Hook
On December 4, 2021, BitMart's hot wallet hemorrhaged roughly $319.5 million. Four to five years later, the exchange is not announcing a recovery. It is announcing a menu. Three options โ one of them cash, two of them tokens that do not yet exist, backed by assets that have not been disclosed, governed by a plan no court has approved. Every timestamp is a potential crime scene, and the 2021 breach is the oldest one in this file.
The proposal landed on a Wednesday. By the following spring, users had already started running โ social-media attacks in May triggered "a series of panic withdrawals," in BitMart's own framing. That is not a market event. That is a verdict. The crowd priced the exchange's solvency before management priced its own plan.
Context
For readers who skipped the FTX and Mt. Gox post-mortems, here is the anatomy. A centralized exchange holds user assets in custodial wallets. When it fails, users become unsecured claimants in a liquidation queue. BitMart sits in the second tier of that market. Founded in 2017, scaled by 2021, it survived the last cycle through a mechanism it now describes without irony: plugging the 2021 hack gap with bull-market profits. When the market softened into 2026, the plug stopped working.
The proposal is a "court-backed" plan offering creditors a three-way election. Path one: immediate proportional cash โ fiat plus USDC, PYUSD, USDT, BTC, ETH, and SOL drawn from BitMart's liquid assets. Path two: a Restitution Token, where $1 of claim converts to one token, backed by assets recovered from the 2021 hack. Path three: a Continuum Token, tradable on a DEX, backed by the disposal of illiquid holdings โ investments, private equity, altcoins โ plus a cut of future profits if BitMart restarts operations.
Advisers include White & Case and Alvarez & Marsal. That is heavy artillery for a second-tier exchange, and heavy artillery is not hired for small, simple estates. A creditor group called Echo Base is circling, weighing an involuntary bankruptcy petition. And an unnamed digital-asset investor offered $10 million to inject; BitMart declined it as "clearly insufficient."
That rejection is the most honest number in the entire file. If $10 million is insufficient, the hole is larger than $10 million. Nobody has told us how much larger.
Core
Let me do what I do: open the mechanism and trace where it fails to bind.
The ledger bleeds where logic fails to bind, and this one bleeds in three places.
First, the three paths are not equivalent choices. They are a risk ladder dressed as a menu. Cash is low-risk, low-return. Restitution Token is contingent on historical recovery. Continuum Token is contingent on future operations. The designer of any payout menu has an incentive to route the least-informed claimant toward the most contingent asset, because that is where liability can be deferred the longest. A menu is not neutrality. A menu is a sorting mechanism, and sorting mechanisms reward the party that built them.
Second, the Restitution Token's collateral logic is broken on its face. If a meaningful portion of the $319.5 million had actually been recovered, BitMart could top up the cash pool directly. Carving out "recovered assets" as a separate token backing implies the recovery is small, slow, or contested. The disclosure language โ "certain pockets of value which are held with certain centralized exchanges" โ provides exactly zero verifiable information. That is not a sentence. That is a hedge wearing a sentence's clothes.
Third, Continuum Token is the most dangerous instrument in the plan, and the one dressed most attractively. It binds historical debt to future operations. Read that again. Creditors are being asked to become angel investors in a restart โ with no equity, no governance, no board seat, and no enforceable legal claim on the "future profits" that supposedly back their tokens. The value of Continuum derives partly from a portion of future profits if BitMart secures financing to restart. That is a structurally Ponzi-adjacent design: later activity paying earlier claimants, with those earlier claimants holding no enforceable claim on the later activity. Code does not lie; it merely waits. So does this structure, and what it waits for is a buyer.
Now the missing data. Has BitMart disclosed the total user claim? No. The asset-liability gap? No. The actual recovered amount from 2021? No. The valuation of its private-equity book? No. It is, in its own words, drafting a report estimating the return percentage of each path โ which is an admission that BitMart itself does not yet know what each path is worth. In my audit experience, when a counterparty asks you to choose between options before it has finished computing their value, the choice is not the product. The delay is the product.
Two tokens, neither audited. No code review disclosed. No third-party custodian named for the recovered assets. No market-maker commitment for the DEX-traded Continuum Token. A tokenized claim listed on a DEX without liquidity guarantees is a pricing void โ easily manipulated, impossible to exit cleanly. Silence in the logs screams louder than alerts.
There is a further structural flaw worth naming. BitMart attributes part of its distress to wash-trading rings exploiting its futures rebates and zero-slippage incentives. That is not just a market complaint; it is a confession about risk-control design. A rebate or zero-slippage mechanism that can be systematically arbitraged is a bug in the incentive layer, and the bug hides in the whitespace you skipped โ the parameter nobody stress-tested because it looked generous.
Then the Howey test. Continuum Token hits all four prongs: money invested (claim converted), common enterprise (shared asset pool), expectation of profit (recovery plus future profits), and reliance on others' efforts (BitMart's recovery and operations). A court blessing the plan is not an SEC exemption. "Court-backed" describes a hoped-for status, not a current fact. No court has approved this proposal. The word is marketing.
And the comparison the source material buried in its final line is the loudest signal of all. Bitget, hit by a zero-day, recovered $388 million and kept operating. BitMart, hit in 2021, is liquidating. Same category of event. Opposite outcome. The difference is not the hack. The difference is what happened in the years between.
Contrarian
Here is where the bulls โ and the reflexive bears โ both miss.
The consensus read is that tokenized payouts are a gimmick and users should take the cash and run. Mostly right. But that read skips the one thing BitMart got correct: immediate cash distribution would exhaust the liquid pool, and an exhausted pool triggers a cascade for everyone, including the people who chose cash. Stretching repayment across a menu, however cynically motivated, buys the estate time to dispose of illiquid holdings at non-fire-sale prices. That is not charity. It is self-interest. But self-interest and creditor interest can temporarily align.
The sharper contrarian point is Echo Base. An involuntary bankruptcy petition looks like escalation. For small creditors, it may be the only path to transparency. A court-appointed trustee would hold subpoena power over the recovered-asset ledger that BitMart has kept private. BitMart's floated "50 largest holders" consultation is a divide-and-conquer move: co-opt the whales, isolate the retail tail. Echo Base breaks that sorting. In my reading of the Terra collapse, the creditors who recovered best were the ones who forced disclosure early, not the ones who trusted the debtor's timeline. Trust is a variable, never a constant.
Takeaway
The three paths are not three futures. They are one future with three delays. Cash is now, discounted. Restitution is maybe, undated. Continuum is later, and someone else's problem โ possibly yours.
Before any creditor taps a path, three questions must be answered in writing: How much was actually recovered? What is the total user claim? How much of the cash pool does the cash path actually cover? Until those numbers exist, "choosing a path" is not a decision. It is a wager placed in the dark, on a table the house built.
Reputation is liquid; solvency is binary. BitMart is no longer trading on the first. It is negotiating the second. Watch the Echo Base filing, not the press release. The ledger will tell you what the menu will not.