Chapter 1Seoul

Foreword

The founders I write about usually make me work for them. The fraud that interests a business reporter is, almost by definition, the part of a company that was never meant to be seen: the side letter, the revised projection, the email sent at two in the morning to a lawyer who had advised against sending it. You reconstruct the crime the way an archaeologist reconstructs a room, from the few hard objects that survived the fire, and you spend years learning to read the silences between them. I had assumed Do Kwon would be more of the same. I was wrong about the method before I was wrong about anything else.

Because Kwon did something almost none of my previous subjects had the nerve, or the carelessness, to do. He narrated it. Not in a confession, and not to me — I have never met the man and make no claim to have — but out loud, in public, in real time, to an audience of more than a million people, on a platform that stamped the date and the hour on every word. The taunts, the bets, the serene contempt for anyone who asked how a stablecoin without reserves could pay twenty percent forever: all of it is still there, timestamped, in his own hand. He called his critics poor. He told a serious economist he had no change on him for her. He accepted a ten-million-dollar wager that his own currency would hold, and lost it while running on a forged passport. When a community formed around him, he named them Lunatics, and they wore the name with pride, and I have come to think that the naming was the whole thing in miniature.

That was the story I could not put down. Not the mechanism — the mechanism was a known failure, a design that had collapsed once before under Kwon's own hands and that the academic literature had buried years earlier. What held me was that the fraud and its language were the same object. The insult was not decoration on the scheme; it was load-bearing. He drove off the skeptics with contempt and kept the believers with confidence, and the community that remained was selected, tweet by tweet, for its willingness to stop asking the one question that would have saved it.

So the research was not what I expected. I went in braced to dig — to file requests, read the unsealed indictment and the bankruptcy record, coax the private man out from behind the public one. The private man is in this book too: the internal emails, the quiet worries about the very reserves he was calling a fortress, the gap between the operator and the performer. But I did not have to excavate the crime. The crime had been broadcast. My work was mostly the humbler kind — to lay the public words beside the private record and the court's findings, to read the timeline against the charges, and to resist the temptation to improve on a story that needed no improvement. He said the cruel thing, and then the cruel thing came true, in the precise sequence that turns arrogance into evidence.

I have written about men who lied in the dark and men who lied to your face, and I thought I understood the range. Kwon is a third kind. He lied in the open, at volume, with timestamps, and the openness was not a lapse in his cover but a feature of the machine: the confidence was the product, and the product was what he was selling. A man who tells you the truth about his own recklessness while calling you poor for noticing it is a genuinely new problem, and new problems are the only honest reason to spend three years inside another person's public record.

What makes him worth a book, in the end, is not the size of the loss, though the loss was enormous and I have tried never to let the reader forget whom it landed on. It is that Kwon left the clearest confession of any founder of his era, and did not know he was making it, because he did not believe the words counted as anything but marketing. They counted. They are why this book could be written at all, and they are, I think, the truest thing he ever gave the public: an unedited record of a man who mistook the sound of his own certainty for the fact of being right.

— Sterling Aldridge, Atlanta

Chapter 1 · Seoul

Kwon Do-hyung was born in Seoul on September 6, 1991. His parents were middle-class professionals. His father worked in finance; his mother was an educator. The family lived in Gangnam, the affluent district south of the Han River where Seoul's striving professional class had been concentrating itself since the 1980s, and where the competitive schooling culture that would shape Kwon's childhood had reached, by the early 1990s, its most intense form. The Gangnam of Kwon's childhood was already the Gangnam of aspirational Korea: the streets that would, in 2012, be celebrated and satirized in equal measure by Psy's global hit, the district that encoded within its apartment blocks and coffee shops and cram-school signs the particular Korean dream of elevation through education, relentless and unsparing and profoundly meritocratic in its stated ideals, profoundly class-stratified in its actual operation.

