Bankman-Fried raised nearly $170 million from a set of investors ascribing to the ‘Effective Altruism’ community – a network of people who try to find the best ways to serve the community, usually by donating or funding causes.

The then 26-year-old SBF intended to invest these funds in the growing and inefficient crypto markets, capturing price differences across markets and creating high-frequency trading (HFT) strategies to pick up pennies every few seconds.

Most of these were losing bets from the start with Alameda losing millions of dollars in its first months. It lost over $500,000 every day throughout one such month, Lewis wrote, while some trading funds had “simply vanished” due to poor fund management.

Another bot called Modelbot, which was programmed to trade nearly 500 tokens on some thirty exchanges, turned out to be yet another dud initially. It made no distinction between deeply-liquid crypto majors such as bitcoin (BTC) and ether (ETH) and very thinly-traded memecoins – sparking concerns among early Alameda staff that it could end up evaporating all of the raised money.

The tides finally changed after Gary Wang and Nishad Singh (both FTX directors who have since pled guilty to fraud in the ongoing trial) joined the firm.

Wang is said to have coded a quantitative trading system that finally started to make Alameda money, while Singh put together the pieces to manage the company – putting it on track to what would eventually become the crypto exchange FTX.