Decisions

The Exit Window Is 31 Days, Not Six Months

BitMart is winding down over six months, but access effectively closes in 31 days. The rule this gives you for any vendor that announces an orderly shutdown.

DecisionsBy Servola Tech Desk2026-07-275 min read

AI-assisted, edited by humans. | Editor-in-Chief: Leon Soliman | Editorial standards

The Exit Window Is 31 Days, Not Six Months

Three dates, and the one that actually binds

On 26 July BitMart, a crypto exchange nine years old, published a notice titled as an orderly cessation of operations. The schedule has three steps. From 01:30 UTC that same day, new registrations and deposits of both crypto and fiat stopped, futures accounts moved to reduce-only, new spot orders were no longer accepted, and automated services including copy trading, grid trading and API trading began to be phased out. From 01:00 UTC on 26 August, all trading services end. At 15:59 UTC on 31 January 2027, the platform formally ceases to operate.

Read casually, that is a six-month runway. Read as an operator, it is not. Positions have to be closed and withdrawal requests submitted by 26 August; the months that follow are a tail during which users can still log in, review trading history and request withdrawals, not a period in which the service works. The company attributed the decision to operating conditions, the market environment and future strategic direction, which is the standard formulation and explains nothing.

The number that should unsettle a supplier-risk reviewer sits elsewhere in the reporting. BitMart recorded 1.6 billion dollars of trading volume in twenty-four hours, up 51 percent on the prior period, with bitcoin close to half of it. This was not a venue starved of activity. Its own token, BMX, fell 58 percent within a day of the notice to around eight cents, a market value near 27 million dollars, after a decline of roughly 70 percent over the preceding year. The token was signalling something the volume was not.

The queue is the constraint, and it was disclosed

The most useful sentence in the notice is the warning attached to withdrawals. BitMart told users that withdrawal requests may face additional scrutiny: identity verification, device and IP checks, address screening, questions about source of funds and sanctions compliance. It added that delays are possible if request volumes surge. That is a company describing, accurately and in advance, the mechanism by which an orderly process becomes a queue.

Consider what that means arithmetically. The capacity of a compliance review function is roughly fixed in the short run; it is staffed for normal operations. Demand for it is now compressed toward a single date by every remaining customer simultaneously. The throughput does not rise to meet the surge, so the effective service level degrades exactly when the deadline approaches. The customer who files on the first day is reviewed by a team with spare capacity. The customer who files in the last week is in a queue whose length was created by everyone who reasoned that they had until 26 August.

This is the general shape of every announced shutdown, and it is why the instinct to use the notice period is wrong. A notice period is the vendor's operational plan for ceasing business. It is not a guarantee of your access, and it is not an entitlement you should consume in full. The disclosure that requests may be delayed is not fine print; it is the vendor telling you that the stated date and the achievable date are different numbers.

The rule this leaves you with

Generalise it away from crypto, because the mechanism has nothing to do with the asset class. Any supplier that announces a wind-down, a product sunset, a regional exit or a platform migration is handing you a date and an implicit queue. The rule that follows is short: when a vendor announces it is stopping, begin your exit in the first week, and treat the published end date as the moment access is gone rather than the moment to start acting. The cost of moving early is inconvenience. The cost of moving late is being in a support queue that no longer has anyone incentivised to clear it.

The second rule concerns what you monitor. This vendor was busy and closing at the same time, which means volume, uptime and transaction counts are not continuity indicators. The signals that carried information here were a collapsing token price and a peer closure days earlier, BitMEX having announced its own shutdown on 23 July. Two venues in one week is a sector signal, and sector signals reach you before vendor-specific ones do. Supplier reviews built only on service metrics will always find out last.

For European businesses there is an existing discipline to attach this to rather than a new process to invent. Financial entities already have to document exit strategies for critical technology providers under DORA, and the reason that requirement exists is precisely this failure mode: an orderly supplier exit that is orderly for the supplier. Whether or not you are in scope, the artefact is worth having. For each critical vendor, one page naming where the data lives, how it comes out, who executes the move and how many days it takes. Written while nothing is wrong, it converts a scramble into a task.

Frequently asked questions

Is this only relevant to companies holding crypto assets?

No. The asset class is incidental to the mechanism. The same sequence appears when a software vendor sunsets a product, when a cloud provider exits a region, when a payment processor drops a merchant category or when a supplier enters administration. In each case there is a stated end date, a shorter effective date, and a support queue whose capacity is fixed while demand for it spikes. If you hold anything at a third party that you would need time and their cooperation to retrieve, the rule applies to you.

How early is early enough when a vendor announces a shutdown?

Within the first week, and preferably the first days, for anything requiring the vendor's active cooperation to release: funds, data exports, licence transfers, custody arrangements. The distinguishing question is whether retrieval needs a human on their side to review or approve something. If it does, you are in a queue and your position in it is decided by when you file. If it does not, and the export is fully self-service, you have more latitude, though vendors in wind-down also reduce engineering support for tooling that breaks.

A calendar that reads six months and an access window that runs 31 days are two different facts about the same notice, and the second is the one that decides whether your assets come back. The company published the reason it will slip, in its own warning about withdrawal scrutiny and volume surges. Whenever a supplier tells you it is stopping, start that week. The queue is shortest on the day the announcement lands, and it never gets shorter again.

BitMartBitMEXvendor wind-downsupplier riskexit planningDORAbusiness continuitycrypto exchanges

16,000 Firms Pooled Their Supplier Risk. You Didn't

Risk Ledger raised 24 million pounds in Series B on 15 July 2026 for a network-first supplier-risk platform used by 16,000 organisations. Why NIS2 and DORA make the questionnaire model your problem, and what European owners should change.

4 min read

Law & Regulation

Four US Clouds Now Report to a UK Regulator

From 13 July 2026 the Bank of England, PRA and FCA directly oversee the resilience of AWS, Microsoft Azure, Google Cloud and Oracle. The liability for an outage still sits with the firm, not the regulator.

4 min read

Law & Regulation

The EA Buyout Now Hinges on Brussels

The largest take-private in history has cleared EA shareholders and US antitrust. The last real gate is the EU - and not classic merger review, but the Foreign Subsidies Regulation, facing its first blockbuster test.

3 min read

← All articles