Part IThe Story
Where Does He Work?
The guard did not recognize the name. In 2014 a BBC business reporter named Irene Caselli walked up to the entrance of MercadoLibre's headquarters in Buenos Aires and asked for the founder. At the time the company operated in twelve countries, from Mexico down to the bottom tip of Chile. It had 109.6 million registered users, $473 million in revenue the previous year, and was the most visited e-commerce site in Latin America. "Marcos Galperin?" the security guard said. "Where does he work?"
He worked upstairs. He had worked there, or in some earlier version of there, for fifteen years. He was the founder, the chief executive, and the owner of a stake then valued at roughly $400 million. He came to the office in jeans and trainers, and he did not arrive in a chauffeur-driven car. Caselli, to her credit, wrote that the guard deserved some defense: the 42-year-old Galperin kept a very low profile. That was understatement. The guard's blank look was the system working as designed.
— Marcos Galperin, to the BBC, October 2014I see many entrepreneurs who use their company as a platform to be famous. For us it's exactly the opposite, we want the company to be famous. The lower our personal profile, the better.
Now jump forward twelve years. The company's name runs down the sidepods of the Formula One cars driven by Franco Colapinto. It is the title sponsor of Argentina's Liga Profesional de Fútbol for the 2026–27 season and sponsors every CONMEBOL club competition, which means it is printed, stitched, or projected onto nearly every surface a South American football fan will look at on a Sunday. MercadoLibre's market value passed $100 billion, making it the most valuable company in Latin America. Its founder is, by the usual estimates, the richest man in Argentina, with a fortune somewhere between $9 billion and $9.5 billion.
And he is still hard to see, though no longer for the same reasons. He lives across the Río de la Plata in Uruguay, between Montevideo and Punta del Este. He gives few interviews. When he returned to Argentine media in 2024 with a long conversation in La Nación, the return itself was treated as news. Yet Argentines know his voice, or a version of it, intimately. Clarín described him in a phrase that is hard to improve: de perfil bajo en lo personal, pero de "teclado rápido" en las redes sociales. Low profile in person, a fast keyboard online. The man who wanted the company to be famous and himself to disappear has become the most argued-about capitalist in a country that has never made up its mind about capitalists.
In January 2026 the Argentine writer Hernán Vanoli published an essay on him in the magazine Iceberg titled "Marcos, el monarca." Vanoli's argument was that Galperin represents a mutation in sovereign power. The old kings displayed themselves in grand ceremonies. The new kind withdraws the body into shadow while the power "expands into the telephones of the populace." It is a hostile reading, a clever one, and it is half right. Galperin did not invent his invisibility in order to rule. He invented it, as far as the record shows, because he believed the founder was the least interesting part of the company. What he did not anticipate, and what the rest of this account follows, is what happens to a man who builds the infrastructure of trust for a continent that has learned, generation by generation, not to trust anything.
By the Numbers
The Machine Behind the Man
$28.9BMercadoLibre revenue, fiscal 2025 (up from $0.65B in 2015)
123,670Employees at year-end 2025
18Countries of operation across Latin America
2.9BItems delivered, trailing twelve months to Q2 2026
7 yearsTime to first profitability; the original plan said two
$289MRaised in the 2007 Nasdaq IPO, the first by a Latin American tech company
26 yearsGalperin's tenure as CEO, August 1999 to December 31, 2025
A Garage on Calle Tronador
MercadoLibre was founded on August 2, 1999, in a parking garage. The detail is so close to American founding myth (Hewlett-Packard's garage in Palo Alto, Apple's garage in Los Altos) that it reads like quotation. It was nothing of the kind. The garage sat on the ground floor of the Panamericana Plaza building on Calle Tronador, in the Buenos Aires neighborhood of Saavedra. It belonged to the Galperin family's leather company, and it had the best broadband connection the founders could find.
"It was a building with terrible offices," Galperin told Fortune in 2024, "but at least we had good connectivity."
That sentence holds most of the company's operating philosophy. Accept the terrible offices. Insist on the connectivity. The founding team built makeshift cubicles in a windowless space. Eventually the building threw them out; the account in Fortune breaks off at the moment it explains the violation, which seems right. Every founding story should contain at least one eviction.
The street turns up again in the record. In July 2007 the U.S. Securities and Exchange Commission sent a comment letter on MercadoLibre's IPO registration statement, routine and dry. It was addressed to "Marcos Galperin, Chief Executive Officer, MercadoLibre, Inc., Tronador 4890, 8th Floor, Buenos Aires." Eight years after the garage, the company was on the same street, eight floors up. By 2016 its 10-K listed a new address, Arias 3751, 7th Floor. By 2015 the Financial Times found Galperin in designer offices spread across three floors at the top of a glass tower overlooking the muddy water of the Río de la Plata. By 2017 the offices were in Vicente López, along the invisible line where Buenos Aires gives way to the northern suburbs, with graffiti on the walls and baskets of fruit in the kitchen. And today the company is headquartered in Montevideo, on the other side of that same river.
Up, then sideways, then across. That is the company's physical trajectory, and in a rough way it is the founder's too.
The Man Who Was Not Warren Buffett
In March 1999, two months before he was due to graduate from Stanford's Graduate School of Business, Galperin sat in a nearly empty classroom waiting for a lecture by Jack McDonald. McDonald was one of his favorite professors. Over a fifty-year career he taught some 10,000 MBA and executive-education students, and he was known for bringing in prominent businesspeople to tell students what the early years had actually been like. A man sat down at the next desk. They talked for perhaps three minutes.
"Hi, I'm Warren Buffett," the man said, smiling. "I've just disguised myself as John Doe."
Galperin had not recognized him. It was 1999. Google Images did not exist, and a business student from Buenos Aires had no particular reason to know the Oracle of Omaha's face. The realization arrived later, when the rest of the class crowded around the man at the next desk. After the lecture Galperin went to McDonald and confessed that he had wasted a golden chance to pitch his idea to Buffett.
McDonald's answer, as Galperin recounted it to La Nación in 2024, was practical. Don't worry. Buffett doesn't invest in Latin America. John Muse does, and he's coming to our class.
