Part IThe Story
The money crossed the room in July 1999, and the room applauded. Representatives of Nokia Ventures and Deutsche Bank, at the launch party for a product called PayPal, sent $3 million in venture funding to the company's chief executive using the software on their PalmPilots. It was a stunt in the way all launch parties are stunts. It also worked, which was the point. Nobody at that party would remember the code. They would remember the gesture: a fortune moving through a handheld organizer the size of a deck of cards, from one hand to another, without a bank in between.
The code belonged to Max Levchin. He had turned twenty-four that month. Eight years earlier he had arrived in Chicago from Kyiv, a city in a country that, as he would later put it, had tanks in its streets. Before that he had been a sickly child whose doctors doubted he would survive, and whose grandmother and parents had put a clarinet in his hands so that he might, through practice, make his lungs larger than they wanted to be.
There is a way to tell this story in which the $3 million is the climax: immigrant boy, garage, money through the air, eBay, billions. Levchin's career does not cooperate with that version. The spectacle at the launch party was the easy part. Over the next three years, nearly everything that mattered at PayPal concerned what happened when the money moved and the wrong people were holding the PalmPilots. Twenty-seven years after the launch party, the company he runs now lends money to 27.8 million people. It is the same problem. At what point can you trust a stranger? And how do you build a machine that knows?
By the Numbers
The Levchin Ledger
1991Year the Levchin family settled in Chicago from Kyiv
2.3%Levchin's PayPal stake at the eBay sale, worth roughly $34 million
26His age when MIT Technology Review named him Innovator of the Year (2002)
$182MPrice Google paid for Slide in August 2010; it shut the product down a year later
~$2.5BValue of his Affirm stake the day after the January 2021 IPO
$50.2BPayments processed by Affirm in fiscal 2026
$0Late fees Affirm charges its borrowers
A body the doctors doubted
Begin with the lungs, because Levchin has.
In an interview with Emily Chang of Bloomberg, he described a childhood shaped by respiratory illness serious enough that doctors questioned whether he would live. The family's answer, steered by his grandmother, was the clarinet. Wind instruments are a kind of training regimen disguised as music; you learn to take in more air than feels natural, to hold it, to release it in controlled, measured portions over the length of a phrase. It is a strange prescription and a very Soviet one, the domestic improvisation of a family that had learned not to wait for the system to fix anything.
The record is thin on what came next. Levchin was born Maksymilian Rafailovych Levchyn on July 11, 1975, into a Ukrainian-Jewish family in Kyiv, then the capital of the Ukrainian SSR. He spent his adolescence watching the state around him come apart. When he described those years publicly, decades later, he did so in the context of an argument about American tax policy, which tells you something about how the memory functions for him: not as nostalgia, but as evidence.
— Max Levchin, on Ukraine at the Soviet collapse, Yahoo Finance, 2019There were tanks on the streets, because we couldn't — couldn't contain the populace from tearing the country apart.
Notice the stutter, the repeated couldn't. It is the kind of thing transcription preserves and memoir would edit out. The we is interesting, too. A teenager in Kyiv was not containing anyone. But the pronoun claims membership in a society that had lost the ability to hold itself together, and the lesson he drew from it was institutional rather than personal. Systems fail. Currencies fail. Promises made by large impersonal bodies fail with a regularity that people raised in stable countries find hard to imagine.
In 1991 the family settled in Chicago. Levchin attended Mather High School on the city's North Side, then the University of Illinois at Urbana-Champaign, where he earned a bachelor's degree in computer science in 1997. He was, by every account that survives, a programmer first: someone who wanted to make machines do precise things. That ambition, with the memory of tanks behind it, turned into something more specific. He wanted to build systems that did not betray the people relying on them. Cryptography, the mathematics of trusting no one, would be his first love.
He was also, in the language of a later era, an immigrant founder. In 2013 he contributed to FWD.us, the Mark Zuckerberg–backed lobbying group for high-skilled immigration, and narrated his own story in a video the group released. The same year he publicly defended the National Security Agency against much of Silicon Valley, arguing that an agency built to protect the country from terrorism deserved support even when it overstepped. Those two positions sit together more comfortably than they first appear. Both belong to a man who has seen what happens when the state stops working, and who has decided, with some reluctance, that he prefers a state that works too hard.
The Champaign pipeline
Silicon Valley's founding myths are set in Palo Alto. PayPal's actually begins two thousand miles east, in the flat cornfield country of central Illinois. Confinity's corporate records placed its headquarters in Champaign County. Its engineering culture was imported from there too.
In the summer of 1995, as an undergraduate, Levchin co-founded a company called SponsorNet New Media with two fellow Illinois students, Luke Nosek and Scott Banister. It does not appear in many histories, which is appropriate; it was apprenticeship rather than achievement. What it produced was a set of relationships that would compound for three decades.
Nosek, an Illinois computer science student like Levchin, was the type of friend who gets you into rooms. He had pitched a web-based calendar to a young investor named Peter Thiel, who put $100,000 into it. The project failed. Nosek then introduced Thiel to his friend Max, who had an idea about cryptography. Nosek would later run marketing and strategy at PayPal and become a partner at Founders Fund, and his life became a long demonstration that a failed company can be more valuable than its balance sheet suggests. Banister, an early PayPal adviser and board member, later joined Thiel and Nosek in backing Ron Paul's 2012 campaign. Three Illinois kids. A summer company. A web of loyalties that would show up again in Founders Fund deals and political action committees.
Then Levchin did something that turned out to be a strategic weapon, though he likely did not think of it that way. When Confinity needed engineers, he recruited them from Urbana-Champaign. Most of the company's early engineers came from the University of Illinois, sourced by Levchin. The business people came from the other pole of the company's personality: Thiel's network, many of them alumni of The Stanford Review, the libertarian student paper he had co-founded in 1987.
