Lunacy Episode 1: Death Spiral
This week, we dig into the rise of Do Kwon, the enormous appeal of his almost magical algorithmic stablecoin – and the far less charmed reality underlying his machinations from the very start.
This week, we dig into the rise of Do Kwon, the enormous appeal of his almost magical algorithmic stablecoin – and the far less charmed reality underlying his machinations from the very start.
"Crypto Crooks" is sponsored by Chainalysis.
Do Kwon said he could print money out of thin air. Who could resist a sell like that?
Welcome back to “Crypto Crooks.” This season, over the course of four episodes plus a bonus fifth on recent developments, we tackle Do Kwon and his so-called algorithmic stablecoin. We’re starting with the seductive appeal of the concept. Everyone, from crypto’s bigwig financiers to hundreds of thousands of retail depositors, was buying into the dream.
But the fatal flaws of this collective fantasy were evident from the very start. It was only a matter of time before the death spiral began.
Chainalysis is the blockchain data platform. We provide data, software, services and research to government agencies, web3 companies, financial institutions and insurance and cybersecurity companies. Our data powers investigation, compliance and business intelligence software that has been used to solve some of the world’s most high-profile criminal cases. For more information, visit www.chainalysis.com.
“Crypto Crooks” is a CoinDesk Podcast Production. The executive producer is Jared Schwartz, with additional production by Eleanor Pahl, Nora Battelle, Jonas Huck, and Moon Beast. Fact-checking is by Amber Von Schassen, and sound design and music are by Altus Noumena. This show is written and voiced by David Z. Morris.
Audio Transcript: This transcript has not been edited and may contain errors.
DAVID Z. MORRIS:
On May 7th of 2022, the end began.
That day, for a few short moments, a cryptocurrency called TerraUSD could be bought for just under ninety-nine cents on crypto exchanges such as Kraken and Binance. For most blockchain-based tokens, this would be no big deal – cryptocurrencies are notoriously volatile.
But TerraUSD was different from Bitcoin, Ethereum, or Dogecoin: it was supposed to maintain a “peg” of one U.S. dollar. It did this without any external backing, such as dollars in a bank.
Instead, TerraUSD was what’s known as an “algorithmic stablecoin.” A pricing algorithm balanced TerraUSD against a second token called Luna. Luna acted as a financial “shock absorber” to keep TerraUSD’s price stable at one dollar.
Both tokens were tracked by a blockchain network, also called Terra. But there was nothing else behind TerraUSD, making its one-dollar value feel like nothing short of free-market magic.
For more than a year, TerraUSD had kept that magic promise of trading at one dollar. Meanwhile, the Terra blockchain, including the Luna token, attracted hordes of both retail speculators and institutional investors. On May 6, nearly $19 billion dollars worth of TerraUSD was in circulation. Luna, its sister token, had a total market value of more than $27 billion.
This left the financial wizard behind the magic riding very high.
Do Kwon, a Stanford graduate based in Singapore, had become a prominent figure in the accelerating world of cryptocurrency. And he wasn’t afraid to flaunt it. Anyone with questions about the design of his algorithmically stabilized blockchain dollar could expect nothing but insults from Do. He called his critics, quote, “stupid,” “re****ed,” and, his favorite insult – “poor.”
CoinDesk reporter Sam Kessler has followed the Terra saga closely. He says this behavior was about much more than being an ***hole for the fun of it.
“An important thing to realize about Do Kwon’s public persona was that … it's almost in the mold of Elon Musk and the Tesla stock where … where he would kind of try to portray himself as a savior in a certain way. But it was also a way to pump up the price of something … And the best way to get people to trust in that currency, so as not to send it plummeting with a bunch of sell offs is to, I guess, portray yourself and the project in this … super bombastic, you know, jokey, you know, tuned for crypto Twitter way that would keep people not only, again, confident in the project, but also excited to be a part of this thing.”
But on May 7th of 2022, starting just after 5pm, something changed. Traders began withdrawing huge amounts of money from the Terra system.
