ATR-based stop loss sizing means setting a stop as a multiple of the Average True Range instead of a fixed number of pips - for example, entry price minus 2x ATR(14), rather than a flat "20 pips" for every trade. ATR measures how much an instrument typically moves over a given number of candles, so a stop built from it automatically widens on a volatile instrument and tightens on a calm one. The problem it solves is simple: a fixed 20-pip stop is tight on a pair that regularly moves 80 pips a day and wide on one that moves 15, even though the number never changes. ATR-based sizing replaces that guess with an actual measurement of the market's own recent behavior, so the stop means roughly the same thing - "outside normal noise" - no matter what's being traded.
What ATR actually measures
ATR is the average, over a chosen number of candles, of the "true range" of each one - the largest of: the current high minus the current low, the current high minus the previous close, or the current low minus the previous close. Using the previous close (not just the current candle's own range) is what makes it account for gaps, not only intra-candle movement. A 14-period ATR on a 1-hour chart, for instance, is simply the average true range of the last 14 hourly candles - a rolling measure of "how much does this thing normally move," recalculated on every new candle.
Why a fixed-pip stop breaks down
A fixed-pip stop has no relationship to the instrument's actual behavior - it's the same number whether volatility just doubled or halved. That creates two failure modes: a stop too tight for current conditions gets clipped by ordinary noise before the trade has a real chance to work, and a stop too wide for current conditions risks far more than intended for no added benefit. Both are corrected automatically by tying the distance to ATR, since the distance is recalculated from real, recent price behavior rather than picked once and left alone.
Fixed-pip vs. ATR-based stops
| Fixed-pip stop | ATR-based stop | |
|---|---|---|
| Adapts to volatility | No - same distance always | Yes - recalculated per candle |
| Consistent across instruments | No - "20 pips" means something different on every pair | Yes - relative to each instrument's own normal range |
| Simplicity | Simpler to reason about | One extra input (the ATR period and multiplier) |
| Best fit | A single instrument under stable conditions | Multiple instruments, or any single one whose volatility genuinely changes |
Choosing a multiplier
1.5x to 3x ATR is a common starting range, not a fixed rule - a tighter multiplier suits a mean-reversion strategy expecting price to turn quickly, while a wider one suits a trend-following strategy that needs room to survive a normal pullback before continuing. The correct number for a given strategy isn't something to guess at once and keep forever; it's something to confirm the same way any other part of a strategy gets confirmed - a real backtest against historical data, then a demo account against a live feed, comparing a few candidate multipliers against each other rather than assuming the first one picked is right.
ATR sizes the stop, not the trade
ATR-based sizing decides how far away the stop sits - it says nothing about how large the position itself should be. Those are two separate decisions: a wider ATR-based stop on a volatile instrument should generally come with a smaller position size, so the actual dollar risk stays the same trade to trade regardless of how far the stop ends up sitting. See position sizing for how that second half of the calculation works, stop loss vs. take profit for the broader picture both pieces fit into, and trailing stop loss for the same ATR-based distance applied to a stop that moves instead of one that sits still. In a block-based builder, this looks like combining an ATR operand with a multiplier directly in the risk block - e.g. "2 x ATR(14)" as the stop distance - rather than typing a static pip value.
Common questions
What does ATR-based stop loss mean?
It means setting a stop loss as a multiple of the Average True Range (ATR) instead of a fixed number of pips - for example, entry price minus 2x ATR(14) - so the distance automatically scales with how much the instrument is actually moving.
What ATR multiplier should I use?
There's no universally correct number - 1.5x to 3x ATR is a common starting range, tighter for faster mean-reversion strategies and wider for trend-following ones that need room to breathe. The right multiplier depends on the strategy and has to be confirmed through backtesting, not assumed.
Is ATR-based sizing better than a fixed-pip stop loss?
It's more consistent across instruments and across changing market conditions, since it's measuring actual volatility rather than guessing at it - but "better" depends on the strategy. A fixed-pip stop is simpler and can be perfectly fine for a single instrument traded under stable conditions.
Does ATR-based stop sizing replace position sizing?
No - they solve different problems. ATR decides how far away the stop sits; position sizing decides how large the trade is given that distance, so the dollar risk stays consistent regardless of how wide or narrow the ATR-based stop turns out to be.