“Banks must be trusted to hold our money and transfer it electronically, but they lend it out in waves of credit bubbles with barely a fraction in reserve.” – Satoshi Nakamoto, circa 2009
“They mistook leverage for genius.” – Steve Eisman, FrontPoint Partners, circa 2008 (portrayed by Steve Carrell in “The Big Short”)
“We are in the process of communicating with relevant parties and fully committed to working this out.” – Zhu Su, Three Arrows Capital, June 15, 2022
Over the past eight weeks a series of seemingly robust pillars of the cryptocurrency industry have fallen. They knocked each other over like dominoes, and as they hit the ground they shattered. Our pillars, it turns out, were made of glass.
A misguided bet nukes the balance sheet of a “hedge fund” that forgot to hedge. Tanking markets trigger margin calls for shadow banks masquerading as crypto lenders, which in turn freeze customer balances. Long-dated illiquid bonds leave nearly everybody in the lurch. Counterparties default on massive loans.
And, just like that, hundreds of billions of dollars’ worth of notional value is gone.
On July 6, FTX exchange founder and CEO Sam Bankman-Fried declared that this catastrophic crypto industry unwind was nearly over. Even if that’s true, the work of understanding what actually happened has barely begun. Thanks to crypto infrastructure, the web of relationships and obligations have been somewhat more visible than would have been the case in traditional markets. But there are still a lot of details to be unearthed.
On the other hand, in broad strokes, we know exactly what happened: leverage, debt and other counterparty relationships based on bullish market bets melted down when those bets didn’t pan out. The cryptocurrency market has dropped by roughly two-thirds, but it was leverage that turned that drop into a death sentence for some companies and many of their customers’ balances.
It's all so tiresome. All so familiar. All so much like the 2008 financial crisis that fueled widespread interest in cryptocurrency in the first place.
The irony is rich and getting richer: Some of the failing entities even got “bailouts” of a sort, in the form of private lines of credit and buyouts. It seems that always and everywhere, capital crises and “Star Wars” prequels are like poetry: they rhyme.
Cryptocurrencies like bitcoin (BTC) were often premised on the elimination or reduction of such leverage from the new financial system. But the Great Crypto Unwind has shown that even a non-counterparty bearer instrument, ideal for preventing such crises, can and will be subverted by financial engineers with warm smiles and treachery in their hearts. However much we think we’ve learned, human frailty still trends towards greed, recklessness and, in the end, chaos.
Read More: Why This Crypto Crash Is Different
There is still much we don’t know about crypto’s Great Unwind. But here’s one attempt at a first draft of history.
Financial Crisis 1.0
First, some context.
The persistent and sometimes frenetic enthusiasm for the Bitcoin project since its creation in 2009 has had a lot to do with timing. Pseudonymous creator Satoshi Nakamoto launched the original cryptocurrency in the wake of the Global Financial Crisis, in which the declining value of a small set of assets caused cascading failures among a rat’s nest of financial institutions tightly linked by shared debt and similar obligations.
Those intertwined obligations sent the entire financial and banking world into a spiral when they went bad, killing or maiming major institutions like Lehman Brothers and AIG, respectively, and leaving others on death’s door. In turn, the resulting credit squeeze sent the global economy into a years-long Ice Age, effectively leaving all of us holding the bag for the big banks’ irresponsible decisions.
The resulting rage was present almost from the earliest days of the Bitcoin project. The original white paper is dry and mostly technical, but pre-launch discussions among developers were much more pointed. By the time the Bitcoin blockchain went live, criticism of banks had become nearly a doctrine: Bitcoin’s first block contained a reference to U.K. banking bailouts.
Thirteen years later, parts of the industry spawned by Satoshi’s brainchild have come to resemble those hated enemies. A wave of collapses has hit centralized “crypto banks” that used customer funds to make leveraged bets. Investment funds exploited the opacity of their centralized balance sheets to hide their risk profiles and harvest loans that have left creditors in the lurch.
Some of those troubled entities, such as the lending platform BlockFi, have even gotten “bailouts” in the form of giant private loans. Those bailouts are mostly better described as buyouts, and don’t present the same immense moral hazard as the taxpayer-funded 2008 bailouts of degenerate gamblers at Citigroup and Bank of America. The entities getting fresh capital this year are the ones private investors (above all, Sam Bankman-Fried’s FTX) see as viable in the long term.