So the company was built from two talent pipes. One carried Stanford-trained contrarians who read Girard and argued about Western Civ. The other carried Midwestern engineers who mostly wanted to make the code run. The PayPal Mafia's Wikipedia entry notes, almost in passing, that most of its members attended either Stanford or Illinois. That note is the whole hiring strategy compressed into a clause. Illinois engineers were not being courted by Stanford money. They were available, they were good, and they owed their opportunity to the person who had found them.
Levchin, the Kyiv-to-Chicago kid recruiting his classmates west, was doing what immigrant communities have always done. He was bringing his people over.
The investor who became a chief executive
Peter Thiel did not want to run anything. That is the detail worth sitting with.
According to the company's own timeline, the idea for the company was Levchin's. Thiel was initially interested only in being an investor, and Levchin convinced him to become CEO. The usual founder story runs the other way. The visionary insists on the throne and the money men grumble. Levchin, at twenty-three, looked at a thirty-one-year-old former derivatives trader and decided that the company needed him at the top more than it needed Levchin there.
Thiel's own biography reads like a variation on Levchin's theme. Born in Frankfurt in 1967, he was carried to Cleveland as an infant. His father was a chemical engineer whose mining work moved the family to South Africa and to what is now Namibia, where Peter changed elementary schools seven times and attended a German-language school in Swakopmund that used corporal punishment. He was remembered there as a smart, withdrawn boy whom other children avoided because they knew he would not stay long. He became a chess prodigy, a Stanford philosophy major, the founding editor of the Review, a law clerk, a Sullivan & Cromwell associate who lasted under a year, and a currency-options trader at Credit Suisse. In 1996 he returned to California and raised $1 million from friends and family to start Thiel Capital Management.
Two displaced children, then. One made rootless by a father's career, the other made stateless by a collapsing empire. One drew from that experience a philosophy of the sovereign individual. The other drew a cryptographer's practical paranoia. They met because a mutual friend's calendar startup had failed.
In 1998 they founded Fieldlink, a security company that let users store encrypted data on PalmPilots and other handheld devices so the devices could serve as "digital wallets." Thiel provided the initial $100,000. In February 1999 they raised $500,000, mostly from friends and family; $35,000 of it came from Thiel's parents. By the middle of that year they had raised $4.5 million, $3 million of it from Nokia Ventures, which is how a Finnish handset company came to be beaming venture money across a Palo Alto room. The company renamed itself Confinity. The first version of the PayPal payments product launched in 1999.
What Thiel thought PayPal was for is on the record, and it is grander than anything Levchin has said:
— Peter Thiel, on PayPal's missionIt will be nearly impossible for corrupt governments to steal wealth from their people through their old means because if they try the people will switch to dollars or Pounds or Yen, in effect dumping the worthless local currency for something more secure.
Read that with Kyiv in mind. Thiel was describing, as theory, a world Levchin had lived through as fact. You can imagine the conversation in which the theorist and the witness found they agreed. You can also imagine that the witness was the more skeptical of the two, because he knew that the people who most needed a stable currency were also the people least likely to own a PalmPilot.
The PalmPilot thesis failed anyway. Confinity found little traction with the handheld business, and the product drifted toward something humbler: payments by email. By November 1999 there was a fully functional website, and staff were emailing friends offering $1 for signing up. Between December 1999 and February 2000, the team noticed where the users actually were. They were on eBay, a marketplace with an enormous concentration of buyers and sellers and no standard way for them to pay each other.
The cryptography company had become a payments company, and the payments company was about to discover that its main problem was crime.
A merger of enemies
There were two companies chasing the same idea, a few blocks apart, and the second one was run by Elon Musk.
Musk had already founded and sold Zip2. In early 1999 he started X.com, an online financial services company, with co-founders Harris Fricker, Christopher Payne, and Ed Ho. In December 1999 X.com, chaired by Musk with Bill Harris as CEO, added online payments to its account service and made that its focus. Confinity and X.com were now building competing products for the same eBay sellers and paying for users in the same expensive way.
Then, on March 1, 2000, eBay announced Billpoint, its in-house payment service built with Wells Fargo. The platform that both startups depended on for growth had decided to compete with them. The next day, March 2, Confinity and X.com announced their merger. It is hard not to read the dates as cause and effect, though the discussions had begun in February. The combined entity took the name X.com, thought to have broader long-term potential than either Confinity or PayPal. Confinity's senior people received vice-presidential titles.
What followed was a year that the participants have spent two decades narrating to each other. Harris was out by April or May 2000, depending on the account. The sources differ, as sources about PayPal tend to. Musk became CEO, Thiel became chairman. In October 2000 Musk decided to close X.com's other banking operations and focus on payments. That same month he was replaced as chief executive by Thiel. The company dropped the X.com brand from its product, though not from its legal name, until June 2001, when it formally became PayPal.
Levchin's role in all of this was technical and, by most accounts, decisive. The merger brought together two codebases, two cultures, and two founders with large egos. One of the engineering questions underneath the fight concerned whose software stack the combined company would run on. The CTO was the person with the most to lose from getting it wrong. He was also the one who could not afford to leave. Founders can be replaced. The person who understands why the system works usually cannot.
The phrase "a merger of enemies," which has since become a YouTube documentary title, is mostly right, but it misses one thing. The enemy that brought them together was not each other. It was eBay. Meg Whitman, eBay's chief executive, later laid out PayPal's growth in three phases: "First, PayPal focused on expanding its service among eBay users in the US. Second, we began expanding PayPal to eBay's international sites. And third, we started to build PayPal's business off eBay." By then we meant eBay. The rival had become the owner.
