The fifth tranche lands. $1.4 billion. Cash, not crypto. The market yawns.
That should be the headline. FTX’s latest distribution—sixth in a series since 2025—pays out to non-convenience class creditors. Another $1.4 billion of liquidity returned to hands that mostly sold their claims years ago. The recovery rate? Above 100%. Up to 120% for some classes. By any traditional metric, this is a success. By the cold logic of cryptographic proof, it is a monument to centralized efficiency.
I spent 2017 auditing ZK-SNARK circuits. I learned that verification is not the same as trust. Here, the verification is the court docket. The trust is in John Ray III. The result: a liquidation that makes Mt.Gox look like a slow-motion car crash. But before I get to the mechanics, let’s be precise about what we are analyzing.
Context: The Protocol of Bankruptcy
FTX filed for Chapter 11 on November 11, 2022. The asset portfolio was a mess: commingled funds, missing keys, fraudulent balance sheets. The court-appointed team, led by restructuring expert John Ray III, began the forensic reconstruction. They recovered assets—crypto, equity stakes (Anthropic), real estate—and liquidated them over three years. By early 2025, they had sufficient cash to start distributions.
The plan was court-approved. It divided creditors into convenience (claims under $50,000) and non-convenience classes. Each class receives cash payments based on the value of their claim as of the petition date—November 2022 prices. That is the key. Not the market price at distribution. The November 2022 price.
As of this fifth distribution, total payouts exceed $10 billion. The total recovery pool is over $14 billion. The percentage recovery for most creditors is between 100% and 120% of their claim value. By any standard, this is unprecedented for a crypto exchange collapse.
Core: The Mechanical Truth
Let’s dissect the distribution engine. It operates on a centralized ledger—the claims database. Every creditor submits proof of claim. The team verifies against FTX’s internal records. Once approved, the claim is assigned a value based on the November 2022 price oracle. Then, periodic distributions are scheduled based on cash availability.
The fifth distribution processed $1.4 billion. The fourth distributed $1.2 billion. The third, $2.1 billion. The pace is aggressive. Compare to Mt.Gox, which took over a decade to begin distributions and is still ongoing. FTX achieved major disbursements within 30 months of filing. Why? Because the team used a centralized authority with full control over the asset pool. No on-chain voting. No multi-sig delays. No community disputes.
Code is law? No. The judge’s order is law. The liquidation team is the sole sequencer of payments. They decide when, how much, and to whom. This is the ultimate centralized sequencer. And it worked.
But here’s the technical trade-off: efficiency comes at the cost of flexibility. Creditors cannot choose to receive crypto instead of cash. They cannot defer distribution to avoid tax events. They cannot negotiate the price peg. The system is deterministic: claim value fixed in November 2022, distribution in cash. That’s it.
From a game-theoretic perspective, this is an optimal solution for maximizing recovery in a closed system. The team faced no conflicting incentives—their compensation likely tied to total recovery. They liquidated assets at market peaks (2024 bull run) and held cash. The resulting surplus allowed them to pay claims above 100%.
Contrarian: The Blind Spot of Success
The narrative is clear: FTX’s liquidation is a win. Creditors get their money back. Some even profit. The market absorbs the cash without disruption. But this story obscures three structural blind spots.
First, the price oracle. By pegging claims to November 2022 prices, the system penalizes creditors who held assets that appreciated. If you had 1 BTC on FTX (worth $16,000 in November 2022), you receive $16,000 cash. That same BTC today is worth over $60,000. The recovery percentage is 100% by claim value, but the opportunity cost is 275%. The distribution is effectively a forced liquidation at a historical low. This is not restitution—it is a settlement that eliminates upside.
Second, the centralization dependency. The entire distribution relies on a single legal and administrative apparatus. If John Ray’s team makes a clerical error—wrong address, misclassified claim—the creditor has no recourse except court litigation. There is no cryptographic proof of payment. No transparent on-chain verification. The process is opaque to all but the administrators. In the 2024 audit of a DeFi lending protocol, I identified a 15% payout error due to a rounding bug. Here, the errors are human, and the cost is identity theft or lost funds.
Third, the fraud surface. FTX’s bankruptcy has already spawned a cottage industry of scammers. The official distribution process requires no wallet connection. Yet phishing sites mimicking the claims portal have stolen millions. The team’s explicit warning—“We will never ask you to connect a wallet”—is a direct admission that the system is vulnerable to social engineering. In a decentralized liquidation, the smart contract is the arbiter. Here, the arbiter is a PDF notice.
We build the rails, then watch the trains derail. But this train didn’t derail—it arrived on time. That’s the dangerous part. It reinforces the belief that centralized trust is acceptable as long as the operator is competent. John Ray III is competent. But competence is not a security model.
Takeaway: The Precedent
FTX’s liquidation sets a new baseline for crypto bankruptcy. Future collapses—and there will be more—will be judged against this 120% recovery metric. Creditors will demand similarly aggressive asset recovery and rapid distributions. That is a positive.
But the precedent also legitimizes a dangerous practice: pricing claims at the low of the crash. This systematically transfers wealth from crypto holders to distressed debt traders who bought claims for pennies on the dollar. The efficiency of the liquidation masks a regressive redistribution.
The final tranche is not announced. The sixth distribution may come in 2027. By then, the narrative will be forgotten. The infrastructure remains—centralized, opaque, and dependent on a single point of trust. That is the real vulnerability.
Code is law, until the oracle lies. Here, the oracle never lied. It just locked the price in a permanent bear snapshot.
Let’s see how long that precedent holds.