LEARN · EN · easyquanttrading.com
ATR-based stops: sizing stops to the market instead of to a round number
A twenty-dollar stop on gold means something completely different in a quiet week and in a week with a central bank meeting. The market does not care where you placed your stop, but it does decide how far it will wander before your idea has a chance. ATR exists to measure that wandering.
- ATR measures the average size of a bar's true range over N periods
- A stop at 2x ATR adapts as volatility changes
- Tighter stops mean more small losses; wider stops mean fewer, larger ones
- Volatility-based stops require volatility-based position sizing
What ATR actually is
True range for a bar is the greatest of three distances: the high minus the low, the high minus the previous close, and the previous close minus the low. The extra two terms exist to capture gaps, which a simple high-minus-low ignores.
Average True Range is simply a moving average of that value. If ATR on gold hourly bars is 4 dollars, the typical bar moves about 4 dollars. It is a measure of typical movement, not of direction, and it says nothing about which way price is likely to go.
Converting ATR into a stop distance
The common choice is a multiple of ATR: two times, or three times. The multiple controls how much normal noise the stop can absorb.
| Stop distance | Behaviour | Cost |
|---|---|---|
| Below 1 x ATR | Exits on ordinary noise | High loss frequency, poor win rate |
| 1.5 - 2 x ATR | Survives typical bar-to-bar movement | Moderate; the common default |
| 3 x ATR and above | Survives most noise | Fewer, larger losses; needs smaller size |
The trade-off is unavoidable: tighter stops reduce the size of each loss and increase how often you take one.
The sizing consequence
This is where ATR stops connect to position size, and where people usually go wrong. If your stop distance changes with volatility and your position size does not, your risk per trade changes with volatility too. In a volatile week you will be risking multiples of what you intended.
The fix is to size from the stop: decide the money you are willing to lose, divide by the stop distance in currency terms, and take the resulting position. Then a wider stop automatically means a smaller position, and the risk stays constant while the stop adapts.
Where ATR stops misbehave
- Around scheduled events, true range spikes and the stop distance computed from the previous days is too tight for the day ahead.
- On very short timeframes, ATR becomes dominated by the spread and stops end up inside the noise band.
- In a strong trend, a trailing ATR stop will frequently be hit on pullbacks that do not threaten the trend.
- On instruments with gaps, the stop may be filled well beyond its level, so the loss exceeds the calculated risk.
A practical setup
Compute ATR on the timeframe you trade, not a lower one. Choose a multiple that reflects how much adverse movement you consider normal — two is a common starting point, three if you want to give the idea more room. Size the position from the resulting stop distance and your risk budget, round down to the broker's minimum, and accept that on small accounts the rounding sometimes means the trade cannot be taken at the intended risk.
Then test the whole thing as a unit. Changing the ATR multiple while holding the position size fixed is not testing the strategy; it is testing a different risk profile that you have not defined.
FAQ
- What ATR period should I use?
- Fourteen is the conventional default and works adequately. Shorter periods make the stop react faster to volatility changes, with more whipsaw; longer periods make it smoother and slower to adapt. As with most parameters, the choice should be made before testing rather than after.
- Is ATR better than a fixed stop?
- For any instrument whose volatility varies meaningfully over time, yes — a fixed distance is either too tight in quiet periods or too wide in active ones. The exception is very short timeframes where spread dominates ATR itself.
- Can I use ATR for profit targets too?
- You can, and it keeps the reward measure consistent with the risk measure. What you should not do is set the target at a fixed number of dollars while setting the stop from ATR, because then the reward-to-risk ratio drifts with volatility without your deciding that it should.
More guides
- How EasyQuant validates strategies — evidence you can filter
- Honest backtesting, not pretty curves
- Gold strategy research that stays honest
- Overfitting detection: catch it before you deploy
- System Forge: design, then prove
- Walk-forward analysis: the only backtest that fights overfitting
- MT5 export without custody
- Glass box, not black box AI signals
Not investment advice. Historical results do not guarantee future performance. EasyQuant is a research factory — you execute on accounts you control.