Prove you are human
Somewhere in this period a new image entered the visual culture of the internet: a box of warped, crowded letters that a user had to read and retype before the machine would let them continue. A machine was asking a stranger to prove that he was a person.
The Gausebeck-Levchin test, named for Levchin and his PayPal colleague David Gausebeck, was one of the first commercial implementations of what became known as a CAPTCHA. It existed because PayPal was being robbed.
As early as 2001, PayPal had serious problems with online fraud. International hackers broke into accounts and siphoned small amounts out of many of them at once, a strategy designed to stay below the threshold of anyone's attention. The conventional defense against payments fraud was the state: find the criminal, prosecute him. That defense did not work across borders. PayPal was losing millions of dollars a month and struggling to get the FBI to pursue international cases. The company was effectively stateless. Nobody was coming.
So the engineers built their own police. From mid-2000 through 2001, Levchin led the company's fraud work. One product was the CAPTCHA, which made it harder for automated scripts to create accounts in bulk. Another was a system called "Igor," which identified online activity that looked fraudulent and froze it. The details of Igor's internals are not in the public record I have seen. Its name alone tells you something about the company's humor: a reference to a lab assistant, a servant who does the dirty work while the master gets the credit.
The timeline is clear on why this mattered. Fraud costs had limited PayPal's ability to offer affordable services. A payments network that cannot tell good actors from bad has two options. It can charge everyone enough to cover the thieves, or it can die. Levchin's systems allowed a third option, and that option is a large part of why a startup that had lost its original thesis survived to an IPO.
BusinessWeek called him an "Online Fraud-Buster" in September 2002. That same year MIT's Technology Review put him on its TR100 list of top innovators under thirty-five and named him Innovator of the Year. He was twenty-six.
The fraud work did not stay inside PayPal. The company's Wikipedia history states the lineage plainly: fraud monitoring software built at PayPal led Peter Thiel to create Palantir, a big-data company whose original mission was to "reduce terrorism while preserving civil liberties." One of the people who carried that knowledge across was Nathan Gettings, PayPal's former director of risk and R&D. Gettings co-founded Palantir, and later co-founded Affirm with Levchin. He is the kind of figure who rarely appears in a magazine headline and appears in every cap table that matters. He spent his career studying how bad actors behave in data, and his work became the plumbing for two companies with very different missions: one hunting terrorists for governments, the other underwriting a sofa purchase for a twenty-eight-year-old in Ohio.
Levchin did not build the CAPTCHA because he wanted to stand between humans and machines. He built it because, at that moment, it was the only way to keep the bank open. A generation of internet users learned to squint at warped letters because a refugee from a failed state could not get the FBI to return his calls.
The consigliere
PayPal went public on February 15, 2002, at $13 a share. The price peaked at $22 and closed at $20 on the first day. Sources put the proceeds anywhere from $61 million to $70 million. On July 8, 2002, eBay announced it would buy the company in an all-stock deal, reported as $1.5 billion by most outlets and $1.3 billion by the Associated Press. The deal closed in October. Levchin's 2.3% stake was worth about $34 million. Thiel's 3.7% was worth about $55 million.
Then almost everyone left. PayPal's people struggled with eBay's more conventional corporate culture. Within four years, all but 12 of the first 50 employees were gone. They stayed in touch socially and professionally and kept starting companies together, and by 2007 there were enough of them in prominent places that Fortune ran a group photo of them dressed as gangsters. The headline was "The PayPal Mafia." The photograph is now famous. The absences are almost more interesting. Fortune reported that Musk wanted to come but was in Chicago receiving an "innovator of the year" award, and that YouTube's Chad Hurley and Steve Chen bowed out "after their Google handler objected to the gangster motif."
The magazine gave Thiel the role of don. "Today they call upon one another when they need money or advice," Fortune wrote, "and when they need both, they go to Thiel, who seems to be at the center of it all." Levchin got the other part.
— Max Levchin, TechCrunch, 2020'PayPal mafia' doesn't bother me much because the original article that used it in 2007 described me as the consigliere, and I always thought the 'consigliere' was the coolest guy in a mafia.
The quote is charming and a little evasive, which may be appropriate for a consigliere. In The Godfather, the consigliere is the outsider in the family, the adopted son, the one whose power comes from judgment rather than blood or muscle. He advises. He does not inherit. For a man whose first company was co-founded with a CEO he recruited, and who would not become chief executive of his eventual great company until two years after he founded it, the role fits better than he may care to admit.
The more revealing part of that 2020 interview came afterward. Levchin said that at PayPal he had deliberately screened for entrepreneurial ambition when hiring. He preferred candidates whose goal was to eventually start their own companies. The PayPal Mafia was not a lucky accident of shared hardship. To some meaningful degree it was the planned result of a recruiting filter. He was hiring people he expected to lose. He said he hoped to replicate the same dynamic at Affirm, and the evidence suggests he has: SentiLink, the identity-verification company founded by Affirm alumni Naftali Harris and Max Blumenfeld; Resolve, spun out of Affirm under Chris Tsai; Jeremy Solomon, an Affirm alumnus who joined the venture firm Nyca Partners through his connection to Levchin. Levchin invested in the first two.