Other entities, including Celsius, Three Arrows Capital and Voyager Digital have been deemed unworthy of new money and are moving towards bankruptcy, collapse or liquidation – and taking a lot of depositor money with them.
Lunatics at the gate
Finance is ultimately a way of turning promises about the future into money. It’s a way to get paid now based on something that might (or might not) happen in two or five or 10 years. This is especially true for crypto, with token and equity values alike often hinging on someone’s huge vision of a far-distant technological revolution.
Few visions of the future have faded as quickly or ignominiously as that of Do Kwon, founder of the Luna ecosystem and its TerraUSD algorithmic “stablecoin.” TerraUSD was ultimately revealed to be, in essence, a Ponzi scheme, propped up by unsustainable interest paid on deposits in a pseudo-bank called Anchor.
In the narrative of the Great Unwind, LUNA is the most dramatic culprit on the asset and technology side. Other assets that weakened crypto institutions caused liquidity problems, but LUNA/terraUSD (UST) were the only crypto assets that simply collapsed to effectively zero thanks to bad tokenomic design. It’s quite impressive – you really have to work to send a crypto token to zero. Even the native token of the largely defunct 2017 ICO-era chain EOS, once worth $4.4 billion, is now worth more than LUNA.
Like most such scams, the LUNA Ponzi only worked as long as markets trended upwards. When those markets began to reverse at the end of 2021, the writing on the wall was clear enough that many early, big-money shills dumped their bags. When LUNA fully unwound in May, it destroyed $68 billion worth of paper value in a matter of days – one of the most dizzying single wealth-destroying events in peacetime history.
That collapse would have been bad enough in isolation. The stories of individual users losing their life savings in Anchor are bleak. But some high-profile and supposedly smart money managers also failed this financial IQ test – and that’s where the real trouble began.
LUNA, GBTC and stETH: The three horsemen of the crypto-pocalypse
If the LUNA collapse was the spark that set off the Great Unwind, Three Arrows Capital was the dynamite factory where the spark landed. And there was a whole lot of dynamite.
With a reported $18 billion in assets under management at its peak, Three Arrows, also known as 3AC, was a mix of venture fund and hedge fund. It has been alleged it was also secretly operating as a prime brokerage service, letting friendly partners trade on its accounts. 3AC co-founders Su Zhu and Kyle Davies were darlings of the crypto world for years, and the fund was generally believed to be working entirely with its own money – a so-called “prop shop.”
But as its various bets went bad, it became increasingly clear that 3AC was engaged in a variety of serious deceptions. Most brazenly, the fund appears to have borrowed large amounts from a variety of counterparties that were each unaware of the fund’s total debt levels. Now, Three Arrows is in liquidation and Zhu and Davies' whereabouts are described as “unknown” in filings by their own legal team – though Zhu did break a lengthy Twitter silence on the morning of July 12.
Such profligate, opaque borrowing is similar to the moves that led Lehman Brothers to the boneyard in 2008. It also echoes the 1998 collapse of Long Term Capital Management, and the more recent leveraged blowout of Bill Hwang’s Archegos fund. During the 2008 crisis, U.S. Federal Reserve Chair Ben Bernanke reportedly saw Lehman as the most reckless of the big investment banks exposed to mortgage-backed securities and credit default swaps. In accounts of the era, there’s a strong sense that Lehman CEO Dick Fuld was as personally arrogant as he was in his financial dealings.
That’s also a common thread with Three Arrows. Much like Lehman’s real estate position, 3AC’s positions amounted to an immense leveraged long, in this case largely on the price of bitcoin. Zhu Su had articulated a “supercycle” theory that he said would send bitcoin to $100,000 or higher. So when the broader market began to slump in November of 2021, Three Arrows was already a dead man walking. The following flood of margin calls and redemption notices simply exposed its failures.
And 3AC had made even higher-risk bets than leveraged BTC longs: You guessed it, it was waist-deep in LUNA. Three Arrows reportedly invested $200 million in LUNA in February. That stake grew to as much as $600 million on paper as LUNA enjoyed a huge run-up. It is now worth effectively zero.