In the space of a few hours, they swapped $180 million worth of TerraUSD for Tether, a stablecoin backed by dollars in a bank. About $200 million worth of TerraUSD were sold off for other tokens on crypto exchanges. Soon, traders had pulled $2.8 billion dollars’ worth of TerraUSD from a digital bank called Anchor, also built by Do Kwon on the Terra blockchain.
But it was barely 7am in Singapore. As the withdrawals and selling continued … Do Kwon was asleep.
People who owned a lot of TerraUSD had decided they’d rather own something else. The withdrawals and selling pushed down the market value of TerraUSD faster than its algorithm could adjust. For a short while on May 7th – around 6:44pm U.S. Eastern time, to be precise – you could buy a Terra dollar for zero point nine eight seven dollars.
Almost no one could have guessed that this slight waver, this tiny wobble, this missing penny – this was the chill wind welcoming the entire cryptocurrency industry to the valley of the shadow of death.
It was not yet 8am in Singapore – and still, Do Kwon was asleep.
The critics – including your host – were about to be proven right. TerraUSD would soon slip much farther from its dollar peg. Its failure would end a three-year tidal wave of public interest in cryptocurrency. It would help bring down a half-dozen interlinked operations selling the dream of effortless crypto riches – operations with names like Celsius, Three Arrows Capital, and – eventually – FTX. It would end one man’s Quixotic quest for an infinite money machine. It would destroy the precious savings of tens of thousands of people in South Korea, and around the world.
But that was still in the future.
On May 7th, at just after 8am, as TerraUSD continued to wobble, Do Kwon awoke. He acted decisively to save his creation in the best way he knew how:
“I’m up ,” he wrote, “amusing morningAnon, you could listen to CT [crypto twitter] influensooors about UST depegging for the 69th time
Or you could remember they’re all now poor, and go for a run instead”
Do Kwon and his creation were about to go for a run, alright.
The design of TerraUSD, you see, included a fatal flaw – a flaw known in finance as a “death spiral.” When everything was going right, the algorithm behind the unbacked token appeared to magically maintain its value of one dollar.
But once the peg slipped, the very same algorithm created a perverse incentive: Traders could make money by driving TerraUSD’s price further away from one dollar. And this is exactly what the free market did.
When he woke up on May 7th, Do Kwon was still a wizard in the eyes of many. He had created money from nothing – a financial miracle that seemed to defy the laws of God and man.
But within a few days, it would become clear that his creation was not magic after all – it was an illusion.
By May 9th, TerraUSD fell even further from its peg, slipping to 71 cents. Do Kwon and his team spent billions of dollars trying to prop up TerraUSD - to maintain the illusion that had made them rich and famous. But it was fruitless – by May 11th, a Terra dollar was worth just 35 cents. At the same time, the supposed balancer token Luna had lost more than 99% of its value.
There was nothing left to prop up TerraUSD. By May 12th, just five days after the death spiral began, the entire system was forced to shut down. There was no illusion left to maintain.
The collapse of Terra wiped out the life savings of tens of thousands of victims worldwide in a matter of days. A few weeks later, Do Kwon would run from international law enforcement, accused of major financial fraud. He would, eventually, run all the way to Serbia to evade capture.
And the run of a lifetime began with one. Missing. Penny.
Welcome to Crypto Crooks, Season 2: Lunacy – The Rise and Fall of Do Kwon.
Hello, I’m David Z. Morris, CoinDesk’s Chief Insights Columnist. I’ve been reporting on cryptocurrency since 2013. I’ve followed crypto so closely because I believe it will profoundly transform how we live.
But I’ve also watched as the real promise of crypto has been undermined again and again by a stream of con-men and hucksters who prey on the uninformed, the naïve, and the desperate. It remains a lightly-regulated and poorly-understood sector, which makes it a perfect hunting ground for scammers. A big part of my job here at CoinDesk is calling out sketchy projects and outright frauds, and I’ve seen enough of them over the years to know what they look like.
Terra was a very familiar bad idea. Others before Do Kwon had chased the dream of a so-called “decentralized stablecoin” – a global crypto-token that would use market incentives to maintain a steady price, without support from any underlying assets. For years, it had been cryptocurrency’s equivalent of the Philosopher’s Stone, the Fountain of Youth, lost Atlantis, or a perpetual motion machine: pure fantasies that people sought desperately, simply because they wanted them to be real.
