Learn · 5 min read

Trailing stops: protecting profit without strangling the trade

ATR-based vs R-step trailing, break-even at TP1, and why tightening the trail cost us money.

Two ways to trail

ATR trailing (also called a Chandelier stop) keeps the stop a multiple of current volatility behind the highest high since entry. It automatically widens in fast markets and tightens in quiet ones.

R-step trailing moves the stop in discrete jumps: at +1R the stop goes to break-even, at +2R it goes to +1R, and so on. It is cruder but completely predictable, and it is easy to replicate manually or in an EA.

Our default and why it changed twice

We currently run ATR trailing at 1.5× activating at 1R, with break-even at the first take-profit. We arrived there by running an optimiser sweep across both modes and multiple multipliers on the same history.

A 1.5× trail combined with loosened entry gates once produced a stretch of losses: the stop was tight enough that ordinary gold noise closed trades that later ran to target. A backtest sweep then showed half that distance — 1.0× ATR plus a 12-pip buffer — cut the worst drawdown while keeping the trade count, so that became the default. Trailing settings and entry settings are not independent.

Break-even is not free

Moving to break-even feels safe, but on gold it converts a meaningful share of eventual winners into scratches. We only do it after the first take-profit has already banked 50% of the position, so the trade is already net positive before the stop tightens.