What a stop limit order is

Two prices on one ticket 

A stop order carries one price and one instruction: when the market reaches this level, transact at whatever price is available. A stop limit carries two prices, and the instruction splits in half at the moment the first is reached. The first price is the trigger. The second is the limit. When the market touches the trigger, nothing is bought and nothing is sold. A limit order appears at the limit price instead and begins to rest, subject to the rule that governs every limit order: it transacts at its price or at a better one, and never at a worse one. The type is a stop order in its first half and a limit order in its second, which is where the name comes from, and why it behaves like neither of them taken alone.

The two prices are set independently, and the distance between them is the whole design of the order. On a buy stop limit the trigger sits above the prevailing market, because a buy stop wakes on a rise; on a sell stop limit it sits below. The limit, in both cases, sits wherever the party writing the ticket puts it. Nothing in the mechanism forces it beyond the trigger in the direction of travel and nothing forces it short of the trigger either, and where it sits determines every behaviour described in the rest of this lesson.

What happens at the trigger 

The order therefore has two events in its life rather than one, separated by however long the market takes to move between them. The first is the trigger being touched, which converts a dormant instruction into a live limit order. The second is that limit order transacting. Between them the market carries on moving, and it can move past the limit price before the resting order has met anything at all. When that happens the order does not chase. It stays where it was written and remains unfilled until the price comes back to it, until it is cancelled, or until its expiry passes.

The two gapping rows are the ones worth rereading. In each, the trigger did its job perfectly: the level was touched and the limit order was created exactly as written. The market then traded straight past it without one price at or inside the limit ever being available, so both tickets end the sequence with a live resting order and no transaction.

Price certainty against execution certainty 

Every order type in this module trades one of two certainties away for the other, and the stop limit makes that trade the most visible. A stop order surrenders price certainty: it transacts, and the price it transacts at is whatever the market offers at that instant, which in a fast market can sit a long way from the trigger. A stop limit surrenders execution certainty instead: the price is bounded by construction, and the cost of the bound is that no transaction may occur at all. Neither certainty is available at the same time as the other, on any venue and in any market, because the two are the same fact stated from opposite ends.

What an unfilled order costs depends entirely on what the order was for. An entry order that does not fill leaves no position, which is a neutral outcome: nothing was risked and nothing was gained. An exit order that does not fill is a different matter, because the position it was written to close is still open, and open in exactly the conditions that made the limit unreachable.

Read across the block, neither version is the good one. In the stop order rows the position is closed, at a price worse than the level written on the ticket. In the stop limit rows the closing price never had the chance to be worse, because there was no closing price, and an open position carries whatever the market does next. The two do not differ in how well they work. They differ in which of those two situations results.

The distance between the two prices 

Because the limit is set independently of the trigger, the gap between them is the one dial the type offers, and execution literature describes its two ends in opposite terms. A wide gap, with the limit set well beyond the trigger in the direction the order travels, makes a transaction very likely, since almost any price reached after the trigger falls inside the limit. It also gives back most of the price certainty the second number was written to buy, so at its extreme a wide gap is a stop order with extra steps. A narrow gap holds the price tightly and raises the frequency with which nothing transacts at all, and setting the limit at the trigger itself is the tightest version of that.

Execution traditions also disagree over what unit the gap is stated in. One convention states it in price, on the reasoning that a fixed number of increments is the only thing a ticket can actually pin down. Another states it as a multiple of the prevailing spread, on the reasoning that the distance to be crossed is widest in precisely the conditions that fire the trigger, so a fixed distance is arbitrary at the moment it matters. The second answers a real objection by giving up the fixed bound that was the point of the type, so neither convention settles the other, and no setting of the gap removes the possibility that nothing transacts.

On the ticket 

Both platforms YAL runs, MetaTrader 5, carry the type as a pending order, and both ask for the same two numbers under different names. The shared requirement is a symbol, a volume, a trigger and a limit, and everything else on either ticket is optional.

• On the MetaTrader ticket the pending order types are named Buy Stop Limit and Sell Stop Limit. The field labelled Price is the trigger; the field labelled Stop Limit price is the level at which the limit order is placed once that trigger is touched. Volume, an optional expiry, and an optional stop loss and take profit belonging to the resulting position complete the ticket.
• The arithmetic is identical to naming a second level, and the field reads as a tolerance rather than as a price.
• On both, a stop limit sits in the pending orders list rather than on the position itself. The stop loss field attached to an open position is a stop order in either platform, so an exit expressed as a stop limit is written as a separate pending order rather than typed into that field.

Where practitioners disagree 

The live argument is whether the type belongs on an exit at all. One tradition holds that the whole purpose of a protective exit is that it transacts, and that a stop limit withdraws that property in the one condition it is being asked to handle, so the type belongs on entries and nowhere else. The opposing tradition answers that a gapping market can transact a stop order at a price no party would have accepted if asked, and that between a transaction at an unacceptable price and no transaction at all, the second at least leaves the outcome undecided rather than settled badly. Both describe the same event correctly. They disagree about which of two bad outcomes is worse, and that turns on the market and the circumstances, so it has no general answer.

A second disagreement concerns how often the unfilled case actually arises. One camp notes that gaps large enough to skip a limit are rare in deep, continuously traded markets, and infers that the type costs little in practice. The other notes that the rare event and the event the exit was written for are the same event, so measuring the type by its behaviour on ordinary days measures it on the days it was not needed. Both arguments are correct and they answer different questions, which is why the disagreement persists rather than resolving.