In April of 2022, I wrote about my concerns about TerraUSD, aided by insights from academics, traders, and my colleague, reporter Sam Kessler. Together, we detailed the many weaknesses that could break TerraUSD’s dollar peg. It wouldn’t be a gentle slide, the experts warned: The incentives built into the Terra system would drive the stablecoin’s value to zero with terrifying speed. This so-called “death spiral,” they warned, would inevitably ensue once TerraUSD slipped from its dollar peg.
The CoinDesk article was titled “Built to Fail,” after research by economist Ryan Clements, and building on earlier insights from crypto lawyer Preston Byrne. It was published on April 22nd of 2022.
Just over two weeks later, TerraUSD’s death spiral began.
TerraUSD fell to ninety cents by May 10th, then to thirty-five cents the day after. The crashing price made it easier for hackers to manipulate the Terra blockchain, so on May 12th, it was halted entirely. Close to $60 billion dollars in paper value had essentially gone to zero – one of the fastest human-caused destructions of wealth in history.
We’re presenting Crypto Crooks to help educate and protect crypto investors, or just the crypto-curious, from this sort of disaster. Each short season of the show focuses on a different major fraud, crime, or bad idea that fleeced investors and the industry. Learning history will help you see the next scam coming – in cryptocurrency, and beyond. Because while crypto is a radical new technology, the fraudsters using it to rope in victims use techniques as old as the financial industry – in some cases, as old as money itself.
This season, we tell the story of a man with a fantastical dream – a dream that turned into a global nightmare.
Part 1: Money for Nothing
The Terra network was launched in January of 2018. While Do Kwon ultimately became the face of the project, he began his round trip to the moon and back with a partner, Daniel Shin.
Terra would be a blockchain, much like Bitcoin, Ethereum, and Litecoin. Blockchains are global public networks, secured and maintained by a decentralized swarm of processors who confirm transactions. These processors add transactions to the chain’s historical database, in exchange for payment in the blockchain’s token. This decentralization helps make blockchains trustworthy and neutral records, without relying on banks or governments.
Terra would have a unique value proposition. Bitcoin’s appeal was simple: it made uncensorable electronic payments possible without a third-party intermediary, like a bank or credit card processor, for the very first time. Ethereum took the Bitcoin concept and added so-called “smart contracts,” allowing more complex financial interactions.
Terra, its creators promised, would add a nearly miraculous third feature to this stack: It would be a home for synthetic dollars, won, or other “fiat” currencies issued by nation-states. But it wouldn’t rely on banks, which risked censorship or seizure by governments.
Instead, instruments like TerraUSD would be “pegged” to their fiat value using only an algorithm that leveraged the power of the free market. In broad strokes, Terra stablecoins would rely on a two-token system. Fiat-pegged stablecoins such as TerraUSD would be paired with a “balancer token” called Luna. Luna could be burned, or destroyed, to create TerraUSD, while TerraUSD could be burned to create Luna.
An automated algorithm would keep TerraUSD’s price at one dollar by adjusting the exchange rate between the two tokens. If the TerraUSD price went above the one dollar peg, indicating excess demand, the algorithm made it cheaper to create TerraUSD by destroying Luna. If TerraUSD went below its peg, the system increased the amount of Luna created by burning TerraUSD.
Put another way, when TerraUSD dropped below its peg, traders could use it to buy more than one dollar worth of Luna. This would create a profitable opportunity for cryptocurrency traders and other financial actors, also known as an arbitrage.
Crypto traders around the world, the theory went, would be quick to exploit this money-making opportunity. Their trades would lower the supply of TerraUSD, driving its value back up towards the one-dollar peg. At the same time, the supply of Luna would expand, theoretically lowering its price.
Luna, in essence, would act as a shock absorber for TerraUSD. This would, the Terra creators argued, create a flexible supply of cryptocurrency that maintained a constant value of one dollar.
And it would work with no centralized authority, such as the central banks that manage most fiat currencies. This meant it wouldn’t be answerable to conventional regulatory authorities – it could, like Bitcoin, be completely uncensorable and decentralized.
