Crypto Exchange Outage: What Happens to Your Open Position in 2026
A crypto exchange outage is the one market risk you cannot trade your way out of. The chart keeps moving somewhere, your position keeps existing, and the button that would close it does nothing. For a few hours you are not a trader. You are a spectator holding size.
Most traders never plan for this, because it feels like an infrastructure problem rather than a trading problem. Then it happens on a Thursday evening, and the only decision that mattered was one they made before the screen froze: how big the position was.
In this guide we will walk through what actually happens during a crypto exchange outage, what it does and does not do to an open position, how the picture changes inside a simulated funded account, and the small set of habits that turn an outage from a disaster into an inconvenience.
Key takeaways
- Assume any position can be frozen. A crypto exchange outage removes your ability to exit, so size is the only lever you still control once it starts.
- Know the four stages. Degraded, dark, cancel only, restored. Each stage allows different actions, and confusing them wastes the window you do have.
- Understand where your stop actually lives. A stop resting at the venue behaves differently from a stop your platform holds locally, and both behave differently again when the venue is down.
- Separate live-market mechanics from simulated ones. In a simulated funded account there is no real counterparty and no real liquidation engine, so the risk is data and platform behavior rather than forced closure.
- Skip the reopen. The first prints after a halt are thin, wide and unrepresentative. There is no rule that says you have to trade them.
What this guide covers
What actually happens during a crypto exchange outage
A crypto exchange outage is a failure somewhere in the venue’s own stack, usually in the matching engine, the API layer or the infrastructure underneath both, that stops the exchange from accepting or processing orders. It is not a market halt. Price is still being discovered elsewhere. You simply lose access to the venue where your position lives.
That distinction matters. In a regulated futures market a halt is a deliberate, announced pause with published rules for how the market reopens. An exchange outage is an unplanned failure, and the venue writes the recovery script as it goes.
The three failure modes
Outages are not all the same, and the type determines what you can still do.
Connectivity failure. The site or app will not load, but the matching engine underneath is alive. Orders placed through an API or a third-party platform may still work. This is the least damaging kind, and also the easiest to panic through.
Matching engine failure. The order book itself stops. Nothing fills, nothing cancels, nothing moves. Your position is frozen exactly as it was, including any stop you had resting at the venue, which is now inert.
Infrastructure failure. Something underneath the exchange breaks and takes multiple systems with it. These are the long ones. Coinbase published a detailed postmortem of its May 7, 2026 outage, in which cooling units failed in a single data hall at a cloud provider and trading, deposits and withdrawals were unavailable or degraded for roughly eight hours, with full recovery taking about twelve hours more.
Why cancel-only mode exists
When a venue comes back, it usually does not open straight into full trading. It opens in cancel-only mode first, where resting orders can be pulled but new ones cannot be added. This is deliberate. It lets everyone clear out stale orders that were placed against a price that no longer exists, before the book reopens for real.
Cancel-only is a gift, and most retail traders sleep through it. If you had a limit order sitting three percent below the market when the venue went dark, and the market moved five percent while it was down, that order is now a trade you never intended to make. Cancel-only mode is the window to remove it.
Your open position while the venue is dark
While a venue is dark, an open position does not close, does not move and does not stop accruing exposure to the price that is being discovered everywhere else. When the venue reopens, your position is marked against a price that may be a long way from where you last saw it.
This is the part that traders underestimate. Nothing bad happens to the position while the screen is frozen. Everything bad happens in the two seconds after it unfreezes.
Spot, perpetuals and the liquidation question
On spot, a frozen position is unpleasant but bounded. You own the asset, and the worst case is that you own it at a worse price when trading resumes.
On leveraged perpetual futures the question is harder, because a liquidation engine may or may not still be running while the customer-facing systems are down. Some venues suspend liquidations during an outage. Others do not, on the argument that the risk engine has to keep the book solvent. Neither answer is universal, and the only way to know is to read the specific venue’s documentation before you need it, not after.
The CFTC has been blunt about this general category of risk for years. Its customer advisory on the risks of virtual currency trading notes that cash markets for virtual currencies are largely unregulated and that platforms may lack the system safeguards and customer protections found in regulated markets. Uptime is one of those safeguards.
Where your stop actually lives
Traders talk about stops as if they are all the same thing. They are not, and an outage is where the difference shows up.
A stop resting at the venue is an instruction sitting in the exchange’s order book. If the matching engine is down, that instruction is not being evaluated. It does not fire.
A stop held locally by your platform or your broker is an instruction that lives on someone else’s server and triggers an order when the condition is met. If the venue cannot accept that order, the trigger fires into nothing.
A mental stop is not a stop. It is a plan, and during an outage it is a plan you cannot execute.
