Trailing stops: a stop that moves one way only
A trailing stop is a stop order that stores a distance instead of a price. For a long position, the trigger sits one trail distance below the highest price reached since the order went live, and because that high can only rise, the trigger can only rise with it. It follows price while the position is winning and then holds still the moment price turns.
A trailing stop is a stop order whose trigger price is not fixed in place. It is recalculated from the best price the market has reached since the order went live, so it follows the position while the position is winning and freezes as soon as price turns against you.
The trail is a distance, and the distance is measured from a high-water mark
A plain stop stores a price. A trailing stop stores a distance and works the price out from it.
For a long position, the venue or the platform keeps a running record of the highest price seen since the order was accepted. Call that the high-water mark. The trigger sits exactly one trail distance below it. Every time a new high prints, the high-water mark updates and the trigger is recomputed. For a short position the mirror applies: the running record is the lowest price seen, and the trigger sits one trail distance above it.
Notice what is absent from that arithmetic. Your entry price never appears. Once price has moved in your favour at all, the stop has no memory of where you got in. This is why a trailing stop on a position that advanced and then reversed can exit you at a profit, at break-even, or at a loss, with no instruction from you about which of those it should be. The outcome is just the high-water mark minus the trail.
Two mechanical details decide how the record is kept. The first is which price counts: the last trade, the bid, the ask, or the mid. A long position is closed by selling, so the side that matters at trigger time is the bid, and a system that trails the last trade but triggers on the bid will behave slightly differently from one that does both on the bid. The second is how often the record updates: on every tick, or only on the close of a bar. A tick-by-tick trail registers every wick. A bar-close trail ignores anything that does not survive to the close. The vocabulary for these pieces is in the glossary.
Why the stop is built so it can only tighten
The reason trailing stops exist is that a fixed stop protects your entry and nothing else. Price can advance a long way, reverse the whole distance, and take out a stop that was never moved, and the account ends where it started. The alternative is to drag the stop upward by hand, which works but requires you to be present and paying attention at the right moments.
A trailing stop automates that dragging, and the one-way rule is not a politeness. It is a consequence of the definition. The trigger is derived from a maximum, and a maximum cannot decrease. If a trail could loosen as price fell away, it would move down ahead of every decline, stay permanently out of reach, and never trigger at all. A protective order that cannot be reached is not a protective order.
What this does to a position, with made-up numbers
The figures below are invented purely so the sum can be followed.
Suppose a long position is opened at 100, and suppose the trail is set to 5 in the same units as the price. Price advances to 118. The high-water mark is 118, so the trigger is 118 minus 5, which is 113. Price then prints a thin spike to 124 and immediately comes back. The high-water mark is now 124, so the trigger is 124 minus 5, which is 119, and it stays at 119. Price settles back to where it had been trading, around 118. That is below 119, so the stop fires.
The position closed above entry, which sounds like the trail working. Look at what actually triggered it. The exit was decided by a spike that lasted moments, and when the order fired, the trigger was about one unit away from the prevailing price, not the five units intended. The trail distance was never respected against normal trading, because the reference point it is measured from was set by an abnormal print.
A percentage trail inherits the same behaviour. Imagine a 4 percent trail instead: at a high-water mark of 124 the trigger sits at 124 times 0.96, which is 119.04. Almost the same place.
The last step is the fill. A trailing stop is a trigger condition, not a price guarantee. When 119 is touched the order usually converts into a market order and takes whatever the book offers, which after a spike can be thinner than usual. If the sell fills at an imagined 118.6, that 0.4 difference is slippage and it is a normal part of the mechanism, not a malfunction. The variant that caps this is the trailing stop limit, which converts into a limit order instead and therefore accepts the risk of not filling at all. The trade-off is the same one described in market order versus limit order and in what a stop-limit order does.
Three ways the trail distance gets defined
| Trail type | How the distance is set | How it behaves as price advances |
|---|---|---|
| Fixed amount | A constant in points or quote currency | Stays the same width throughout the position |
| Percentage | A share of the high-water mark | Widens in absolute terms as the mark rises on a long |
| Volatility multiple | A multiple of a measured range | Widens in choppy conditions, narrows in quiet ones |
The mistake is choosing the trail from your tolerance instead of the instrument’s noise
The common error is to pick the trail distance by asking how much loss feels acceptable, then setting the trail to that. The number ends up tight, because tolerance is usually smaller than market movement. A trail narrower than the range the instrument routinely oscillates through is not a protective distance. It is a near-certain exit, triggered by ordinary back-and-forth that has nothing to do with the position being wrong.
The ratchet makes this worse rather than better over time. Every spike that touches a new extreme tightens the effective distance between the trigger and where price is actually trading, and it never gives that room back. A trail that was already marginal when placed becomes tighter with every wick.
The order to decide things in is therefore the reverse of what most people do. Start from how far price normally travels against a position without the idea being broken, and set the trail wider than that. Then adjust the position size so that the loss implied by that honest distance is one you can carry. The position size calculator does that arithmetic, and the pip value calculator converts a trail expressed in pips into money. Shrinking the trail to make the loss small does not reduce risk. It converts a sized loss into a stream of small, repeated, noise-driven exits.
One more detail worth confirming before relying on a trail overnight: whether it is tracked at the venue or inside your terminal. A trail recalculated on the venue’s side keeps ratcheting with your machine switched off. A trail simulated locally stops the instant the platform does, leaving either a plain stop at its last position or no stop at all.
Education only, not financial advice, and trading carries a risk of loss.
Key facts
- A trailing stop stores a distance and derives its trigger price from the best price reached since the order was accepted.
- For a long position the trigger sits one trail distance below the running high; for a short it sits one trail distance above the running low.
- The entry price plays no part in the arithmetic, so the trigger can sit above or below entry depending only on where price has travelled.
- The trigger can only tighten, because the high-water mark it is derived from is a maximum and a maximum never decreases.
- A brief spike raises the high-water mark permanently, leaving the trigger closer to normal trading levels than was intended.
- When it fires, a trailing stop usually becomes a market order, so the fill price is not the trigger price.
Frequently asked
Does the trailing stop move back down if price falls?
No, and that is the defining property rather than a quirk. The trigger is calculated from the best price reached since the order was placed, and that best price cannot get worse. A stop that loosened as price retreated would retreat ahead of price forever and could never be reached, which would make it useless as a protective order.
Is a fixed-amount trail better than a percentage trail?
They behave the same at a single price level and diverge as the position runs. A fixed amount keeps the same absolute distance no matter how far price advances. A percentage recalculates against the high-water mark, so the absolute distance widens as price rises on a long, which gives a long runner more room than a fixed trail does.
Does a trailing stop keep working if I close my platform?
Only if the trail is tracked at the venue rather than on your machine. Some order books accept a trailing instruction directly and recalculate it on their own side. Other setups simulate the trail inside the terminal, watching the feed locally and sending a plain order when the condition is met, which means the ratchet stops when the terminal stops.
Why did my trailing stop fire when price had barely moved?
Usually because the high-water mark was set by a short spike rather than by sustained trade. The spike lifts the trigger and never gives it back, so price returning to the level it was trading at before the spike can now be below the trigger. A sudden widening of the spread does the same thing when the trail is tracked against the side you would exit on.
Risk warning: trading is high risk and most retail accounts lose money. Education only, not financial advice.