This proposition was incredibly attractive to investors. There were many reasons for the enthusiasm, but one aspect of the appeal was simple: What if you could print dollar bills, and the U.S. government couldn’t do anything to stop you?
The team behind Terra was also a major selling point. Daniel Shin was already a technology rock star in South Korea for founding the online ticket sales platform Ticket Monster. Ticket Monster was sold to Groupon for $260 million in 2014.
Daniel Shin was not mainly focused on the details of the Terra blockchain. Instead, he planned to use Terra as a payments backbone for an ecommerce payments application called Chai. Chai’s goal was to eventually compete with payment services like AliPay and Paypal, but to track and send payments via the Terra blockchain network and its algorithmic stablecoins.
That was a potentially lucrative vision. Along with Daniel Shin’s pedigree, it helped bring together a coalition of 15 large Asian e-commerce organizations, including names like Megabox and Musinsa, to form the Terra Alliance.
Ecommerce players may have been enticed in part by the longstanding prospect that cryptocurrency networks could become a less expensive form of digital payment than credit cards. Credit cards currently take around one to three percent of most online transactions, a huge bite out of retailers’ profit margins.
But Terra promised even more, according to a 2019 Medium post by Don Kim, then the Head of Business at Terra and most recently Head of Planning and Operations at Chai. According to Kim, “inflation in [Terra’s] internal cryptoeconomics leads to the issuance of new money supply to stabilize prices … and in the beginning phases of adoption, the majority of seigniorage will be used to fund steep discounts for customers using Terra to pay for goods and services on our e-commerce partner platforms.”
In retrospect, this public statement should have been a clear red flag. Not only was Terra proposing to create dollars out of thin air – another way to say what Kim here calls “seignorage.” Terra was also claiming its system could print enough money to hand out discounts to users.
Most blockchains have some function for generating new currency. But in long-running blockchains like Bitcoin and Ethereum, this seigniorage is used almost exclusively to pay for security: in Bitcoin, it’s called the “block reward” and it’s used to pay the miners who process transactions. It’s a cost of maintaining the chain, not a free bonus you can distribute to bribe retail partners or users.
This was barely the beginning of a deeper misunderstanding of cryptocurrency shared, not just by Do Kwon and his team, but also by many of the other chaos agents who crashed and burned in 2022.
Cryptocurrency is a new form of monetary technology – but it doesn’t rewrite the rules of finance. While it may look that way at first glance, cryptocurrency cannot create money from nothing – at least, not money that lasts very long.
Ryan Clements, who wrote the “Built to Fail” paper that detailed the reasons TerraUSD could not last, is very clear on this distinction.
“This is a huge deviation from the ethos of Satoshi and there's valid reasons to disintermediate and decentralize … This now turns into something that not just is self dealing, and self referential, but very opaque and heavily financialized.”
After designing and launching the Terra network, Shin and Kwon formed an entity called Terraform Labs in April 2018. Daniel Shin’s track record and ecommerce allies, along with Do Kwon’s Stanford connections, helped Terraform raise $32 million by August 2018.
But it wasn’t until April 2019 that the team publicly released a so-called “whitepaper” describing the functioning of the Terra blockchain. This was at least a little odd – most cryptocurrencies had appeared first in whitepaper form well before they raised money or launched as a live service. The Bitcoin whitepaper appeared about three months before the Bitcoin network. The Ethereum whitepaper was released about a year before the software.
This reversed order of operations hinted at what would later become all too clear. Terra’s investors, both early on and as the system grew, were betting on who Do Kwon and Daniel Shin were, more than on the content of their ideas. This can be a valid approach in traditional venture capital, where relationships and status can have huge impacts. But it seems alien to the professed goal of crypto, where transparency and decentralization are the only paths to long-term success for new blockchains.
The Terra whitepaper was coauthored by Evan Kereiakes, Do Kwon, Marco Di Maggio, and Nicholas Platias. Platias was a longtime associate of Kwon and would have a prominent place at Terraform Labs. Kereiakes was part of the research team at Terraform.