John Muse was a Texan and co-founder of the private equity firm Hicks, Muse, Tate & Furst. He already held stakes in Latin American businesses, including the cable operator Cablevisión. He was the right man. The question was how to get time with him. McDonald arranged for Galperin to drive Muse to his plane after the lecture. Galperin brought along two classmates, Marcelo Benítez and Ignacio Giraldo. This was a mistake. The classmates would not stop talking. In one later account Galperin recalled, laughing, that a Colombian friend nearly killed MercadoLibre before it was born simply by holding the floor. With the airport approaching and the pitch still undelivered, Galperin did what any rehearsed and desperate twenty-seven-year-old might do. He got lost on purpose.
He has said he rehearsed that pitch 150 times. In another telling the number is 700. Such numbers tend to grow with retelling, and either way the point stands: he had prepared for the conversation obsessively, and the circumstances of it were almost entirely luck. Muse listened. Muse became, in Galperin's word, embalado, carried away. Then Muse said something Galperin had not let himself expect: "I love your idea. Let me know if you need money."
"I went back home and told my girlfriend," Galperin later told the Financial Times, "'You remember that trip we were going to do to Europe after graduation? Well, it's not going to happen.'" On March 11, 2000, he married Karina, with whom he has three children. Whether she was the girlfriend whose European trip was cancelled is the kind of thing he does not volunteer, and I have not found it confirmed.
The car he drove to the airport was a green Volkswagen Golf. By his own account to Fortune, he also sold a Volkswagen Golf at Stanford through an early online classified ad. That was one of the small experiences that convinced him the internet would change how ordinary people bought and sold things. The vehicle that carried the pitch and the vehicle that proved the thesis may well have been the same car.
The thesis itself had been tested before Muse ever heard it, and it had failed the test. At Stanford, Galperin surveyed twenty of his Latin American classmates about the plan for an eBay-style marketplace for the region. In a 2015 talk at the Universidad de San Andrés he said all twenty told him it was unviable. La Nación, telling the story nine years later, put it at nineteen of twenty. The difference hardly matters. The people best placed to understand the market he wanted to serve thought he was wrong.
— Marcos Galperin, Stanford GSB View From The Top, November 2019Everyone said this is never going to work in Latin America — no one will buy a product they haven't touched, not in that culture. But I believed, and just knew we had to build a system that worked well.
He had spent long hours in the library working on the business plan, which made at least one classmate suspicious. Stelleo Tolda, a Brazilian in the same MBA class, said later that he wasn't used to seeing Galperin there and concluded that "you're on to something." Tolda became a co-founder and, for two decades, the company's chief operating officer and its Brazilian anchor. Hernán Kazah, another member of the class of 1999, also signed on; Galperin, by his own account, convinced him. Kazah would later leave to co-found Kaszek Ventures with Nicolás Szekasy, who served as MercadoLibre's chief financial officer shortly after the founding. Kaszek raised $95 million for its first fund in 2011 and $135 million for its second in early 2014, with Sequoia Heritage and Horsley Bridge among the backers. The garage, it turned out, had produced not just a company but much of the leadership of the Latin American venture industry that followed.
One more Stanford detail belongs here. According to Galperin's own account on the 20VC podcast, part of what freed him to start the company was the acquisition of the firm he had been expected to return to after business school. That firm was YPF, Argentina's oil company, which had sponsored his path to the Stanford MBA. A plan that would have put him back on an Argentine corporate ladder was dissolved by someone else's deal. He has rarely made much of this. It is worth noting all the same.
Grain Sacks and Cattle Hides
Galperin's paternal grandfather arrived in Argentina from Russia very poor. He had not finished primary school and did not speak Spanish. He found work loading and unloading sacks of grain from railway cars. With nowhere else to stay, he lodged at the Hotel de Inmigrantes, the immigrants' hotel near the port, and ended up living there for months. Galperin's maternal grandfather came from Germany during the crisis of the 1930s, with somewhat more resources. Galperin told these stories to the Colombian magazine BOCAS in 2019, when he was 47. He told them plainly, the way people tell stories they have heard all their lives and still haven't quite finished thinking about.
Within two generations the family owned SADESA, a tannery that describes itself as one of the largest leather manufacturers in the world. His father, Ernesto Galperin, was a prominent figure in the leather industry. His brother Miguel now runs the business. Another brother, Nicolás, sits on MercadoLibre's board as a Class II director. Marcos, born in Buenos Aires on October 31, 1971, was the fourth of five sons of Ernesto and Silvia Julia Lebach. He was born, as Clarín put it, into comfort, but raised in competition.
This is the part of the story that his critics treat as decisive and his admirers prefer to skip. Galperin was not a scrappy outsider. He attended St. Andrew's Scots School in Olivos, a bilingual school founded in 1838 by Scottish immigrants to educate children in English. There he played rugby, toured Australia and New Zealand, and was selected for Argentina's junior national team. He gave up the dream of playing professionally. In 1990 he left for the Wharton School at the University of Pennsylvania and graduated with honors in 1994. When the family garage was needed, there was a family garage. Vanoli's essay, otherwise unsparing, insists that Galperin is not a nepobaby, not someone living off inherited visibility. Vanoli calls him "an unmistakable child of the social fracture" that separates those born into business families from everyone else. Both things are true, and the story requires both.
Two smaller details from his childhood complicate the portrait of a polished heir. When he was ten, Ernesto brought home the boy's first personal computer, an unusual gift in an Argentina where PCs were not yet common. Ernesto also brought him books about Bobby Fischer, the brilliant, difficult American chess champion, whom the teenage Galperin admired. For a while he imagined becoming a scientist. Wharton changed his mind.
After graduating he went home to the Argentina of the 1990s, which, under convertibility, looked briefly like a country with a future. Through a Wharton friend, José Estenssoro, son of YPF's president, he joined the state oil company. From 1994 to 1997 he worked there as a futures and options associate, running the company's currency and oil derivatives program. In the summer of 1998, between his two years at Stanford, he worked in the fixed-income department of J.P. Morgan Securities in New York. Rugby, derivatives, a Bobby Fischer biography. A man learning, separately, how to take a hit, how to price risk, and what it costs to be brilliant and impossible. Years later he told BOCAS he was a fan of athletes' biographies. He coached his son's rugby team at the same school in Olivos.
