8 minute read · Ultimo Research Desk · Reviewed 5 Sept 2026
ATR and Volatility Stops: What They Measure and When They Lie
ATR is the average size of a bar — high to low, with any overnight gap counted in — over the last 14 bars, in the instrument's own price units. It is the most practical indicator in this series because it is not a signal at all. It is a ruler. It tells you how much room a stop needs in current conditions, and how unusual today's move is compared with the last two weeks.
It lies when it is read as direction (it has none: a crash has a high ATR and so does a melt-up) or as a guarantee that a stop set at "2 × ATR" will contain the loss. ATR is the recent past; the next bar can be three times it.
What it measures
Wilder introduced true range in 1978 because high-minus-low misses something: if a market opens far from yesterday's close, that gap was movement too. True range includes it. ATR averages true range over 14 periods by default.
Because it is in price units, an ATR of 2 means different things at 20 and at 200. Divide by price to compare instruments — ATR as a percentage — but that is a separate measure and should be labelled as one. ATR has no direction and no opinion. It is usually paired with a directional rule, and the direction comes entirely from the other rule.
The formula, in words
For each bar, true range is the largest of: high − low; |high − previous close|; |low − previous close|.
The first ATR is the simple average of the first 14 true ranges. After that, Wilder's smoothing: new ATR = ((n − 1) × previous ATR + today's TR) ÷ n.
Some platforms offer a "simple" ATR that re-averages the last 14 true ranges each bar. It is a different, twitchier line. Know which one you are looking at.
Worked example
Previous close 100; today's high 108, low 101. Candidates: 108 − 101 = 7; |108 − 100| = 8; |101 − 100| = 1. True range = 8 — the gap up from 100 to the day's low is included, which the plain range would have missed.
If the previous ATR was 6.50: new ATR = (13 × 6.50 + 8) ÷ 14 = 92.5 ÷ 14 = 6.607.
Volatility stops
A volatility stop puts the stop a multiple of ATR away from a reference price instead of a fixed number of points:
- Long: stop = reference − k × ATR
- Short: stop = reference + k × ATR
With reference 100, ATR 6.607 and k = 2, the long stop sits at 100 − 13.21 = 86.79. A trailing version uses the highest close since entry as the reference and only ever moves the stop up. k = 2 is a convention; 1.5, 2.5 and 3 are all in common use, and the right one depends on the timeframe and how much noise you are willing to sit through.
What this buys you is consistency: the same rule gives a wider stop on a wild instrument and a narrower one on a calm one, instead of "50 pips everywhere". What it does not buy you is protection. An ATR stop is still a stop order, and the order-types guide explains why a stop is executed at the next available price, not the price on the ticket, when the market gaps through it.
ATR also normalises: a 10-unit move is huge if ATR is 2 and routine if ATR is 12. That is description, not probability.
When it lies
It cannot forecast a range. It is an average of past bars; the next bar does what it does. After a shock ATR stays elevated for weeks while the market has already calmed, so stops built on it are too wide; before a shock it is lowest of all, so stops are too tight at the worst moment.
One multiple copied across instruments with different tick sizes, session hours and liquidity is a guess dressed as a rule. A daily ATR and an hourly ATR measure different movement. A feed that drops overnight gaps understates it.
A wide stop reduces routine stop-outs and increases the loss when it is hit — unless the position is made smaller to compensate. That trade-off is the whole point of the position size calculator: the stop distance goes in, the lot size comes out, and the account risk stays fixed. ATR gives the distance; nothing about ATR chooses the risk.
What it does not tell you
Direction, fair value, the likelihood of a gap, the maximum loss, liquidity at the stop price, or position size. A stop level is where you would like to exit, not where you will.
What the evidence actually says
Three papers come up in every serious discussion of indicators, so it is worth knowing what they found rather than what people say they found. Brock, Lakonishok and LeBaron (1992) tested simple moving-average and trading-range rules on ninety years of the Dow and found they carried information relative to a random benchmark. Sullivan, Timmermann and White (1999) then re-ran that idea across nearly eight thousand rule variants and showed that once you account for how many rules were tried, the best-looking one is far less impressive — the "data-snooping" result. Park and Irwin (2007) reviewed ninety-five later studies and found roughly half positive, a quarter negative and the rest mixed, with results weakening after transaction costs and risk adjustment.
The fair summary is not "indicators work" and not "indicators are astrology". It is that a fully specified rule — entry, exit, size, costs — can be tested, and most rules that look good on a chart do not survive the test. Whatever you build on this indicator, test it as a complete rule on data it has not seen.
No study isolates ATR, because ATR on its own makes no prediction to test. It appears inside complete rules — Wilder's own systems, the "turtle" rules, most modern trend-following — as the sizing component. Sullivan, Timmermann and White's warning applies to the k multiple: try enough of them and one will look brilliant on last year's data.
Where it fits
ATR is not a vote on our signal pages, for the reason above: it has no direction. It is the number to have in front of you when you read the pivots on those pages and decide how far outside a level a stop should sit, and when you use Bollinger Bands, which measure the same thing in standard deviations.
Risk line: an ATR stop is a planning distance, not an exit price. The only thing that fixes the loss in advance is the size of the position.
Sources
- Wilder, J. Welles, New Concepts in Technical Trading Systems (1978): Internet Archive
- Park and Irwin (2007): DOI
- Sullivan, Timmermann and White (1999): DOI
- Lo, Mamaysky and Wang (2000): DOI
- Brock, Lakonishok and LeBaron (1992): DOI
- Murphy, John J., Technical Analysis of the Financial Markets (1999): Google Books
Written by the Ultimo Research Desk and checked against our own contract specifications and client agreement before publication; reviewed again when those change. Educational only — nothing here is a recommendation to trade. Spotted an error? Tell us.
