Ninety-eight days is the only number in CoinEx's farewell letter worth trusting.
The announcement reads like a eulogy. Nine years of operation. A committed community. A graceful exit attributed to "sustained market downturn" and "rising compliance costs." Every sentence works to soften one hard fact: the exchange will stop moving user funds on December 22, and after that date the only thing left is a closed door.
I have dissected shutdown notices before. The code whispered what the pitch deck screamed, and a wind-down letter is just a pitch deck played backward. When a centralized exchange voluntarily opens a withdrawal window instead of freezing accounts overnight, it is telling you something precise — and the something is not gratitude. It is a solvency statement. It is also, historically, the least reliable statement in this industry.
So before the sentiment, the arithmetic. CoinEx went live on December 22, 2017. The headline claims "almost nine years." Nine years from December 22, 2017 lands on December 22, 2026 — a date apparently mirrored by the withdrawal deadline. But eight years is December 22, 2025. The notice never states the year attached to its own withdrawal clause. In a document whose entire purpose is to define a deadline, an unlabeled year is not a typo. It is an artifact, and artifacts are the first place I look.
The distinction matters. If the shutdown is happening in 2025, we are reading a structural death inside a bull market — the most damning possible framing for the rest of the industry. If it is 2026, the cycle context shifts and so does the meaning of "market downturn." I am flagging this because too many analysts swallowed the number whole. Confidence: moderate. But comfortable readers should not be comfortable.
Context: a mid-tier exchange, a real founder, and a five-stage exit
CoinEx is not a scam shell. That is what makes it interesting.
It was founded by Yang Haipo, the same engineer behind ViaBTC, one of the largest Bitcoin and Bitcoin Cash mining pools in the world. Unlike the anonymous teams that populate exchange graveyards, Yang has a decade of verifiable engineering history. CoinEx ran spot, contracts, and wealth-management products. It issued a platform token, CET. It maintained a long, open affinity with the Bitcoin Cash ecosystem, where its founder was an active and vocal builder. It was, by any honest accounting, a legitimate mid-tier venue.
That legitimacy is exactly why the wind-down deserves examination rather than dismissal. Rug pulls are boring forensically — the code was always a lie and everyone simply failed to read it. An orderly shutdown executed by a competent team is far more revealing, because it forces us to ask what a well-run small exchange looks like when it decides to die.
The mechanics are staged in five phases. New registrations stop. Contracts go close-only. Non-spot services terminate. Spot markets halt. Withdrawals close. The sequence moves users from lowest-liquidity exposure to highest, giving a clear exit path at each rung. Compared to Mt. Gox, which froze withdrawals and left creditors in court for a decade, and FTX, which collapsed into bankruptcy with commingled funds, this is a genuinely superior design. I would score the process design itself as close to industry best practice.
But process design is not solvency. And the letter never once mentions a proof of reserves.
Core: the gaps the letter chose not to fill
Truth hides in the assembly, not the press release. So let me walk the assembly.
Start with the withdrawal mechanism, because it is the single technical risk vector that defines this event. On a centralized exchange, user balances are not assets — they are liabilities. The exchange owes you coins it claims to hold. That means every CEX ultimately runs on the same latent structure as a bank, minus the deposit insurance and the capital requirements. A 98-day window where every user can demand settlement simultaneously is, stripped of marketing, a bank run test — except the bank gets to schedule the run and set the tempo.
Here is the part nobody wants printed: a voluntarily opened 98-day window is an indirect solvency signal, and it points, weakly, toward the positive. Organizations with empty wallets do not offer three months of orderly exit. They freeze, or they vanish. The very fact that CoinEx chose a long, staged timeline implies that management believes on-chain assets can largely cover the liability book — at least in aggregate, at least on paper. Confidence: moderate. But notice what I did not say. I did not say the coverage is verified, because it is not.
The letter discloses no cold-to-hot wallet ratio. No third-party custody arrangement. No on-chain reserve attestation. No auditor. For a company asking the public to trust a three-month queue, that is an extraordinary omission. Every major exchange that survived the post-FTX reckoning — Binance, OKX, the survivors that wanted to be believed — published proof of reserves specifically to prevent exactly this moment of doubt. CoinEx published a calendar.
Then there is CET.
CET is where the wind-down turns from orderly to brutal.
A platform token's value capture is 100% dependent on the exchange's cash flow and ecosystem privileges. Fee discounts. Launchpad allocations. Governance votes on listing. A buyback-and-burn pool historically funded by trading revenue. Every one of those mechanisms has exactly one upstream dependency: the exchange keeps operating. When the exchange stops, the dependency graph collapses.
Fee discount on a market that no longer trades: worthless. Voting power over listings that no longer happen: no object. Buyback funded by revenue that has ceased: no fuel. Governance rights over a protocol that is being closed: governance over nothing.
I ran a similar death-screen during the FTX collapse in 2022, when I spent a stretch auditing that exchange's multi-signature structure against 200 terabytes of transaction logs. The pattern is consistent: when a platform dies, the platform token is the last liability in the queue and the first asset to zero. Users' spot balances at least have a withdrawal path. CET holders have, in this notice, absolutely nothing. No redemption schedule. No conversion ratio. No liquidation distribution formula. No burn announcement. The letter treats the token like it never existed.
