The FTX bankruptcy estate was able to return significant funds to creditors, but some deals now look like a painful loss of future value. Assets that were sold for quick payouts later increased in value by tens of times or became part of multi-billion dollar deals.
The irony is that it was Sam Bankman-Fried’s venture bets, which once helped build the FTX empire, that after the collapse became an example of forced sales at the worst moment. Cursor, Anthropic, Robinhood, Solana, and Sui all show the same problem: liquidation lacks the patience of a venture fund.
Cursor Was the Loudest Missed Opportunity
One of the sharpest examples is related to Anysphere, the developer of the Cursor AI tool for programmers. In 2022, Alameda Research invested $200,000 in an early round and, according to Forbes, received about 5% of the company.
After FTX’s bankruptcy, this stake was sold back at the entry price. At that time, the asset looked secondary: a small check, early stage, a niche product for developers. For the liquidators, it was a position that was easier to close and convert into a clear sum.
Later, the situation changed dramatically. After news of SpaceX’s deal to buy Cursor for $60 billion, the same stake could have been worth about $3 billion. In other words, the position that was sold for $200,000 on paper would have become one of the most valuable assets ever associated with FTX.
The Cursor Mistake Looks Especially Harsh
In the case of Cursor, the gap between the sale price and the potential value became almost symbolic. This is not a typical venture revaluation of two or three times. This is a scenario where an early bet could have generated a return of thousands of times.
Social media quickly noticed this contrast. Users wrote that Bankman-Fried was good at finding deals but lost time and control over the empire. Others noted that the bankruptcy estate simply could not wait for an early-stage startup to become a giant.
This is the main conflict. Creditors need money on a court schedule. Venture assets need time. These two logics almost always mix poorly.
Anthropic Became an Even Bigger Loss
The most expensive story is related to Anthropic. FTX invested about $500 million in the AI company in 2021, before the artificial intelligence market became a mainstream investment theme.
This bet gave the company a stake of about 8% in Anthropic, founded by former OpenAI employees Dario and Daniela Amodei. After the bankruptcy, the asset was sold in 2024 in two parts with court approval.
In March, a group of institutional investors bought a stake worth about $884 million. In June, there was another sale for about $452 million. In total, FTX received about $1.3 billion.
Today, This Stake Could Be Worth More Than $30 Billion
After new funding rounds, Anthropic was valued at about $380 billion after raising capital. At this valuation, FTX’s previous 8% could have been worth more than $30 billion.
This is the clearest example of how the bankruptcy process clashed with the rapid growth of the AI sector. The sale helped increase payments to creditors, but the difference between the amount received and the possible current value became enormous.
For FTX, this looks especially paradoxical. The company collapsed due to abuses and a lack of liquidity, but its venture portfolio contained assets that could have made the former empire much richer even after the collapse.
Robinhood Was Also Sold Before Major Growth
Another example is the Robinhood stake. Through Emergent Fidelity Technologies, a structure associated with Bankman-Fried, a 7.6% stake in Robinhood was purchased for about $648 million in 2022.
After FTX’s collapse, the shares were seized by U.S. prosecutors. In 2023, the U.S. Marshals Service sold 55.3 million shares back to Robinhood at $10.96 per share. The deal amounted to about $605.7 million.
Now, the same stake would be worth more than $5 billion at Robinhood’s market valuation of about $87 billion. Unlike Cursor and Anthropic, this was already a public and more understandable asset, but the result was similar: a forced sale occurred long before a strong price recovery.
Solana Became a Controversial Example
The story with Solana is more complicated. Alameda was one of the early and major supporters of the project, and SOL was long considered one of the main assets associated with the FTX ecosystem.
In 2024, the bankruptcy estate sold about 30 million locked SOL at about $64 per coin. Galaxy Digital and Pantera Capital were named among the major buyers.
Later, SOL rose to almost $293 at the beginning of 2025, then dropped to about $74. Therefore, this sale looks especially unfortunate relative to the highs, but not so dramatic at the current price. Nevertheless, the asset showed that even locked tokens could have delivered a much greater result with longer holding.
Sui Was Closed Almost at Entry Price
A similar story happened with Mysten Labs. In 2023, the company bought back FTX’s stake and rights to Sui tokens for about $96 million.
This was close to the amount the exchange paid a year earlier, about $101 million. For the bankruptcy estate, such a deal looked logical: the asset was closed with almost no capital loss and added funds for settlements with creditors.
But in the context of subsequent growth in part of FTX’s venture portfolio, this sale is seen as another example of giving up future potential for quick liquidity.
Why Liquidators Sold So Early
At first glance, all these stories look like a series of investment mistakes. But the team managing the bankruptcy had a different task.
John J. Ray III and his team had to raise money, stabilize the process, and return as much as possible to affected clients. They were not managing a venture fund that could wait five or ten years.
The legal process requires predictability. Liquidators often choose the real price today, not the potential value tomorrow. That is why some sales may look terrible in hindsight, even though at the time of the deal they seemed reasonable from the perspective of returning funds.
The FTX Portfolio Turned Out Better Than the Company’s Reputation
A separate conclusion concerns Bankman-Fried himself. Despite the collapse of FTX and the criminal case, his early investments in Anthropic, Cursor, Solana, Robinhood, and Sui show that the former empire had strong access to promising assets.
Some investors on social media are already calling him one of the luckiest venture players of his time, if you look only at asset selection. According to some commentators, if all the stakes had been kept, his fortune could have been measured in tens of billions of dollars.
But this assessment does not cancel the main point. The assets could have been strong, but the company’s management was catastrophic. In the end, good bets did not save FTX from collapse, and after bankruptcy, many of them were sold before they revealed their main value.
Creditors Got Their Money, but Missed the Upside
The sales of Anthropic, Robinhood, and other assets helped bring creditor payouts close to full recovery. For affected users, this was more important than waiting for future rounds and the risks of private companies.
But now every new Anthropic round, Robinhood’s growth, or a deal around Cursor strengthens the sense of missed opportunity. The higher FTX’s former assets rise, the sharper the question becomes: was it worth selling so quickly?
The answer is not obvious. If the markets had turned down, the liquidators would have been praised for their caution. But against the backdrop of the AI boom and the growth of certain tech companies, their decisions now look too early.
What Happens Next?
The story of FTX’s former portfolio will resurface every time one of the sold stakes gets a new valuation or becomes part of a major deal. Especially if it’s about AI, fintech, and blockchain.
For creditors, the main result has already been achieved: the assets helped return money. For the market, the main lesson is different. Forced sales are almost always worse than patient capital, especially if the portfolio contains early bets on fast-growing sectors.
The main takeaway is simple. FTX collapsed due to internal violations, but its venture assets turned out to be surprisingly strong. Anthropic, Cursor, Robinhood, Solana, and Sui show that the bankruptcy estate did not so much make mistakes in asset selection as it was forced to sell them before the market revealed their true value.
Read More: Russia Will Add USDC to the List of Permitted Crypto Assets
