ATR Stop-Loss
Learn how Average True Range can scale stop distance to current volatility, how ATR is calculated conceptually and why ATR stops still need structural and regime context.
Reading progress — saved on this device
An ATR stop uses Average True Range as a volatility unit, placing the exit a chosen multiple of recent trading range away from entry or a trailing reference.
Learning objectives
- Explain what ATR measures and why true range includes gaps.
- Calculate an ATR-multiple stop and translate it into position size.
- Recognise when volatility regime changes make historical ATR unreliable.
What it is
True range is the largest of the current high-low range, the distance from the current high to the previous close, and the distance from the current low to the previous close. ATR averages or smooths true range over a chosen lookback.
A 2×ATR stop does not claim price cannot move two ATR. It creates a volatility-scaled distance intended to reduce exits from ordinary noise compared with a fixed percentage.
ATR can be measured in price units or converted to a percentage of price. Comparing ATR percentages is more useful across assets with different prices.
How it works
If ETH trades at £2,000 with a 14-day ATR of £100, ATR is 5% of price. A 2×ATR stop is £200 away, or 10%. With a £400 risk budget, raw size is 2 ETH before slippage.
The chosen multiple controls a trade-off: a tighter multiple reduces loss per unit but increases the chance normal volatility hits the stop; a wider multiple reduces noise exits but requires smaller size.
ATR stops can be anchored to entry, a swing point or a trailing high/low. These are different systems and should not be mixed casually.
Crypto's 24/7 market means daily ATR has no exchange close gap in the traditional sense, but large moves between sampling periods still matter. Data source and bar construction must be consistent.
How to analyse and apply it
| Check | Why it matters | What to verify |
|---|---|---|
| ATR lookback | Controls responsiveness. | Use a period suited to the strategy horizon. |
| ATR multiple | Sets stop width. | Test sensitivity rather than optimising one perfect value. |
| Anchor | Determines what the stop follows. | Specify entry, structure or trailing extreme. |
| Volatility shock | Can make the estimate stale. | Use stress floors/caps or pause rules. |
Risk rules should be written before a live position is opened and evaluated across many trades or scenarios. A control that is changed only after losses appear is discretionary damage control, not a repeatable risk system.
Worked example and thought exercise
BTC trades at £60,000 and 14-day ATR is £2,400. A long strategy uses a 1.5×ATR initial stop, so raw stop distance is £3,600. With £900 risk, size is 0.25 BTC.
If ATR rises to £4,000 the next day, mechanically moving the stop farther away would increase the monetary loss unless position size is reduced. Many systems therefore do not widen an existing protective stop simply because ATR rises.
Thought exercise: why might a 2×ATR stop behave very differently when ATR is 2% of price versus 10%?
Common mistakes and practical workflow
- Treating ATR as a forecast of maximum movement.
- Changing the ATR multiple after every losing trade.
- Widening an existing stop when volatility rises without reducing size.
- Comparing raw ATR values across differently priced assets.
Practical workflow
- Choose a consistent data source and ATR lookback.
- Select and test an ATR multiple and anchor rule.
- Translate stop distance into stressed monetary loss per unit.
- Size the position from the risk budget.
- Define how the stop behaves if ATR changes after entry.
✅ Knowledge checkpoint
- What three values can determine true range?
- What does a 2×ATR stop mean economically?
- Why can rising ATR create a risk-control problem after entry?
- How should position size change when ATR-based stop distance doubles?
FAQs
❓ Is ATR directional?
No. It measures range/volatility, not whether price should rise or fall.
❓ What ATR period is best?
There is no universal best period. It should match the strategy horizon and be tested across regimes.
❓ Should I move a stop farther away if ATR rises?
Not automatically. Doing so increases loss at the stop unless size changes; many systems allow only tightening after entry.
❓ Can ATR replace market structure?
It can define a volatility stop, but structural context may still improve whether the chosen distance corresponds to thesis invalidation.
📋 Summary
ATR stops scale risk distance to recent market variability. They are useful for consistency across assets and regimes, but only when the lookback, multiple, anchor and post-entry adjustment rules are explicit and protected against volatility shocks.
Want this in a personalised order?
Take the crypto assessment and get a custom path of 10 modules matched to what you already know. Free, no card required.
Build my path →