Deep Behavioral Lessons From The Saga Of Sam Bankman-Fried
By Hersh Shefrin
The focus on fraud in the trial of has overshadowed many important psychological issues about markets and financial management. There are lessons to be learned, which extend well beyond fraud, from the saga of SBF and the firms he founded.
Currently, crypto markets are part of the real economy in much the same way as casinos and lotteries are part of the real economy. They are entities which use real resources to provide people with opportunities for satisfying the psychological need to make large, speculative bets. In addition to needs, human psychology features vulnerabilities and pitfalls such as gambling addiction, fear of missing out, biased assessments of risk, and misplaced trust. Investor pitfalls were a key part of the events which transpired at SBF’s firms, crypto exchange FTX and hedge fund Alameda Research.
The managers who run cryptocurrency firms are vulnerable to the same psychological pitfalls as traders and investors. These pitfalls can lead managers to make serious errors in the way they run their firms, and this is what happened at FTX and Alameda Research.
A key lesson from the saga of SBF is that the combination of investor pitfalls and manager pitfalls produced a crypto disaster.
Fast and Slow Thinking
SBF’s psychological profile lies at the heart of his meteoric rise and swift fall. Economists employ a framework in which they assume that people are very skilled at making probability judgments about risks, evaluating potential consequences, scoring consequences using a scale, and making decisions to maximize “expected utility.”
Very few people actually make decisions this way. But SBF was an exception.
Most people make decisions by relying on a combination of feeling and thoughts, what psychologist Daniel Kahneman calls fast and slow thinking. Most people heavily depend on their emotions to help them feel what constitutes a good decision and what does not. SBF was an exception.
SBF’s emotional system was seemingly deficient relative to the norm, especially in regard to feeling empathy. The academic literature documents that most entrepreneurs are high in agreeableness, meaning they with other people. SBF was an exception, being more of a loner.
Emotions are especially important when it comes to following rules and being disciplined. People who habitually follow rules do so by habit, which is emotionally regulated. Most drivers automatically stop at intersections when the traffic light is red. Their first impulse is not to make an expected utility computation to decide whether or not to go through a red light. We rely on our emotions for self-regulation and self-control.
SBF has a tendency to act as if rules do not apply to him. After his indictment, he posted bail in order to be able to live with his parents. Bail came with rules, which SBF breached. As a result, in August 2023, the judge overseeing his case his bail, and sent him to jail, where he resided in poor conditions until his trial. In a related vein, financial institutions face rules about , which was the basis for the charges he faced about fraud.
Gray Area Between Black and White Portrayals of Intent
Those who followed SBF’s trial about his having committed fraud heard two versions of the saga. One was the prosecution’s version of SBF as a fraudulent con artist, and the other was the defense’s version as a nerdy CEO who unintentionally made serious mistakes.
Between the black and the white lies gray, and in his new book, , Michael Lewis documents many of the nuances underlying the rise and fall of SBF, Alameda Research and FTX. There is much to learn from Lewis’ account, beginning but not ending with fraud.
Lewis makes clear that more than $10 billion of FTX customers’ money found its way into SBF’s private trading funds at Alameda Research, and this was clearly illegal. Lewis writes at length about trying to ascertain if SBF himself transferred the funds, and states that the closest SBF came to admitting that he was responsible was in not objecting to a statement made to him that he had done so. Cognitive dissonance is a distinct possibility here.
Given Lewis’ account of events, it seems reasonable to attach a high probability to SBF having transferred the funds to his private trading funds. Relatedly, Caroline Ellison, who had been Alameda Research’s CEO, at the trial that SBF had instructed her to borrow customer deposits in order to fund major financial transactions.
Behavioral Corporate Finance
Behavioral finance, the application of psychology to financial decision-making and financial markets, focuses both on investors and on corporate managers. SBF was both an investor and a manager. Although he demonstrated that he had the skills to be a successful investor, he was a very poor manager.
FTX had no real board of directors, no Chief Financial Officer and no Chief Risk Officer. Not wanting a board of directors is an extreme example of the psychological trait known as desire for control. This trait in combination with high overconfidence is extremely dangerous. It is not that overconfident people lack intelligence. They can be very smart, not just as smart as they judge themselves to be.
SBF contended that he could do the job of CFO, especially tracking money and making financial projections. Given the way things turned out at FTX, SBF was overconfident about his ability to perform the CFO’s function. For instance, a CFO would have understood the implications of having a sizeable portion of Alameda’s balance sheet consisting of the digital token FTT, created by FTX, and similar to cryptocurrencies like bitcoin.