The consigliere's other job is to place bets for the family. Levchin was a key early investor in Yelp, founded in 2004 by PayPal alumni Jeremy Stoppelman and Russel Simmons. He served as chairman of Yelp's board from its founding until July 2015, and by 2012 he was its largest shareholder, with more than 7 million shares. That week Forbes ran the headline "Who Got Rich This Week: Chief Yelper Levchin." Suhail Doshi, who founded Mixpanel at twenty-six, credits Levchin with the company's survival. Levchin was a seed investor in WePay, SmartThings, and Zendrive, an investor in Square, TransferWise, Pinterest, and AdRoll, and a backer of AngelList alongside Chad Hurley and David Sacks. He served on the boards of Evernote from August 7, 2006, to 2016, Yahoo from December 2012 to December 2015, and Kaggle, Quid, Unity Technologies, and the San Francisco Symphony.
The Symphony is the one that catches the eye. A boy who learned the clarinet as therapy ended up governing an orchestra.
The family's most important adviser may not have been in the photograph at all. Nellie Minkova worked in financial-system product management at PayPal. She and Levchin married in 2008, after a long relationship. In 2010 they co-founded SciFi VC, which describes itself as an early-stage generalist firm focused on fintech, science, and AI. Its portfolio includes Uber, Stripe, and Anduril. Levchin credits her with many of the fund's most successful deals and many of his angel investments. When Business Insider interviewed the women of early PayPal, the "PayPalians" who built the company but were rarely listed among its mafiosi, Nellie Levchin noted that bias makes it hard for women leaders to raise early funding. She was speaking as someone whose judgment had produced returns that are often credited to her husband. The consigliere had a consigliere. The 2007 photograph, like most corporate origin stories, is accurate about who stood in the frame and quiet about who chose the investments.
Slide, or the uses of failure
There is a photograph of Max Levchin playing Guitar Hero at a conference. It's worth keeping in mind, because the next chapter of his career looks, in hindsight, like a man trying hard to have fun and finding that fun does not compound.
In 2004 he founded Slide, a personal media-sharing service built for the social networks of the time, MySpace first and later Facebook. Slide made slideshows, widgets, and the small decorative objects that teenagers used to cover their profile pages. In April 2008, AllThingsD covered him under the headline "Max Levchin Becomes the Internet's New Wacky Pix Guy!" The exclamation point is the reporter's, and it does a lot of work. The PayPal fraud-buster was now making novelty images.
In the same period he served as a producer on Thank You for Smoking, the satire about a tobacco lobbyist who can argue anything. It is an odd credit for a cryptographer. I have not found anything in the record in which he explains it. The detail lingers anyway: a man who made his name building systems to catch liars helped bring to the screen a comedy about the art of persuasive deception.
He also talked publicly about mistakes. At Y Combinator's 2007 Startup School, he described his entrepreneurial path, the errors he had made, and what he had learned. The New York Times profiled him that October, under a headline that captured the moment: "After Succeeding, Young Tycoons Try, Try Again." The Israeli newspaper Haaretz sent a reporter to Silicon Valley around then, and the line it filed has a nice dryness to it: "Levchin, 32, Jewish of course and born in Kyiv, refused to discuss money."
In August 2010, Google bought Slide for $182 million. On August 25, Levchin joined Google as a vice president of engineering. On August 26, 2011, Google announced it was shutting Slide down and that Levchin was leaving the company.
That is not quite a year, to the day. It is a remarkably clean ending for an acquisition. There was no lingering, no years spent drifting inside the parent company collecting vested shares. Founders of acquired startups often stay on as prisoners of a retention package. Levchin had already lived through that once, in watching PayPal's founding team leave eBay within four years. This time the whole thing took one.
Was Slide a failure? Not to anyone who bought shares in it. A $182 million exit is the dream of most founders. But measured against the arc of Levchin's career, it was a detour. He had spent seven years building a company in a category where switching costs were close to zero and where the platforms — MySpace, then Facebook — controlled the terms. It was the eBay problem again, without a payments business underneath to make it worth enduring. A product that decorates other people's platforms lives at their discretion. He had learned that once with Billpoint. Slide taught it to him again, at a higher price.
What came next was explicitly a reaction.
Hard, valuable, fun
In late 2011, weeks after leaving Google, Levchin started a company called HVF. The letters stood for "Hard, Valuable, and Fun." HVF was meant to explore and fund projects built on data, including data from analog sensors, the physical world's version of the signals he had once mined for fraud.
The name is a filter, and filters are Levchin's genre. He had hired at PayPal for future founders. He had built CAPTCHA to sort humans from scripts. Now he set out three conditions for his own time. Slide had been fun and probably valuable; it was not hard, at least not in the way that matters for building something defensible. PayPal had been hard and valuable; it was fun only in the way that war is fun for people who survive it. HVF tried to require all three at once.
What came out of the studio is a portrait of a restless mind. In 2013 HVF launched Glow, a fertility app to help couples conceive naturally. Quartz greeted it with a headline Levchin presumably did not enjoy: "The co-founder of PayPal wants to put a baby in you." HVF invested in Emerald Therapeutics, which applied cloud computing to drug discovery. The common thread was the one that ran through Igor and the warped letters: the conviction that patterns hidden in data could tell you something important about a person that the person might not be able to say for themselves. Is this transaction fraudulent? Is this cycle fertile? Is this borrower good for the money?
That last question became Affirm.
In early 2012, Affirm spun out of HVF with the goal of building a next-generation credit network. Its founders, depending on which record you consult, were Levchin, Nathan Gettings, and Jeff Kaditz of First Data — or those three plus Alex Rampell. The ambiguity is very PayPal. Confinity's founding list varies by source too; one names Russel Simmons, others leave him off. Credit, in the attributional sense, has always been as contested in Levchin's world as credit in the financial one.
