Part IThe Story
The door that would not open
The door came first. Before the queue, the forms, the call center, and the months, there was a bulletproof door at the entrance of a bank branch in São Paulo, flanked by armed guards. For a moment it held a young Colombian financier between the street and the bank, the way such doors are designed to hold anyone who might turn out to be a thief. He was told to leave his backpack outside. He waited, by one telling, forty minutes, and by another an hour. He came back, he has said, about six times. The process took five months. At the end of it, he wrote later, "a bank account that would charge hundreds of reais per year in fees was approved in my name."
David Vélez has told this story so many times that it has become a kind of liturgy, and like most liturgies it has drifted in the retelling. Stanford's business school, in a 2022 write-up, places the ordeal in 2008, when he moved to Brazil to open a regional office for a private equity fund. In a founder's letter quoted by the banking analyst Marc Rubinstein, Vélez dates it to "mid-2012," and the duration shrinks to four months. Onstage in New York in May 2022, he said: "In 2013, I went to try to open a single simple bank account in Brazil, and it was a five-month process." The years move around. The door never does. In every version he is standing in that vestibule while men with guns look at him, as he put it, "as a criminal," and he is trying to square what is happening to him with a fact he already knew from his day job: he was standing in the lobby of one of the most profitable industries on earth.
That is the hinge of the story, and Vélez has always framed it as an intellectual puzzle more than a grievance. "It was one of the worst experiences I ever imagined," he told Fortune in 2023, and then, almost immediately, the analyst took over: "that was hard to understand in the context of this being the most profitable industry in Brazil and these being the largest companies in Brazil. I couldn't understand why these big companies were treating consumers that way and also leaving about 60 million people completely unbanked."
Most people who are humiliated by a bank change banks, or complain, or forget about it. Vélez treated the humiliation as data. The bulletproof door showed him how the business was built. It was a security measure, and it was also a statement about who the customer was assumed to be. An institution that treats every person who walks in as a potential robber has decided, in its architecture, that it does not need to be liked. In a working market that decision would be fatal. In Brazil's market it was simply the cost of entry, and the customer was the one who paid it.
He did not yet know that he would spend the next decade building a bank with no door at all: no branch, no guard, no vestibule, nowhere to leave your backpack. He did know, because he had grown up in a family that kept score in buttons and cows, that a contradiction this large is rarely an accident. Usually somebody is being paid for it.
By the Numbers
The Bank Without a Branch
5 monthsTime Vélez says it took to open his first Brazilian bank account
$2MSeed capital from Sequoia and Kaszek Ventures in 2013
$45BValuation on Nubank's first NYSE trading day, December 2021
135M+Customers across Brazil, Mexico and Colombia, May 2026
$2.87BNu Holdings net income, 2025
$74.9BTotal assets, 2025
1 cowVélez's first investment, at age 12
An industry that grew in good times and in bad
Around the same time, in the early 2010s, a London fund manager named Marc Rubinstein was also in São Paulo looking at Brazilian banks. He has described being driven between meetings in an armored vehicle through some of the worst traffic in the world, and learning only later that the truly rich commuted by helicopter. The executives he met laid out an extraordinary business. A history of hyperinflation had kept interest rates high, so simply buying and holding government bonds was profitable. Corporate relationships fed the consumer franchise, because employees tended to bank wherever their employer banked. The sector was concentrating fast: Itaú and Unibanco had recently merged, Citigroup would later sell its consumer business to Itaú Unibanco, and HSBC would sell to Bradesco. By the end of that wave, Rubinstein calculated, the five largest banks held more than 80 percent of Brazilian banking assets, up from 50 percent in 2000.
Rubinstein went home and bought. Bank of America, short of capital after the financial crisis, was a forced seller of its Itaú Unibanco stake and offered an attractive price. "One of the best—and easiest—trades I ever did," he wrote. Then, a few lines later, he made the wry observation that gives his essay its point: "It turns out I wasn't the only one looking at Brazilian banks at the time." Vélez was looking too. He had the same facts and the same oligopoly, and he reached the opposite conclusion.
The five were Itaú, Bradesco, Santander, Banco do Brasil, and Caixa Econômica Federal. Vélez listed their virtues as if he were reading charges in court. Brazil, he told Fortune, had "one of the most concentrated banking industries in the world, one of the highest interest rates in the world, one of the highest profitability ratios in the world, in an industry that grew consistently in good times and in bad times." Five of the ten largest companies in Brazil were banks. The largest companies in Mexico and in Colombia were banks as well. A JPMorgan analysis later cited by Forbes found that in 2019 fees made up nearly 40 percent of Brazilian banks' revenue, against 15 to 20 percent in Mexico, Argentina, Peru, and Chile. "All of these banks had grown through a lot of acquisitions for decades," Vélez said, "and had effectively become a very big oligopoly. It was an industry that was very complacent because there was really no competition."
A public-markets investor looks at that picture and sees a moat. A venture investor sees a moat that has stopped being defended because nobody has tried to cross it in a generation. The two readings are not really in conflict. They are bets on different time horizons, and both men were paid for theirs. But only one of them had stood in the bulletproof door. Rubinstein rode in the armored car. Vélez was the man the armor was designed to keep out.
He was born in Medellín in 1981, at the start of the decade that made the city's name a byword for violence. His family were makers and sellers. His father owned a button factory, and, according to a Stanford profile, all of his father's siblings had built businesses of their own; one account counts eleven of them. The sources disagree on where the boy started work. One has him at age three sorting buttons at an uncle's jean factory; another puts him in his father's button factory. Either way, his first job was sorting buttons: small, cheap, identical things that only make money in very large quantities.