South Korean education, for a bright child of Gangnam parents in the 1990s, was a kind of hell. The school day was long. The after-school academies, the hagwons, extended the day until ten or eleven at night. Children were tested constantly. Their results were ranked. Their rankings were posted. Parents compared their children's rankings with the rankings of their colleagues' children, with the exhausting precision that competitive cultures reserve for the things that most definitively sort the winners from the losers. The hagwon system had, by the early 1990s, grown into a billion-dollar industry premised on the assumption that the margin between a child who gained admission to an elite university and a child who did not was a margin that could be purchased: that sufficient hours of supplemental instruction could compress the natural distribution of ability into the shape that the university entrance examination required. By the time a Gangnam child was twelve, he had been trained in a set of skills: intense focus, rapid memorization, comfort with rank-ordered competition, and the ability to work for eighteen hours a day on problems chosen by someone else.

Kwon was good at all of this. He tested well. He was quick. He had the kind of analytical intelligence that Korean education rewarded: pattern recognition across domains, speed with abstract problems, the capacity to see the structure of a question faster than his peers could finish reading it. By his early teens he had been identified as a candidate for the country's elite academic tracks. He attended Daewon Foreign Language High School in Seoul, one of the most selective secondary schools in Korea, a school whose name was synonymous, in Korean educational culture, with the kind of concentrated talent that the country's system produced at its upper end, and from which a substantial fraction of graduates went on to Stanford, Harvard, or Seoul National University. Daewon's student body had been selected from the entire country through an entrance process that filtered tens of thousands of applicants down to approximately four hundred students per year. To be at Daewon was to have already demonstrated something. The question Daewon asked was what you would demonstrate next.

At Daewon, in a class of approximately four hundred, Kwon was in the top ten percent. He was intense. He was articulate. He was, in the phrase his classmates used years later, ambitious in the way Korean parents wanted their children to be ambitious: not in the passive or deferential way that Korean social conventions encouraged, but driven, visibly hungry, oriented toward the kind of achievement that would make the competitive ledger of Gangnam unambiguous. He was also notably skilled in English. Daewon's emphasis on foreign languages gave him a fluency that went beyond the functional competence most Korean students achieved; he was capable, by his late teens, of writing in English at a level that approximated native fluency, and of arguing in English with the precision and aggression that would, later, characterize his Twitter persona.

He also had a trait that was less common in the Daewon environment. He was arrogant. Not in the quiet, deferential way that Korean teenagers sometimes masked ambition with humility, but openly. He told classmates what he thought their weaknesses were. He corrected teachers when he believed they were wrong. He argued. He did not back down from arguments. The classmates who remembered him from Daewon would describe, years later, a boy who was enormously capable and enormously certain of his capability, and who was, at seventeen, already operating with the conviction that the rules that applied to other people did not apply to him. In a culture that valued deference to authority and the collective suppression of individual ego for the sake of group harmony, Kwon's particular brand of confident insubordination was unusual. It was not punished, because the performance justified it. But it was noted.

The Daewon years also gave Kwon his first sustained exposure to international peers. Foreign Language High Schools in Korea, by design, brought together students with exceptional linguistic aptitude and exposed them to global educational frameworks more than conventional Korean high schools did. The culture was slightly more cosmopolitan, slightly less bound by the strictest hierarchies of the Korean educational establishment. For Kwon, this exposure to a world slightly larger than the Gangnam parents' imagination had mapped out for their children provided an early template: it was possible to be Korean, to have been shaped by the Korean educational machine, and still to orient your ambitions toward something that was not Seoul National University and a government ministry position. It was possible, in other words, to think globally from the Korean starting point.

The Korean context of Kwon's childhood and education deserves some further development, because the features of Korean society in the 1990s and 2000s that shaped him were not simply biographical background. They were the cultural infrastructure from which he drew the skills, ambitions, and pathologies that would produce both Terra's success and its fraud.