The family had moved, in one lifetime, from carrying other people's goods off trains to processing hides for export to the world. The grandson would build something that carried other people's goods across a continent. Nothing in the record suggests he planned it as a family rhyme. It is one anyway.
The first thing to understand about MercadoLibre in 1999 is that it was not original. Kazah has been especially blunt about this. "We weren't very innovative," he said at Stanford in 2019. "At the time there were like 80 other companies throughout Latin America following the same business model. We may have had a great team and long-term vision. But the idea was not unique. Even our friends sometimes got confused about which business was ours."
The most formidable of the eighty was DeRemate, founded by a group of Harvard Business School MBAs under an Argentine named Oxenford. DeRemate launched the same year. It raised more first-round money than MercadoLibre's $52 million and spent it the way the decade expected money to be spent: on mass-media advertising, aggressive growth, and an early public offering. "Their focus was on marketing and going public very soon," Kazah said. "Ours was around building organic engines for growth."
Then the Nasdaq broke. Within months of the founding, the dot-com bubble burst and the venture capital that had chased Latin American internet companies disappeared. Kazah remembers about ten investors seriously interested in MercadoLibre at one point. One by one they all dropped out. Some of the investors already in the company, investment banks that had bought in at the top, lost their nerve and wanted to shut it down. "We had to go to a shareholder vote," Galperin said, "and almost lost."
Then Argentina broke too. In late 2001 the country defaulted on its sovereign debt, at the time the largest default in history. The government froze bank deposits under the corralito, and the middle class watched its savings become theoretical. Galperin was not sleeping well, or rather, he was not getting up. In a conversation with Endeavor years later, he described those mornings with unusual candor:
"There were times, especially in the beginning, when I remember that I did not want to get out of bed. The alarm clock rang in the morning and said: No, I do not want to get out of bed because I have to face all my employees and … the company is going to melt, it is not going to work, we do not have money to pay salaries, all these people who left their jobs or who believed in me and my idea."
The company survived for two reasons. The first is in the arithmetic. MercadoLibre had raised serious money from Hicks Muse, JPMorgan Partners, Flatiron Partners, Goldman Sachs, GE Capital, and Banco Santander Central Hispano, and it had not spent that money on television. "We were the only ones that didn't throw the money away doing mass media advertising," Galperin told the FT. "Instead, we focused on getting a team together and a good product." Speaking at Stanford in 2011, he was more specific: "We really saved the money for the rainy days. So, when our competitors ran out of cash, we still had $40 million in the bank. That was important."
The second reason is stranger, and it belongs to Argentina. When the banks froze, people who had lost access to their savings turned to MercadoLibre to sell what they owned: cameras, furniture, collections, cars. A country's financial collapse became one of the platform's early growth engines. A marketplace named mercado libre, free market, rose on the wreckage of a state that had stopped letting its citizens reach their own money. The irony was available to anyone who wanted it. Galperin, who would later become one of the loudest free-market voices in Argentine public life, has never needed it underlined.
In September 2001, just before the default, eBay bought a 19.5 percent stake. The structure deserves attention. eBay paid in part with its recently acquired Brazilian subsidiary, iBazar, which folded MercadoLibre into the largest market in the region. And, as the BBC reported, eBay pledged not to enter Latin America for at least five years. The biggest e-commerce company in the world had looked at the region, looked at the Argentine kid's company, and paid to stay out. Without the global giant as a competitor, MercadoLibre was free to dominate. eBay held its stake for fifteen years and sold it in 2016.
The company broke even in 2005. In 2008, with the bubble long gone, it spent $50 million to buy the remaining half of DeRemate, its Harvard rival, for a fraction of what DeRemate had raised. Recounting this to Stanford students in 2011, Galperin allowed himself a joke: "So, Stanford is better than Harvard." Then he gave them the real lesson, which was about cash.
Escrow for a Country Without Trust
On paper, the business was eBay's: let strangers sell things to strangers and take a cut. In practice, the main obstacle in Latin America was never software. It was faith. Buyers would not send money to sellers they had never met. Sellers would not ship to buyers who might never pay. Transactions failed, as one later analysis of the company put it, "not because of lack of demand, but because of lack of confidence."
So MercadoLibre stepped into the gap. The first version of Mercado Pago, built in the early 2000s, was not a fintech product in any sense the word would later carry. It was escrow. The buyer paid MercadoLibre. The seller shipped. Only when the buyer confirmed receipt did the money move. It was a narrow fix for a specific failure. In February 2004 Osvaldo Giménez, an engineer of the founding generation, took over Mercado Pago's operations. He has run it, in one title or another, ever since, and became Fintech President in August 2020. What started as an escrow layer now processes payment volume measured in hundreds of billions of dollars a year, offers digital accounts, insurance, investment funds, and credit, and for millions of Latin Americans is effectively the financial system.
The company's logo is a handshake: two hands meeting over a deal. During the pandemic Galperin changed it on the company's channels to an elbow bump. It was a small joke about social distancing, and an unintended admission that the gesture at the center of the brand had always been a stand-in. A handshake is what people do when they cannot rely on institutions. MercadoLibre's whole history can be read as an attempt to replace the handshake with something more dependable.
That attempt kept forcing the company to build things it had never meant to build. Galperin explained the logic most directly to the FT in 2015, and the sentence reads almost like a theorem: "The more inefficient retail is, the more value we can provide to a society." To Fortune he described the problem in starker terms. Brick-and-mortar retail reached only the big cities, and the countryside was cut off. "People would rather live in a slum next to a big city than live in better conditions in the countryside," he said. "So I was convinced this was going to work really, really well in Latin America."
Payments came first, then shipping. For years MercadoLibre relied on the region's postal and carrier networks. In 2019 it opened its first distribution centers in Argentina, Brazil, and Mexico, and the next year announced centers in Chile and Colombia. By 2025 it had more than 25 fulfillment centers with 30 million square feet of capacity, hundreds of service centers, the largest electric-vehicle fleet in Latin America, and six Boeing 737-800 converted freighters flown by the Brazilian airline Gol. In 2024 it handled 1.8 billion items, 49 percent of them delivered the same day or the next, in a region where, as Galperin admitted at Stanford, "the stereotypes about that market are true. Infrastructure is not as developed."