Then an uncle was kidnapped. The family made what the Stanford profile calls "a difficult choice: leave everything behind for Costa Rica." Vélez was around nine. Harvard Business School's Ranjay Gulati, introducing him on a podcast, described the move as a flight from "rampant drug gang violence." Vélez rarely dwells on it in public. The Stanford account says the years of turmoil "taught him to be perceptive and focused," which is the sort of thing an alumni office writes, and may also be true. He holds both Colombian and Costa Rican citizenship. He was an outsider for the first time before he was ten.
On the family farm in Costa Rica, at twelve, he made his first investment. He told Gulati about it in a nearly unbroken paragraph that is worth quoting whole, because the whole later company is in it:
— David Vélez, on the Deep Purpose podcast with Ranjay Gulati, 2022I decided that the best investment and the best startup I could do was buying a pair of cows, because the cows multiply themselves. My dad had six cows, and I convinced him to let me buy one. I bought one, and then from 12 to 18, that cow became another six cows, and that was kind of the entryway or understanding of compound interest.
The money came from birthdays and summer jobs. Between twelve and eighteen the one cow became six, roughly a 35 percent annual return if you do the arithmetic, which the boy almost certainly did. Then comes the clause that makes the anecdote: "Obviously, my dad was nice enough that he didn't make me pay any expenses, no veterinary, no food, so ultimately, this became actually a really good business." At eighteen he sold the herd and used the proceeds toward Stanford.
There is a fair amount of autobiographical honesty in that sentence. The returns were real because the fees had been waived. Someone had absorbed the carrying cost, quietly, and the beneficiary only noticed later, when he was explaining to a Harvard professor why the business had been so good.
The interview about his mother
Stanford gave him a bachelor's degree in management science and engineering in 2005. Then came the standard résumé of the ambitious Latin American financier. He worked at Goldman Sachs and Morgan Stanley, where he learned how institutions are valued, and at General Atlantic, where he learned how they grow. Partway through, a growth-equity assignment took him to Brazil. In 2010 he went back to Stanford for an MBA. In late 2010, according to Sequoia Capital's own account, the firm "recruited David Vélez from Stanford GSB to help us evaluate whether to open a Sequoia office in Brazil."
The way Sequoia recruited him has become one of his favorite stories, and it reveals more about him than about Sequoia. "The first interview was with the head of the firm, not with the youngest analyst," he has said. "In most firms, you start from the bottom and then you go up. At Sequoia, it's the opposite. You go from the top and then you go down." The head of the firm was Doug Leone. He spent the meeting asking almost nothing about Vélez's CV and a great deal about his family: Tell me about your dad. What does he do? How's your mom? How's your relationship with your mother? Is she pretty strict?
"There's no way to game those questions," Vélez said later. "There's not a right answer." When he left, he walked to his car, started it, and found an email from Michael Moritz already waiting. Leone had apparently taken about a minute to brief his partner, and Moritz wanted a meeting.
Leone is a figure Vélez keeps coming back to, the way people come back to a teacher whose methods they have quietly adopted. In Vélez's account, Leone's distinguishing gift was a "people radar," a theory, built up over decades, of what a person's answers about their parents reveal about how they will behave under pressure. Leone was blunt and fast, with what Vélez has called a no-BS directness. He also treated a business-school intern as a partner. He sat him in investment committee meetings and asked his opinion of deals.
The schedule that followed sounds like something invented, and Vélez tells it with a faint disbelief. He woke at four in the morning to prospect Brazilian companies until eight. He went to class, was back at the Sequoia office by two, worked until eight at night, and then did his homework. Some weeks, in the middle of the week, Leone would pick him up at three in the afternoon, "like Dad was picking me up from school," and drive to San Francisco airport. They flew fourteen hours to São Paulo on Leone's plane, met seven companies in a day, signed term sheets, and flew back in time for an 8 a.m. Thursday class. "I remember just waking up almost like lost where I was," Vélez said, "and I was finding myself like flying to Brazil, about landing in São Paulo, pinching myself like, what is going on here?"
He graduated in 2012 as a Sequoia partner responsible for Latin America. The practice he was meant to build lasted about as long as his MBA.
In October 2012, by his account, Leone told him directly that Sequoia would not be opening in Latin America. The reasons were straightforward. Distance was one. Talent was another; Wikipedia's account cites a shortage of local software engineers. The third, the one Vélez has stressed, was that the startups they kept meeting were copies of Silicon Valley models, food delivery and the like, rather than companies attacking the region's structural problems. Sequoia offered to bring him to San Francisco as a growth-fund partner. He turned it down. "I didn't think twice," he told Rest of World. "It felt liberating."
Sequoia's own retrospective is generous and a little rueful: "While we ended up passing on opening the office, we had stumbled upon something else." The firm had sent Vélez to find a great Brazilian company and he had not found one. So he would have to build one himself.
The company he could not find
Nubank was incorporated on May 6, 2013. Vélez was thirty-one. Its first office was a house he rented in São Paulo, reportedly for about $500 a month, and the seed money came from Sequoia and Kaszek Ventures. Kaszek is a Latin American firm founded, the Stanford write-up points out, by two other graduates of the business school. The seed round was about $2 million. The pitch deck was short. Years later Sequoia published a page from it with an observation that is almost unheard of in venture capital: Nubank went on to execute exactly the vision in its seed deck, and "the only thing that changed was their name."
