Lewis traces how Bankman-Fried treated trust as an expected-value bet

1

Flat affect, single equation

SBF's near-flat emotional register and utilitarian upbringing made him willing to treat trust, money, and relationships as interchangeable units in one expected-value equation.

Lewis traces the childhood to a household of Stanford law professors whose dinner talk ran on utilitarian logic, not sentiment. Bankman-Fried described his own feelings as close to flat, at times in the language of anhedonia. Most people check choices against guilt, loyalty, or comfort. He checked them against calculated outcomes alone.

As a teenager he preferred puzzles and games with clean payoffs over social approval. At MIT he studied physics and found the constant social performance of college exhausting and largely pointless. He kept a small circle of peers who spoke in odds and utility rather than right and wrong.

That wiring shaped the rest of the arc. Every later decision, from founding a firm to spending deposits, traces back to a mind trained to ask only whether a move maximized expected value. Instincts that stop most people simply did not fire.

In practice: He treated as straightforwardly correct a coin-flip thought experiment that could destroy the world if it also had a slightly better-than-even chance of doubling it.

2

Earning to give

Effective altruism gave him a moral frame in which becoming as rich as possible, by any legal means, counted as the most ethical career available.

The pivot, as Lewis tells it, was a talk by Oxford philosopher William MacAskill that Bankman-Fried heard as an MIT undergraduate. MacAskill held that a person could do more good in a high-paying job and donating the proceeds than by working directly in a nonprofit, because dollars convert efficiently into saved lives or reduced existential risk.

Earning to give reframed the pursuit of wealth as duty rather than indulgence. Bankman-Fried calculated how many lives a dollar at a Wall Street desk could save compared with a dollar earned anywhere else. He concluded finance was the highest-leverage path and chose Jane Street over academia or direct charity.

Lewis notes the ideology was sincere; friends from the effective altruism community back that up. Yet the same frame that justified extreme ambition also supplied a ready moral alibi for almost any means, since the ends were defined as saving the world. Once profit and virtue fused, internal guardrails against cutting corners with other people's money were already gone.

In practice: His post-graduation choice of Jane Street Capital, made by comparing lives saved per dollar across possible careers.

3

Jane Street's edge

Trading-floor training taught him to treat every decision, including loyalty and institutional trust, as a bet with a calculable payoff.

Between 2013 and 2017 he worked at Jane Street Capital, a quantitative firm known for hiring physicists and mathematicians to price risk instantly. The culture ran on games of expected value. Traders learned to accept any bet with a positive mathematical edge and to distrust intuitions that contradicted the math.

He learned to make ETF markets off tiny price gaps across exchanges. The core lesson stuck: emotion is noise, the number is the truth. Lewis shows this produced a genuinely skilled trader. It also produced someone who came to see structure, hierarchy, and disclosure as negotiable variables rather than fixed obligations.

Frustration grew with rules he considered suboptimal rather than necessary. He felt the firm's caution left expected value on the table. That push is what sent him out in 2017 to build something with no one else's risk limits left to defer to. The same indifference to sentiment that works on a trading desk becomes dangerous once the trader controls other people's deposits.

In practice: His rising irritation at Jane Street risk limits he read as leaving edge unclaimed rather than protecting the firm.

4

Alameda's early fracture

Shared ideology could not hold the trading firm once his secrecy and risk tolerance collided with colleagues who wanted basic transparency.

In late 2017 he founded Alameda Research with a group drawn largely from the effective altruism community, including early leader Tara Mac Aulay. The founding trade was the kimchi premium—the gap between bitcoin prices in Korea and elsewhere—exploited by moving capital across borders faster than rivals. Early funding came partly from loans inside the EA network, people who trusted the philosophy enough to trust him with money.

Lewis describes 2018 as the year that trust broke. Colleagues grew alarmed at sloppy accounting, unclear position sizing, and his habit of withholding information even from his own risk managers. A near-collapse followed, with large losses through positions no one outside SBF fully understood.

Several founding employees left en masse. The rupture fractured relationships inside the effective altruism movement itself, because many had joined on shared moral commitments rather than a contract. Lewis treats the episode as preview, not aberration: concentrated opaque control, defended as necessary for speed and edge, would return at far larger scale.

In practice: The 2018 mass exit after positions only SBF fully understood nearly sank the firm and split EA friendships.

5

Two names, one balance sheet

FTX's rapid rise came less from technology than from backing the exchange with Alameda capital and quietly erasing the line an exchange and a trading firm should never cross.

After stabilizing Alameda, he identified a concrete inefficiency: existing crypto exchanges offered clumsy margin and derivatives products, and market makers were losing money to bad execution rather than bad predictions. In 2019 he and Gary Wang founded FTX, first based in Hong Kong, to build the exchange he wanted as a trader.

