The withdrawal was supposed to take 20 minutes. It’s been three weeks, and the BitMart UI still shows a spinning wheel. The chat is a graveyard of angry emojis and unanswered support tickets. This isn’t a hack. This is a slow bleed.
Over the past six weeks, four centralized exchanges—ABFinance, BitMart, BitMEX, and AscendEX—have announced their closure. The pattern is not random. It’s a mechanical, predictable failure of a business model that was already broken before the bear market arrived. I’ve been tracking the on-chain signals since the 2021 NFT peak, and this feels different. The pace is accelerating. The industry is entering a phase of structural re-calibration, and the casualties are not just small players.
Why now? The market is in a sideways grind, trading volume is dropping, and regulatory compliance costs are rising. The math for mid-tier CEXs has become impossible. They need to cover server costs, licensing fees, legal teams, and insurance premiums—all while competing with Binance’s liquidity and Uniswap’s zero-friction model. The margin for error is zero. When a user requests a withdrawal, the system must return the asset. If the asset is missing—because it was lent to a market maker, or used for yield farming, or simply lost in the shuffle—the machine breaks.
ABFinance was the canary. Founded by Helen Liu, the former co-CEO of ByBit, it was supposed to be a new chapter. She launched it in March 2026, resigned from ByBit on April 30, and by the end of the sixth week, the exchange was shutting down. It never even went live. The company stated it would conduct an “orderly wind-down.” But the cost of the licensing application, the legal fees, the office space—it’s all sunk. The lesson is brutal: in a contracting market, even a top-tier founder with a known brand cannot build a new CEX from scratch without burning through millions before the first trade is executed.
Tracing the trail from NFT peaks to DeFi valleys, the real story is about the architecture of trust. A centralized exchange is a black box. When you deposit funds, you receive a database entry. There is no smart contract enforcing the reserve ratio. There is no on-chain proof that the exchange holds the assets. The model relies entirely on the goodwill and solvency of the operator. When the market turns, the first thing to break is the withdrawal pipeline.
BitMart is the textbook example. Users began reporting “extremely slow” withdrawals. The chief product officer resigned. Then the founder started threatening legal action against anyone who asked for transparency. The message was clear: the company was in defensive mode. The withdrawal speed is the canary indicator for a CEX. When it slows down, it means the treasury is being squeezed. The exchange is either unable to access the hot wallet, or—more likely—the hot wallet is empty, and they are waiting for cold wallet funds to be transferred, which itself is a sign that the asset-liability mismatch is severe.
AscendEX went further. On-chain detective ZachXBT flagged the exchange’s reserves as missing significant amounts of ETH, USDT, and SOL. The chain doesn’t lie. The data is public. If a CEX claims to have $100 million in user deposits, but the on-chain addresses hold only $60 million, there is a gap. That gap is usually filled by the exchange’s own trading profits, or by user funds that have been rehypothecated. When the gap is exposed, the trust collapses instantly. AscendEX’s shutdown was not a surprise. It was a foregone conclusion from the moment ZachXBT published his findings.
BitMEX is the most symbolic closure. The exchange that invented the perpetual swap, the platform that defined the derivatives market for a decade, is shutting down in September. It has a $270 million insurance fund, but users are worried about what happens to that money. The fund is not user property. It’s the exchange’s own capital, set aside for contingencies. In a bankruptcy scenario, it might be used to pay lawyers, not customers. The insurance fund is a mirage: it protects the exchange from its own trading losses, not the user from the exchange’s collapse.
Chasing the alpha through the noise, I’ve been looking at the market structure. The aggregate effect is a redistribution of trust. The users who lose access to their funds on BitMart or AscendEX will not leave crypto. They will migrate to self-custody wallets or to the top-tier CEXs that have published proof-of-reserves and undergone third-party audits. Coinbase, Binance, OKX are the relative winners. The DEX ecosystem—Uniswap, dYdX, Synthetix—is the structural beneficiary. The narrative is shifting from “CEX is convenient” to “CEX is a single point of failure.”
The contrarian angle is uncomfortable. The market is celebrating the “death of the CEX” narrative, but the reality is more nuanced. The four closures are not a sign of a healthy ecosystem’s evolution. They are a sign of a systemic fragility that is not being addressed. The industry is still building on the same model: a centralized entity that controls user funds. The only difference is that the survivors are better capitalized. But the underlying risk—the possibility of a sudden, arbitrary freeze—remains.
Hype, heartbeats, and hard data—the on-chain data tells a story of contagion. The withdrawal addresses of the failing exchanges are being monitored by liquidation bots and analytic firms. Every time a large wallet moves, the market interprets it as a sign of distress. The speculative pressure is self-reinforcing. I saw this in the 2022 DeFi crash, when the collapse of one protocol triggered a cascade of margin calls across the entire ecosystem. The same dynamic is playing out in the CEX space, but slower. The fuse is longer, but the explosion is just as final.
From the peak to the pit: a survivor’s guide—the key takeaway for the next six months is clear: do not use a centralized exchange that has not published a verifiable proof-of-reserves. The technology exists. It’s not expensive. If a CEX refuses to show its on-chain addresses, assume it is insolvent. The industry is entering a period of “trust de-concentration,” where the assumption is that every CEX is a potential failure until proven otherwise.
The race isn’t over—it’s just entering a new phase. The exchanges that survive will be the ones that treat transparency as a competitive advantage, not a regulatory burden. The ones that fail will be the ones that tried to hide in the shadows. Six weeks, four closures. The tunnel is dark, and the exit is not guaranteed. But the lights are on for those who look at the chain.