The vision was this: a credit card first, managed entirely from a smartphone, with no annual fee. Everything else, the accounts, loans, insurance, and investments, would follow. "After a couple of meetings," Vélez recalled, "we decided that we wanted to change the industry, as a fully digital bank, consumer obsessed, and building a technology company that will provide financial services for the bank and the unbanked consumers. So, our goal was to have 1 million credit cards in five years."
He did not do it alone, and the two people he recruited, on the advice of his investors, are the clearest evidence of how he saw his own limits.
Cristina Junqueira was the insider. She is Brazilian and trained as an engineer at the Universidade de São Paulo, where she also earned a master's degree, and she later took an MBA at Northwestern's Kellogg School. She had spent years at Itaú Unibanco, one of the five, working in consumer credit, card products, and marketing. She knew from the inside how the business Vélez wanted to attack actually made its money. After the IPO, her 2.9 percent stake, worth about $1.3 billion, made her only the second woman in Brazil to become a self-made billionaire. When the stock cooled, Forbes reported that she no longer was one. Today she is the group's chief growth officer and is expected to lead Nu's banking operation in the United States as its CEO. So the insider from Itaú is now the one being sent somewhere unfamiliar.
Edward Wible was the engineer, an American who served as chief technology officer for eight years. In April 2021 he gave up the title, in the company's phrasing, "to contribute more directly to the construction of the systems and infrastructure of the company." It is unusual for a co-founder to step down in order to get closer to the machinery. That tells you most of what you need to know about him.
So the company had three founders from three countries. One was an outsider with capital, one an insider with credit expertise, and one an engineer without either. The skeptics, of whom there were many, kept identifying Vélez's nationality as the fatal flaw.
— David Vélez, Goldman Sachs Talks with John WaldronEverybody told me, you know, you're Colombian. You have no idea what you're doing in Brazil. They're going to crush you. Like, don't. Just don't do it. And I did it mostly because everybody told me I couldn't do it.
"We talked to 30 CEOs, experts of the biggest banks in Brazil," he said in New York. In his ten-year letter he put it more simply: "The idea seemed absurd to the experts." Forbes, reporting the company's rise years later, recorded the warnings he had received from Brazilians that the banking establishment would block him, "or worse." The two words "or worse" carry a lot in a sentence about a man whose uncle had been kidnapped.
The rented house filled up. Around twenty people worked inside it. Vélez has said candidates who cared about how the job would look on their LinkedIn profiles took one look and left, and the ones worth hiring sat down on the floor and said they would be there Monday. The company's first customers, by its own account, were its founder and its employees. The first transaction on a Nubank card took place on April 1, 2014. Nobody seems to have commented on the date.
Nobody ignored it
The card was purple.
That decision now seems so obviously right that it is hard to remember how odd it was. Brazilian bank cards were the colors of institutions: navy, gold, black, the shades of a marble lobby. Purple was the color of nothing in finance. That was why Vélez chose it. He wanted the card to be, as he told Rest of World, "the most anti-bank possible." Then he described the reaction in a sentence that works as a complete marketing philosophy:
— David Vélez, to Rest of World, 2020Some people absolutely hated it … some people absolutely loved it, but nobody ignored it.
Brazilians nicknamed the card Roxinho, "the little purple one." When the company reached Colombia it became La Moradita. People give nicknames to things they have chosen to have a relationship with, and this was the first piece of Brazilian banking plastic in anyone's memory to get one. In 2018, after a redesign, Nubank set a Guinness world record for the largest simultaneous unboxing. Thousands of people opened envelopes at the same time, in a stunt that turned a bank card's arrival into an event. A bank had once made its customers wait months at a door. Now people stood in line, willingly, to open one of its envelopes.
The waitlist grew faster than the plan. Vélez has said Nubank built a waitlist of more than a million people within its first eighteen months. The five-year target of a million cards was reached, by one account, in about twenty-four. Growth came almost entirely through word of mouth. Jag Duggal, a former Google, Facebook, and Quantcast product executive who later became Nubank's chief product officer, put the figure at 80 to 90 percent of all growth. When Morgan Stanley's Latin America bank analyst Jorge Kuri sat down with Vélez, the conversation was framed around how the company had won more than 60 million customers almost entirely through referrals. A bank that does not spend on advertising is unusual. A bank that does not need to is something else.
Something stricter than a color sat underneath the purple. Sequoia's account of the company's early culture describes a decision that seems minor until you think about what it cost. When Nubank failed to send customers payment reminders, it reversed the late fees those customers had incurred, even though the money was legitimately owed and even though the reversal hurt profits. The logic is simple and expensive. If the company's mistake caused the customer's penalty, the company pays. The incumbents had built a business in which fees were nearly 40 percent of revenue. Nubank was quietly establishing that fees caused by its own failures were unacceptable. Employees noticed. So did customers.
The incumbents noticed as well. Sequoia's telling of the founding is a "David vs. Goliath" story, with the obvious pun, about competitors trying to crush the upstart. In 2016 the threat became legislative. A regulatory change put Nubank's existence at risk, and by the founders' account it was beaten back through a large public mobilization, with customers who had been using the purple card for barely two years speaking up for it. Inside the company Vélez chose what Sequoia calls "brutal honesty." He told employees how serious the threat was before he knew how it would turn out. Telling people you do not have an answer is a particular kind of courage, and it buys trust you cannot buy any other way.
With regulators he took the opposite stance from a typical disruptor. Nubank, by the Sequoia account, aimed to be "the kid in the front row," the student getting top marks rather than the rebel at the back. That eventually meant pursuing full banking licenses early, which slows growth and deepens it. The purple card was the rebellion, and the compliance file was the strategy. Vélez appears to have understood that you cannot be anti-bank for very long without becoming a bank.
