Past midnight in the Bahamas, where he was reportedly still living, the founder was answering his direct messages.
The exchange he had built was already in bankruptcy. Billions of dollars of customer deposits were apparently gone. There were, by one filing's count, possibly more than a million creditors, and by his own account one job left for the rest of his life: to find the roughly eight billion dollars it would take to make those customers whole. Still, for more than an hour, he kept typing into a chat window. Blue and gray bubbles, short replies, timestamps a little before ten on the screenshots.
A reporter he had spoken with over the summer came back to a question from that earlier conversation. Months before, she had asked whether there was a line a person should not cross, even to do enormous good. His answer then had been careful, almost stern. "There is some line," he had said. "The answer can't be there is no line. Or else, you know, you could end up doing massively more damage than good."
Now she asked whether that careful answer had been the real one, or the public relations one. He typed back: "man all the dumb shit I said." Then: "it's not true, not really."
She pressed. So the ethics, she wrote, mostly a front? "yeah," he replied. "I mean that's not all of it." Then: "but it's a lot." He had been good at talking about right and wrong, he added, because "I had to be." It was, he wrote, "what reputations are made of, to some extent." He said he felt bad "for those who get fucked by it," by what he called "this dumb game ... where we say all the right shiboleths and so everyone likes us."
For a couple of years, this had been one of the most admired reputations in finance. Here was a young man who, the story went, had set out at twenty to "make as much money as he could, in order to give away everything he earned to charity." He drove an old Corolla. He put his face on magazine covers and said he wanted to spend his fortune stopping the next pandemic and keeping the human race alive long enough to have a future. The philanthropic fund, launched in February and paid for mostly by him, promised to move at least a hundred million dollars that year and said that in principle it could deploy a billion. It told applicants there was "no limit on how much you can apply for."
Here is the uncomfortable part. That mission was not a costume the trouble happened to wear. It was collateral.
A promise to do immense good does not make a balance sheet any safer. What it changes is the people standing around the balance sheet. It makes the ordinary questions feel rude. You are slow to interrogate the books of a man who is trying to save the world. It feels small, almost indecent, like frisking a surgeon on his way into the operating room. So you audit the mission, which is inspiring, and you forget to audit the money, which is the only thing that was ever actually at risk.
Look at what the questions would have found. A balance sheet from the founder's trading firm, dated the middle of the year, showed about $14.6 billion in assets. The single largest piece, $3.66 billion, was the exchange's own token, a coin the exchange had invented. Another $2.16 billion was more of that same token, pledged as collateral. Cash: $134 million. Against all of it sat $8 billion in liabilities, most of it loans. The token was a receipt printed by the same store that needed the receipt to be worth something. It looked valuable exactly as long as everyone trusted the store.
Days before the collapse, with rumors spreading, the founder tweeted four words: "FTX is fine. Assets are fine." The fine print customers had agreed to said their coins stayed theirs and were never loaned to the exchange. Then the withdrawals came, the token collapsed, and the receipts were suddenly worth what receipts are worth. In the days that followed, it was reported that eight billion dollars or more of customer money had quietly moved to the trading firm. He disputed the wording. "We didn't secretly transfer," he said, and blamed confusing internal labeling.
A good enough reason to give money away is also a good enough reason to stop counting it.
None of that was hidden by the philanthropy. The philanthropy did not forge the books. No one can measure how much of the trust came from the halo and how much from the venture money, the trading volume, the yield, the plain fear of being left out. But the good story widened the zone where ordinary proof felt less urgent. A good enough reason to give money away is also a good enough reason to stop counting it.
Think about what a spotless cosigner does to a loan. The borrower's numbers do not change. But the bank relaxes, the rate drops, the questions that would have been asked go unasked, because a trusted name is standing behind the deal. That name is being pledged as an asset even though nobody writes it into the contract. A mission to save future lives is a cosigner with a very good name.
When the name failed, you could see exactly who had been standing behind it.
The fund was run by a team of five. On November 10 they quit in a single public note. "We are now unable to perform our work or process grants," they wrote. They had "fundamental questions about the legitimacy and integrity" of the business that had been funding them. It looked likely, they said, that "there are many committed grants that the Future Fund will be unable to honor." Then: "We deeply regret the difficult, painful, and stressful position that many of you are now in."
A children's charity in India had been given two hundred thousand dollars in the spring and had counted on more; its director had planned to widen a multi-year program, and now had to rethink the expansion and where the next grant would come from. A man who received a hundred and eighty-two thousand dollars to build a public forecasting website wrote that "likely the fraud was already happening when the money was donated." "That weighs on me." A chemistry professor got half a million dollars to work on a thin protective film for the next pandemic and, once the money arrived, never heard from anyone again. An investigative newsroom had two-thirds of a five-million-dollar grant frozen.
Now put the numbers next to each other. The fund had committed a bit over a hundred and sixty million dollars. The hole reported at the exchange was around eight billion. The most the fund had ever promised to hand out in a single year, in principle, was one billion. The good the money was meant to do was a rounding error against the reported shortfall.
The obvious objection is that this kind of failure will wear any costume, and this one happened to wear a halo. Strip the mission away and the machinery is depressingly ordinary: customer money reportedly treated as house money, a trading arm dressed up as a stranger, collateral the house printed for itself, a run once the doubt spread. The sophisticated money did not wire in because it admired the charity. A sovereign wealth fund, a teachers' pension, and some of the most respected venture firms in the world all bought in, and one of them wrote that it had done "extensive research and thorough diligence," pointing to a billion dollars of revenue the year before. They were not chasing a good deed. They were chasing a growth story, and they got fooled by the same numbers everyone else did. The movement he had been the young face of said as much when it was over: if he lied to and defrauded people, one of its central groups wrote, "they were not acting in accordance with the principles of effective altruism." On that reading, if the reports hold up, the philosophy was robbed, not guilty.
That is a fair reading. It is also incomplete, and the most honest voice against it belongs to the person with the most to lose by admitting it.
A decade ago, a visiting philosopher had lunch with a vegan undergraduate who wanted a career helping animals, and told him he might do more good by earning a fortune and giving it away than by working the cause directly. That philosopher went on to become the intellectual face of the whole enterprise. After the collapse, he did not hide behind the principle that fraud breaks the rules. He wrote something harder. The founder had goodwill, he said, and some of that goodwill came from ideas he himself had spent years promoting. "If that goodwill laundered fraud, I am ashamed."
Laundered is the exact word. Goodwill can become collateral before anyone notices it has been pledged. The good story did not require the fraud. It bought something more useful than money: the benefit of the doubt, at scale, from careful people who were sure they were being careful.
By the time he was disowning that moral language in a chat window, it had already done its work. None of this is really about crypto.
We treat a person's stated goodness as if it were evidence about their books. The better the story about where the money is going, the less we feel entitled to ask where the money actually is. So make the list of the causes you believe in so completely that it would feel indecent to audit them. That list is not a measure of what you value most. It is a map of where you are easiest to rob.