Gettings's presence is the significant fact. The man who ran risk at PayPal, and then helped turn PayPal's fraud tools into Palantir's intelligence software, came back to apply the same instincts to lending. Fraud detection and credit underwriting are cousins. Both try to predict, from a trail of behavior, whether a stranger will honor an obligation. PayPal asked: is this person who they say they are? Affirm asks: will this person do what they say they'll do? The second question is harder. It is also more valuable, and if you are the sort of person who builds fraud engines for pleasure, more fun.
Levchin did not take the CEO title right away. He became Affirm's chief executive in 2014, the same year the company announced it had raised $45 million. Once again he let the company exist before deciding he should run it. This time he decided he should. In July 2015 he stepped down as Yelp's chairman. The Silicon Beat blog read it as "a sign Affirm is taking off." The consigliere was becoming a don.
Paying over time
The product is simple to describe and hard to build. At checkout, online or at a store kiosk or inside a digital wallet, Affirm offers a shopper a fixed installment loan and decides in real time whether to approve it. Merchants pay Affirm a fee, sometimes as high as 12.5%. Borrowers pay interest, from 0% to 36% APR, on loans up to $25,000, over periods of one to sixty months. Sometimes both sides pay. Affirm says it charges no compound interest, caps its interest, and charges no late fees. A borrower who falls behind is not fined. They may be refused the next loan, and the missed payment may affect their credit score.
That last design choice deserves attention, because it is where the philosophy lives. The modern credit card business earns a large share of its money from customers who fail: penalty rates, late fees, interest compounding on interest. Its incentives point toward the borrower's trouble. Affirm's structure tries to make the lender's profit depend on the borrower succeeding, by deciding at the start how much to lend and then declining to profit from the borrower's slip. Whether it works that way in every case is a question for regulators and plaintiffs. As an engineering thesis, it is Igor turned around. PayPal's systems existed to catch people who would cheat the company. Affirm's pricing is meant to show the borrower that the company will not cheat them.
A USA Today headline from December 2014 summarized Levchin's market read in four words: Young people "dislike big banks." He had grown up in a country whose institutions had failed spectacularly. His customers had grown up after 2008, in a country whose institutions had failed less dramatically and more expensively. The emotional texture was not the same. The distrust rhymed.
The growth came through the same door PayPal's had: other people's platforms. Affirm's service was available on Shopify, Magento, and Salesforce Commerce Cloud by 2016. Walmart signed on in February 2019, at self-checkout kiosks and online. Then came the pandemic years. Affirm filed to go public on November 18, 2020, postponed in December, and listed on Nasdaq on January 13, 2021, raising about $1.2 billion. By the next day the share price had doubled, and Levchin's stake was worth about $2.5 billion. A man who had sold his first company for a 2.3% slice of $1.5 billion now owned a fortune in a company he ran.
Partnerships piled up. Amazon in August 2021, then as Amazon's exclusive U.S. buy-now-pay-later partner through January 2023. Target and Apple. Stripe in May 2022. Booking.com, American Airlines, Best Buy. In June 2024, Apple announced that Affirm loans would be available inside Apple Pay in the U.S.; shortly afterward it said it would shut down its own Pay Later service after a year. The largest company on earth had tried to build Affirm's product, and then decided to rent it. Shopify made Affirm its exclusive pay-over-time provider and took the partnership to Canada and then Australia. JPMorgan Chase's payments unit, FIS's banking clients, Costco, Intuit's QuickBooks, Expedia. Affirm Edge now lets banks embed Affirm's loans directly in their own apps.
The platforms can also take things away. In March 2025, Walmart replaced Affirm with Klarna as its exclusive provider after Walmart-backed OnePay was permitted to take a stake in Klarna. Affirm disclosed that the Walmart relationship had accounted for about 5% of its gross merchandise volume. That was a real loss, and a familiar one. Billpoint had been eBay deciding it would rather own its payments than rent them. A quarter-century later, the retailer had done the same thing through an equity side door.
Hard times came on schedule. In mid-2022, as interest rates rose, the securitizations that funded about a third of Affirm's lending got cheaper for investors to buy and more expensive for Affirm to issue. In February 2023 the company cut 19% of its workforce and shut down its crypto unit. There was no drawn-out wind-down. The pattern from Slide held: when a thing should end, end it.
Then came the regulator, and here Levchin did something most fintech founders do not do. In 2015 he was appointed to the Consumer Financial Protection Bureau's advisory board for a three-year term, the first Silicon Valley executive given that seat. That same year, PayPal Credit, which was the old Bill Me Later business that eBay had bought for $945 million in 2008 and rebranded, agreed to pay $25 million to settle a CFPB complaint alleging that consumers had been signed up for credit accounts without their knowledge. The company Levchin co-founded was settling with the bureau in the same year the bureau was asking him for advice. If he noted the irony in public, I have not found it.
When the CFPB opened an inquiry into buy-now-pay-later lenders in December 2021, Affirm said it welcomed the investigation. In 2021 Levchin argued publicly that the tech industry needed to engage more with regulators. In May 2024, the bureau applied some of the credit-card rules under the Truth in Lending Act to buy-now-pay-later companies: dispute investigations, refunds for returned products, periodic statements. In late January 2026, Affirm applied to Nevada's Financial Institutions Division and to the FDIC to charter an industrial loan company called Affirm Bank. Levchin described the application as a bid for regulatory clarity for Affirm and its partners. He called it a long-term investment, and he said approval could take many years, with an uncertain outcome.
So the young man who once could not get the FBI to chase his thieves is now asking the federal government to examine his books. That is an evolution in posture, though not in philosophy. At PayPal, the state was absent, and so the company had to build its own police. At Affirm, the state is present, and the company has decided to join it rather than route around it. Either way, the underlying belief is the same one a teenager formed in Kyiv. The system you rely on must be made trustworthy, by someone, on purpose.