Credit came after logistics. In December 2013 an Argentine business publication reported, with a slightly alarmed parenthesis, that MercadoLibre was moving toward becoming "(also) a bank." Mercado Crédito was offering loans of up to 30,000 pesos. In 2018 the company partnered with the BIND group so that wallet balances could earn returns in a mutual fund. A marketplace that began by holding a buyer's money for a few days in escrow was now lending money to buyers and sellers, underwriting them on data no traditional bank had.
Some of the craft lived at a much smaller scale. Until late 2012, a seller with one style of high-heeled shoe in ten sizes and ten colors had to create a hundred separate listings, one per combination. Collapsing them into a single listing with dropdown menus took, as Stanford's reporting put it with some understatement, "serious reprogramming." It was not glamorous work. It was the kind of work Galperin cared about. "In the past, we did better by focusing on the product and a really good user experience and not so much with building our brand through mass marketing," he said in 2014. "We saved most of our money, our competitors ran out of money, and we had a better product."
The phrase for this in strategy decks is vertical integration. Inside MercadoLibre it looked more like a series of reluctant admissions: nobody else is going to fix this, so we have to.
The Film, Not the Photograph
At the Universidad de San Andrés in June 2015, Galperin said something that has stayed central to how the company understands itself. "There is nothing more dangerous in this industry than looking at your photo and falling in love," he told the students. "It's full of stories of successful companies that in a few years stopped being so. I like to look at the whole film, not the photo."
The film was long. The original business plan said MercadoLibre would break even in two years. It took seven to become profitable, a fact Galperin repeats so often that it works as an inoculation. In 2006, just before the IPO, the company recorded $52 million in revenue and $1.1 million in net income, a profit so thin it was almost symbolic. In August 2007 it became the first Latin American technology company to list on the Nasdaq, under the ticker MELI, raising $289 million. The timing was bad in a way that turned out to be good. The global financial crisis was beginning that same month, Lehman Brothers would collapse the following year, and MercadoLibre went into the storm with fresh capital that none of its rivals had.
By 2011 gross merchandise volume had reached $4.8 billion, up 41.5 percent, and net income was $76.8 million. Internet penetration in Latin America, 3 percent when the company was founded, had passed 30 percent. The company was profitable, dominant, paying a dividend. This is exactly the photograph Galperin warned against falling in love with.
The threat arrived in 2012, when Amazon entered Latin America. In March 2014 Galperin told Stanford students that margins had come down and the company faced "a lot of new competitors, so we will need to adapt." Then he said the thing that defined the next decade: "Profitability for us is never the driving force. … In this stage of the internet, we think market share is more important than having very high margins."
The numbers bear him out, and they are painful to read if you held the stock. Revenue rose from $1.29 billion in 2017 to $2.29 billion in 2019. Operating income went the other way: $0.18 billion in 2016, $0.05 billion in 2017, then losses of $0.07 billion in 2018 and $0.11 billion in 2019. A Stanford case study written in 2020 described the bet without decoration. Galperin expected to lose money in the short term as spending rose, and he "hoped investors would recognize this." The company was building warehouses and fleets and credit books in a region where most investors were still not sure e-commerce would work.
Somewhere in this period, the company also turned down an acquisition offer. Galperin has described it only obliquely: an offer that looked lucrative next to the share price of the day, rejected because he believed the company could keep growing for another twenty years. He has called it one of the most painful decisions he ever made, alongside the layoffs of good people the company could not afford to keep. The buyer has never been named in the sources available to me. The company is now worth far more than the offer.
Then came 2020. The warehouses were built just in time for a pandemic that closed tens of thousands of shops across the region. "The pandemic has moved us forward maybe five years in the last 12 months," Galperin said. Volumes at Mercado Envíos grew roughly fivefold almost overnight. Revenue went from $2.29 billion in 2019 to $3.97 billion in 2020 and $7.06 billion in 2021. In 2022, in the middle of a global tech contraction, the company announced it would add 13,000 employees. It went on to post more than thirty percent year-over-year revenue growth for twenty-seven consecutive quarters, a streak one analysis believed no other public company had matched. In 2025 revenue was $28.89 billion and operating income $2.84 billion.
None of this changed how he talked. "We're always paranoid," he told Fortune in early 2024, with his company second only to Petrobras in Latin American market value. "We don't think it's game over by any means." In 2015 he had told the FT almost the same thing in more personal terms: "I feel exactly the same way as I did 15 years ago. I feel a desperation that there is so much to do, that this is just starting." Desperation is an odd word for a man worth hundreds of millions of dollars, sitting, as the FT found him, in a hooded top with his feet up on a sofa and a laptop on his knees. It was not an affectation. Some people's anxiety is a form of competitive advantage. His best ideas, he has said, come around five in the morning, half asleep, which is why he keeps a notepad on his nightstand.
Across the River
In February 2020, Galperin moved to Uruguay.
He had lived there before; he briefly returned to Argentina during Mauricio Macri's presidency and then left again. Uruguay is less than an hour's flight from Buenos Aires, across the estuary his former office tower overlooked, and for Argentines it has long been the place the wealthy go when the home country becomes unbearable. He cited a friendlier business environment. Many other Argentine businessmen followed. MercadoLibre's headquarters followed too. The company is now based in Montevideo, incorporated in Delaware, earns most of its revenue in Brazil, and is described in its own encyclopedia entry as "a company of Argentine origin." The phrase is technically accurate and a little chilly.
The move did not lower his profile. If anything it emboldened him. Argentina had been governed for most of his company's life by the Peronist coalition in its Kirchnerist form, and Galperin's relationship with that government was, in Clarín's word, gélida, frozen, marked by lawsuits and public exchanges of insults. He said what he thought of the national economy to The Economist, and it was not diplomatic. Argentina, he said, was a patient who "is now morbidly obese, a drug addict, has cancer, AIDS and is an alcoholic."