South Korea in the 1990s was a country in the middle of a transformation that its own citizens experienced as both exhilarating and disorienting. The Korean economic miracle, the rapid industrialization and export-driven growth that had taken the country from the poverty of the post-Korean-War years to the middle-income status it occupied by the 1990s, had produced a society oriented around economic competition and upward mobility in ways that were more intense and more explicit than comparable societies elsewhere. The Korean concept of "ppal-li ppal-li" — hurry hurry, the cultural imperative toward speed and efficiency and getting-it-done — was not merely a work style. It was a worldview, an understanding that progress was a race and that the race demanded everything.

This worldview had produced, by Kwon's childhood, the hagwon system, the exam culture, and the particular form of Gangnam parenting that measured children against each other with the precision of industrial quality control. It had also produced, more diffusely, a tolerance for risk-taking and even for aggression in the pursuit of advancement: a cultural environment in which the person who moved fastest, who claimed the most space, who refused to defer when deference would have cost momentum, was not penalized but rewarded. The boldness that Kwon displayed at Daewon — correcting teachers, refusing to back down from arguments, operating with a certainty that his classmates found unusual — was not simply a personal temperament. It was a temperament that the Korean competitive culture had selected for, at least in the schools and the households where competition was most intense.

The 2008 financial crisis, which Kwon observed during his upper-secondary years at Daewon, was also formative in its own way. Korea experienced the crisis through the filter of its own financial vulnerabilities. The Korean won declined sharply against the dollar, Korean banks required government support, and the export-dependent Korean economy contracted in ways that produced visible economic distress for Korean families. The crisis had two effects on the generation that would become, in the next decade, cryptocurrency's most enthusiastic adopters. It produced distrust of conventional financial institutions and of the fiat currencies those institutions managed. And it produced an appetite for financial systems that operated outside the control of central banks and governments, the appetite that Bitcoin would satisfy when it launched in 2009 and that Terra would capitalize on when it launched in 2018.

In 2010 Kwon was admitted to Stanford University. He flew to California that August. He declared his major in Computer Science. His years at Stanford followed the template that had produced a generation of Silicon Valley entrepreneurs. He took the standard CS sequence: algorithms, operating systems, distributed systems, machine learning. He joined extracurricular groups: the entrepreneurship club, the Korean student association, occasional hackathons. He interned at Apple during one summer, at Microsoft during another. He graduated in 2015 with a BS in Computer Science.

What Stanford gave him, beyond the credential, was an intellectual framework for grandiosity. Stanford, by 2010, had become the American institution most directly responsible for producing the technology industry's professional class: not simply an educational factory that trained engineers, but a cultural institution that transmitted a set of beliefs about what extraordinary individuals could achieve when they refused to be constrained by conventional institutional thinking. It had also become the incubator, in recent years, for a kind of entrepreneurial arrogance that would produce some of the most notorious business failures of the coming decade. Elizabeth Holmes had dropped out of Stanford in 2004 to found Theranos. Sam Bankman-Fried, nearly the same age as Kwon but a Stanford faculty brat, his parents being Stanford law professors, would enter the same Silicon Valley ecosystem a few years later. Adam Neumann, though not Stanford-educated, would be funded by Stanford-educated venture capitalists who applied to his WeWork the same logic they applied to everything: that the right founder, sufficiently committed to a vision, could bend reality.

The Stanford founder-myth had a particular content. It held that disruption was not just a strategy but a moral obligation: that incumbents were, by definition, entrenched and complacent, and that any young person with sufficient intelligence and will could outcompete them. It held that failure was not catastrophic but educational, the necessary precondition for eventual success. It held that the best founders bet large, not because the odds favored them, but because the magnitude of what they were building justified the risk they were asking other people to absorb. The myth was, in its academic form, a theory of creative destruction. In its applied form, in the minds of the undergraduates who absorbed it over four years, it was moral license for the arrogance that produced the industry's most characteristic pathology: the founder who believed his vision so thoroughly that he stopped caring whether the mechanics through which he pursued it were honest.