A language he wished he spoke
In 2018 Sequoia asked Vélez, for a short company Q&A, what he didn't know that he wished he knew. His answer was not about credit risk, regulation, or Brazilian politics. It was a programming language.
"I wish I knew more about Clojure, which is the programming language we use most at Nubank," he said. "If I did, I could better understand some of the more complex problems we're trying to solve, and I think I could ask better questions when we're setting priorities." He had taken a couple of online courses and written a little Clojure, "but it wasn't enough. I want to dive in again at some point."
Clojure is a niche language, admired by those who use it and unknown to almost everyone else. A bank betting its core systems on it was either eccentric or making a deliberate claim about what kind of company it was. Nubank built its software in-house, all of it, at a time when incumbent banks rented theirs. In January 2020 it made its first acquisition, Plataformatec, a Brazilian engineering and agile-methods firm, partly to hire engineers. In July 2020 it bought Cognitect, the company behind Datomic and the steward of the Clojure language itself. Two years after Vélez said he wished he spoke the language better, his company owned the organization that maintained it. That is one way to close a skills gap.
The same Q&A described his calendar. "When your day is packed with nonstop appointments and meetings, it's easy to be reactive instead of proactive," he said. "It's like you're held hostage by your own schedule." So he blocked out two hours of thinking time each day, one in the morning and one in the afternoon, with Slack and email off. He would read, write in a journal, or walk. In early 2018, during those hours, he had been reading about Chinese financial services and decided to take several team members to see them firsthand. "That trip ended up being transformational. Nubank now has a very different strategy, based in part on what we learned." Later that year, in October, the Chinese technology company Tencent bought a $180 million minority stake.
He had moved, he said, so that his commute was a six-minute walk instead of an hour or more in São Paulo traffic. "I have a really good time just walking—sometimes I even wish it was a little longer." He trained for marathons, and ran Buenos Aires in 2019. "While he runs, he thinks," Rest of World reported, without irony. At home he had three children under three and made up bedtime stories for them every night. The reading pile on his nightstand at the time is almost too fitting: Ray Dalio's Principles, a biography of Walt Disney, Andre Agassi's Open, and Nassim Nicholas Taleb's Antifragile. That is a systems manual, a story about building a world of wonder, a confession about competing at a sport you hate, and a theory of benefiting from disorder.
The disorder was real. The data was as well. By 2020, Rest of World reported, Nubank held about 10,000 data points on each customer. Its reporters noted, with some unease, that "in the hands of a different boss, that data could be used for nefarious purposes." Vélez, they wrote, was "an optimist at heart." Both observations are probably true. A bank that knows its customers well enough to lend to people the incumbents ignored also knows them well enough to exploit them. The difference is not technical. It is the same decision as reversing the late fees, made again every day.
The branches close
Then the branches closed, all of them, everywhere.
The pandemic did for Nubank what years of persuasion could not. "The pandemic, in a way, forced that behavior to change among most of the population, because branches were closed and it wasn't even an option," Vélez told CNBC. "People started trying us and realized we were a better option—we charge no fees, better customer experience all in your smartphone, then suddenly we started seeing people in their 60s, in their 70s, in their 80s, people all across Brazil, in different municipalities, start using us." A generation that had never trusted money it could not see in a building had no building to go to. The bulletproof doors were locked from the inside.
The company had already crossed borders. Mexico, its first market outside Brazil, was announced in May 2019. Colombia was announced in November 2020. In September 2020 it bought Easynvest, Brazil's largest independent brokerage. By then it had raised $820 million across seven rounds and was valued at $10 billion. Fifteen months later it was worth four times that.
On December 9, 2021, Nu Holdings listed on the New York Stock Exchange. The offering raised about $2.6 billion, the largest Latin American IPO since PagSeguro's in 2018, according to Latinvex. On its first trading day the company was valued at roughly $45 billion, more than any bank in Latin America. It had 48.1 million customers, up from six million three years earlier, and was adding 2.1 million a month. The company claimed it had saved its customers $4.8 billion in fees and about 113 million hours of waiting. That last number translates the bulletproof door into hours. It was also still, at that point, a company that had only just reported a small profit after years of losses. Berkshire Hathaway, not usually a buyer of unprofitable fintech, invested a total of $1 billion before and after the listing. Before the IPO, Nubank had run a share program that brought millions of its own customers in as part-owners. The people who had once waited in queues now held stock in the alternative.
The listing made Vélez the richest person in Colombia. The man he displaced is worth a short sketch. Luis Carlos Sarmiento Angulo, founder of Grupo Aval, was the Colombian who had held the top spot on Forbes's list more often than anyone, a banker whose fortune had been built over a long lifetime in the old way: acquisitions, branches, an empire run from Bogotá. He was ninety in 2023. The top of Colombia's rich list now passed between two bankers born half a century apart, and the younger one did not own a single branch.
The two traded places repeatedly. Forbes Colombia noted the paradox in April 2023: when Vélez first became the richest Colombian in December 2021, Nubank "was not yet profitable in any of its operations." The stock then fell about 45 percent in a year of inflation and rising rates, his fortune dropped by more than $2 billion to $4.5 billion, and Sarmiento was back on top. By October 2023 Nubank was profitable, Morgan Stanley was forecasting a $100 billion market capitalization by 2026, and the shares closed on October 24 at $8.08. Forbes estimated Vélez's fortune at $8 billion against Sarmiento's $6.9 billion. In 2025 the magazine ranked him 236th in the world at $10.7 billion.