When Javier Milei won the presidency in late 2023, Galperin posted one word on X: Libres. Free. Given the name of his company, the word had two meanings, which was presumably the point. Over the following years he became one of the most prominent business validators of Milei's libertarian program, to international investors and to the Argentine elite. In April 2025 he told CNBC in Miami that he was bullish on the reforms, that a U.S. trade war was "a very big opportunity" for Latin America, and that Mexico might be its biggest winner.
His fights with the state have become more local and more literal. In 2025 MercadoLibre announced it would close its software center in Córdoba, opened in 2016 and doubled in size soon after, citing municipal taxes. Staff would keep working remotely. In La Matanza, the vast Peronist district of Greater Buenos Aires, the company announced a $75 million expansion of its storage and distribution center in Tapiales, meant to double capacity. Less than two months later the municipality shut down the construction site for lack of a permit. Galperin posted photographs of the notices, obra paralizada and clausurado, and wrote: "Peronism, 80 years fighting capital and generating poverty." He added: "We have more than 20 distribution centers across America, only in La Matanza do they close them." The municipality's planning secretary, Miguel Saredi, answered that the project violated the law, and then went further. He noted that the land-use law in question dated from 1977, "when Peronism was persecuted, imprisoned and tortured by Galperin's friends." Argentine arguments over zoning tend to end up in the 1970s.
The man who did not want to be famous had become, in his home country, a political object. Partly this was his choice: the fast keyboard was his own. Partly it was structural. When your company handles a large share of a nation's commerce, payments, and consumer credit, you cannot stay out of politics, because your business is now part of what politics is about.
The Usurer and the Monarch
In August 2026 the political argument arrived at the newest and most sensitive part of the business, which was lending.
On August 20, Axel Kicillof, the Peronist governor of Buenos Aires Province, and his production minister, Augusto Costa, announced that the province's consumer-protection directorate had opened an investigation into Mercado Pago and several other digital wallets, among them Ualá, Naranja Digital, and Personal Pay. The directorate had received more than 2,240 complaints in the first half of 2026. The complaints alleged harassment of families behind on loan payments, unclear disclosure of the total financial cost of credit, missing withdrawal buttons, and unilateral changes to terms. Possible fines ran up to 1,883 million pesos. Kicillof blamed the national government directly: "What happened is Milei's economic policy, pure and simple."
At almost the same time, Gabriel Solano, the leader of the Trotskyist Partido Obrero, expanded a criminal complaint before Criminal and Correctional Court No. 53 against Galperin and MercadoLibre's board. It accused Mercado Pago of usury, alleging effective total annual financial costs above 1,375 percent, and asked that loan collections be suspended pending investigation. Galperin answered online. He argued that banks, not fintechs, account for most of the arrears, and that Peronists talked about defaults but never about the growth in loans that were being repaid on time. On August 26, 2026, Solano replied with a numbered list. Mercado Pago, he wrote, accounted for 19 percent of all defaulters in the country, about two million people. Those people were poor, owed "little" money in absolute terms and a great deal relative to their incomes, and had borrowed for food, utilities, and medicine, not televisions for the World Cup. He began the post with an epithet: Al usurero Galperin. To the usurer Galperin.
It is a measure of how far things had come that the escrow layer, originally designed so that strangers could trust each other enough to complete a sale, had become the instrument through which two million Argentines were in arrears to a single company. The same data that let Mercado Pago lend to people banks had never served also let it reach people banks had never served when they could not pay. "Financial inclusion" and "usury" can describe the same balance sheet, depending on where you stand in relation to it. The investigation and the complaint had not been resolved at the time of writing. Their outcome will matter less to the business than to the argument about what kind of power this is.
That argument is Vanoli's subject. His sharpest point concerns how this power is collected. The old kings, he wrote, levied tributes that often put their legitimacy at risk. "Galperin prefers to accumulate information, receive commissions and produce interest." The monarch's body disappears and his presence becomes "microphysical," spread across the phones of millions. Read in August 2026, against the provincial complaints and the criminal filing, the essay seemed less like theory and more like a forecast.
There were smaller incidents as well. At one point Galperin reacted on social media to a video of a retiree describing her struggle to make ends meet, and the popular streamer Coscu read the reaction as mockery. "Does he laugh at a retiree?" Coscu asked. The clip went viral. Galperin did not respond publicly. No single incident defines him. Together, though, they show the cost of the fast keyboard for a man whose founding principle was that the company, not the founder, should be famous.
Better Than Me
On May 21, 2025, Galperin sent a letter to MercadoLibre's employees announcing that he would step down as chief executive on January 1, 2026, after twenty-six years, and become executive chairman. His successor would be Ariel Szarfsztejn, the company's president of commerce.
The letter explained the timing with dry humor. "When I turned 30, I thought I'd take this step at 40. When I turned 40, I thought it would be at 45. When I turned 50, I decided it was better not to make any more predictions!" Then it turned serious and stated a rule he said had guided his hardest decisions since 1999: always put what is best for the company ahead of his own feelings.
In the same letter he allowed himself what he had avoided for a quarter-century, a list of what he was proud of. He had written the business plan in college, hired a formidable team, many of whom were still with him, achieved profitability, gone public on the Nasdaq, built the largest technology team and the largest e-commerce logistics network in Latin America, and become the most valuable company in the region, "and, above all … helping hundreds of millions of people live better lives." It reads like a résumé from someone who never needed one.
Szarfsztejn's own résumé matched the company's history closely. He graduated cum laude in economics from the University of Buenos Aires, worked at Boston Consulting Group, ran the hotels business at Despegar, the Latin American travel company, and then earned a Stanford MBA, the same degree and the same campus. He joined MercadoLibre in 2017 as vice president of strategy and corporate development. From 2018 to 2021 he ran Mercado Envíos, the logistics arm, during the pandemic years when volumes multiplied. Then he led the marketplace. In Galperin's account, the preparation was deliberate. Szarfsztejn's role was changed to make him effectively the marketplace CEO, and he was asked to start meeting the company's major shareholders every quarter. "Today," Galperin told Exame in November 2025, "they know him better than me." To make room, Stelleo Tolda, Galperin's Brazilian co-pilot of more than twenty years and the classmate who had grown suspicious of all that library time in 1999, had been moved aside several years earlier. At least one board member, Henrique Dubugras, reportedly asked Galperin why he was leaving at all.
