Was Bankman-Fried Wrongly Convicted?
Sam Bankman-Fried’s widely anticipated guilty verdict has been rendered. But Michael Lewis’ book raises many questions, including the existence and extent of an actual crime.
A central and recurring theme of Michael Lewis’s new book, Going Infinite: The Rise and Fall of a New Tycoon, about the cryptocurrency exchange FTX and its sister hedge fund Alameda Research, and its founder and CEO Sam Bankman-Fried – apart from its exploration of Bankman-Fried himself – is a riddle: Where did the money go? By the end of the book, Lewis thinks he has the answer. But like lifting a rug to see where a bug came from and seeing dozens more stream out, solving that riddle merely exposes the festering infestation.
Most reviews of Lewis’s book focus on the character of Bankman-Fried. Some of them think that Lewis was snookered to believe that Bankman-Fried was less of a crook than he is and to defend him. But that’s not what I see in the book. Lewis simply delivers a straight story without layering in moral judgment. And that straight story makes it clear that Bankman-Fried cut corners and did things that are fraud to most of us, as confirmed by his guilty verdict last week.
But that’s not what I want to focus on. I read the book as if it were a kind of whodunnit – or rather, a what-dunnit. I knew that Bankman-Fried’s cryptocurrency exchange, FTX, would go bankrupt in the end. But how exactly?
I was assuming before and as I read it that the denouement would be similar to so many other investment boom-bust stories, for example Long-Term Capital Management, the 1990s hedge fund run by and staffed by financial “masters of the universe” including two Nobel Prize winners in economics. In that case – like so many others – the managers of the investments thought they had the key to winning, achieving “alpha.” And they had in fact achieved alpha, in the case of LTCM for four strong years in a row. Then … boom! (or rather, bust!) … suddenly they didn’t win anymore. In fact, in a stroke they lost in value everything they had gained and more. They were heavily leveraged so the requirement to pay down their debts to the banks that loaned to them bankrupted them – or would have if the banks hadn’t been strongarmed by the New York Fed to bail the fund out. (In the next year, the banks realized a 10% return on their investment, according to one of the LTCM partners, Victor Haghani.)
I don’t remember if I knew the basic facts even before reading Lewis’s book or learned them near the beginning of the book. Sam Bankman-Fried, the son of two Stanford law professors and a graduate of the Massachusetts Institute of Technology, had set up a hedge fund to arbitrage cryptocurrencies called Alameda Research. Then he set up a cryptocurrency exchange called FTX. But funds owned by customers of FTX that were held on FTX were allowed to leak over to Alameda Research, where they could be invested in the risky way a hedge fund invests – which means that the customer funds that were deposited on FTX for safekeeping (but let’s talk about that again later) would no longer be safe at all. This is basically the crime for which Bankman-Fried was found guilty (and to which several of his FTX and Alameda underlings have already pleaded guilty).
(Just to be clear: a hedge fund is, well, a hedge fund, that is, an investment fund. A cryptocurrency exchange is a company that brokers exchanges of fiat currencies for cryptocurrencies and cryptocurrencies for other cryptocurrencies, taking a fee for the service.)
And then what happened? … and this is where I assumed the story would be like that of LTCM and so many others. Alameda Research was run by yet another set of “masters of the universe,” who would of course be geniuses at investing, and who had exhibited that genius both at Alameda Research and previously at high-frequency trading firms – in the cases of both Bankman-Fried and his compatriot, business partner, sometime personal partner, and Alameda Research CEO Caroline Ellison, herself a Stanford graduate and the daughter of MIT professors and lecturers (elite universities are presumably suffering from acute embarrassment, yet again; see Elizabeth Holmes), at the high-frequency trading firm Jane Street Capital.
I assumed it would be the same story. Alameda Research, run by such investing experts, would suddenly see its investments fall off a cliff, taking the customer funds from FTX with them.
But that may not be what happened. Instead, it appears to be essentially a bank-run story. Alameda and FTX might have been solvent if it weren’t for the precipitous run on it in the first week of November 2022. This is what Bankman-Fried claims. By itself that wouldn’t absolve him of fraud. Nevertheless, if it hadn’t been for that bank run, Bankman-Fried might still be a revered, if quirky (or rather, because he is quirky) international celebrity and multibillionaire.
The paragraph in Lewis’s book that runs counter to all the other narratives
The standard explanation of FTX’s collapse and bankruptcy is simple.
FTX was a cryptocurrency exchange – to be precise, a centralized cryptocurrency exchange, or CEX1. Hence, when customers sent currencies (fiat or crypto) to it to exchange for other currencies, it held those currencies in its own account – i.e., for crypto, at its own crypto address – rather than the customer’s. Customers sent funds denominated in either a fiat currency (such as U.S. dollars, Hong Kong dollars, euros, yen, etc.) to FTX, or in a cryptocurrency (such as bitcoin, ether, etc.), along with an order to exchange them for another currency, either immediately (a spot trade) or at a specified time in the future (a futures or options contract). If the funds that customers sent to FTX were collateral for a derivatives contract such as a futures contract on bitcoin, then the amount deposited would not be enough to pay for the future transaction; hence, FTX was in effect loaning the customer money borrowed from other customers in anticipation of the future transaction. This is standard practice. Thus, Bankman-Fried’s argument that it was normal to lend customers’ money to other customers – and therefore it was not necessarily abnormal to lend customers’ money to Alameda Research.