The financial statements tell the same story more slowly. Net losses in 2019, 2020, 2021, and 2022, the last of them about $400 million. Then $1 billion of profit in 2023, $2 billion in 2024, and $2.87 billion in 2025. Total assets grew from $2.8 billion in 2018 to $74.9 billion. The losses went on for years and the profits arrived late but in size. Anyone who had once watched a single cow become six would recognize the curve.
Return trips
"I returned to my country," Vélez wrote in his ten-year letter, "now with 10 other Nubankers."
The sentence describes a homecoming staffed like an expedition. Nubank started operating in Colombia in 2021, and the letter, written on the eve of another visit, is unusually sentimental for a CEO's newsroom post. "Despite leaving my country at a young age, I carry my roots with pride and nostalgia: the bandeja paisa, salsa, vallenato, and all the flavors and colors I experienced there." Paisa is the word for people from Antioquia, the region around Medellín, and Colombian magazines use it of him routinely, as if to reclaim him. Within two years Nu Colombia had more than 700,000 customers. It was the fastest-growing card issuer in the country. For 30 percent of those customers, La Moradita was the first credit card they had ever held. The company calculated it had saved Colombians 246 billion pesos, about $63 million, in management fees. "Seeing that impact in my country of origin," he wrote, "is undoubtedly rewarding."
There is a small coincidence in the record here. Search for David Vélez of Medellín and you also find another one: David Vélez Gómez, an audiovisual communicator who survived a gas explosion that burned much of his body and in 2017 published the first volume of a fantasy saga called El Caminante de Mundos, "The Walker of Worlds." The two men have nothing to do with each other. But the title fits the banker too.
He was raised a Colombian in Costa Rica, educated as a Latin American in California, made a partner as an outsider in São Paulo, and became a billionaire through a company listed in New York. By 2023, Forbes Colombia reported, he and his family were living in Uruguay. He holds about 75 percent of Nu's voting power. He kept selling stock: $191 million in August 2023 and roughly 31 million shares, about $404 million, in August 2024, leaving him with roughly a 20 percent stake. In September 2024 the Colombian Congress gave him the Order of Democracy Simón Bolívar at the rank of Grand Knight. The honor is named for the general whose project was a Latin America without internal borders.
The company now walks the same way. Nu applied for a full banking license in Mexico in October 2023 and expects its total Mexican investment to reach $4.2 billion by 2030. Late in 2025 it applied to the U.S. Office of the Comptroller of the Currency, and in January 2026 it received conditional approval for a national bank. On March 4, 2026, it announced that Inter Miami's new stadium at Miami Freedom Park would be called Nu Stadium. By May 2026 it reported more than 135 million customers: about 115 million in Brazil, 15 million in Mexico, and nearly 5 million in Colombia. In September it announced its U.S. launch, with U.S. Hispanic consumers identified as an important early market.
It is worth pausing on that. The bank was founded because a Colombian could not get through a door in Brazil. Its American entry point is the population most familiar with the experience of being a foreigner at the counter. The company that was the outsider in São Paulo is now the outsider in Miami, and its strategy assumes that, as before, being the outsider is an advantage.
Not every scene in the expansion is so neat. Twenty years after Vélez left investment banking, a Morgan Stanley analyst flew to São Paulo to interview him. In 2023 Endeavor named him Business Leader of the Year, and the man asking questions on stage was Martin Escobari, co-president of General Atlantic, Vélez's old firm. In May 2022, at a fintech conference in New York, his interviewer was Nigel Morris, co-founder of Capital One, the American credit card upstart that had shown in the 1990s that a new entrant could take on the incumbents with data and a card. Capital One had been an explicit model for Nubank. Morris's firm, QED, had backed Nubank's Series B in 2015. Morris opened by noting that Time had named Vélez one of the 100 most influential people of the year that morning. Vélez brushed it aside and started telling the story about the door.
Eight generations
In August 2021, four months before the IPO made him the richest man in Colombia, Vélez and his wife signed the Giving Pledge, committing to give away most of their fortune during their lifetimes.
Mariel Reyes Milk is a Peruvian-American entrepreneur. The sources say little more about her than that, along with her place in the pledge and the family: four children, the youngest born in 2022. She and Vélez together are another three-country partnership, Colombian, Peruvian, and American. In 2022 they launched VelezReyes+, a philanthropic platform whose first grants went to eleven organizations in Brazil, Colombia, and Peru. In 2023, with the Alquería dairy executive Carlos Enrique Cavelier, whose business card reportedly reads coordinador de sueños, "coordinator of dreams," they announced training in web development and artificial intelligence for 1,400 young people in Cundinamarca.
The model they chose was Chuck Feeney, the co-founder of Duty Free Shoppers, who gave away essentially his entire fortune, much of it anonymously, and lived modestly while doing it. They consulted Bill Gates. They also made a decision that ran directly against regional custom. "We saw that in Latin America philanthropy is done anonymously," Vélez told Forbes Colombia, "but we concluded that it was good to make it public, to inspire and show others how it can be done." Afterward he was struck by the emails he received from founders of other unicorns who wanted to follow. "People used to only want to be on the Forbes list as an indicator of success," he said. "Now there is a lot of interest in being on the Giving Pledge."
It would be easy to treat this as reputation work, and some of it is: anyone who announces a gift is also announcing himself. But the way Vélez explains it has the same logic as his company. Speaking in Spanish with Iván Duque, the former president of Colombia, he put it in actuarial terms: Se requieren ocho generaciones para alguien salir de la clase más baja a la clase media. It takes eight generations for someone to climb from the lowest class to the middle class. Sistemas donde se necesitan ocho generaciones realmente son sistemas muy injustos. Systems that require eight generations are deeply unjust systems.
