What a stop out is

The moment it happens 

Consider an account holding two open positions with prices moving against both. Most of the panel does not change shape while that happens. Balance sits where it was, because nothing has been closed. Used margin sits where it was, because it was fixed when each position opened and does not respond to price. Only the unrealised result moves, and since equity is balance plus that result, equity falls with it. Margin level, being equity measured against used margin, falls in step. Somewhere in that descent a margin call level is passed and the platform reports it. Prices carry on. Then, at a lower percentage, the reporting ends and something categorically different begins: the platform closes one of the positions itself.

The structural fact underneath that is where the check runs. It is performed by the firm's server, which is where the positions actually live, and not by the terminal on a desk or a phone. It does not depend on a terminal being open, on a person being awake, or on anything at all being acknowledged. A margin call is a notification. A close out is an action, and the account is its object rather than its author.

The level, and the arithmetic that reaches it 

The stop out level is a percentage, and the quantity it is compared against is the margin level built two lessons ago: equity divided by used margin. Nothing new is calculated at the moment of a close out. The same comparison that produced a comfortable reading when the positions were opened produces the trigger once it has fallen far enough, and the arithmetic between those two states is continuous rather than stepped. The level is a published term of the account rather than a market variable, so it is known before any position is opened. The stop out level at YAL is 50%. The worked blocks below use a different figure, chosen only to keep their arithmetic legible, and each states that it is an assumption rather than a term.

Two features of that block carry the whole mechanism. Used margin does not respond to price, so the denominator of the comparison is fixed while the positions are open, and everything that moves the margin level moves it through equity. And the same instrument, the same size and a move of the same distance produce a reading far above the level in one direction and a reading exactly at it in the other.

The order positions are closed in 

Describing a close out as the account being wiped out overstates it, because it is usually not a single event. What runs is a loop. One position is closed, the figures are recalculated on what remains, and the comparison is made again. If the recalculated margin level is back above the level, the loop ends and the remaining positions stay open. If it is not, the next position is closed and the check repeats. An account holding several positions can come out of a close out with some of them intact, and which ones survive is settled by the order the loop worked in.

1. The check runs, and the margin level is at or below the stop out level.
2. One open position is selected, by whichever rule the firm has published for that account.
3. That position is closed at the price the market is showing when the closing order reaches it.
4. Its used margin is released, and its unrealised result becomes a realised one that lands in the balance.
5. Equity and used margin are recalculated, and the margin level is taken again on the new figures.
6. If the recalculated level is above the stop out level the loop ends. If it is not, the loop returns to the second step.

The selection rule in the second step is where firms genuinely differ, and the differences are structural rather than cosmetic. One convention closes the position carrying the largest unrealised loss first, on the reasoning that it is draining equity fastest. Another closes the position holding the largest used margin first, on the reasoning that it releases the most collateral per closure. A third works in the order the positions were opened. Each carries an obvious criticism: the first crystallises the worst position at its worst recorded moment, the second can close a position that is currently in profit and is not the source of the problem, and the third is indifferent to both size and result. None of the three is standard across the industry, which is why the rule is published as a term of the account rather than reasoned out while the event is running.

Why it happens without any further action 

A close out needs no involvement because collateral, not permission, is what keeps a position open. A position exists on the condition that an amount of the account is held against it, and it continues while the account can support it. When the supporting arithmetic fails its own test, that condition has stopped being met, and closing the position is the mechanical consequence rather than a decision anyone takes at the time.

The practical reach of that is easy to underestimate. Markets move outside the hours anybody is watching them. A position held across a weekend meets the first price of the following week at whatever level that price opens, and the check runs against it before any terminal has reconnected. A device switched off, a dropped connection and an alert nobody saw change nothing about the sequence. The margin call that typically precedes a close out reports a state that has been reached, and it is not a gate the process has to pass through.

The reason the loss can run that far is that it is calculated on the full size of each position while only a percentage of that size was ever posted against it. The collateral sets the level at which positions become liable to be closed. It does not bound what the arithmetic can produce, and losses are not limited to the amount deposited.

Why the realised loss can be larger than the level implies 

The level and the price are two different things, and conflating them is the most common misreading of this mechanism. The level is a condition on a ratio: it settles when the closing begins. It says nothing about the price the closing gets, because that price is set by the market at the instant the order arrives in it. In a continuously quoted, liquid market the distance between the two is usually small. In a market that has stopped trading and reopened somewhere else, no price exists between the last one before the break and the first one after it, so the equity reading that defined the trigger may never have corresponded to a tradeable price at all. The close out then runs at the first price available, which can sit a long way past it.

The adverse row is why a close out level cannot be read as a floor underneath a loss. The account passed through the trigger without any trade being possible at it, and what is recorded is the result at the first price that existed, not the result at the level. A close out running across several positions carries a second version of the same exposure: each closure takes the price available when its turn comes up.

Where a close out completes far enough past the trigger, the arithmetic can finish below zero. The realised losses exceed the balance that was there to absorb them, and the account records a negative figure, which is an amount owed to the firm. What happens to such a balance afterwards is a matter of the account documentation and the client agreement, which this page does not state.

Where practitioners disagree 

Whether a close out level is best described as a protection or as a hazard is genuinely unsettled, and both descriptions are stated with equal confidence. One tradition treats it as a backstop: an exposure that collateral no longer supports is ended mechanically, and the alternative is an open position with nothing behind it. Another points out that it forces a closure at the least favourable point of an adverse move, at a price nobody chose, and that a position closed by the loop stays closed regardless of what the price does an hour later. Both are accurate about different sequences of prices, which is why the argument persists rather than resolving.

The second disagreement concerns what reaching the level is evidence of. One view treats it as an ordinary mechanical feature of a margined account, no more remarkable than the margin call before it. Another treats it as a fact about position size rather than about the market: the size opened embodied an assumption about how far a price could travel before its collateral ran out, and a close out is that assumption failing. Neither view is a rule, and this page puts forward no figure for the relationship between position size and equity.