The transition, Galperin told Exame, was announced as a months-long process, "which in practice lasted a week." By the time the two men sat for the interview in the São Paulo office, Szarfsztejn had already run the previous day's management meeting. Then Galperin said the most revealing thing he has said in public:
— Marcos Galperin, to Exame, November 2025For me, if it were someone from the outside, it would've been a sign of personal failure. Because I have always found it very important to work with people who I think are better than me, thus, if I have to step aside and no one is there to fill the gap, I would have failed in hiring people who are better than me.
"I'm very perfectionistic," he added, "and I want Mercado Livre to be the best company in Latin America. For that, it must be able to function without its founder."
He was 54. As chairman, he said, he would focus on strategy, product, culture, capital allocation, and how artificial intelligence would change the business. He kept investing elsewhere: a $66 million Series A for the HR-software startup Humand in February 2026, the Chilean food company NotCo before that. In 2023 he co-founded and backed the Miami Sharks, a professional rugby union team in Florida, a quiet return to the sport he gave up for Wharton. He supports Club Atlético Independiente, and Independiente once arranged for him to meet Ricardo Bochini, the club's greatest player, the way any fan would want.
The company keeps moving. It delivers 2.9 billion items a year to 131 million unique buyers, from Tijuana to Ushuaia, through warehouses and electric vans and six converted Boeing freighters. Somewhere in that system, on any given morning, a driver whose name no customer knows is lifting a package off a truck.
A century ago, at the port of Buenos Aires, a Russian boy with no Spanish and no schooling spent months in the immigrants' hotel. He loaded and unloaded sacks of grain from the railway cars, other people's goods, for people who never learned his name.
Part IIThe Playbook
What follows are twelve principles drawn from the record of Galperin's twenty-six years as chief executive. They are not virtues, though some of them look like virtues from a distance. They are the operating habits of a founder who built in a region where every assumption Silicon Valley takes for granted (working payments, functioning roads, stable currencies, institutional trust) had to be built, bought, or done without. Some of them translate directly to other markets. Some translate only to founders willing to stay in one place long enough for the constraint to become the moat.
The principles
- 1.Make the company famous, not the founder.
- 2.Save the money for the rainy season.
- 3.Treat the constraint as the product spec.
- 4.Build trust infrastructure before growth infrastructure.
- 5.Watch the film, not the photograph.
- 6.Trade equity for time, and time for territory.
- 7.Over-prepare for the conversation you cannot schedule.
- 8.Read unanimous skepticism as a pricing signal.
- 9.Price inefficiency as addressable market.
- 10.Institutionalize paranoia at the top.
- 11.Hire your successor a decade early.
- 12.When you become infrastructure, you become politics.
Principle 1
Make the company famous, not the founder
The security guard who did not know Galperin's name in 2014 is usually told as a charming anecdote. It was a strategy. Galperin's stated reasoning to the BBC was a deliberate inversion of the founder-as-brand model: "we want the company to be famous. The lower our personal profile, the better." The jeans, the trainers, the absence of a chauffeur, the FT finding him slouched on an office sofa: these were not modesty for show. They were a refusal to let the brand depend on one person.
The payoff came twenty-six years later. A company built around a founder's persona struggles to survive the founder's departure, because investors, employees, and customers have been taught to trust the man rather than the system. MercadoLibre's succession reportedly took a week in practice. Its shareholders, by Galperin's own account, knew the incoming CEO better than the outgoing one. That is only possible if the founder's personal brand was never the main thing holding the company up.
The record also shows the limit of this principle. In the second half of his tenure Galperin's online persona became loud and polemical, and the company's political exposure grew with it. The original discipline (no founder fame) proved easier to keep inside the company than in the public square.
Tactic: Audit which of your company's relationships (with investors, key customers, senior hires) would weaken if you disappeared tomorrow, and start moving each one to a colleague now.
Principle 2
Save the money for the rainy season
MercadoLibre was one of perhaps eighty clones and not the best funded. Its Harvard-founded rival DeRemate raised more and spent it on advertising and a fast path to the public markets. When the bubble burst, the difference was decisive. "When our competitors ran out of cash, we still had $40 million in the bank," Galperin said. A few years later he bought DeRemate's remaining half for $50 million, a fraction of what its founders had raised.
The lesson is not frugality for its own sake. Capital raised in good times is most valuable when nobody else has any. Galperin treated his early fundraising as insurance against a crash he could not predict but assumed would come. He later gave founders blunt advice: "Manage your own destiny and don't depend on investors to begin your business."
⚖
Two Theories of 1999
MercadoLibre and its best-funded rival launched the same year with the same model. The divergence came from capital discipline.
| Dimension | DeRemate (Harvard founders) | MercadoLibre (Stanford founders) |
|---|---|---|
| First-round capital | More than $52M | $52M |
| Primary spend | Mass marketing | Team and product |
| Exit thesis | Go public quickly | Build "organic engines for growth" |
| Position after the crash | Capital exhausted | $40M in the bank |
| Outcome | Remaining half acquired by MercadoLibre, 2008 | Break-even 2005; Nasdaq IPO 2007 |
Tactic: Model your next eighteen months on the assumption that no new capital arrives, and set your marketing budget by what survives that model rather than by what your competitors are spending.
Principle 3
Treat the constraint as the product spec
The garage on Calle Tronador was chosen for one reason: it had the best broadband. "Terrible offices, but at least we had good connectivity." That small trade-off contains the whole method. Identify the one constraint that actually binds, satisfy it fully, and tolerate everything else.
The method scaled. The constraint on Latin American e-commerce was not demand but trust, so the company built escrow. The next constraint was delivery, so it built fulfillment centers, an electric fleet, and an air cargo operation. Then working capital, so it built credit. Each piece of the ecosystem started as an answer to a specific failure in the transaction, not as part of a grand plan. As one analysis of the company concluded, its ecosystem was "built out of necessity."
The opposite approach, deciding what to build and then looking for a problem to justify it, produces companies that look complete on a slide and fragile in the field. Galperin's company looked fragile on slides for years, and lost money in 2018 and 2019, because it was spending on the actual constraints.