That is the bulletproof door again, measured in lifetimes instead of months. A financial system that makes a newcomer wait five months for an account and a society that makes a poor family wait eight generations for the middle class are the same kind of machine. Each charges its fee in time, and the fee falls hardest on the people with the least of it. The founder who built a bank to cut five months down to five minutes is now trying to apply the same arithmetic to the longer wait.
"There is a limit," he told Duque, "to how much satisfaction or happiness money can bring." In another interview he compared the startup life to the Odyssey: the point is the journey, because ninety percent of startups fail and a founder who measures himself only by the destination will almost always feel he has failed. It is an unusually rueful thing for a man worth about $10 billion to say. It is easier to believe once you remember that his first fortune was six cows.
He sold them at eighteen. The money went toward Stanford, then Goldman, then Sequoia, then a rented house in São Paulo with twenty people on the floor, then a purple card, then a hundred million customers who would never have to stand in a bulletproof door. But the herd, in his own account, was only a good business because of a waiver. On a farm in Costa Rica a man with six cows let his son buy a seventh and quietly paid for the feed and the vet for six years. He never sent the boy a bill.
Part IIThe Playbook
What follows is an attempt to extract the operating logic from the story: not the mythology of the purple card, but the decisions underneath it that can be repeated. Vélez's playbook is less about technology than it looks. The software mattered, but the durable advantages came from three things the incumbents could have copied and chose not to: treating the customer's anger as a map, absorbing costs the industry insisted on passing on, and being very patient about where and how fast to grow. The twelve principles below are taken from the evidence in Part I.
The principles
- 1.Read the outrage as a market map.
- 2.Position in the scarcity, not the oversupply.
- 3.Pick the problem that makes smart people flinch.
- 4.Recruit the insider you are not.
- 5.Set the culture before you can afford to.
- 6.Interview the person, not the résumé; hire the spike.
- 7.Be impossible to ignore.
- 8.Make fairness expensive on purpose.
- 9.Rebel in the product, sit in the front row with the regulator.
- 10.Go deep before you go wide.
- 11.Own the language your company thinks in.
- 12.Schedule the thinking, then go see the future in person.
Principle 1
Read the outrage as a market map
The bulletproof door was not a revelation in itself. Millions of Brazilians had been through it. What Vélez added was a second piece of information that most of them did not have: he knew, professionally, how profitable the institution behind the door was. Bad service combined with extraordinary margins is not a sign of incompetence. It means there is no competition, and that means there is an opening.
Familiarity hides this. A Brazilian banker would have experienced the door as normal. An investor who arrived by armored car, like Rubinstein, would have experienced the margins without the door. Vélez had both experiences at once, and the tension between them was his thesis. The skill is less creativity than holding two facts together and refusing to accept that they can both be true.
Tactic: List the experiences in your industry that customers hate, then put each one next to the margin it protects; where you find the worst experience next to the fattest margin, that is the opportunity.
Principle 2
Position in the scarcity, not the oversupply
"You want to position yourself in the scarcity of the market, not in the oversupply," Vélez told Stanford students in 2022. He learned it as an investor. Sequoia closed its Latin American effort partly because the startups it met were copying Silicon Valley categories, food delivery and the like, in markets already full of copies. Capital, ambition, and imitation were plentiful. What was scarce was anyone willing to take on the region's most important and most protected industry.
The principle applies to careers as well as companies. Vélez left Wall Street credentials, which plenty of people have, for a combination very few people had: Silicon Valley venture training, firsthand Brazilian market experience, and a native's understanding of Latin American customers. Nobody else was holding all three.
Tactic: Before committing to a market, count the credible teams already attacking it; if you can name more than five, look for the adjacent problem nobody has been willing to take on.
Principle 3
Pick the problem that makes smart people flinch
"Do something that most people think is hard," Vélez advised in 2018. "When you tell someone what you're starting, their reaction should be, 'Are you sure you want to do that? That's too hard.'" His reasoning is mechanical rather than heroic. Easy ideas attract five competitors within two months. Hard ideas filter out competition, attract people who enjoy hard problems, and so make fundraising easier, because better teams with fewer rivals are easier to fund.
The thirty bank executives who told him it couldn't be done were, without meaning to, protecting him. Every expert who declined to compete was one fewer competitor. Vélez has said that he went ahead "mostly because everybody told me I couldn't do it." That sounds like stubbornness. It is also a reasonable reading of the odds: unanimous expert skepticism about a market with fat margins means the field is empty.
Tactic: When ten domain experts tell you an idea is impossible, ask each of them why they personally would not attempt it; if the answers are about difficulty and not about economics, keep going.
Principle 4
Recruit the insider you are not
Vélez's biggest liability, being a Colombian with no Brazilian banking experience, was the one every skeptic pointed to. He answered it by hiring it. Cristina Junqueira had spent years inside Itaú Unibanco, one of the five banks Nubank was built to beat, in consumer credit and card products. Edward Wible supplied the engineering foundation that neither of the other two could. The founding team was built around Vélez's gaps, not his strengths.
The broader lesson is that the outsider's advantage only lasts if someone on the team understands the inside. The outsider sees what is wrong. The insider knows why it was built that way and where the regulatory and operational traps are. Without the insider, the outsider's ideas fail on details. Without the outsider, the insider never questions the details at all.
N
The Founding Triangle
Three founders, three countries, three distinct gaps closed.
| Founder | Origin | What they supplied | Role |
|---|---|---|---|
| David Vélez | Colombia / Costa Rica | Capital access, outsider thesis, investor network | Vision |
| Cristina Junqueira | Brazil | Itaú Unibanco credit and card expertise | Insider |
| Edward Wible | United States | In-house technology and infrastructure | Build |
Tactic: Write down the single objection skeptics raise most often about you, and make your next senior hire the person who makes that objection irrelevant.
