Trailing Stops: The Complete Guide (And Why Your Exit Rule Is a Market Bet)

This site has fifty-four articles on risk management and not one on the trailing stop. That is an accurate reflection of the industry, and it is embarrassing, because your trailing rule shapes your results more than your entry does.

Here is the proof, and then the guide.

One entry, five exits, fifty thousand paths

Take a single trade. Enter, place the stop two percent below, and call that distance 1R. Now vary nothing except what you do afterwards.

Run it through a trending market.

Fourteen times the expectancy, from the same entry, purely by changing what you do after the fill. And look at the win rate column: the best rule wins one trade in five.

Now run the identical five rules through a mean-reverting market. No drift, and a gentle pull back toward your entry.

The five rules, and what each one actually buys

Fixed target

Exit at a pre-set multiple of R. Highest win rate, zero variance in the win size, and a hard ceiling on the upside. Notice from the second table that it is the only rule that survives a mean-reverting market. It is also the rule that would have destroyed the trend trader in the first table, capping a 13R average winner at 2R.

Buys: certainty and psychological ease. Costs: the tail.

Breakeven stop

Move the stop to entry once the trade is 1R in profit. Universally recommended, rarely examined. It converts your loss distribution from “sometimes −1R” to “sometimes 0R,” which feels wonderful and does something quite specific to your statistics: it collapses the win rate, because a large fraction of trades that would have recovered are now scratched.

In a trend that is a magnificent trade. In a range it is fatal, because in a range almost every trade returns to entry, and you have installed a rule that guarantees you exit there.

Buys: elimination of the loss. Costs: the trades that needed room.

Structural trail

Stop rides beneath the most recent swing low. The market defines the level, not your arithmetic. This is Darvas’s box, seventy years old, and it remains the most honest trailing rule available because it exits only when the structure that justified the trade has broken.

Buys: alignment with your thesis. Costs: variable stop distance, so position size must be computed per trade.

Volatility trail (ATR, chandelier)

Stop sits a fixed number of ATRs beneath the highest close since entry. Adapts automatically to the instrument and the conditions. The Turtles used it, and it is the correct default for anyone trading more than one market.

Buys: comparability across instruments. Costs: it ignores structure entirely, and will happily sit you through a broken thesis inside a wide band.

Time trail

Exit after N bars, or at the session close, regardless of price. Almost nobody uses it, and it is the only rule on this list that addresses the cost of a thesis that has neither worked nor failed. A position that is going nowhere is a position you are financing.

Buys: a bound on carry and attention. Costs: it exits winners for no reason at all.

The trade-off nobody names

There is no setting that improves both columns. Every trailing stop is a point on this curve, and every trader who “optimises” his trail is really just walking along it in the direction of whichever feeling he finds least tolerable.

Choosing, properly

Four ways traders ruin a good trail

Moving it in. Tightening the trail because the profit is large is not risk management. It is the reflection effect, and it is the mechanism by which a 13R winner becomes a 2R winner.

Breakeven too early. A breakeven stop at 0.5R, in a market with any noise at all, is a rule that scratches your winners and keeps your losers. It converts the payoff distribution and it does not tell you.

Trailing on the wrong timeframe. A stop trailing the five-minute swing lows in a daily-timeframe trade is not a trailing stop. It is a scalping exit attached to a swing entry.

Mixing trail and scale-out. Scaling out reduces the size that the trail is protecting. Do both aggressively and you have engineered a system in which the tail, when it finally arrives, arrives on a quarter position.

The honest summary

A trailing stop does not protect profits. It defines what a profit is, and in doing so it decides which trader you are.

The fixed target makes you a trader with a high win rate, a smooth curve, and no tail. The wide trail makes you a trader who loses most of the time and is paid for the year in a handful of trades. Neither is superior. Each is correct in exactly one market, and the market is not obliged to tell you which it is until afterwards.

What you can do is answer the regime question first, write the rule down, and then leave it alone, because the version of you holding the position is the last person who should be adjusting it.