Kwon fit the pattern exactly. By the time he graduated in 2015, he had absorbed the Stanford founder-myth: that a smart young man with a good idea could rewrite an industry, that traditional regulations were obstacles to be navigated rather than constraints to be respected, that the future belonged to founders who refused to accept what the existing system told them was possible. He believed all of it. He would continue to believe all of it, even after the belief had cost tens of billions of dollars and destroyed the lives of hundreds of thousands of people who had trusted him because the Stanford credential and the confident Twitter persona had seemed, from the outside, like evidence that the belief was deserved.

His first professional jobs were brief and not particularly notable. He worked at Microsoft for a few months after graduation as a software engineer, assigned to a team working on the Bing search product. He did not, by the accounts of colleagues who later spoke with reporters, distinguish himself. He complained about the pace of the work. He complained about the bureaucracy that made it impossible to move as fast as he thought the situation demanded. As many young engineers at Microsoft have wanted before him, he wanted to be doing something more important, something where the constraints on what he could build were his own ambitions rather than someone else's approval chains. After roughly six months, he left.

He moved to Apple, where he worked briefly on iOS engineering. The tenure was similarly short. He left in late 2015. Apple's engineering culture, famously demanding, was also famously collaborative and famously shaped by a hierarchy of judgment: products were approved by committees of senior engineers and designers and product managers whose collective experience filtered the impulses of individual contributors. For Kwon, who was by this point firmly convinced that he was smarter than most of the people around him, the filtering was intolerable. He lasted only a few months.

In January 2016 he co-founded Anyfi, a Seoul-based mesh-networking startup whose proposition was a peer-to-peer wireless communications layer that would let smartphones route traffic among themselves without depending on cellular carriers or the broader internet backbone. Anyfi raised approximately one million dollars from a combination of venture investors and a bureau of the South Korean government interested in disaster-resilient communications infrastructure. Kwon was the chief executive. The company never produced a commercially deployable product. He left in 2017, in circumstances that the existing public record does not fully describe. Anyfi is the part of his career history that he afterward tended to omit from the founder-narrative he constructed around Terraform Labs, in which his trajectory was always Stanford to crypto with the Microsoft and Apple stints folded in as brief impatient detours and the eighteen Anyfi months elided entirely. The omission was strategic. Anyfi had not worked. The founder-narrative could not accommodate a previous startup that had not worked, because the narrative depended on the premise that Kwon's technical judgment was so demonstrably superior that the conventional risk of an early-career failure did not apply to him.

His subsequent CV listed, after Microsoft and Apple and Anyfi, a brief period of self-directed project work on various concepts that did not produce commercial outputs. During it, Kwon was, by the accounts of friends who saw him during the period, increasingly convinced that cryptocurrency was the industry where his ambitions could find their appropriate scale: the industry where the rules were still being written, where the incumbents were insufficiently imagined, where a technically brilliant young man could build something transformative and claim, credibly, that the transformative thing was his alone.

Cryptocurrency, in 2016 and 2017, was entering its first major bull market. Bitcoin had climbed from approximately $400 in early 2016 to nearly $20,000 by December 2017. Ethereum, launched in 2015, had climbed from $8 to over $800. The initial coin offering boom had begun. Hundreds of new tokens were being launched, many of them with business plans that were speculative to the point of fraud, each raising tens or hundreds of millions of dollars from retail investors who believed they were participating in the next phase of financial history. The ICO boom had created a class of overnight cryptocurrency multimillionaires who had, in some cases, done nothing more technically demanding than write a few pages of a whitepaper and deploy a Solidity contract on Ethereum. The rewards had been enormous. The accountability had been essentially zero. For someone with Kwon's technical skills, his ambition, and his particular relationship with the idea that rules applied to other people, the combination was irresistible.

