a down round, and this is the part people skip. Morris has portfolio companies funded at "apex maximus" in 2021 at 20× revenue that have since grown five to eight times and are still marked below their 2021 valuation. The second-order consequences ran for years inside those companies: underwater options, repricing, the cultural work of resetting expectations. He includes himself — a big pile of 2021 options that took five years to approach the money. The lesson for founders is that the highest number available is not the right number; the wrong valuation becomes an operating liability.

Why don't banks buy fintechs?

Morris calls it the Galapagos Effect. A bank is good at not messing up — conservative on credit, diversified, growing 3–5% a year, opening its earnings call with how much equity it holds. What it is bad at is innovation; he invites anyone to name something substantial a large bank has innovated in twenty years. Meanwhile, directly in front of it, is an archipelago of organisms competing to reproduce, funded by people like him. That is outsourced R&D, and banks should be watching it, partnering with it, learning from it and selectively buying from it.

The reason they don't is a timing trap he describes precisely. While a fintech target is small, the legal department objects that its KYC and AML aren't industrial strength. By the time it has proven itself, the CFO objects to the dilution. That leaves a narrow window around $300–400 million where a deal is actually possible. His exceptions are Capital One's move on Brex, which he calls audacious and sensible, and the Discover acquisition, which he calls the crowning glory of Fairbank's leadership — prompting the line that JPMorgan and Capital One are playing chess while the rest of the banks play checkers.

What has actually changed in Latin America?

Not capital. Not regulation. Not technology. Belief.

Morris's argument is that in Nubank's early days the incumbents simply did not believe a new bank could be built — the same disbelief Citi and Chase had about Capital One until it was too late. What Nubank did was demonstrate that it can be done, which changes what a generation of founders considers attempting, and what an investor considers underwriting. He pairs it with the mechanism he has watched twice: great companies throw off talent. Capital One seeded a diaspora that now runs Zopa, ClearScore and Revolut's US business. He expects Nubank to do the same in Latin America.

And then, a few minutes later, answering a different question about Nigeria, he produces the sentence that is really the answer to this one:

"Nubank did more than just create this juggernaut. It made Brazil investable.”

Nigel Morris

What about currency volatility and the lack of exits?

This is the standard objection to Latin American venture and Morris does not dodge it. Yes, currencies are volatile — though they have moved favourably against the dollar this past year. Go in with your eyes open; it is not the US or the UK. But he argues the trade-offs cut both ways: the Brazilian regulator has been genuinely forward-looking, in the way India's was on UPI; labour costs are lower; incumbents are weaker.

His actual diagnosis relocates the problem, and it is the most useful thing in the episode for anyone allocating:

"Where there is a vulnerability is the lack of growth capital. Not so much on the venture side. When I talk to my PE friends, it's not clear to me why they are reticent. There's a petrification with a lot of private equity at the moment, and they're looking for reasons not to do things."

Nigel Morris

He extends it with the Nigeria case: a well-run lending company there struggles to get a private equity firm to take the call, on a view he says is "not grounded in empirics." Which is precisely why "it made Brazil investable" is doing so much work — the constraint is a belief held by late-stage capital, not an arithmetic fact about the region.

Where is AI actually being deployed in fintech?

Three places, in Morris's account. Code — across essentially all 140 active portfolio companies, with machines now writing 60 to 90% of it, which has collapsed the premium on the engineering hire. Anything call-centre or repeatable-manual, which is going away. And third, businesses that don't exist yet: Capital One used to claim the right product to the right customer at the right time and right price, but was really doing it through clunky batch direct mail. Targeting at N equals one is now real.

The portfolio has followed: six of QED's last eight investments were vertical SaaS tools serving fintechs rather than fintechs themselves.

What is "lazy money," and why is agentic AI coming for it?

A regional bank analyst told Morris that 40% of regional bank profitability comes from what he called lazy money: deposits paid 30 basis points or nothing that are worth 400 to the bank. That spread survives on customer inertia — and Morris quotes his colleague Amias Gerety's line that bankers confuse loyalty with inertia.

His point is that inertia is a technology-dependent quantity. Nobody switches banks themselves for ten dollars a month. But in a world of open banking and agentic AI, where a robot will do it for you at no cost, everybody switches. He expects a tsunami against that 40%, and thinks the beneficiaries are precisely the players with national charters and state-of-the-art tooling: Revolut, Klarna, Nubank.

Note what this does structurally. It closes the loop back to the opening: the same fintechs coming to the US now have a specific, quantified pool of incumbent profit to attack.

Rapid fire: five questions

One question founders should ask VCs but almost never do. "Am I going to get the partner I've worked with?" Morris's view is that the difference between a good partner and a bad partner is larger than the difference between firms — so ask directly whether they are staying, and whether they are thinking of leaving or retiring. He asks the reciprocal question of founders: if the company were worth $100 million and you owned 19%, would you sell? He says he has watched people sell two or three years too early, and that the pull toward early liquidity is stronger in emerging markets.

What founders in emerging markets underestimate. That capital is more capricious than efficient-markets theory suggests, and the less developed the country the more capricious it gets. See the Nigeria example above.

Highest-ROI use of his own time in the past year. A full day of diagnostic testing — MRI, extensive bloodwork, probabilistic modelling of what he's likely to get. Five grandchildren; the goal is living better for longer, not merely longer.

The habit that most improved his decision-making. Unit economics, hands down, and horizontally rather than vertically. What does it cost to acquire this customer, what cash flows will this customer produce, what is the NPV. Stack enough positive customer-level economics and company-level economics follow. Capital One was fanatical about it and QED still is.

Most under-the-radar company in Latin America. Two: Félix Pago, using stablecoins to move money and growing fast, and Plazo, doing buy-now-pay-later in Mexico — which Morris singles out as an example of succeeding without the regulatory tailwind Brazil has enjoyed.

COMPANIES AND INSTITUTIONS

⠀PEOPLE

⠀IDEAS AND FRAMEWORKS

Co-founder and Managing Partner, QED Investors. Co-founded Capital One with Rich Fairbank, spinning the credit card business out of Signet Bank in 1994 and serving as President and COO until 2004. QED has made more than 200 fintech investments in 20 years, including Nubank, Credit Karma, Remitly, Flywire, AvidXchange, Creditas and Bitso. Also an investor in Swansea City. Previously on the boards of The Economist, Brookings, National Geographic and London Business School.

Hernán Kazah, Kaszek — building Latin America's $100B+ companies · Stelleo Tolda — inside MercadoLibre's rise · Brian Requarth, Latitud — why ~50% of LatAm founders now go global

🎙 The J Curve  is where LATAM's boldest founders & investors come to talk real strategy, opportunity and leadership.