Roundup: Where to put the stop, MAE, ATR and time stops from across the board
Posted: Thu Oct 15, 2026 4:34 pm
Roundup: Where to put the stop, MAE, ATR and time stops from across the board
Stop placement threads on this board come at it from different angles: data with MAE, volatility with ATR, time stops for trades that go nowhere, and the metals threads where stops just don't behave. Here's the combined picture.
Stop first, then entry
The stop-first thread makes a simple process change. Decide where the idea is wrong before looking for the entry, not after. Otherwise the stop ends up wherever is convenient: too close to flatter the reward to risk, or too far because you don't want to be wrong. Once the stop comes first, size becomes what adjusts. A reply adds the underrated part: if the stop has to sit so far out that the target stops making sense, the honest call is no trade.
Let your own trades set the distance
The MAE thread suggests recording how far each trade went against you before closing, then plotting MAE for winners only. When nearly all winners stayed within 4 pips of entry, a 10 pip stop is paying for room you don't use. When plenty of went 6 or 7 pips against, a 5 pip stop is cutting good trades. It's the most direct evidence you'll get, and it needs a decent sample.
ATR, with a scalping warning
The ATR thread argues against fixed stops like "always 20 pips", because the same distance is fine in quiet conditions and far too tight in volatile ones. The reply worth reading points out a trap for scalpers: right at the London open an M5 ATR(14) still reflects roughly seventy minutes of quiet Asian candles, so at 07:55 it can read 3 pips just before the market starts moving 10. ATR on low timeframes lags the session change.
The gold session ATR thread uses a different approach. Measure the average London range on XAUUSD over 10 to 20 days and compare the planned stop with it. If a setup needs more than about a quarter of London ATR, it's closer to a swing than a scalp.
Time stops
Two threads here. The time stops thread suggests checking how long your winners usually take to reach 1R and using that as the window. A reply warns not to use time stops on setups meant to be slow. The EURUSD time vs price stop thread splits the jobs cleanly: the price stop is for invalidation and never moves further away, the time stop is for ideas that go nowhere. If a continuation trade hasn't moved at least 0.5R in its favour within the expected time, it's out.
Metals
Silver gets a thread of its own because it moves in jumps. Structure stops that make sense on EURUSD get taken out by normal silver noise, and the answer in that thread was sizing down rather than tightening up. Same R, wider stop, smaller lot.
The slipped gold stop thread covers what happens when it goes wrong anyway. Flatten related dollar risk first, then write down the facts while fresh: server time, stop level, fill, spread. One bad fill is enough for the moment.
Where they differ
The MAE and ATR approaches can give different answers. MAE comes from your past trades, ATR from the current market. Using MAE for the baseline and ATR to adjust on unusual days seems like a reasonable compromise, but nobody in these threads tested that directly.
Do you set stops from structure, ATR or your own MAE data? Has anyone compared them on the same set of trades?
Threads this pulls from
Stop placement threads on this board come at it from different angles: data with MAE, volatility with ATR, time stops for trades that go nowhere, and the metals threads where stops just don't behave. Here's the combined picture.
Stop first, then entry
The stop-first thread makes a simple process change. Decide where the idea is wrong before looking for the entry, not after. Otherwise the stop ends up wherever is convenient: too close to flatter the reward to risk, or too far because you don't want to be wrong. Once the stop comes first, size becomes what adjusts. A reply adds the underrated part: if the stop has to sit so far out that the target stops making sense, the honest call is no trade.
Let your own trades set the distance
The MAE thread suggests recording how far each trade went against you before closing, then plotting MAE for winners only. When nearly all winners stayed within 4 pips of entry, a 10 pip stop is paying for room you don't use. When plenty of went 6 or 7 pips against, a 5 pip stop is cutting good trades. It's the most direct evidence you'll get, and it needs a decent sample.
ATR, with a scalping warning
The ATR thread argues against fixed stops like "always 20 pips", because the same distance is fine in quiet conditions and far too tight in volatile ones. The reply worth reading points out a trap for scalpers: right at the London open an M5 ATR(14) still reflects roughly seventy minutes of quiet Asian candles, so at 07:55 it can read 3 pips just before the market starts moving 10. ATR on low timeframes lags the session change.
The gold session ATR thread uses a different approach. Measure the average London range on XAUUSD over 10 to 20 days and compare the planned stop with it. If a setup needs more than about a quarter of London ATR, it's closer to a swing than a scalp.
Time stops
Two threads here. The time stops thread suggests checking how long your winners usually take to reach 1R and using that as the window. A reply warns not to use time stops on setups meant to be slow. The EURUSD time vs price stop thread splits the jobs cleanly: the price stop is for invalidation and never moves further away, the time stop is for ideas that go nowhere. If a continuation trade hasn't moved at least 0.5R in its favour within the expected time, it's out.
Metals
Silver gets a thread of its own because it moves in jumps. Structure stops that make sense on EURUSD get taken out by normal silver noise, and the answer in that thread was sizing down rather than tightening up. Same R, wider stop, smaller lot.
The slipped gold stop thread covers what happens when it goes wrong anyway. Flatten related dollar risk first, then write down the facts while fresh: server time, stop level, fill, spread. One bad fill is enough for the moment.
Where they differ
The MAE and ATR approaches can give different answers. MAE comes from your past trades, ATR from the current market. Using MAE for the baseline and ATR to adjust on unusual days seems like a reasonable compromise, but nobody in these threads tested that directly.
Do you set stops from structure, ATR or your own MAE data? Has anyone compared them on the same set of trades?
Threads this pulls from
- Maximum Adverse Excursion: Using MAE to Set Better Stops
- Time Stops, Exiting a Trade That Goes Nowhere
- Time stop vs price stop on EURUSD scalps
- Deciding my stop before I look at the chart for the entry, not after
- ATR, The Tool That Fixes Your Stop Loss Problem
- XAUUSD session ATR bands for stop placement discussion
- XAGUSD stop placement: how I handle the width
- After a slipped gold stop: what I do next