Early growth came from better product design, tighter spreads, and faster liquidations that attracted retail traders and professional firms. Lewis frames a structural decision as the real hinge. Alameda, still owned and effectively run behind the scenes by SBF, received privileged access to FTX systems and eventually to customer funds sitting on the exchange.

What looked from outside like two separate companies was in practice one operation with two names. Ordinary deposits became working capital for Alameda's trades. Lewis argues this did not begin as fraud in SBF's own mind; he believed overall positions were sound enough to make customers whole eventually. The practical effect was that FTX never simply held customer money the way it told customers and regulators it did.

In practice: Alameda's privileged system access turning customer deposits into trading capital while the public still saw two firms.

6

Penthouse without walls

Communal living in the Bahamas mistook informality and shared quarters for accountability and removed ordinary boundaries between personal judgment and corporate oversight.

In 2021 FTX relocated to the Bahamas partly for regulatory convenience. Bankman-Fried moved into a penthouse at the Albany resort with roughly ten senior colleagues, including Alameda CEO Caroline Ellison, with whom he had an on-and-off romantic relationship. The boundary between office and household essentially disappeared.

Major financial decisions were discussed over shared meals. Staff played League of Legends or Storybook Brawl during meetings. SBF kept a famously disheveled appearance, drove a Toyota Corolla, and wore the same shorts and t-shirt regardless of audience. Outsiders read this as authenticity and focus. It became part of the mythology that helped FTX raise money at a valuation near 32 billion dollars in early 2022 from venture firms and pension funds that performed only limited due diligence.

Lewis treats the culture as structurally dangerous, not merely quirky. Conventional firms use independent boards, financial controls, and formal reporting lines precisely to catch problems founders cannot see or will not admit. FTX's dormitory-style management left almost no one inside positioned to challenge his decisions. The people closest to the actual numbers were also the most personally entangled with him.

In practice: The Albany penthouse where senior staff lived, ate, gamed, and decided major financial moves under one roof.

7

Purchased blue-chip status

Stadium naming rights, celebrity deals, and political donations were a calculated purchase of legitimacy that made a barely regulated offshore exchange resemble an established financial firm.

Lewis catalogs a spending pattern that looked like ordinary marketing but functioned closer to reputation laundering. FTX paid roughly 135 million dollars for naming rights to the Miami arena that had housed the Heat. It signed endorsement deals with Tom Brady and Gisele Bundchen. It ran a Super Bowl advertisement in February 2022 featuring Larry David mocking historical figures who missed good ideas, implicitly casting crypto skepticism as the same error.

Politically, Bankman-Fried became one of the largest donors of the 2020 election cycle, giving tens of millions of dollars largely to Democratic causes while also funding Republican-aligned dark money groups. Lewis reads the dual track as an attempt to buy access across the spectrum rather than commitment to one side. He testified before Congress on crypto regulation, positioning FTX as the industry's responsible actor, which inclined regulators and journalists to treat him as a good-faith partner rather than a subject of scrutiny.

None of this spending built revenue. Its entire value was reputational. It converted a young, opaque, offshore company into something that felt, to investors, politicians, and customers, like an established institution. That borrowed legitimacy is part of why so many sophisticated parties failed to ask where customer money actually sat.

In practice: The roughly 135 million dollar Miami arena naming deal and the Larry David Super Bowl spot that made an offshore exchange feel institutional.

8

The hole already there

FTX's collapse was not a classic bank run but the sudden public exposure of a hole that had existed for months, where customer deposits had already been spent by Alameda.

The proximate trigger in November 2022 was a CoinDesk report revealing that a large share of Alameda Research's balance sheet consisted of FTT, the token issued by FTX itself, rather than independent assets. That raised an obvious solvency question in public. Binance CEO Changpeng Zhao then announced he would sell Binance's FTT holdings, a move Lewis frames as retaliatory given prior tension between the exchanges.

The announcement triggered a wave of customer withdrawal requests. Within days it became clear FTX could not honor them. A large portion of customer deposits, by some estimates around 8 billion dollars, had already been transferred to Alameda and used for trading, investments, and loans rather than held in reserve. FTX filed for bankruptcy on November 11, 2022. Bankman-Fried was arrested in the Bahamas the following month.

Lewis emphasizes that the run did not create the insolvency. It only revealed one that had existed for an extended period. Failure was close to inevitable once anyone looked closely at the assets actually backing customer accounts. Blurred boundaries visible since Alameda's founding had reached their logical endpoint.

In practice: CoinDesk's FTT revelation plus Zhao's sale announcement exposing that deposits were already gone into Alameda's positions.