Two traders use the same 30 pip stop. One is trading a major FX pair in a quiet August week, where 30 pips is a meaningful move and the stop sits well outside routine noise. The other is trading gold during a session where the metal covers several times that distance in an hour. Same number, completely different decision. The first trader is running a real stop. The second is donating.
Volatility measurement fixes this by converting stop distance from a fixed number into a fraction of what the instrument normally does.
True range, and why the previous close matters
Average True Range is the standard tool. It averages the true range of the last N bars, usually fourteen, and true range is the largest of three distances: the current bar's high to low, the distance from the current high to the previous close, and the distance from the current low to the previous close.
Those last two exist because of gaps. If a market closes at 1.0850 and reopens at 1.0790 before trading in a narrow band, the bar's own high minus low understates what happened. Measuring against the previous close captures the jump. That is the whole reason ATR is preferred to a plain range average, and it matters most on exactly the instruments where it is easy to forget, such as anything holding over a weekend break.
ATR is reported in the instrument's price units, which means it needs converting before you can compare anything. An ATR of 0.0062 on a pair quoted to five decimals is 62 pips. An ATR of 14.80 on gold is 1,480 points at the usual quotation. Our note on what a pip actually is covers the conversion, and getting it wrong by a factor of ten is one of the most common sizing errors there is.
Sizing so that risk stays constant
The sequence is fixed: decide the money at risk first, then the stop distance, then let those two determine the size. Never the other way round.
Take an account of 10,000 units of account currency and a rule of 1% risk, which is 100 per trade. Read the daily ATR, choose a multiple that keeps the stop outside routine movement, and divide.
| Step | Quiet conditions | Active conditions |
|---|---|---|
| Risk per trade | 100 | 100 |
| Daily ATR | 50 pips | 120 pips |
| Stop at 1.5 x ATR | 75 pips | 180 pips |
| Value per pip needed | 100 / 75 = 1.33 | 100 / 180 = 0.56 |
| Resulting position | Larger | Roughly 40% of the quiet-market size |
The money at risk is identical in both columns. What changed is the size, and it changed automatically because the measurement changed. A trader using a fixed lot with a fixed pip stop would have been risking the same amount in the first case and a multiple of it in the second, without deciding to. The mechanics of turning a pip value into lots are in lots and position sizes.
An ATR multiple sets a minimum sensible distance, not a location. Put the stop somewhere structurally meaningful, beyond a swing high or below the level you are trading against, and use the ATR floor to check that the level gives the trade enough room. If structure wants a stop tighter than the floor, the setup is usually too tight to take.
Volatility has a schedule
Daily ATR hides the fact that movement is concentrated. Most of the day's range in FX is produced in the London morning and the overlap with New York, while the late Asian hours are frequently flat. A stop sized on the daily average will feel enormous at 03:00 and thin at 09:30, which is why intraday traders read ATR on the timeframe they trade rather than on the daily chart.
The calendar does the same thing at a larger scale. Rate decisions, inflation prints and employment data compress a normal day's movement into minutes. Volatility around scheduled events is not a reason to avoid trading, but it is a reason to size against the event's likely range instead of the week's average, and to expect wider spreads while it happens.
The measurement also travels across correlated positions. If you hold three trades that all express the same dollar view, their volatilities do not diversify, they add. That is the point made in currency correlations, and it is where ATR sizing done position by position still leaves a book that is too big.
Other ways to read it
ATR is the workhorse, and a few other measures answer different questions. Standard deviation of returns is the statistical version, which is what Bollinger Bands display when they widen and contract. Realised range over a fixed lookback, for example the average daily high to low over the last twenty sessions, is easier to hold in your head and is a reasonable proxy for a plausible adverse move. Implied volatility from the options market, where it is available for the underlying, is a forward-looking estimate rather than a backward-looking one.
All of them share a limitation. Every measure of volatility is a description of the recent past. Regimes change faster than a fourteen-period average can follow, so ATR is at its lowest just before the breakout that triples it, and at its highest just as the move exhausts. Using it as a timing signal is a mistake. Using it to decide how much room a trade needs and how much of it to hold is what it is for. Keep that decision inside the fixed limits described in your risk rules, and treat the measurement as an input to those rules rather than a replacement for them.
"Decide the money first, then how much room the trade needs, then let the size fall out of the arithmetic. Anyone picking lot size before stop distance is guessing at their own risk."
— Alex Onta, Executive Director, SINGUARD
Key Takeaways
- ATR counts gaps by measuring against the previous close, which is why it beats a plain high minus low average.
- Fix the money at risk, derive the stop from volatility, then let position size fall out of the division.
- Read ATR on the timeframe you trade. Daily averages hide how concentrated movement is inside the session.
- Volatility measures describe the past. Use them for sizing and stop room, never as an entry timing signal.
Frequently Asked Questions
What does ATR actually measure?
Average True Range is the average of the true range over a set number of bars, normally fourteen. True range is the largest of three distances: the current bar's high to low, the high to the previous close, and the low to the previous close. Including the previous close means ATR counts overnight gaps as movement, which a simple high minus low calculation would miss.
Should a stop be placed at a multiple of ATR?
An ATR multiple is a sensible floor rather than a location. It tells you how much room the instrument normally needs so the stop is not hit by routine movement. The level itself should still be somewhere structurally meaningful, such as beyond a swing point. When structure asks for less room than the ATR floor, the setup is usually too tight to trade.
Does volatility sizing reduce risk?
It keeps the monetary risk per trade stable rather than reducing it. As ATR rises the stop widens and the position shrinks, so the same percentage of the account is exposed. What it prevents is the common failure where a fixed lot size quietly doubles real risk when conditions change. Leveraged trading remains high risk under any sizing method.