Whether or not you buy that argument, FTX did lend its customers’ money to Alameda Research, which then used that money to make investments.
The crypto company Coinbase revealed on November 2, 2022, that as much as a third or more of Alameda’s $15 billion of assets were in the crypto token FTT, a token issued by FTX. Coinbase said, “The situation adds to evidence that the ties between FTX and Alameda are unusually close.”
This, together with the cashing in of a massive amount of FTTs by Changpeng Zhao, the CEO of rival exchange Binance, triggered a run on FTX. In the first week of November, billions of customer money were withdrawn. At some point, neither FTX nor Alameda could honor the requests to withdraw funds from FTX.
The shortfall is generally cited in the financial media as $8 billion, the “$8 billion hole.”
John Ray’s revelation
On November 12, Bankman-Fried signed FTX into bankruptcy. A man named John Ray, who had previously handled the Enron bankruptcy in the early 2000s, became CEO with a mission to resolve FTX’s bankruptcy and compensate its claimants, insofar as possible.
Michael Lewis wondered whether the hole might have been caused by trading losses in Alameda – which is what I assumed, a priori, to have been the case. But this was firmly denied by Ellison, Bankman-Fried, and others, who said Alameda was highly profitable. Lewis accepts this, because Ellison and Bankman-Fried were seasoned investors with backgrounds at a sophisticated trading firm, Jane Street Capital. But Lewis should not have accepted this assurance so readily, but perhaps it is true that the cause was not trading losses in Alameda.
Lewis also considers the possibility that the money was lost because of hacks. There had, in fact, been, hundreds of millions of dollars lost to hacks on FTX, but that was, believe it or not, relatively “small potatoes.” FTX was absurdly profitable due to the cryptocurrency craze that drove a million traders to gamble untold amounts on the FTX casino.
But Lewis says that in a report filed by Ray at the end of June 2023, Ray claims, “To date, the Debtors have recovered approximately $7 billion in liquid assets, and they anticipate additional recoveries.” Lewis says that an investor who was hoping to bid for the remaining portfolio said it should go for at least $2 billion.
That comes to $9 billion, more than enough to cover the “$8 billion hole.” And it doesn’t include other assets that are more difficult to value. Hence, none of FTX’s claimants should lose.
This revelation makes me think that Bankman-Fried was right. FTX was solvent but couldn’t liquidate assets fast enough to deal with the run on it in early November 2023.
But it also leaves many questions that are unanswered by Lewis’s book.
If Bankman-Fried is right, why have so many of his co-workers pleaded guilty to fraud and made heart-rending confessions? If they’re so sure it was fraud, who was defrauded? If it really was just a bank run, you wouldn’t accuse the bank of fraud (aside from the fact that Bankman-Fried, Ellison, et al were clearly less than honest about it).
What exactly were those $9 billion of assets that couldn’t be liquidated quickly enough?
Bankman-Fried says that at one point, Alameda was worth as much as $100 billion. If that is true, why was it only worth about $16 billion at time of liquidation? Were there, after all, massive losses in the fund, in spite of his and Ellison’s assurances to the contrary?
FTX wouldn’t allow U.S. citizens to open accounts because of strict U.S. regulations. Therefore, apart from account holders who, in one way or another, faked otherwise, its claimants were not U.S. citizens. Lewis strongly implies that Bankman-Fried was convinced while in custody in the Bahamas to surrender to the U.S. instead, by partisans of FTX’s law firm advisor Sullivan & Cromwell, because it would bring hundreds of millions of dollars in fees to the firm. But even so, what was the legal reason for trying the case in the U.S.?
What legally binding assurances did FTX give to customers that their funds would not be loaned to Alameda Research? Why are we so sure it broke trust with its customers and committed fraud? After all, in an ordinary bank, customer funds are loaned out all the time and it is expected. While a conventional financial exchange might be indicted if it loaned out customer funds, would that hold for a cryptocurrency exchange whose business was regulated not by the U.S. for U.S. investors but by foreign jurisdictions with very light or non-existent cryptocurrency regulations?
Many years ago, the financial firm EF Hutton was one of the most recognized and respected brokerages in the world. (Some may remember the ads, “When EF Hutton talks, people listen.”) Then suddenly, it was bankrupt.
I never understood why until I read the distinguished Boston University emerita professor Tamar Frankel’s case study, EF Hutton’s Slide into Oblivion, which made it all crystal clear.
While I enjoyed reading Lewis’s book, for the real FTX story I’ll wait for the law review article.
Economist and mathematician Michael Edesess is adjunct professor and visiting faculty at the Hong Kong University of Science and Technology. In 2007, he authored a book about the investment services industry titled The Big Investment Lie, published by Berrett-Koehler. His new book, The Three Simple Rules of Investing, co-authored with Kwok L. Tsui, Carol Fabbri and George Peacock, was published by Berrett-Koehler in June 2014.
1 A centralized cryptocurrency exchange (CEX) holds cryptocurrencies and currencies for its clients, like a bank holds its customers’ deposits. Therefore, it must abide by KYC (know your customer) rules. Hence, CEX users cannot trade anonymously. A decentralized currency exchange, by contrast, facilitates a trade but the crypto funds go to the customer’s crypto address, not the exchange’s. Most of the large crypto exchanges, like Binance and Coinbase and (formerly) FTX, are CEXes.
A message from Advisor Perspectives and VettaFi: To learn more about this and other topics, check out our podcasts.