In a bit of striking poetry, though, LUNA was just one of three mortal wounds that took out 3AC. The other two were not frauds, but illiquid bonds the fund couldn’t sell when it needed cash to cover its loans: stETH (Lido staked ETH) and GBTC, the Grayscale Bitcoin Trust. (Grayscale and CoinDesk have the same parent company, Digital Currency Group.)
Back in January 2021, 3AC announced a $1.2 billion position in GBTC. GBTC, essentially a kludgy legal workaround for institutions that want BTC exposure, cannot be easily swapped for its underlying bitcoin. This isn’t much of a problem in bullish periods – when 3AC bought in, GBTC was trading at a premium of nearly 20% above bitcoin, arguably because it had few competitors for institutional access.
But almost as soon as Three Arrows bought in, GBTC holders started taking profits off that premium, and competitors arrived in the form of Canadian bitcoin ETFs. By August 2021, the trust had flipped from trading at a premium to trading at a discount of roughly 15%. This meant Three Arrows was deeply underwater on its massive GBTC position.
The GBTC discount has now widened to more than 30%, a potentially insane buying opportunity for long-term BTC bulls with cash on hand. But Castle Island’s Matt Walsh has speculated that the GBTC disconnect was the first shoe to drop for Three Arrows, which found itself locked into a degraded GBTC position and took bigger and bigger risks – including in LUNA – as it desperately tried to fill the hole.
Then, following that plague of locusts, came stETH. StETH is a token issued in exchange for ETH staked through the Lido protocol. It will convert to ETH at an uncertain future date when the Ethereum proof-of-stake Merge is complete.
That makes it essentially a tokenized, zero-coupon ETH bond with an unknown maturity. StETH can be bought and sold but it can’t be used on Ethereum, so it’s fairly natural that it trades at a discount against ETH. 3AC’s apparently forced stETH sale came when that discount was ranging at around 10%, a potentially devastating haircut even ignoring ETH’s own swooning price (it has fallen about two-thirds in 2022 so far), compared to returns if 3AC had been liquid enough to wait for the Merge.
Three Arrows to the knee
Between these illiquid bets, its massive screwup on LUNA and the macro-driven decline in bitcoin, 3AC faced an insurmountable wave of margin calls from exchanges where it had traded. It also had large, sometimes unsecured loans.
It now appears that, despite a cryptic tweet from Su Zhu on June 14, 3AC was not in fact “committed to working this out” with its array of creditors. Instead, it has ghosted exchanges and other counterparties. Genesis Trading, another CoinDesk sibling company, was taken for “hundreds of millions of dollars.” Other exchanges were able to liquidate 3AC’s books and made it out more or less intact – BitMEX, for instance, says 3AC still owes it $6 million.
Those victims, it turns out, got off easy.
Lending platform Voyager Digital, in a demonstration of just how trusted Three Arrows was, gave the fund a reported $650 million in unsecured loans – that is, loans without collateral that could be called in if things went south. Voyager declared bankruptcy this week largely because, with Three Arrows now in liquidation, it’s unclear if or when it will get even a portion of those funds back. It seems Voyager’s retail depositors shouldn’t expect all of their money back.
Lenders and exchanges were far from the only entities that got “rugged” by Three Arrows’ bizarre disappearing act. A spokesperson for a protocol 3AC invested in claimed that the fund had also managed the project’s treasury, which is now just as gone as all the rest of the money. Moreover, according to the same source, 3AC enticed projects to deposit with it by offering 8% interest. That high fixed rate suggests this may have been less a generous service than a way for Zhu and Davies to get more chips with which to gamble.
This and other behaviors have led at least some to conclude Three Arrows was not simply a failed hedge fund but an outright scam. Castle Island’s Walsh, on his firm’s “On the Brink” podcast, vociferously condemned Three Arrows’s alleged actions.
“This is not just some bad investing,” said Walsh. “Based on what we’re starting to learn about Three Arrows, this is a massive fraud. This is a Bernie Madoff-style fraud. These are bad people.”
That’s not what we meant by "unbanked"
3AC’s shenanigans will have echoing long-term effects as the lenders and startups who lost funds run out of money entirely. But it took another class of dingus to turn the same speculative screwups into immediate, searing pain for individual, crypto-novice retail investors.