The broader cryptocurrency ecosystem of 2016-2017 also had a cultural texture that appealed to Kwon's temperament. It was a culture that celebrated heterodox thinking and contempt for established institutions. Bitcoin was, in its founding mythology, a rebellion against central banks and fiat currency and the financial establishment that had produced the 2008 crisis. The ICO boom had extended this mythology into a general principle: that smart young people building outside the established regulatory framework were, by definition, the forces of progress, while the regulators who tried to constrain them were the forces of reaction. The culture rewarded exactly the combination of technical confidence and institutional contempt that Kwon had spent his entire life developing.

Kwon saw what this represented. A technology ecosystem in its infancy, essentially unregulated, flooded with capital, and oriented around founders with grand visions had produced, through 2017, the greatest concentration of wealth transfer to young technical entrepreneurs in the history of the industry. It had also produced enormous fraud. The combination — the opportunity to build something legitimate, the opportunity to commit fraud and claim you were building something legitimate, the ambiguity between the two — was, for someone with Kwon's confidence and technical skills, an irresistible field. He would later describe, in interviews, the moment he decided to go into cryptocurrency as a clarity experience: a sudden recognition that this was where the most important work of the next decade was going to happen, and that he was the person who should be doing it. The clarity was probably real. The certainty it produced about his own importance was the thing that would eventually destroy him.

In January 2018, Kwon co-founded Terraform Labs with Daniel Shin, a Korean-American entrepreneur several years older who had previously co-founded Ticket Monster, a South Korean e-commerce site that had been acquired by Groupon for several hundred million dollars. Ticket Monster's acquisition had been one of the notable Korean startup success stories of the early 2010s, and Shin had, through it, accumulated both the financial resources and the industry credibility that a cryptocurrency venture required. He brought commercial legitimacy. He had actually run a business that had generated revenue and been sold, and he was the more externally-facing of the two founders in the company's early years. Kwon was the technical founder. The company was incorporated in Singapore, a jurisdiction favored by cryptocurrency entrepreneurs because of its relatively light regulatory regime, its English-language legal framework, its time zone position between Asian and European financial centers, and its distance from the Korean Financial Services Commission, which was, in 2018, still developing its approach to cryptocurrency oversight.

The initial proposition of Terraform Labs was a blockchain-based payments network. The Terra blockchain would host a family of stablecoins — cryptocurrencies pegged to the values of various national currencies — that could be used for commerce on partner platforms, including Shin's existing e-commerce network CHAI in Korea. The initial funding round, completed in early 2018, raised approximately $32 million from a consortium of cryptocurrency investors including Binance, Polychain Capital, and Hashed, the Korean venture fund that had become one of the most active investors in the Korean cryptocurrency ecosystem. The investors were buying into the thesis that algorithmic stablecoins — stablecoins whose dollar peg was maintained through mechanical supply adjustments rather than through reserves of actual dollars — were the future of digital commerce. They were also buying into Kwon: his Stanford credential, his technical confidence, his articulate and aggressive account of why the existing financial infrastructure was inadequate and why Terra was going to replace it. The combination was compelling. The $32 million arrived.

Whether algorithmic stablecoins were actually the future was, and remains, contested among cryptocurrency technologists. The mechanism was simple to describe and complex to defend. A stablecoin called TerraUSD (UST) would be issued. Its price would be maintained at $1 through an arbitrage mechanism involving a sister token called Luna. When UST was trading above $1, users could burn Luna to mint new UST, receiving the difference as profit and increasing the supply of UST until the price dropped back to $1. When UST was trading below $1, users could burn UST to mint new Luna, receiving the difference and reducing the UST supply until the price rose to $1. The mechanism depended, at every step, on participants believing that the system would continue functioning. As long as they believed, the arbitrage worked. If they stopped believing, the arbitrage reversed and accelerated the collapse.

The concept of an algorithmic stablecoin was not original to Kwon. Several prior attempts (Basis, NuBits, Steem Dollars) had been made since 2014, and all of them had failed. The failures were, in the cryptocurrency research community, well-documented. The failure mode was also well-documented: under sufficient selling pressure, the algorithmic mechanism would enter a death spiral in which falling prices triggered further minting of the sister token, which further reduced its price, which accelerated the selling pressure, which further broke the stablecoin peg. The mechanism, once broken, could not recover.