The four stages of a crypto exchange outage
Most venue failures move through the same sequence. Knowing which stage you are in tells you what you can still do with an open position, and what you cannot.
Degraded
Orders are slow or rejected. Charts still print. Easy to mistake for your own connection.
Dark
The matching engine is down. No new orders, no cancels, no exits. Positions sit exactly as they were.
Cancel only
Resting orders can be pulled but not added. This is the window to clean up stale working orders.
Restored
Trading reopens, often into a repriced market. The first prints can be wide.
Trading, deposits and withdrawals unavailable or degraded, per the exchange’s own published postmortem. Full system recovery took roughly twelve hours more.
Crypto has no closing bell, so there is no session boundary that naturally flattens you before a venue fails. Every hour is an hour you can be holding.
Outside your control
- Whether the venue accepts your exit
- Where price reopens after the halt
- How long cancel-only mode lasts
- Whether a liquidation engine keeps running
Inside your control
- Position size before the outage starts
- Whether a stop is resting at the venue
- How many venues you are exposed to
- Whether you re-enter into the reopen
Size for the freezeAssume any position could be locked in place for hours, and pick a size you can sit with.
Verify, then waitCheck the venue status page before you assume the problem is yours. Do not spam orders into a dead engine.
Skip the reopenThe first minutes back are thin and wide. Sitting out costs nothing.
tradefundrr.comIllustrative example only. Simulated trading environment. Figures describe published incidents and general venue behavior, not a projection of any result.
How a crypto exchange outage works in a simulated funded account
Inside a simulated funded account, a crypto exchange outage does not hit you the way it hits a live venue account, because no real order is being sent to a real exchange and no real position exists to be liquidated. What can still hit you is the data feed and the platform, and that is a different problem with different consequences.
This is worth stating plainly rather than glossing over, because a lot of prop firm content quietly implies that everything that happens in the live market happens identically in a simulation. It does not.
What is genuinely different
There is no counterparty holding your position, no real margin balance and no exchange risk engine that can force-close you because its own book needs to stay solvent. The simulated account is a rules environment running against market data.
What that removes is the worst-case scenario of the live market: being liquidated by a machine you cannot reach while the venue you would use to defend the position is down.
What is genuinely the same
Market data comes from somewhere. If the venue or the data provider feeding your simulated account has a problem, the price on your chart can stall, gap or print badly. Your platform may reject orders. Your open simulated position still carries an unrealized number that moves when the feed comes back.
And the account rules keep applying. A daily loss limit and a maximum drawdown do not pause because a feed hiccuped. If the reopen moves against an oversized simulated position, the rule consequence is real even though the position was not.
| What happens during an outage | Live venue account | Simulated funded account |
|---|---|---|
| Position can be force-closed by an exchange risk engine | Possible, depends on the venue | No, there is no real counterparty |
| Order entry can fail | Yes | Yes, platform and feed dependent |
| Chart data can stall or gap | Yes | Yes |
| Resting stop may not evaluate | Yes | Yes, depends where the stop is held |
| Withdrawals or transfers blocked | Yes, during the incident | Not applicable, payouts follow the program schedule |
| Account rules keep applying | Broker terms apply | Yes, loss limits and drawdown still count |
A simulated environment removes the forced-closure risk. It does not remove the data risk, the platform risk or the rule consequences of a bad reopen.
Why the skill still transfers
This is the honest case for practicing outage discipline in a simulation. The habits that protect you are not exotic. They are position sizing, knowing where your stop lives, and refusing to trade a chaotic reopen. Those are cheap to build when the position is simulated and expensive to build when it is not.
A trader who has already sat through three frozen platforms in a funded evaluation reacts very differently the first time it happens with their own money on the line. That is the point of the environment.
Building an outage plan before you need one
An outage plan is short. It has to be, because you will be executing it while mildly panicking. Three decisions made in advance cover most of the damage.
Decide your maximum frozen size
Ask a specific question: if this position could not be closed for eight hours, would I still be comfortable? Not "would I be happy", just "would I still be functional". Whatever size passes that test is your ceiling for holding through low-liquidity hours, weekends and anything else where a venue failure would be most awkward.
For a funded trader this maps directly onto the account’s published limits. A per-position risk cap and a daily loss limit already define an outer boundary. The outage question just asks whether you are comfortable sitting at that boundary without an exit.
Know the status page before the incident
Every serious venue publishes a status page. Bookmark it. During the first two minutes of a suspected outage, the only useful action is establishing whether the problem is you or them, because the answers diverge completely. If it is your connection, switching networks might restore access. If it is the venue, nothing you do at your desk matters and hammering the order button just queues up surprises.