Opaque and centralized “lending platforms” including Celsius and Babel Finance were exposed to 3AC, LUNA, stETH, and other risks, largely because they were chasing outsized returns and taking huge risks to deliver them. Voyager’s $650 million unsecured loan to Three Arrows may have also been yield-chasing, and in retrospect looks like the equivalent of jumping out of a plane without a parachute. Amid all this, another lending platform, BlockFi, managed to only wipe out $1 billion in investor funds. So kudos to them, I guess.
These risks were concealed from depositors, for whom Celsius’s sales pitch – “Unbank Yourself” – turned out to be terrifyingly accurate.
Voyager’s huge unsecured loan to Three Arrows may go down as the single stupidest transaction of the 2022 crypto unwind. But the case of the lending platforms is a good bit more nuanced and interesting.
It's helpful to understand the difference between Celsius and its ilk and “decentralized finance,” or DeFi, a category with which the lenders often misleadingly affiliated themselves. At the highest level, DeFi protocols run entirely on the blockchain while Celsius, BlockFi, Babel and the like were traditional centralized financial entities that would put your money in DeFi and other investments and reap profits on your behalf. They sold themselves as an alternative to traditional banks that still had “DeFi” cool, vastly increasing the eventual damage inflicted. This is why they’re sometimes referred to derogatorily as “CeFi” – centralized wolves in decentralized sheep’s clothing.
Because behind the scenes they were closer to being some kind of fixed-return hedge fund, engaging in constant high-risk gambling with depositors’ money. And, yet again, these hedge funds forgot to hedge.
By contrast, decentralized finance trading, stablecoin and lending protocols including MakerDAO, Compound and Curve have been basically unscathed by the big crash. They’ve lost a lot of users and volume, but they haven’t wound up in massive debt because the protocols themselves can’t really take on leverage. Instead, all loans are overcollateralized – that is, you have to post something like $2,000 worth of ETH to borrow $1,000 worth of something else, and you’ll get automatically liquidated if your collateral drops below or near your borrowing.
These rules can’t be bent. It would have been completely impossible, then, for 3AC to take out a huge unsecured loan from Compound, which is why Compound is still running and Voyager Digital is roadkill. In a grim bit of comedy, Celsius just repaid a $183 million loan from Compound to redeem its collateral, even as withdrawals from Celsius remain locked and Voyager prays to every God in heaven to get back pennies on its insane unbacked loan to 3AC.
So as crazy as it might sound to use a bank or asset exchange with no owners, that very lack of a backstop forced the systems to build with extremely conservative financial assumptions. The humans, it turns out, were always the dangerous ones.
And perhaps no human lender was more dangerous than Celsius CEO Alex Mashinsky, who matched or outdid Do Kwon for bluster, bombast and bad ideas, right up until the day before his operation imploded under his feet.
Celsius was founded on the premise that it would accept retail crypto deposits, then loan those tokens to institutional traders at high rates. For a time, there was enough loan demand to make this work. But eventually Celsius started engaging in much riskier behavior as it searched for high yields for all of its reported $12 billion under management.
The shift to higher-risk strategies was likely thanks to increased deposits in response to its high yield offerings, combined with declines in institutional interest as the crypto market started to cool in late 2021. Celsius started out making mostly collateralized loans to institutional traders, though they misrepresented this, even early on. From there, they eventually progressed to engaging in degen-level yield farming by staking customer funds to every DeFi protocol under the sun.
This was a much higher-risk strategy, and in some cases, as with an estimated $50 million lost in the hack of BadgerDAO, it blew up in Celsius’s face. In another case, Mashinsky got ridiculously lucky. Celsius reportedly began depositing on LUNA’s Anchor system in December, where it would have collected something like 20% nominal yield – more than enough to cover its payouts to depositors, with a bit extra off the top. But Celsius just barely managed to pull out $500 million before that system completely melted down in May. That’s nearly the dictionary definition of “picking up nickels in front of a steamroller” – taking immense, opaque risks for modestly above-average returns. And, in this case, doing it with other people’s money.
Celsius managed to dodge that steamroller. But if you play stupid games, you eventually win stupid prizes. If his representations of Celsius as safe and stable are determined to have been deceptive to depositors, Alex Mashinsky could wind up with the same prize already awarded to another former Celsius officer: a pair of handcuffs.
Though we’ve inevitably left out a few details, that’s the basic outline of the Great Crypto Unwind. But its impacts were felt far beyond those who were directly exposed to entities like Three Arrows, Celsius or Voyager Digital. For much the same reason that a deflating real estate market locked up the entire economy in 2008, problems at a few crypto entities have echoed far and wide.