Di Maggio, stunningly, was a professor at Harvard, where Business Development head Don Kim had also graduated. That’s just one of the elite connections that seem to have eased the way to Terra’s later immense growth – and immense fallout. Despite his name and endorsement having fueled one of the biggest and most damaging crypto failures of all time, in July of 2022 Di Maggio would be named Director of the Harvard Business School’s new Fintech, Crypto, and Web3 lab.
Part 2: Drawn into the Death Spiral
The Terra whitepaper described, among other things, why a decentralized stablecoin was needed in crypto. It’s a vital question for understanding the bigger picture. Why did Do Kwon build TerraUSD? More importantly, why were big-name cryptocurrency investors and everyday folks so excited about it?
Cryptocurrency is very volatile relative to government-backed currencies, at least for now. Even very established systems like Bitcoin and Ethereum are still in relatively early stages of adoption – a large portion of holders buy them as speculative investments, much like stock in early-stage technology companies. This means that changes in sentiment, hour to hour or month to month, can have a big impact on investor interest and asset prices.
This volatility helps attract active crypto traders, who try to make money by predicting these swings in sentiment. Critics see this as one big casino, and in some sense that’s fair – day trading is an extremely risky activity, and most individuals lose money trying it, whether in crypto or the stock market. But for better or worse, there’s huge demand for seats at this casino: that’s why crypto exchanges like Binance and BitMEX, which offer advanced trading and derivatives tools, make so much money.
But crypto’s volatility is also a big problem for traders.
Converting cryptocurrency into actual government currency in a proper bank is slow and expensive. On many global crypto exchanges, it’s outright impossible. This means there is no “safe haven” for traders. There’s nowhere inside the crypto ecosystem to park their funds without the risk of suddenly losing value – even Bitcoin can still drop by 10 percent or more in a matter of a few hours. And there’s nowhere to safely park money when markets as a whole are crashing.
The first major attempt to offer a solution to this problem was called Tether. Tether is what’s known as a “backed” stablecoin. While it is tracked and moved on a blockchain, a one-dollar Tether token is a claim on real dollar assets in a bank or similar account. Tether has not made its backing transparent, however, which has long invited skepticism about its stability. Another stablecoin, Circle’s USDC, grew from 16% to 29% of all stablecoins between early 2021 and early 2022 by copying Tether’s model in a more transparent and regulated way.
But these “backed” stablecoins have a significant shortcoming: They’re run by relatively small, centralized companies. Because those companies need access to bank accounts and other parts of the traditional finance world, they are forced to comply with government demands to shut down some users – what’s referred to in crypto as “financial censorship.” Both Tether and Circle regularly freeze tokens at the behest of governments, creating opaque risk for holders.
This problem is the main driver of the quest for a “decentralized stablecoin.” A truly decentralized stablecoin would be tradable on blockchains like Ethereum, and would maintain a fixed value in fiat terms. But it wouldn’t rely on banks or the traditional financial system – and so its users wouldn’t be at risk of having their money frozen or seized. A decentralized stablecoin would be as trustworthy as Bitcoin, but without the price volatility.
There is one successful stablecoin that is substantially, if not quite completely, decentralized. It’s called DAI (D-A-I), and it’s created by a system called MakerDAO. DAI was launched in December of 2017, and has run reliably through two crypto bubbles and collapses since.
But MakerDAO’s stability comes at a huge cost - literally. To create one dollar worth of DAI, users must deposit as much as a dollar and seventy-five cents worth of cryptoassets like Ether in a digital vault. This is known as “overcollateralization,” and it helps make sure every DAI is actually backed by a dollar worth of crypto, even if the crypto declines significantly in price. But it also means DAI users lose access to a big chunk of their crypto assets, which can’t be invested, lent, or spent elsewhere. MakerDAO and DAI sacrifice capital efficiency for stability. But for traders in particular, this so-called “opportunity cost” can be huge.
From that shortcoming came the dream that would enthrall, and ultimately destroy, Do Kwon: the dream of the algorithmic stablecoin.
