The BitMart shutdown announcement hit on July 26, 2025. Nine-year-old exchange. Stops registrations. Ends trading August 26. Withdrawals close four hours later. Platform fully dead by January 2027. Clean timeline. Professional.
Then the founder called it a hack.
Sheldon Xia, via his own X account, claims the Chinese-language open letter was fabricated by an attacker. The letter demanded he and Nancy Li disclose all wallets, liabilities, and reserves by August 19. Pay unpaid salaries. Xia says he will file a police report and send a lawyer letter to X.
Two narratives. Same event. Only one can survive basic forensic analysis.
Immutable logic: if the shutdown was real, the founder would have nothing to hide. If the letter was fake, the exchange would have published a counter-statement with wallet addresses to prove solvency. Neither happened.
Context: A CEX Built on Sand
BitMart is a fifth-tier centralized exchange. Not a protocol. Not a DeFi primitive. A simple order book with a database of user balances. The technical architecture is trivial: hot wallets, cold wallets, a matching engine. No innovation. No proof-of-reserves implementation.
In December 2021, a hot wallet exploit drained $196 million. The exchange survived. But it never rebuilt trust. No Merkle-tree audit. No third-party attestation. Just promises.
Now, the shutdown reveals the structural flaw: a CEX without transparent reserves is a black box. Users deposit assets. They receive a database entry. The exchange holds the private keys. This is the fundamental risk of centralized custody. Code is not law here. Trust is.
Based on my experience auditing smart contracts in 2017, I learned that any system that cannot be independently verified is a liability. BitMart is the textbook example.
Core: The Order Flow Tells the Story
Let’s follow the on-chain data. Arkham Intelligence marks one BitMart wallet. As of the shutdown announcement, it held approximately $70 million. By the time the open letter surfaced, the balance had dropped to $36 million. A net outflow of $34 million.
Two possibilities.
First: users are withdrawing in panic. The exchange is processing requests, but the outflow is faster than the system can handle. This is plausible. But the withdrawal window is only open for four hours after trading ends. That restriction suggests the exchange wants to limit the outflow, not facilitate it.
Second: the exchange is moving funds to unmarked addresses. This is the more dangerous scenario. If the $36 million is all that remains in known wallets, and the total liabilities exceed that, the exchange is insolvent. The open letter’s demand for a full wallet disclosure implies the employees suspect this.
Immutable logic: a solvent exchange would publish wallet addresses to stop the panic. BitMart did not.
Compare to FTX. The same pattern. The same silence. The same “we are working on it” language. The difference is that BitMart’s failed proof-of-reserves is not a surprise—it was never implemented.
I profited from the 2022 Terra collapse because I studied the code. The algorithm was broken. The same analytical framework applies here: the reserve mechanism is broken. The only question is the magnitude of the shortfall.
Contrarian: The Hack Claim is a Diversion
Retail users are clinging to the founder’s narrative. “The account was hacked.” “The shutdown is fake.” “Our funds are safe.”
This is emotional betting. Smart money reads the signal.
If the account was genuinely hacked, why didn’t Xia immediately publish a video statement or a signed message from a verified address? Why didn’t he provide a single wallet address to prove the exchange still has funds?
Instead, he threatened legal action. He went silent. The only official communication since the shutdown is the timeline. No wallet. No proof.
Contrast this with the leaked CEO speech from an internal meeting. The CEO reportedly said the exchange would “guarantee to pay users.” But guarantees without proof are worthless.
Immutable logic: a claim of a hack is the most convenient excuse for a failing exchange. It shifts blame. It buys time. It allows the founders to move assets without scrutiny.
I saw the same playbook in 2020 when a yield farm lost its TVL and blamed a “compromised admin key.” The real story was a rug pull. BitMart may not be a rug, but the opacity is identical.
Takeaway: The Only Actionable Step
If you have funds on BitMart, assume they are at risk. The withdrawal window is a narrow exit. Do not wait for the founder’s next statement. Do not hope for a miracle. Execute the withdrawal. Immediately.
For everyone else, this is a hard lesson: centralized exchanges without proof-of-reserves are speculative counterparties. The market is pricing this risk into the premium on decentralized alternatives. The next time you see a CEX with no transparent reserves, assume the worst.
The BitMart shutdown is not a black swan. It is a predictable outcome of a system that failed to implement the most basic security verification. The code—or the lack of it—was always the truth.
Immutable logic: if you cannot verify the reserves, they do not exist.