The easiest way to see this is simply in the price of blue chip crypto assets like bitcoin and ether. Granted, those prices began steadily declining starting back in November 2021 because of macroeconomic conditions, and those pressures have continued in parallel with the impact of the credit unwind.
But if you look at specific dates, it’s not hard to pin steep, sudden price declines on C-Suite frauds and jumped-up degens.
The depegging of UST became publicly obvious between May 9 and May 12 of this year, while BTC dropped a staggering 27% between May 5 and May 11. Though the UST depeg likely didn’t involve bitcoin directly, it would have created a lot of granular BTC sell-pressure for those covering LUNA losses. It also, more obviously, created a kind of confidence contagion, since many new crypto entrants didn’t seem to understand the distinction between LUNA and other assets.
Just over a month later, problems at Three Arrows became public as early as June 14. This coincided with another gigantic macro dump, as BTC cratered a further 26% between June 12 and June 17. Again, this involved some vague loss of confidence, but we know that Three Arrows was in the process of various liquidation strategies, so in this case there was likely some direct sell pressure.
In both cases, you’ll notice that the dumps began shortly before bad news made it to public sites like CoinDesk. That points to a difficult truth not so much about crypto as about financial markets overall – whatever the rules say, insiders often know what’s happening first thanks to whisper networks that mere plebs can never hope to access. Sometimes a dump is just a dump, but other times it’s a hint at very bad news around the corner.
The leverage is too damn high
It’s a bit early to look for many lessons in this mess because there may yet be other shoes to drop – other entities that were overleveraged or made reckless loans to the overleveraged, but have managed to hang on for a few weeks or months without revealing their mistake.
One important insight for those trying to dodge falling shoes: Companies that were late entrants to the crypto game were far more likely to be making leveraged bets as they tried to catch up to more established players. One reason Alameda Research is in a position to rescue flailing partners is that it was founded in 2017 and has apparently relied mostly on organic growth.
Longer term, I recommend filing away the names of individual culprits and holding on to the list for future reference. Character and judgment obviously count when people are asking to manage your money, and figures like Do Kwon, Alex Mashinsky, Su Zhu and Kyle Davies clearly came up short on one or both of those measures. They deserve to be banished from the finance industry for the rest of their lives. At least in some cases, that will happen – I’d bet it’s tough to run a shadow bank from prison.
But the more important questions are systemic. Obviously, there was way too much borrowing going on, and widely trusted individuals seem to have lied about their debt. That’s theoretically a regulatory issue. But even if it were universally desired, a global crypto regulatory regime doesn’t seem very plausible right now.
There’s a lot more potential in a social solution that doubles down on crypto’s original vision of self-custody. Remember, depositors in Celsius actually sent their bitcoin to a guy on the internet, and that’s why they don’t have it now. By a similar token (ahem), a percentage of Celsius users got rooked by collecting their “yield” in a corporate s**tcoin, CEL.
Many retail victims of these cons and failures will be turned off crypto for a long time, if not permanently. That’s a harm to the vision we came here to pursue, and it will take principled operators to rebuild trust overall. On the other hand, for many people, there’s no education quite like getting ripped off. We have to hope they’ll spend some time thinking about where exactly they went wrong, and help educate others about what crypto really has to offer.
That’s been a big part of the aftermath of previous crypto cycles. But the truth is, thanks to leverage and a few centralized players, this one likely hurts far worse than even the brutal 2018-2019 Crypto Winter. At least when your garbage 2017 ICO became valueless, you got to keep the tokens and vaguely think, “well, maybe someday!”
But thanks to malign and incompetent figures like Alex Mashinsky and Do Kwon, a lot of potential crypto converts have been left with nothing at all.
The leader in news and information on cryptocurrency, digital assets and the future of money, CoinDesk is a media outlet that strives for the highest journalistic standards and abides by a strict set of editorial policies. CoinDesk is an independent operating subsidiary of Digital Currency Group, which invests in cryptocurrencies and blockchain startups. As part of their compensation, certain CoinDesk employees, including editorial employees, may receive exposure to DCG equity in the form of stock appreciation rights, which vest over a multi-year period. CoinDesk journalists are not allowed to purchase stock outright in DCG.