The first known algorithmic stablecoin was launched in 2014 – that’s right, most of a decade before TerraUSD imploded. BitUSD was launched on the BitShares blockchain – an Ethereum precursor with some of the same complex financial features. Other algorithmic stablecoins have appeared regularly since, with names like Iron Finance, Basis Cash, and neutrino USD.
Much like TerraUSD, these algorithmic stablecoins used two tokens – the stablecoin itself, and what’s called a “balancer token.” NeutrinoUSD’s balancer token was called WAVES, and IRON’s balancer was called Titan.
Some algorithmic stablecoins are partially backed, or “undercollateralized” – IRON was backed 75% by the fully-backed stablecoin USDC, for instance. But other algorithmic stablecoins, including TerraUSD, don’t use any collateral at all. Luna and TerraUSD were simply created on a blockchain, for a minimal computational cost. If the main problem with an overcollateralized stablecoin like MakerDAO’s DAI is that it doesn’t use capital efficiently, then an algorithmic stablecoin can go to the opposite extreme – it’s hard to think of anything more capital efficient than printing dollars from thin air.
That’s one reason the algorithmic stablecoin is a very seductive proposal, particularly for those fond of math and economics. It relies heavily on 18th century economist Adam Smith’s theory of the “invisible hand” – the argument that rational self-interest in a free marketplace can create the best outcomes for all.
Adam Smith’s economic assumptions held immense sway for centuries – but they have come under increasing scrutiny. In particular, it is less and less taken for granted that market actors have all the information necessary to make good economic decisions, or even that market actors are consistently rational.
These may seem like very abstract theoretical concerns, but they help explain a very important reality: algorithmic stablecoins, with their reliance on pure rationality and perfect information, don’t work.
The track record could not be more clear. Back in 2014, BitUSD lost its peg five days after launching. Basis Cash crashed in early 2021. Iron Finance, despite being 75% backed by more or less real dollars, experienced a death spiral in June of 2021. We’ll dig into this deeper in a future episode, but the structure of a partially-backed algorithmic stablecoin may make it more fragile than one with no external backing at all.
That history of failure is clear, and yet another reason it’s so staggering that professional investors bet big on TerraUSD. And really, anyone who looked at the design of algorithmic stablecoins should have intuitively asked – where is the value coming from?
Serious academic thinkers have spelled out the flaws in these ideas more formally, but come to the same incredulous conclusion. Way back in 2017, Preston Byrne, an early economic and legal thinker about crypto, declared that such projects were inevitably “doomed to fail.” Four years later, economist Dr. Ryan Clements would pen an academic paper explaining why in more formal economic terms.
“In Built to Fail, I argued that algorithmic stablecoins are inherently fragile, because of three factors that they rely on which history has shown isn’t reliable. Number one, they need a support level of demand in order to function effectively. Number two, they need a permanent supply of independent actors to perform price stabilizing arbitrage. And number three, they need near perfect information for those actors to trade on. In prior financial products in history, we’ve seen those three assumptions fail. When you need a product that requires perpetual baseline demand to function at all, and you don’t have that demand, if the market loses confidence in the ecosystem, [then] it’s going to fail.”
These assumptions – perfect information, constant demand, and a permanent pool of traders – are left out of the descriptions that make systems like TerraUSD look so appealing on paper. They are simply assumed to be permanently true, without comment. But in fact, in the real world, they’re frequently untrue – especially when markets are going down.
Despite both these theoretical doubts, and a long history of real-world failures, the allure of an algorithmic stablecoin hypnotized both amateur crypto investors and supposed professionals. Mike Novogratz was for years one of the most respected investors in crypto. But the hallucinatory promise of TerraUSD drew him in, like a cobra staring down a rat.
“You know, I met Do three or four years ago, he is a brilliant young man. And he had this idea for an algorithmic stablecoin, digital money for a digitally native world. With no central bank, no ties, no dollars saved.”
Novogratz even got a large Luna-themed tattoo, and proudly displayed it on Twitter. He was hardly alone in this behavior – by the end of 2021, both small-time amateurs and big-name financial operators were declaring themselves “Lunatics” en masse.