Principle 5
Set the culture before you can afford to
"The culture of a business is built by the first 10 to 15 employees in the first six months of life," Vélez has said, a lesson he credits to Sequoia. He is dismissive of founders who say they will deal with culture once they become a unicorn. "It's a little bit like saying, 'I'm too busy to educate my kids. By the time they turn 18, I'll start working educating.' Forget it. It's too late."
The rented house was a cultural filter as much as an office. The candidates who left were not the people Nubank needed, and the ones who sat on the floor were. Culture there was not a document. It was the decisions people saw being made in the first months, the obvious ones and, more importantly, the costly ones. "Culture allows you to hire people, people build products, products bring you customers," Vélez told Stanford. The chain only works in that order.
Tactic: Before your tenth hire, write down the three decisions you made that cost the company money in order to keep a value, and make sure every new employee hears those stories in their first week.
Principle 6
Interview the person, not the résumé; hire the spike
Doug Leone's interview about Vélez's mother became Nubank's hiring method. "My interviews today are not interviews about career experience," Vélez has said. "They're all about character traits. I don't really want to spend time asking you about what school did you go to." The aim is to find out what actually drives a person, which a CV cannot show.
The second half of the method goes further. Vélez prefers candidates with large strengths and large weaknesses to candidates who are competent at everything. "I'd rather have the people that have this incredible strengths and the weaknesses than the people that are okay in just certain areas. Those tend to be, if you manage them adequately, much more powerful type of hires." The qualifier, "if you manage them adequately," carries much of the weight. Hiring for spikes only works if the organization is designed to cover the weaknesses, which is the same reasoning that produced the founding team.
Tactic: Spend at least half of every senior interview on the candidate's family, upbringing, and formative failures, and reject candidates whose only selling point is the absence of weaknesses.
Principle 7
Be impossible to ignore
A purple card in an industry of navy and gold was a positioning decision. Vélez wanted the product to be "the most anti-bank possible," and he accepted that some customers would hate it as the cost of making others love it. "Nobody ignored it" is the right measure for a challenger, because an incumbent's best defense is that the alternative is invisible.
Distinctiveness also made growth cheaper. A card people nickname is a card people show their friends. Word of mouth accounted for 80 to 90 percent of Nubank's growth, by its product chief's estimate. That is what happens when the product itself is the advertisement. Jag Duggal's phrase for it, "fundamentally different, not incrementally better," describes the same choice. A product that is ten percent better is something people consider. A product that is visibly different is something they talk about.
Tactic: Test every major brand or product decision by asking whether a meaningful share of customers will dislike it; if nobody would, it probably isn't different enough to be talked about.
Principle 8
Make fairness expensive on purpose
The incumbents earned nearly 40 percent of their revenue from fees. Nubank reversed late fees whenever its own failure to send a payment reminder had caused them, even though the money was technically owed. That decision did more for the culture than any slogan, because it cost something. Values that cost nothing are not tested. Values that cost revenue are, and both employees and customers can tell the difference.
Sequoia's summary calls customer love "the ultimate moat." The moat comes from accumulated sacrifices of exactly this kind. Each fee waived or service hour absorbed builds trust that a competitor cannot acquire quickly, because trust only comes from a history of decisions. Vélez's own early profits from the cows depended on a father who covered the costs. Nubank built its business by covering costs the industry had always passed on.
Tactic: Find the fee in your business that is legal but that your customers would call unfair, eliminate it, and tell your team why you are giving up the revenue.
Principle 9
Rebel in the product, sit in the front row with the regulator
Nubank's public identity was rebellious. Its regulatory posture was the opposite. By its founders' account, the company aimed to be "the kid in the front row," the model student, and it pursued banking licenses earlier than its growth strictly required. When the 2016 legislative threat came, the combination worked: a product popular enough to mobilize customers, and a compliance record clean enough that regulators had no reason to side against it.
Inside the company, crisis meant candor. When the threat was serious, Vélez told employees so, without pretending to have a solution. The reasoning is simple. Comfort that later turns out to be false destroys more trust than honesty, and a team that knows the real stakes will work harder on the real problem.
R
Two Postures, One Company
Where Nubank chose to rebel, and where it chose to conform.
| Domain | Industry norm | Nubank posture | Mode |
|---|---|---|---|
| Brand | Institutional colors, marble lobbies | Purple, "the most anti-bank possible" | Rebel |
| Fees | Nearly 40% of revenue | No annual fees; reverse self-caused charges | Rebel |
| Distribution | Branches and guarded doors | Smartphone only, word of mouth | Rebel |
| Regulation | Challengers fight regulators | "The kid in the front row"; early licenses | Conform |
| Crisis communication | Reassure the team | Disclose the stakes before the answer | Candor |
Tactic: Separate where you will defy convention, which should be the customer experience, from where you will exceed it, which should be compliance, and never mix them up.
Principle 10
Go deep before you go wide
The seed deck described a global ambition and began with one product in one country. Nubank spent six years in Brazil before announcing Mexico, and it remains concentrated in three large markets rather than thinly spread across many. By Vélez's account, the company chose to become people's primary bank, which requires licenses and builds slowly, rather than an extra wallet they rarely open.
He has also been candid about how much remains. In 2023, with nearly half of Brazilian adults as customers, he noted that only about 60 percent of them used Nubank as their main account, and that its share of credit, loans, investments, and insurance was still small. A founder who thinks the company has already won stops looking for the next layer of depth. The financials show what patience costs and what it pays: four years of losses, then profits that tripled within two.