FTT entitled its holders to a claim on FTX revenues, and in this respect was similar to equity. Alameda acquired the FTT by repurchasing it in the market. However, when firms repurchase equity, the treasury stock does not sit on the balance sheet as an asset. In the end, the FTT position was highlighted in an article published by CoinDesk. This article was the catalyst for the subsequent run on FTX. With nobody having true CFO skills at FTX and Alameda, the leadership failed to see the threat.
SBF was certainly overconfident about not needing a CRO, and eventually admitted that this was a big mistake. His FTX associates viewed the firm’s major risk as SBF being kidnapped. But they seriously underestimated the liquidity risk that eventually took FTX into a bank run followed by bankruptcy. An effective CRO would have focused on liquidity risk long before the liquidity event.
SBF ran FTX so that he and he alone understood the big picture. Everyone else could only see a small piece of the puzzle. In respect to business continuity, this was a big mistake, and an example of excessive optimism bias. If SBF were kidnapped, or run over by a bus, nobody else would know how all the pieces fit together. This became apparent when an interim-CEO took over after FTX declared bankruptcy.
Most of the time, venture capitalists who invest in a company accept preferred stock in a company they fund and provide adult supervision in the form of a board of directors, an experienced C-suite, and guidance. They also understand when it is time for the founding CEO to step aside and be replaced by a CEO who knows how to manage growth. For a variety of reasons, this did not happen at FTX.
In the end, what got SBF convicted of fraud was his psychology, especially weak mental accounting. Conceptually, mental accounting is about boundaries. Mentally, SBF could not keep FTX and Alameda Research in separate mental compartments. Michael Lewis tells us that SBF called this an accounting error. If so, it was fundamentally a mental accounting error. SBF managed both firms simultaneously when he operated out of Hong Kong, and never separated them properly when he moved his operation to the Bahamas. This, perhaps, was the major cost of not having a CFO. A proper CFO does not just track money and engage in planning, but puts structures and policies in place.
SBF had an enormous appetite for risk. Part of this might have been high ambition and the strong need to feel like a winner. When crypto prices were soaring, FTX thrived with profit margins between 40% and 50%. Yet SBF aspired for more.
Another part of SBF’s risk appetite might have stemmed from an addiction to gaming. He routinely played video games at the same time he was conducting conversations and making important decisions. He also admitted to feeling sad most of the time, which suggests that he experienced difficulty generating dopamine flows. Problems regulating dopamine surface in people with addictions.
In the end, none of SBF’s psychological traits and poor practices were destined to bring down the company. Instead, it was investor psychology that induced a fatal run on FTX, based on rumors that the brokerage firm was financially fragile.
There is no evidence to support the claim that FTX was actually insolvent, just a rumor that began with the CoinDesk article, and was amplified by Changpeng Zhao, the CEO of SBF’s rival, Binance. Indeed, nobody at FTX or Alameda Research took the rumors seriously, until forced to do so. Their initial reaction was that FTX was on a sound footing, and that investors were rational enough to recognize that this was so. SBF and his fellow officers learned the hard way about excessive rationality assumption bias. And by that time, it was too late.
Crypto markets play a role in the real economy. For example, economic instability and inflation in Argentina is inducing many of its citizens to store their wealth using crypto assets in place of local currency. Nevertheless, globally, crypto market values mostly reflect sentiment rather than economic fundamentals. In the main, crypto markets enable investors to bet on changes in sentiment. Economist John Maynard Keynes described this state of affairs as a beauty contest in which judges bet on which contestant other judges will find most attractive. That feature makes for volatile dynamics, and that is certainly the case with crypto markets.
An additional contributor to volatile dynamics is the testosterone effect. Most of FTX’s customers were young males. This demographic feature is the reason SBF contracted with major sports figures such as Tom Brady and Stephen Curry, to promote FTX. There is good reason to believe that testosterone also impacted the rivalry between SBF and Binance’s CZ.
SBF’s conviction on seven counts of fraud is highly salient. Less salient but perhaps more important are the behavioral lessons to be learned about the managers of crypto firms and the investors who are their customers. Having both investors and managers fail to address their psychological vulnerabilities creates a recipe for disaster.
This column originally appeared in the Nov. 5, 2023 online edition of Forbes.
Nov 6, 2023
Photo of Sam Bankman-Fried courtesy of Shutterstock.