This is the befuddling core of the TerraUSD/Luna mystery. How did an idea that had been proven so wrong, so many times before, wind up worth $60 billion dollars? Why did so many ignore the intuitive absurdity of printing money from thin air?
But this may point to a dirty secret of Terra and Do Kwon’s rise: what if many of his supporters, especially the high-profile professionals, didn’t actually believe the algorithmic stablecoin hype at all? What if, instead, they thought that Do Kwon’s headline-generating personality would let them quickly sell their Luna investment on to naïve retail buyers who actually believed his nonsensical pronouncements?
Whatever they said out loud, that certainly seems to be how some major investors behaved. Pantera Capital told CNBC that they turned a $1.7 million Terra investment into $170 million – and cashed out to real dollars before the collapse. In the first three months of 2022, the months leading up to the TerraUSD collapse, Mike Novogratz and Galaxy Digital sold $355 million dollars’ worth of Luna. In fact, it was the investment firm’s single biggest source of gains that quarter.
More dramatic still – what if Do Kwon didn’t actually believe his own hype? Whatever his public pronouncements, he too behaved as if he had some doubts. Just before Terra’s collapse, Do announced he would build a reserve of $10 billion worth of Bitcoin to be used as backing for TerraUSD – strongly suggesting that he didn’t believe his own promises about the durability of an unbacked algorithmic stablecoin.
In the end, Terra’s algorithmic promise didn’t actually have to be true: it just had to sound convincing to a future buyer. Investors didn’t need Do Kwon to actually build a durable system: they just needed him to tell a good story, and tell it with absolute conviction.
Part 3: I Don’t Debate the Poor
There’s another reason to think both Do Kwon and his investors were being less than forthright about their faith in the stability of TerraUSD and the Terra blockchain. Vocal, qualified, and articulate critics began to poke holes in the Terra concept as it grew – but its supporters rarely seemed to have clear answers. For years, Do Kwon’s main response was bullying and intimidating critics with personal insults via Twitter.
“I hate to use the word religious, but it almost seemed like that a little bit in that there's the right way to act and the wrong way to act. And the right way to act is to just believe there's always going to go up and this is perfectly stable, and the wrong way to act is to … engage in any form of criticism.”
As early as April of 2018, a crypto analyst named Cyrus Younessi was asked to evaluate Terra as a potential investment by his employer, the venture capital fund Scalar Capital. In a message later shared by Scalar cofounder Linda Xie, Younessi wrote that:
“Unless I’m missing something, this project isn’t gonna work … it seems to me that Terra suffers from the death spiral … if [TerraUSD] were to fall and break the peg, then it would depend on [the balancer token] Luna to save Terra,” Younessi wrote. “But Luna,” he went on, “would fall as investors would panic, and then [TerraUSD] would continue to fall, and they would just each keep contributing to each other’s demise.”
This was an internal warning to Scalar Capital, and wouldn’t become public until after Terra’s collapse. But it was almost exactly how the unwind would play out, more than four years later.
But there were other clear and accurate warnings made publicly. In July of 2021, nearly a year before the Terra death spiral began, banking and finance guru Frances Coppola wrote a series of prescient Tweets questioning systems like TerraUSD.
“Self-correction mechanisms that rely on financial incentives,” she wrote in part, “Do not work when panicking humans are stampeding for the exit.”
Do Kwon replied to this entirely accurate critique with the misguided, empty bluster and vicious condescension that defined his public, and by many accounts private, persona.
“I don’t debate the poor on Twitter,” he wrote. “And sorry I don’t have any change on me for her at the moment.”
This dogged resistance to any critical reflection would lead to both the growth, and the ultimate downfall, of Terra. Both individual and institutional investors seemed attracted to the commitment and intensity of its founder – it was certainly easier to understand than the substance of his system. But it turns out that intense commitment to defending a terrible idea only makes the pain worse when that idea inevitably fails.
Still, these interactions highlight a major quandary at the heart of the Terra story: What exactly did Do Kwon do wrong?
You see, the algorithmic system for maintaining TerraUSD’s dollar peg was deeply flawed, but those flaws weren’t hidden. Critics were able to highlight them, well before the system grew to its eventual huge size, because they were described more or less accurately in public documents and statements.