The numbers for fiscal 2026, reported in August of that year, show what that belief became when it compounded: $50.2 billion in payments processed, $4.26 billion in revenue, $147 million in operating income, $1.93 billion in net income, 27.8 million customers, 5.2 million card holders, 571,000 merchants, and 2,358 employees. In October 2025 the Coca-Cola Company elected him to its board. In April 2026, Forbes ranked him among its "Self-Made 250: The Greatest Living Self-Made Americans." He has funded the Levchin Prize since 2016, awarded each year for advances in cryptography with real-world impact. Among its early honorees were the programmers behind AES and the Signal messaging app: people who build the mathematics that lets strangers trust one another.
For a full narrative of the life, see Bold Impact- Max Levchin: A True Story of Grit, Genius, and the Fight to Make Tech More Human. The PayPal years are covered in Jimmy Soni's The Founders and Eric Jackson's The PayPal Wars. Soni ends his book with a note to his young daughter: "Your life will be shaped by the things that you create and the people you make them with."
Measured breath
Return to the launch party, to the $3 million moving across a Palo Alto room on a PalmPilot in July 1999. Everyone watched the money. Almost no one thought about what had to be true for it to arrive. Someone had to be sure the sender was the sender. Someone had to be sure the receiver was the receiver. Someone had to be sure that the people the system served were people at all, and that they would not take more than they were owed.
Levchin has spent his life on those questions. He asked them as a cryptographer, as a fraud-buster, as an investor who hired people he expected to leave, as a lender who refused to profit from the borrower's slip, and finally as an applicant to the institutions he once had to build around. He has never been the loudest person in the photograph. In the most famous one, he is the consigliere: the adviser who sits slightly to the side and is trusted with the family's secrets.
His own origin story, the one he tells when asked, is not about tanks or money. It is about a child in Kyiv whose lungs were not supposed to be strong enough. His grandmother had an idea. The family found an instrument.
The record does not say what he played, or whether he was any good. It says only that he learned, day after day, to draw in more air than his body wanted to hold, and to let it out in measured portions.
Part IIThe Playbook
What follows is an attempt to extract the operating logic behind Levchin's career: the recurring decisions that connect a Soviet childhood, a fraud war, a failed social widget, and a $50 billion lending network. These are not his rules; he has not published a list. They are patterns visible in what he built, whom he hired, and what he chose to end. The register here is more analytical, but the evidence comes from Part I.
The Levchin Playbook
- 1.Treat the adversary as your product spec.
- 2.Make verification a design surface, not a tollbooth.
- 3.Recruit from your own hinterland.
- 4.Hire the people who will leave you.
- 5.Separate the founding role from the ruling role.
- 6.Keep the core competence; abandon the thesis.
- 7.Write your filter before you choose your problem.
- 8.Profit from the customer's success, not their stumble.
- 9.Walk toward the regulator.
- 10.Rent the platform, but never mistake it for home.
- 11.End things on the day they should end.
- 12.Play the second chair until you're ready for the first.
L
The Levchin Arc
Every chapter returns to the question of how strangers can trust each other through a machine.
1991
Family settles in Chicago from Kyiv amid the Soviet collapse.
1995
Co-founds SponsorNet New Media with Luke Nosek and Scott Banister at Illinois.
1998
Founds Fieldlink with Peter Thiel; convinces the investor to become CEO.
1999
PayPal launches; Nokia Ventures sends $3M via PalmPilot at the launch party.
2000
Confinity merges with X.com; Levchin leads the anti-fraud effort through 2001.
2002
PayPal IPO in February; eBay acquisition closes in October. Named Innovator of the Year at 26.
2004
Founds Slide; becomes an early backer and chairman of Yelp.
2010
Slide sold to Google for $182M; SciFi VC co-founded with Nellie Levchin.
2011
Leaves Google as Slide shuts down; founds HVF.
2012
Affirm spins out of HVF; Levchin becomes CEO in 2014.
2015
Joins CFPB advisory board; steps down as Yelp chairman.
2021
Affirm IPO raises ~$1.2B; his stake reaches ~$2.5B.
2026
Affirm applies for a bank charter; processes $50.2B in fiscal-year payments.
Principle 1
Treat the adversary as your product spec
Most companies treat fraud as a cost center, a line item to contain once the real product is built. At PayPal, the losses ran to millions of dollars a month, the FBI could not pursue international cases, and fraud costs limited the company's ability to price its service affordably. Levchin's response was to treat the attacker as the most informative user the company had. Igor, the account-freezing system, and the Gausebeck-Levchin test were not add-ons. They defined what PayPal could charge and therefore whether it could exist.
The deeper lesson is that adversaries tell you the true shape of your business. Every fraudster probing PayPal was running an experiment on the edges of its trust model. Gettings, Levchin's risk lead, carried that adversarial approach into Palantir and then into Affirm's underwriting. In each case the core intellectual property was a model of how bad actors behave, and it turned out to be portable.
Tactic: Assign your best engineer, not your most junior one, to study whoever is trying to exploit your product, and treat their findings as roadmap input.
Principle 2
Make verification a design surface, not a tollbooth
The CAPTCHA is perhaps the most widely seen piece of Levchin's work, and it is a piece of friction. But it was friction in the right place, with a clear rationale, and simple enough for any honest user to clear in seconds. It turned an engineering problem — scripts creating accounts in bulk — into a small user interaction.
Affirm's real-time approval at checkout follows the same logic. The verification happens inside the purchase, takes moments, and replaces the long credit-card application with a decision tied to one transaction. Affirm says it looks beyond credit scores, at factors as fine-grained as the time of day. Done well, trust-checking stops being an obstacle in front of the product. It becomes part of the product.