Tactic: Write down the single reason your last hundred failed transactions failed, and make fixing that reason, not the next feature, your quarter's top engineering priority.
Principle 4
Build trust infrastructure before growth infrastructure
Everyone told Galperin that Latin Americans would not buy products they had not touched. They were right that the barrier existed. They were wrong that it was cultural and permanent. It was a trust problem, and trust problems can be engineered. The first Mercado Pago held the buyer's money until the buyer confirmed the goods had arrived. In a region where the handshake was often the only enforcement available, MercadoLibre put itself between the two hands.
The sequence matters. Escrow came before the logistics network, before credit, before the advertising business, before the Nasdaq listing. Every later business rests on the willingness of strangers to transact through MercadoLibre, which means each of them depends on that earliest and least glamorous one.
◷
The Stack, Layer by Layer
Each business unit answered a failure exposed by the one before it.
1999
Marketplace launches from a garage on Calle Tronador, August 2.
2001
eBay takes a 19.5% stake, contributes its Brazilian unit, and pledges to stay out of the region for at least five years.
2004
Mercado Pago operates as an escrow layer; Osvaldo Giménez takes over its operations in February.
2007
Nasdaq IPO raises $289M, the first by a Latin American technology company.
2012
MercadoShops lets small businesses run stores on their own sites; Amazon enters Latin America.
2013
Mercado Crédito begins offering loans of up to 30,000 pesos.
2014
Mercado Pago's card-reader app lets merchants accept payments by phone or tablet.
2019
First owned distribution centers open in Argentina, Brazil, and Mexico.
2026
Ariel Szarfsztejn, former head of Mercado Envíos, becomes CEO on January 1.
Tactic: Before spending on acquisition, identify the moment in your transaction where one side has to trust the other blindly, and put your own balance sheet or process in that gap.
Principle 5
Watch the film, not the photograph
"There is nothing more dangerous in this industry than looking at your photo and falling in love," Galperin said in 2015. In 2011, MercadoLibre had a lovely photograph: profitable, dominant, paying a dividend. When Amazon arrived, Galperin chose to make it uglier. Margins fell. Operating income went negative in 2018 and 2019 while revenue nearly doubled. "Profitability for us is never the driving force," he told Stanford students. "In this stage of the internet, we think market share is more important than having very high margins."
That is easy to say and hard to do as a public company with quarterly shareholders. Galperin's credibility for the decision came from history. He had already said in public that profitability took seven years instead of two. He had already been through one shareholder vote that nearly shut the company. Investors who had watched the earlier part of the film were more willing to sit through the next one.
The vindication came in numbers: revenue rose from $2.29 billion in 2019 to $28.89 billion in 2025, and operating income from a $0.11 billion loss to $2.84 billion. But the principle also shows in a decision with no number attached: an acquisition offer turned down because the company had two more decades of growth in it.
Tactic: When you present results to your board, show the five-year trajectory alongside the quarter so the conversation is about direction, not just position.
Principle 6
Trade equity for time, and time for territory
The eBay deal of September 2001 is usually remembered as a validation: the global giant bought in. Its real value was in its terms. eBay paid partly with its Brazilian subsidiary and agreed not to enter Latin America for at least five years. Galperin gave up 19.5 percent of the company and got a five-year head start in the largest e-commerce market in the region, protected from the only competitor that could have beaten him.
Five years was enough. By the time eBay's non-compete expired, MercadoLibre was profitable, preparing its IPO, and entrenched. eBay kept the stake for fifteen years and sold it in 2016, and even then the two companies kept working together. The deal turned a potential competitor into a shareholder with an interest in the company's dominance.
The principle generalizes. The most dangerous competitor is often the one who would be happiest as an investor. Equity can buy more than cash. It can buy time, a market, or a competitor's absence.
Tactic: List the two companies that could most plausibly enter your market in the next three years, and ask what each would accept in exchange for staying out.
Principle 7
Over-prepare for the conversation you cannot schedule
Galperin did not schedule his meeting with John Muse. He had a ride to the airport, two talkative classmates, and a few minutes. He had rehearsed the pitch, by his own varying accounts, either 150 or 700 times. When the window kept shrinking, he got lost on purpose. Muse's "I love your idea, let me know if you need money" is the kind of luck that follows that much preparation.
The Buffett episode a few weeks earlier is the counter-example. Galperin sat next to one of the most important investors in the world for three minutes and did not recognize him. The lesson he took was not to recognize faces better. It was to make sure he was ready the next time the right person was in range, and to find someone, in this case Jack McDonald, who could put him there.
Tactic: Write and rehearse a ninety-second version of your most important pitch until you can deliver it in a moving car with interruptions, then ask three people who know your field whom you should be driving to the airport.
Principle 8
Read unanimous skepticism as a pricing signal
Galperin surveyed twenty of his Latin American classmates at Stanford, people who understood both business and the region, and they told him the idea would not work. In one telling all twenty said so; in another, nineteen. Either way, the people best placed to judge said no.
That consensus was not proof he was right. It was proof the opportunity was cheap. If the people with the best information were unanimously skeptical, capital and talent would not be crowding in, and anyone who got the model working would have the field almost to themselves. The eighty clones that did crowd in mostly misjudged how to execute, not whether the market existed. Galperin's own advice puts the two halves together: "If you have a view of something that is a contrarian view but makes sense to you, pursue it, but you need to be resilient and long-term focused." Contrarian is the easy part. The resilience is the hard part.
He has drawn the comparison himself. Speaking at an Endeavor event, he said the crypto believers of recent years reminded him of his own team: a small group convinced they would change the world while most people thought they were crazy. "That was us too, in '99."
Tactic: When knowledgeable people unanimously reject your idea, separate their objections into "this market doesn't exist" and "this is hard to execute," and only abandon the idea if the first category survives scrutiny.
Principle 9
Price inefficiency as addressable market
"The more inefficient retail is, the more value we can provide to a society." Most foreign entrants saw Latin America's broken logistics, unbanked populations, and unstable currencies as costs. Galperin treated them as the size of the opportunity. Every problem the state or the incumbents failed to solve was something MercadoLibre could charge for once it had solved it.
