A trailing stop follows price at a fixed distance, locking in gains as the trade moves. For scalpers, the main trouble is that the trailing distance has to be wider than normal noise, or the trade gets stopped out by ordinary wiggles.
If typical one-minute noise is 2 pips, a 2 pip trail will get tagged constantly. A 4 pip trail survives, but on a 6 pip target, it surrenders a lot of profit.
Alternatives exist. Structure-based trailing places the stop beneath the most recent swing low in a long trade, updating as new swings form. ATR-based trailing uses a multiple of average range. Both adapt better than a fixed distance.
Another issue is platform behaviour. Some trailing stops run on the trader's computer and stop working if the connection drops. Server-side trailing, where supported, continues regardless.
Which trailing method has been tested, and does it run on the platform or the server?
Trailing stops and the noise problem
-
FXS London Scalper
- Posts: 978
- Joined: Sun Oct 18, 2026 12:25 pm
Re: Trailing stops and the noise problem
Two pips of one-minute noise with a 2 pip trail is a guaranteed early exit. Using an ATR-based trail, perhaps 1.5 times the M1 ATR, gives the trade space to breathe without giving back too much.
FXS Official Team | London Session Desk
-
Amanhasnoname
- Posts: 2060
- Joined: Fri Oct 30, 2026 9:45 am
Re: Trailing stops and the noise problem
I spent a while testing fixed trails on EURJPY and gave up on them fairly quickly. Anything tight enough to protect a scalp got clipped by normal wobble, and anything wide enough to survive handed back most of the move. On my numbers the fixed trail was worse than simply using a fixed target, which surprised me at first.
What has worked better for me is a two-stage approach. I leave the original stop alone until price has moved about one full initial risk in my favour, then move it to break even. After that I trail behind the last completed one-minute swing, only updating when a new swing has clearly formed. It looks clumsy on the chart, but it lets the occasional runner pay for several small losers.
On the server question, I learned that one the hard way. My platform's trailing stop was client side, and a router restart during a fast move left me with no trailing at all for several minutes. Now I move the stop manually as a real order on the server instead.
What has worked better for me is a two-stage approach. I leave the original stop alone until price has moved about one full initial risk in my favour, then move it to break even. After that I trail behind the last completed one-minute swing, only updating when a new swing has clearly formed. It looks clumsy on the chart, but it lets the occasional runner pay for several small losers.
On the server question, I learned that one the hard way. My platform's trailing stop was client side, and a router restart during a fast move left me with no trailing at all for several minutes. Now I move the stop manually as a real order on the server instead.