Kwon was aware of all of this. He read the papers. He participated in the technical discussions. He had, years earlier, been involved in the design of Basis Cash, an earlier failed algorithmic stablecoin: a detail that would not become public until after the Terra collapse and that would become, in retrospect, one of the central facts of the fraud case against him. He had watched, from inside a development team, the mechanism he was now building at vastly larger scale fail in precisely the way the academic literature had predicted it would fail. He had processed this experience not as a warning but as an invitation. The prior failure had been a matter of insufficient scale, insufficient market adoption, insufficient liquidity. He would provide all three. He launched Terra anyway.

The company's first years produced steady growth. Terra's blockchain was technically solid. The CHAI partnership in Korea produced real commercial adoption — by 2019, CHAI was processing tens of thousands of daily transactions denominated in Terra's Korean-won-pegged stablecoin, giving the project something that the vast majority of cryptocurrency ventures lacked: an actual user base conducting actual commerce. The token economy attracted cryptocurrency speculators. By 2020, Terraform Labs was one of the mid-sized cryptocurrency projects, with a market capitalization for its Luna token of roughly $1 billion. Kwon, at twenty-nine, had become a person in the cryptocurrency industry: not a major figure yet, but a credible one, a Korean Stanford graduate running a well-funded blockchain project with legitimate commercial partnerships and a growing technical team.

The explosion came in 2021. Two events coincided to produce it. First, the broader cryptocurrency market entered its second major bull market, with Bitcoin reaching $60,000 and total crypto market capitalization exceeding $3 trillion: numbers that, even in the cryptocurrency industry's habituation to improbable valuations, produced a collective sense that something historically significant was occurring. Second, Terraform Labs launched Anchor Protocol, a lending platform built on Terra that offered approximately 20% annual yield on UST deposits. The 20% yield was, by the standards of conventional financial markets, impossibly high. It was being funded, in its early months, by subsidies from the Terraform Labs treasury. The yield was not actually being generated by legitimate lending activities; it was being paid out of a reserve that the company was spending down, on the calculated gamble that the inflowing deposits would eventually generate the commercial activity needed to sustain the yield organically.

The gamble never resolved. The yield never became self-sustaining. But it remained high for long enough to attract, between 2021 and early 2022, approximately $14 billion in UST deposits to Anchor Protocol, a gravitational pull created by the simple mathematics of a 20% annual return on what was marketed as a stable, dollar-pegged asset. Between 2021 and early 2022, Anchor deposits grew from under $1 billion to over $15 billion. Luna's price, tied to the broader growth of the Terra ecosystem, climbed from roughly $1 in early 2021 to a peak of $119 in April 2022. At the peak, Luna had a market capitalization of approximately $40 billion. UST had become the third-largest stablecoin in the world. Kwon had become, on paper, a billionaire many times over.

The next chapter traces what Kwon was doing, publicly and privately, during the fifteen months when Terra was becoming an apparent cryptocurrency juggernaut. In public he was building a Twitter persona as the voice of algorithmic-stablecoin evangelism: brash, dismissive, explicitly contemptuous of critics, performing for the million followers who had come to treat his confidence as a form of product endorsement. In private he was, according to the subsequent federal indictment and the internal documents it relied upon, aware that the yield was not sustainable, that the peg mechanism could not withstand significant selling pressure, and that the system he was selling to millions of retail investors was structurally identical to the earlier failed algorithmic stablecoins that had taught the cryptocurrency industry, in theory, what not to do.

He knew. He built it anyway. He sold it anyway. He attacked the people who questioned it. And when it collapsed, in May 2022, erasing in five days the savings of hundreds of thousands of ordinary people who had trusted the Stanford graduate from Gangnam, he ran.