The 2001–2002 crisis was the purest example. When Argentina froze its bank deposits, MercadoLibre's traffic grew as people sold what they owned to raise cash. The worst moment for the national economy became a growth engine for a platform that offered liquidity when institutions would not. The pattern repeated in the pandemic: as tens of thousands of shops closed, the platform became one of the only ways to buy many products, and the newly built warehouses filled up.
This principle has a sharp edge. A business that profits from institutional failure has an interest in that failure continuing, or at least can be accused of having one. The 2026 usury complaints rest on exactly this accusation: that credit offered where banks would not lend becomes extraction when borrowers cannot pay.
Tactic: Map the three biggest failures of your market's infrastructure, estimate what customers already pay to work around each one, and treat the largest as your next product.
Principle 10
Institutionalize paranoia at the top
"We're always paranoid," Galperin said in 2024, when MercadoLibre was worth about $85 billion. "We don't think it's game over by any means." In 2015, already dominant, he described "a desperation that there is so much to do, that this is just starting." People inside the company report that the line "only the paranoid survive" is built into its quarterly reviews and its focus on customer satisfaction scores.
Paranoia from a founder can be a personality trait or a management system. At MercadoLibre it became a system. The company kept treating itself as an underdog well after it had stopped being one, and that habit is what led it to lose money building logistics in the face of Amazon rather than defend its margins. Complacency at the top filters down quickly. So does anxiety, and in a company that needs to keep reinventing itself, anxiety is the more useful of the two.
His routine supports this. The notepad by the bed for ideas at five in the morning, the hours of exercise during which he says he thinks best: the paranoia is not reserved for board meetings.
Tactic: Once a quarter, have your leadership team write the memo a well-funded competitor would write about how to beat you, and fund at least one of the defenses it implies.
Principle 11
Hire your successor a decade early
Galperin's standard for his own departure was severe: an outside CEO would have been "a sign of personal failure." The reasoning was that his job, for twenty-six years, had been to hire people better than himself. If no one inside could replace him, he had failed at that job.
The succession was designed long before it was announced. Szarfsztejn joined in 2017 in strategy, was given the hardest operating job in the company (building logistics during a pandemic), then the marketplace, then quarterly meetings with every major shareholder. A long-time co-founder was moved aside to give him room. By the time the CEO title passed, it described a role Szarfsztejn was already doing. "The transition," Galperin said, "in practice lasted a week."
His letter's humor about the timing (forty, then forty-five, then no more predictions) is a warning to other founders. The right moment to leave never feels obvious from the inside. The way to make the decision possible is to have the successor ready long before you feel ready to go.
→
Founder Succession: The Common Pattern vs. Galperin's
Most founder transitions fail because they begin at the announcement. MercadoLibre's began roughly eight years earlier.
| Stage | Common pattern | MercadoLibre |
|---|---|---|
| Candidate identification | At announcement, often external | Years earlier Internal, hired 2017 |
| Operating proof | Assumed from résumé | Tested Logistics during the pandemic surge |
| Investor exposure | Introduced at transition | Quarterly meetings "They know him better than me" |
| Founder role after | Undefined or ceremonial | Executive chair: strategy, culture, capital, AI |
Tactic: Name, privately, the two people inside your company who could run it in five years, and give each one a job this year that you would otherwise keep for yourself.
Principle 12
When you become infrastructure, you become politics
A marketplace can stay out of politics. A company that handles a large share of a country's commerce, payments, and consumer lending cannot. Galperin's record in the 2020s shows what happens when that line is crossed. The Córdoba office closed over municipal taxes. The La Matanza warehouse was shut down over permits. A provincial consumer investigation built on 2,240 complaints. A criminal usury complaint alleging annual costs above 1,375 percent and claiming Mercado Pago held 19 percent of the country's defaulters. Each of these is a political fight, whatever the legal merits.
Galperin's response has been to fight openly: posting photographs of the closure notices, taking on the Peronist party by name, publicly backing Milei. Whether that is wise is the most contested question about him. It has made him a clear ally of the national government and an obvious target for the opposition, which is a risky position for a company that needs to operate under every future administration. What the record shows clearly is the cost of not planning for it. A business that becomes essential to a country's daily life will be held to political, not just commercial, standards of accountability, and the founder's personal statements will be read as company policy.
The earliest principle in this playbook, keep the founder's profile low, turns out to matter most at this stage, when it is hardest to keep.
Tactic: Once your product touches a population's money or livelihoods at scale, separate the founder's public voice from the company's regulatory posture and assume every post will be quoted in a courtroom.
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
— Marcos Galperin, Stanford Graduate School of Business, March 2011We really saved the money for the rainy days. So, when our competitors ran out of cash, we still had $40 million in the bank. That was important.
— Marcos Galperin, to the Financial Times, 2015The more inefficient retail is, the more value we can provide to a society.
— Marcos Galperin, Universidad de San Andrés, June 2015 (translated from Spanish)There is nothing more dangerous in this industry than looking at your photo and falling in love. … I like to look at the whole film, not the photo.
— Marcos Galperin, to Fortune, 2024We're always paranoid. We don't think it's game over by any means.
— Marcos Galperin, Stanford GSB View From The TopTry to take risks. If you have a view of something that is a contrarian view but makes sense to you, pursue it, but you need to be resilient and long-term focused.
Maxims
- The company is the celebrity. A founder's fame is a liability the company inherits on the day he leaves.
- Satisfy the binding constraint and tolerate the rest. Terrible offices are fine if the connectivity is good.
- Cash is a weapon only when everyone else is out of it. Raise in good times as insurance against the crash you assume is coming.
- Trust is infrastructure. Before a market can grow, someone has to stand between strangers and hold the money.
- Equity can buy time and territory. The right investor may be worth most for where he agrees not to go.
- Unanimous skepticism is a price, not a verdict. When the people who know best all say no, the opportunity is uncrowded.
- Profitability is a phase, not a purpose. Sacrifice the margin to defend the market when the market is still being decided.
- Paranoia should outlast dominance. The day the company feels safe is the day it starts to decline.
- An outside successor is a hiring failure. If no one inside can replace the founder, the founder did not hire well enough.
- Scale turns a business into a political actor. Once you hold a country's money, every word the founder posts is policy.