Tactic: Map every place your product asks a user to prove something, and redesign each one so the honest user experiences it as speed rather than suspicion.
Principle 3
Recruit from your own hinterland
Confinity's engineers came mostly from the University of Illinois, recruited by a CTO who had studied there. That talent pool was overlooked by Silicon Valley's usual recruiting channels, deeply technical, and personally connected to the person hiring. The business side came from Thiel's Stanford Review network. Two pipes, each controlled by a founder who knew its people well.
This is less romantic than it sounds and more strategic. Recruiting where competitors are not looking lowers the price of talent and raises its loyalty. The people you bring over from your own place arrive with trust already built in, and in a company whose survival depends on a few hundred thousand lines of code, trust among the people writing it is not a soft benefit.
Tactic: Identify the one talent pool you understand better than any competitor, whether a school, a city, or a community, and make it your primary source before you go to the open market.
Principle 4
Hire the people who will leave you
Levchin said he deliberately screened PayPal candidates for entrepreneurial ambition, favoring those who wanted to start their own companies. The predictable result was a mass exodus after the eBay sale. All but 12 of the first 50 employees left within four years. The less predictable result was the PayPal Mafia: a network that founded or funded YouTube, Yelp, LinkedIn, Palantir, and many others, and that has fed deal flow, co-investment, and advice back to its members for two decades.
He has kept doing it. SentiLink, founded by Affirm alumni, and Resolve, an Affirm spinout, both count him as an investor. An employee who leaves to found a company is not lost. He becomes part of your network, and often your portfolio. The cost is retention. The return is a reputation that draws the most ambitious candidates in.
Tactic: In interviews, ask candidates what company they hope to found someday, and treat a strong answer as a reason to hire rather than a flight risk.
Principle 5
Separate the founding role from the ruling role
The company's timeline is explicit. The idea for Confinity was Levchin's, Thiel initially wanted only to invest, and Levchin persuaded him to be CEO. At Affirm, Levchin co-founded the company in 2012 and took the CEO role in 2014. In both cases the person who conceived the company did not automatically assume command. The question of who should lead was answered separately from the question of whose idea it was.
That separation is rare and valuable. Founders often confuse authorship with fitness to govern. Levchin seems to have asked, each time, a more practical question: who should hold the title right now? At twenty-three, the answer was a former derivatives trader with a network. At thirty-nine, with a fraud war, a social-media company, and a sale to Google behind him, it was himself.
Tactic: Before assigning yourself the CEO title, write down which of the company's next eighteen months of problems you are unusually well-suited to solve, and ask whether someone else's list is longer.
Principle 6
Keep the core competence; abandon the thesis
Fieldlink's thesis was encrypted digital wallets on PalmPilots. The thesis failed. The competence underneath it — cryptography, security engineering, secure movement of value — turned into email payments on eBay and then into the fraud systems that saved PayPal. The handheld was a wrapper. Levchin let the wrapper go and kept what was inside.
Affirm shows the same pattern in reverse. The competence is pattern recognition about trust in data, a lineage that runs from Igor through Gettings to Affirm's underwriting. The products built on top of it have changed repeatedly: checkout loans, a virtual card, a debit card, a savings account, a white-label product for banks. The crypto unit was shut down in 2023 without visible regret. The thesis is negotiable. The competence is not.
Tactic: Write one sentence describing what your team is uniquely good at that does not mention your current product, and evaluate every pivot against that sentence rather than against the product.
Principle 7
Write your filter before you choose your problem
After Slide, Levchin did not pick his next company. He named his criteria first: Hard, Valuable, and Fun. HVF was a studio built around a filter, and its spinouts — Affirm, Glow, an investment in Emerald Therapeutics — were what got through. The acronym reads as a diagnosis of Slide, which was arguably fun and valuable but not hard enough to defend.
Most founders choose an opportunity and then rationalize it. Writing criteria in advance makes the decision legible to yourself and to co-founders, and protects against the pull of whatever is fashionable. It also reveals what you got wrong last time, which may be the most useful thing an experienced founder can carry forward.
Tactic: Before evaluating your next venture, write three non-negotiable criteria that your last project failed to meet, and reject any idea that misses one of them.
Principle 8
Profit from the customer's success, not their stumble
Affirm charges no late fees and no compound interest. It earns money from merchant fees and from interest disclosed upfront. When a borrower falls behind, Affirm's main remedy is to stop lending to that person. This places the company's profit on the side of the borrower who pays, not the one who struggles, in contrast to a credit-card model that earns heavily from penalty charges.
This is not charity. It is a bet that aligned incentives produce better long-run economics: lower acquisition costs, more repeat customers, merchants who trust that the financing will not anger their shoppers, and a stronger position with regulators. The fiscal 2026 results, $1.93 billion in net income on $4.26 billion in revenue, suggest the bet can work at scale.
⚖
Two Machines for Trust
PayPal fought strangers who would cheat the company. Affirm tries to prove to strangers that the company will not cheat them.
| Dimension | PayPal (1999–2002) | Affirm (2012–present) |
|---|---|---|
| Core question | Is this person who they claim to be? | Will this person do what they promise? |
| Primary adversary | International fraud rings Risk | Credit loss and rate cycles Risk |
| Distribution host | eBay, which built a rival (Billpoint) | Shopify, Amazon, Apple Pay; lost Walmart to Klarna |
| Posture toward the state | Build your own police; FBI unavailable | CFPB adviser; bank charter application |
| Revenue logic | Shifted from float interest to service fees | Merchant fees plus upfront interest; no late fees |
| Levchin's title | CTO Second chair | Chair and CEO First chair |
Tactic: Audit your revenue lines and identify any that grow when your customer has a bad outcome. Replace at least one of them with pricing that grows when the customer succeeds.