- You know which venue or data provider your platform depends on
- You have the venue status page saved and can reach it from your phone
- You know whether your stops rest at the venue or are held by the platform
- Your position size passes the "frozen for eight hours" test
- You have a written rule about not trading the first minutes of a reopen
- You know your account’s daily loss limit and drawdown in dollars, not percentages
- You have screenshots or an export of your open positions from before the incident
Write down what you saw
Take a screenshot when things go wrong. Timestamps, open positions, working orders, error messages. If there is any later dispute about what your account looked like at the moment of the failure, contemporaneous records are worth far more than a recollection. This is standard practice for professionals and almost nonexistent among retail traders.
What outages should change about venue selection
Outages should move venue reliability from a footnote to a real selection criterion, roughly on par with fees and liquidity. A venue with tighter spreads and worse uptime is not obviously the better venue, and most traders have never done that comparison honestly.
The uncomfortable part is that reliability is hard to shop for. Exchanges do not advertise their downtime, and the incidents that get reported are the large public ones rather than the many small ones.
Reliability is not a crypto-only problem
It is worth saying clearly, because the crypto industry gets accused of being uniquely fragile: regulated venues fail too. CME Group has had multiple Globex disruptions, including a halt in metals and natural gas futures during a contract expiry window in February 2026 and a broader halt after a data center cooling failure in late 2025. The difference is not that regulated venues never break. It is that they publish rules for what happens when they do.
Concentration is the real exposure
If every position you hold sits at one venue, an outage there is a total outage for you. Traders who spread across two venues are not doubling their operational headache for nothing. They are buying the ability to hedge or reduce elsewhere when one goes dark.
That said, spreading across venues has its own costs: split liquidity, more accounts to secure, more fee schedules to track. It is a trade-off, not a free improvement, and for a trader running a single funded account it may not be available at all. In that case the answer defaults back to size.
Frequently asked questions
What happens to my open position during a crypto exchange outage?
Nothing happens to it during the outage. The position stays exactly as it was, unchanged and unclosable, until the venue restores trading. The risk is concentrated at the reopen, when the position is marked against whatever price the market has moved to in the meantime.
Can I be liquidated while an exchange is down?
Possibly, on leveraged products at a live venue. Some exchanges suspend their liquidation engine during an incident and some keep it running to protect the book. There is no universal rule, so check the specific venue’s documentation before you hold leverage through low-liquidity periods.
Do stop-loss orders work during an exchange outage?
Generally no. A stop resting at the venue cannot be evaluated if the matching engine is down, and a stop held by your platform will trigger an order the venue cannot accept. Treat any stop as unavailable for the duration of an outage.
Can a crypto exchange outage breach my funded account rules?
Indirectly, yes. The outage itself is not a rule breach, but a position that reopens badly can still push a simulated account through its daily loss limit or into its maximum drawdown. Rules keep applying through an incident, so confirm how your own program handles platform disruptions in its written terms.
Does an exchange outage affect a simulated funded account the same way?
No. In a simulated account there is no real order at a real exchange and no exchange risk engine that can force-close you, so the forced-liquidation risk is absent. What remains is data feed disruption, platform order rejection and the ordinary account rules, which continue to apply.
Should I trade the reopen after an outage?
Usually not. The first prints after a halt tend to be thin, wide and unrepresentative of where the market settles minutes later. Waiting for the book to refill costs you nothing but a missed trade, and it removes the single worst execution environment of the day.
How long do crypto exchange outages usually last?
They vary enormously, from a few minutes of degraded API access to most of a day. The Coinbase incident of May 2026 ran roughly eight hours for core trading, with full system recovery taking about twelve hours beyond that. Plan for hours, not minutes.
Does TradeFundrr allow crypto trading through an outage?
TradeFundrr provides a simulated crypto trading environment with published rules covering daily loss limits, maximum drawdown and per-position risk. Platform and data availability depend on the provider, and the account rules apply throughout. Confirm the current written terms of your own account.
The outage is not the risk. The size is.
Every trader who has been badly hurt by a crypto exchange outage was hurt by a position they would not have held if they had thought about it beforehand. The outage was the trigger. The size was the cause.
You cannot control whether a venue stays up. You can control whether an eight hour freeze is an inconvenience or a career event. That decision is made in advance, calmly, in the position sizing box, which is where most of the good decisions in trading get made.
If you want to build that habit somewhere the consequences are educational rather than financial, that is exactly what a structured simulated environment is for. Related reading: crypto slippage and sizing and crypto liquidation cascades.
Trade a structured crypto program with published rules
TradeFundrr runs a simulated environment with clear daily loss limits, drawdown and position rules, so you always know where the boundaries are.
Explore the programs →