The problem, at least at first glance, wasn’t deception: the problem was that even many supposed professional investors seemed to ignore financial common sense – and the gaping flaws in Do Kwon’s plan.
Even after Terra’s collapse, Kwon himself has repeatedly insisted that he still believes in the fundamental concept of an algorithmic stablecoin. He has repeatedly claimed that TerraUSD’s depegging wasn’t caused by a flawed design, but by a malicious attack from a few large antagonists.
Kwon and his allies spent billions of dollars worth of Bitcoin to prop up TerraUSD’s price during the May 9th to May 11th collapse. This was Do Kwon quite literally putting his money where his mouth was.
So Do Kwon may have truly believed in his bad idea, burning mountains of investor money in a misguided but sincere effort to change the world. Yet South Korean prosecutors issued a warrant for Kwon’s arrest on September 14th of 2022, on charges related to South Korea’s financial fraud laws.
Based on his own statements, one might think Kwon would be happy to defend himself from those charges. Kwon had told the Wall Street Journal in June of 2022 that “I made confident bets and made confident statements on behalf of TerraUSD because I believed in its resilience and its value proposition … There is a difference,” he continued, “between failing and running a fraud.”
This defense has become increasingly common both in cryptocurrency scams and in mainstream corporate frauds. Trevor Milton, founder of a purported electric truck company called Nikola, would claim incompetence rather than malfeasance when it was discovered that Nikola had faked demonstrations of its flagship product. In a whirlwind media tour in the days before his arrest, FTX founder Sam Bankman-Fried would say he “did not do a good job” and “made a lot of mistakes” – an attempt to downplay what the U.S. Department of Justice would later argue was a massive and premeditated crime.
But Do Kwon’s story most closely mirrors that of another infamous figure: Elizabeth Holmes. Holmes founded the ambitious biotech startup Theranos. Like Kwon, she started with a dream and didn’t worry too much about whether it was actually achievable. When biotech experts told Holmes the technology she wanted to create was a far-future fantasy, she, like Kwon, denigrated and dismissed them.
For Holmes, the dream was a miniaturized, automated blood testing machine. When she couldn’t make it work with billions of investor dollars, she could have shut the company down and chalked it up to a failed experiment. Instead, she faked test results for partner companies – and wound up sentenced to more than a decade in prison.
For Do Kwon, the dream was a decentralized dollar, backed by nothing but the forces of the free market. Just as with Elizabeth Holmes, seasoned experts told him it could never work. But Do Kwon, like Holmes, was very good at ignoring what he didn’t want to hear.
There was at least one strong sign that Do Kwon did more than make an honest mistake. Elizabeth Holmes and even Sam Bankman-Fried voluntarily turned themselves in when their arrests were ordered.
But in September of 2022, when a warrant was issued for his arrest, Do Kwon did not behave like an innocent man. Instead, he ran away.
He would live on the run for months before finally winding up, most recently, in Serbia – a country that does not have a bilateral extradition treaty with South Korea.
This international flight leads us to the most obviously criminal accusations leveled against Do Kwon. Kwon has said the collapse of Terra left him nearly broke – but if that’s true, how was he moving around the world without being spotted?
Financial investigators claim they’ve found the answer: transactions and paper trails that they allege show Do Kwon and other members of the Terra team extracted large sums of money from the system before, and even during, its collapse.
There is also evidence that Do Kwon misrepresented specific elements of Terra’s operations – including his involvement in a previous failed attempt to build an algorithmic stablecoin. On a deeper level, some argue that Terra’s structure – particularly the inflated rewards offered to depositors in the bank-like Anchor protocol – made it into little more than an elaborate Ponzi scheme.
So Terra and its magical dollar stablecoin may have started life as nothing more than a terrible idea. But at some point, it seems, it became something much worse – a fraud, on a massive scale.
How did Do Kwon craft this elaborate illusion of stability? Did finance professionals really fall for it – or were they effectively in on the confidence game? And who escaped with the last few drops of real money as the system died?
All that and more, when we return with Episode 2 of Lunacy: The Rise and Fall of Do Kwon.
Coming soon … on Crypto Crooks.