If CET has a wrapped on-chain version — and most platform tokens of its vintage do, on Ethereum or BNB Chain — the situation is even stranger. The exchange closes, but the on-chain contract does not. It keeps trading. It just trades against a buyer pool that has just been told the issuer is shutting its doors. That is a liquidity vacuum with a ticker attached. The token will not be delisted by anyone, because there is no one left to delist it. Confidence on the wrapped-contract point: moderate, based on the standard architecture of tokens from that era rather than any disclosure here. But the direction of the price action is a near-certainty: toward zero, with or without a formal announcement.
There is a second, uglier question. Does the team's own CET sit in lock or unlock? Will insiders exit through the same 98-day queue as retail, or ahead of it? The notice does not say. Silence here is not neutral. It is a disclosure failure, and in my experience disclosure failures cluster.
The market is not depressed. The middle is.
Now the part the letter wants you to accept without checking.
CoinEx attributes its death to "sustained market downturn" and "declining trading volume and liquidity." Read that against the actual state of the leaders. Binance remains profitable. Coinbase remains profitable. OKX and Bybit are expanding. So the diagnosis cannot simply be "the market is bad." The leaders are eating. The patient is dying anyway.
That means the correct diagnosis is structural, not cyclical. Two forces are doing the killing.
The first is liquidity concentration. Trading volume, market depth, and institutional order flow have spent several years consolidating toward a handful of top venues. A mid-tier exchange sits in the most exposed position in the entire market: too small to enjoy the network effects that make the giants unassailable, too regulated to run as a grey-market venue, too established to pivot to a niche the giants have ignored. When volume leaves the middle, it does not return to the middle. It goes up or down — to the big venues or to decentralized protocols. The middle hollows.
The second force is compliance cost. MiCA in the European Union fully applied and raised the floor on licensing, capital, and disclosure. The FATF Travel Rule now demands originator and beneficiary data on transfers, which requires infrastructure that a mid-tier exchange must build and audit. The SEC has continued expanding its enforcement reach. Hong Kong's VASP regime turned registration into a gatekeeping cost. Individually each is survivable. Together they form a fixed cost that scales with nothing. The giants absorb it across billions in revenue. A CoinEx absorbs it across a fraction of that, and the math stops working.
Public memory adds a note that the letter omits: CoinEx previously settled with the State of New York over operating as an unregistered money transmitter and exited that market. If that history is accurate — and my confidence on the general outline is moderate — then "rising compliance cost" is not an abstraction in this notice. It is a scar. It is the polite phrasing for "the legal perimeter finally closed around us."
This reframes the whole event. An orderly shutdown under compliance pressure is not a market signal — it is a regulatory exit dressed as a market exit. And that distinction should terrify every founder running a venue between the top ten and the long tail.
Contrarian: the bears are right about CET and wrong about the industry
Here is where the easy narrative breaks, and where I have to defend the people everyone wants to dismiss.
The reflexive take — "another exchange down, crypto is collapsing" — is lazy. CoinEx does not move Bitcoin. Its market share is a rounding error against the leaders. The systemic risk from an FTX was that the exchange sat inside the plumbing of the entire credit market. CoinEx sits at the edge of the plumbing. Its failure does not threaten BTC or ETH pricing. It threatens a specific cohort of users and a specific token, and nothing more. Anyone using this shutdown to justify panic in majors has not done the arithmetic.
The more uncomfortable contrarian point is this: the orderly wind-down is the single most honest thing this industry has produced in years, and it should be rewarded with analysis, not mockery. The team chose the harder corporate path. They opened a three-month window rather than a 48-hour freeze. They staged the phases rather than pulling the rug. In a sector where the modal exit is a hard freeze and a vanishing founding team, competence that ends in a quiet queue is a strange, real form of integrity. The aesthetics of a graceful exit are not a mask here — they reflect an architecture that was actually run by adults.
Now the rebuttal, because there is always one.
Beauty is the most sophisticated rug pull, and "orderly" can itself be a performance. A staged withdrawal is orderly only if the money is actually there. The 98-day window is a promise, and promises about CEX solvency are the asset class with the worst track record in this industry's history. The letter describes a process — it does not prove a balance sheet. Until someone publishes an on-chain reserve attestation matching the liability book, "orderly" is a mood, not a fact. Confidence in the money being there: genuinely uncertain, and I refuse to launder that uncertainty into comfort.
There is a third angle the bulls miss, and it is the darkest. A high-profile mid-tier shutdown is a contagion seed, not an isolated event. Users reading this notice will, rationally or not, log into their own mid-tier accounts and ask whether the same letter could arrive in their inbox. Some will withdraw. If enough withdraw from venues with thinner reserves, the panic becomes self-fulfilling — and the next headline is written by the fear this one generated. The quiet killers of small exchanges are never the dying exchange. They are the survivors who panic and drain the next one.
Takeaway: the queue is a mirror
Silence is the only honest consensus mechanism, and this notice is loud. It shouts its cause list — markets, regulation, cost — while staying silent on the two things that would actually reassure anyone: proof that the reserves exist, and a stated plan for the token it issued.
The forward-looking question is not whether CoinEx will finish its wind-down. It is what its completion teaches the next hundred mid-tier venues. If an orderly, competent, well-funded exit still ends with retail users holding a token that has no redemption path, then the real failure is not CoinEx's balance sheet. It is the entire industry's refusal to standardize what a platform token is worth when the platform stops existing.
Watch the CET liquidity pool after December 22. Watch whether a single wrapped-CET venue posts a notice. Watch whether any mid-tier exchange publishes a fresh proof of reserves in the following thirty days. That last chart will tell you more about what the industry learned from this shutdown than any of the eulogies being written about it tonight.