$
The Patience Curve
Nu Holdings net profit and deposits, in US$ billions.
| Year | Net profit | Total deposits | Phase |
|---|---|---|---|
| 2019 | -0.1 | 2.7 | Build |
| 2020 | -0.2 | 5.6 | Build |
| 2021 | -0.2 | 9.7 | Build |
| 2022 | -0.4 | 15.8 | Trough |
| 2023 | 1.0 | 23.7 | Inflect |
| 2024 | 2.0 | 28.9 | Compound |
| 2025 | 2.9 | 41.9 | Compound |
Tactic: Before entering a new market, measure what share of your current customers treat you as their primary provider; until that number is a majority, the next dollar probably belongs in depth, not breadth.
Principle 11
Own the language your company thinks in
Incumbent banks rented their technology. Nubank built its own, in a niche programming language, and then bought the companies that would deepen that capability: Plataformatec for engineering talent and Cognitect, steward of Clojure itself. Vélez's stated regret in 2018, that he did not know Clojure well enough to ask his engineers better questions, shows a CEO who regarded technical fluency as part of the job.
The point is not that every founder should learn to code. It is that when the core system is the product, the people setting priorities have to understand it closely enough to know what is hard. A CEO who cannot follow the technical argument ends up deciding by politics. Ownership compounds as well. Every system built in-house is one an incumbent would have to rebuild from scratch to match.
Tactic: Identify the one technical domain your competitive advantage depends on, and spend enough personal time in it, through courses, code reviews, or acquisitions, that your engineers stop simplifying their explanations for you.
Principle 12
Schedule the thinking, then go see the future in person
Vélez blocks two hours a day for unstructured thinking, with Slack and email off, because otherwise "you're held hostage by your own schedule." Those hours led directly to strategy. Reading about Chinese finance turned into a team trip to China, which he called "transformational," and which helped reshape Nubank's direction in the same year Tencent invested.
Two habits are combined here: protecting time to think, and refusing to rely only on what you read. The trip mattered because reading about something is not the same as seeing it, which Vélez had learned firsthand in a bank lobby. His Sequoia years, flying fourteen hours to São Paulo to meet seven companies in a day, taught the same thing. Physical presence surfaces things reports leave out.
T
Decision Timeline
The key moves, from the cow to the stadium.
c. 1993
At 12, buys one cow in Costa Rica; the herd grows to six by age 18.
2005
Graduates Stanford in management science and engineering.
2010
Recruited by Sequoia from Stanford GSB to evaluate a Brazil office.
2012
In October, Sequoia shelves Latin America; Vélez declines a move to San Francisco.
2013
Nubank founded May 6 with Junqueira and Wible; seed from Sequoia and Kaszek.
2014
First Nubank card transaction, April 1.
2016
Legislative threat beaten back through customer mobilization.
2018
China trip reshapes strategy; Tencent invests $180 million.
2019
Mexico announced, the first market outside Brazil.
2020
Acquires Plataformatec, Cognitect, and Easynvest; Colombia announced.
2021
Giving Pledge in August; NYSE IPO in December at about $45 billion.
2023
First full year of profit: about $1 billion.
2026
Conditional US bank approval in January; Nu Stadium announced March 4.
Tactic: Put two recurring one-hour blocks in your calendar with every communication tool off, and commit to visiting in person, within ninety days, any market or model you read about during those hours that seems to point to the future.
Free playbook
Get The Business Model Playbook
58 business models, one visual page each: how the money flows, the metrics that matter, and who runs it. Free when you join the Faster Than Normal email.
Free. No spam. Unsubscribe anytime.
Part IIIQuotes and Maxims
In their words
— David Vélez, Stanford GSB View From The Top, 2022You want to position yourself in the scarcity of the market, not in the oversupply.
— David Vélez, Deep Purpose podcast with Ranjay Gulati, 2022The culture of a business is built by the first 10 to 15 employees in the first six months of life.
— David Vélez, Sequoia Capital, Seven Questions, 2018If you try something easy, there will be five other companies doing the same thing two months later. But if you try something that's difficult at first, everything gets much easier as soon as you make it through those initial challenges.
— David Vélez, Stanford GSB View From The Top, 2022Culture is the driving force, there's nothing more important, because culture allows you to hire people, people build products, products bring you customers.
— David Vélez, in conversation with Iván Duque, Wilson CenterSe requieren ocho generaciones para alguien salir de la clase más baja a la clase media. Sistemas donde se necesitan ocho generaciones realmente son sistemas muy injustos.
Maxims
- The contradiction is the company. When the worst service sits next to the fattest margin, you have found a market that nobody is defending.
- Outsiders see the door. Familiarity makes dysfunction invisible; the person kept waiting in the vestibule is the one who notices the vestibule.
- Hire the insider you are not. An outsider's thesis needs someone who knows where the traps are.
- Difficulty is a filter. Problems that make experts flinch keep competitors away and attract people who like hard work.
- Being ignored is the real risk. A product some people hate and others love will spread; a product everyone tolerates will not.
- Values have to cost something. A fee you waive at your own expense teaches more culture than a value statement.
- Rebel with customers, conform with regulators. Challenge the experience and comply on the rules.
- Depth compounds, breadth dilutes. Become the primary relationship before becoming a presence everywhere.
- Interview the mother, not the résumé. Character is what lasts; credentials are what get described.
- Waived costs are the real gift. The best returns often rest on fees someone else chose not to charge.
