Michael Lewis has a rare gift for taking something complex in the world and writing a great book about it. In this particular case, he also happened to be in the right place at the right time, capturing the rise of Sam Bankman-Fried (SBF) and friends, just before they got all consumed up in managing their downfall.
The story starts with SBF attending an exclusive high school — Crystal Springs — in the San Francisco Bay area and being a gifted awkward student that by itself, would have been commonplace in Silicon Valley. He then went to MIT and graduated with a degree in Physics and a minor in Mathematics. His formative years in the workplace were at Jane Street Capital, one of the top global trading firms credited with trading trillions of dollars of security annually. The first few chapters in the book describe the quirky hiring practices and workplace culture at Jane Street that attracted the brightest young graduates from the top schools in the US. They paid them handsomely as well. One of their main tenets was evaluating the Expected Value (EV) of trades and decision making was often reduced to maximizing the EV around trades, positions, etc.
From Michael’s description it is clear that SBF and friends (Caroline Ellison, Nishad Singh, Gary Wang, Ramnik Arora, …) tried to use EV as a general framework for all sorts of decisions that they were faced with. It’s amusing to read how random probabilities were thrown around and used to compute rough Expected Values to conveniently arrive at decisions.
The whole story can be broken down into a few key points
1. Sam was heavily influenced by the Effective Altruism Movement and had William MacAskill retained as an unpaid advisor.
2. He got tired of working at Jane Street Capital and figured out that he could make a lot of money in the arbitrage of of cryptocurrency prices in different countries. He started a company in Berkeley California, where he recruited more than a dozen young effective altruitsts, and named it Alameda Research.
3. They were making money hand over fist and SBF started getting noticed at crypto conferences in Macau and then Hong Kong.
4. Being one of the largest traders in the crypto sphere, he realized that the existing crypto exchanges were really crappy in terms of their user interface, glitchy and had little to no customer support. Along with Gary Wang, he decided to build a much better crypto derivative exchange and FTX was born in 2019.
5. Some of the key innovations with FTX was a system where margin accounts were liquidated within 30 seconds of their going negative. This greatly limited FTX from holding the bag in case one side of a trade went insolvent and protected the other customers from being hurt in the process.
6. Alameda Research was now a sister company that also had a trading account of FTX. It typically was the market maker on the exchange and was the last resort for liquidity, buying up the contents of insolvent margin accounts whenever needed.
7. To ensure that Alameda was always around, FTX had a special “do_not_liquidate” flag for the Alameda Research account that allowed it to go negative in terms of its balance on FTX. This little loop-hole, turned into a $65B line of credit that ended up sucking out customer money from FTX.
8. Alameda Research siphoned out ~$10B worth of customer funds and that is what ulitmately triggered the run on FTX and its ensuing bankruptcy.
9. One of the big mysteries that is left open in the book is where did this money go? Other than fraud, there doesn’t seem to be any glaring trading loss from Alameda Research, etc. that can explain this vast hole in the bank account.
In addition to telling the above story there’s a lot of detail of the many characters at the top of FTX and Alameda Research. The strange romance between SBF and Caroline Ellison is also explained with little fanfare by Michael. It is made clear to the reader that Sam’s quirky behavior is consistently plastered over all aspects of his life.
While Sam was clearly a poor manager — he abhorred one-on-ones and simply didn’t like talking to people, unless he was playing a video game on the side — it turns out that he wrote a bunch of memos to his team that demonstrated that he not only thought through managerial things, but also was able to articulate it well. Typically he did this by writing a memo with a few succinct bullet points. Here’s an example of a memo he wrote on elevated titles in the workplace:
1. Having a title makes people feel less willing to take advice from those without titles.
2. Having a title makes people less likely to put in the effort to learn how to do well at the base-level jobs of people they're managing. They end up trying to manage people whose jobs they couldn't do, and that always goes poorly.
3. Having titles can create significant conflicts between your ego and the company.
4. Having titles can piss off colleagues.
The book also has some pithy observation credited to Sam like “People like you more if you agree with them”. He apparently embraced this aphorism personally and used to respond with a “Yup” as a standard response to most attempts at conversation.
It’s too bad that Michael published his book before the US court proceedings. I am sure, he would have got a lot more insight into FTX/Alameda and would have a more detailed explanation of the many acounting mis-deeds that brought the company down.
One glaring absence from the book is Gary Wang. He is the clear architect and developer of the FTX software, but Michael could not get even a couple of words out of Gary. If you want the technical details of how Alameda Research was able to draw down billions of dollars from FTX, check out this Verge article which describes Gary Wang’s testimony in court. For a more in-depth review of the fradulent code, Molly White includes the damning evidence from the code.
All quibbles aside, this is a great piece of writing that is quite entertaining and sheds light on the characters that made and broke FTX and Alameda Research. It makes for a quick and fascinating read.