Principle 9
Walk toward the regulator
At PayPal, the state was mostly absent. The FBI could not pursue international fraud, and the company built its own enforcement. At Affirm, Levchin has gone the other way. He accepted a seat on the CFPB's advisory board in 2015, the first Silicon Valley executive to do so. Affirm welcomed the bureau's 2021 inquiry into buy-now-pay-later companies. He argued in 2021 that the tech industry should engage more with regulators. In 2026, the company applied for a bank charter.
This is not deference. It is a recognition that in lending, regulatory clarity is a competitive moat. A charter that takes years to obtain is a barrier for anyone who comes after. A founder who helped shape the rules understands them better than competitors who resisted. And the alternative was on display in the same year: PayPal Credit paying $25 million to settle CFPB allegations while Levchin sat on the bureau's advisory board.
Tactic: If your industry has a regulator, get someone senior in a room with it this quarter, before the regulator decides what your industry is.
Principle 10
Rent the platform, but never mistake it for home
PayPal grew on eBay, and eBay responded by launching Billpoint the day before Confinity and X.com announced their merger. eBay eventually bought PayPal, then spun it off in 2015 and moved card processing to Adyen. Slide grew on MySpace and Facebook and lived at their discretion. Affirm grew on Shopify, Amazon, Walmart, and Apple Pay, and lost Walmart to Klarna in 2025, at a cost of about 5% of its volume.
Levchin's career shows both how necessary platforms are and how dangerous dependence on them is. The platforms provide reach that no startup can match. They also hold leverage that no startup can match. Affirm's response has been to diversify across many hosts and to build direct channels: the Affirm Card, the savings account, the bank charter.
P
Platform Dependence, Then and Now
Each host offered reach. Each held leverage.
| Host | What it gave | What it took or threatened | Status |
|---|---|---|---|
| eBay | PayPal's first mass market | Launched Billpoint; later acquired PayPal | Absorbed |
| MySpace / Facebook | Slide's audience | Controlled distribution terms | Shut down |
| Walmart | ~5% of Affirm GMV | Switched to Klarna in 2025 | Lost |
| Apple Pay | Embedded distribution on iPhone | Apple had built a rival, then closed it | Won |
| Shopify | Exclusive pay-over-time provider, expanding abroad | Concentration risk | Expanding |
Tactic: For every platform that supplies more than 10% of your volume, keep a written plan for the day it builds your product itself.
Principle 11
End things on the day they should end
Google bought Slide in August 2010. Levchin joined on August 25. On August 26, 2011, Google announced it was shutting Slide down and that Levchin was leaving. He did not stay to collect retention packages or drift inside the parent company. Within weeks he had started HVF. In February 2023, facing higher rates, Affirm cut 19% of staff and closed its crypto unit in a single announcement.
Clean endings preserve the founder's most limited resource, which is time. They also preserve credibility. Teams and investors are more willing to follow a leader into the next thing when they have watched him close the last thing honestly. The temptation to let a declining project linger, out of sentiment or to avoid an admission, eats years that do not come back.
Tactic: For each initiative you run, write the specific condition under which you would shut it down, and review that list on a fixed date every quarter.
Principle 12
Play the second chair until you're ready for the first
Levchin has spent much of his career in the adviser's seat. He was CTO to Thiel's CEO at PayPal. He was the "consigliere" of the PayPal Mafia, a label he embraced. He was chairman of Yelp but not its chief executive, the angel who kept Mixpanel alive, a board member at Evernote, Yahoo, Unity, and now Coca-Cola. The second chair gave him visibility into dozens of companies without the operational burden of any of them. It also gave him pattern recognition he could not have bought.
When he took the first chair at Affirm, he arrived with that accumulated judgment. The consigliere is valuable because he has watched many families make the same mistakes. A founder who has spent years advising other companies through crises has, in effect, rehearsed his own.
Tactic: Take a board or advisory seat at a company one stage ahead of your own, and keep a private log of every decision you would have made differently.
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
— Peter Thiel, on PayPal's missionEveryone in the world needs money – to get paid, to trade, to live. Paper money is an ancient technology and an inconvenient means of payment. You can run out of it. It wears out. It can get lost or stolen.
— Fortune, 'The PayPal Mafia,' 2007Today they call upon one another when they need money or advice—and when they need both, they go to Thiel, who seems to be at the center of it all.
— Jimmy Soni, The FoundersYour life will be shaped by the things that you create and the people you make them with. We tend to sweat the former. We don't worry enough about the latter.
Maxims
- The attacker writes your spec. Whoever is trying to cheat your system knows its real boundaries better than your product team does.
- Make proof feel like speed. Verification that honest users barely notice is a moat. Verification that punishes them is a leak.
- Bring your people over. The talent pool you understand better than anyone else is cheaper, more loyal, and overlooked by competitors.
- Hire for the exit. People who want to found companies make the strongest employees and the most valuable alumni.
- Authorship is not command. Decide who should run the company separately from deciding whose idea it was.
- Name the filter first. Write down what your next project must be before you fall in love with any particular one.
- Get paid when they win. Revenue that grows with customer failure eventually turns customers into adversaries.
- Meet the state early. In regulated industries, the founder who helps shape the rules gets to live inside them.
- Every host is a future rival. Use the platform's reach, and keep a plan for the day it builds your product itself.
- Close cleanly, then begin. Ending a project on time is what earns you the trust to start the next one.
