How to Use the ATR
Michael Whitman
·
Trading since 2017, full-time since 2020
·
Published 2026-09-07
·
Updated
·
Reviewed
To use the average true range, read it as how far this instrument normally travels in one bar, gap included. Set your stop a chosen multiple of it away from the entry, then divide your dollar risk by that distance to get the position size. It says nothing about direction.
The average true range answers one question: how far does this instrument usually travel in one bar. Used for that and nothing else, it turns a stop from a guess into a distance and a position size into arithmetic. Measured on SPY since 1994, a stop set in ATRs behaved about the same in quiet and busy markets, which a fixed-percentage stop did not.
Before you start
The indicator’s period chosen on purpose, with 14 as Wilder’s default. A shorter period follows recent bars closely; a longer one describes a broader normal. Pick one and keep it, so readings compare with each other.
Your account risk per trade in dollars. The ATR gives you a distance; the size comes from dividing this dollar figure by it, so it has to exist before the chart is opened.
A stop multiple of the ATR written down before the trade. One, two, three: chosen in advance and applied the same way every time, not fitted to the chart in front of you.
The steps
1. Read the ATR as a distance in price
A reading of $6.77 on SPY means its recent true ranges averaged about $6.77 a day. A high reading means large bars in either direction; it carries no information about the next bar’s direction.
2. Check that your platform’s true range includes the gap
True range is the largest of the bar’s high minus its low, the high minus the prior close, and the prior close minus the low. The last two terms count an overnight gap, which a plain high-minus-low misses.
3. Measure the stop from the entry in ATRs
Multiply the current ATR by your chosen multiple and subtract it from a long entry, or add it to a short one. That is the stop price.
4. Divide your dollar risk by the stop distance
Shares = dollar risk / (multiple x ATR), rounded down. When the ATR is high the position comes out smaller for the same dollar risk.
5. Compare today’s reading with this instrument’s own history
Divide the ATR by the price to get a percentage, then compare it with the same instrument’s past readings. A different instrument’s ATR is in different units and tells you nothing about this one.
6. Leave the stop alone once the trade is open
A rising ATR changes the distance for the next trade. Widening the stop on an open position increases a risk you already set.
A worked example on SPY
SPY closed at $769.64 on 2 Oct 2026 with a 14-session ATR of $6.7706, 0.880% of the price. A long entry at that close with a two-ATR stop puts the stop $13.54 lower, at $756.10.
Size it for a hypothetical $10,000 account risking 1%, which is $100. $100 / $13.5412 = 7.38, so 7 shares, about $5,387 of SPY at that close.
The same rule six months earlier gives a different trade. On 31 Mar 2026 SPY closed at $650.34 with an ATR of $10.6715 (1.641% of price, the highest share of 2026). Two ATRs was $21.34, and $100 bought 4 shares. The rule kept the dollar risk the same while the distance and the size changed with the market.
How to tell it worked
Pull your last 30 trades from at least 60 days of history.
Write each stop distance in ATRs at entry. If the multiple you planned was 2 and the distances run from 0.8 to 3.5, the multiple you used is not the one you chose.
Check the dollar loss on every stopped trade. Each should come within a few percent of your planned dollar risk; one twice that size means the size step was skipped or the stop was moved.
Count stops reached in the first 2 sessions. If more than a third of all stops were hit that fast, compare the multiple you used against the reach rates in the data below before changing anything else.
How often a stop measured in ATRs gets reached
On SPY since 1994, a stop below the close was reached within ten sessions this often: 58.82% at one ATR, 45.22% at one and a half, 33.57% at two and 18.13% at three. Those are reach rates, not losses: a long position can recover after touching a level a stop would have closed.
The same two-ATR stop behaved alike in calm and busy markets. Split every session into thirds by ATR as a share of price, it was reached 35.24% of the time in the calmest third, 33.27% in the middle and 32.20% in the busiest. A fixed 2% stop was reached 26.53%, 40.16% and 60.29% of the time across the same thirds. Measuring the distance in ATRs is what held the rate steady.
Why one fixed distance cannot stay the same risk
SPY’s ATR moved across a twenty-fold range. The lowest reading since 1994 was 0.375% of price on 18 Oct 2017 and the highest 7.913% on 27 Oct 2008. Year by year, the median ran from 0.603% in 2017 to 2.068% in 2002.
Inside a single year the swing was large too. In 27 of 32 full years the year’s highest ATR was at least double its lowest, with a median ratio of 2.60; the steadiest year, 2005, still moved by a factor of 1.70, and 2020 by 10.71. A stop set at a fixed number of dollars in a quiet month is a fraction of an ordinary bar in a busy one.
The original data
Sample: SPY daily bars, 3 Jan 1994 to 2 Oct 2026, 8,243 sessions (Yahoo Finance, split-adjusted highs, lows and closes; opens not used). ATR per Wilder: first value the mean of the first 14 true ranges from February 1993, then (prior ATR x 13 + today’s true range) / 14. Stop test: for each session with ten sessions after it (8,233), was any low in those ten at or below the close minus k x that day’s ATR.
| Stop below the close | Reached within 10 sessions |
|---|---|
| 1 x ATR | 4,843 of 8,233 (58.82%) |
| 1.5 x ATR | 3,723 (45.22%) |
| 2 x ATR | 2,764 (33.57%) |
| 3 x ATR | 1,493 (18.13%) |
The thirds: 2,744 sessions with ATR from 0.375% to 0.981% of price, 2,744 from 0.981% to 1.517%, and 2,745 from 1.517% to 7.913%. Every session’s true range, ATR and next-ten-session low is in the SPY ATR stop file; yearly medians, lows and highs are in the SPY ATR by year file.
What this does not show. Reach rates describe how far SPY’s lows traveled, not the result of any trading rule, and they are for one index fund on the daily chart. A single stock gaps more often and would reach the same multiples at different rates.
When it fails
It is an average of the past, so it is smallest just before volatility arrives. A long quiet stretch pulls the reading down and stops sized from it get tight. SPY’s lowest ATR of the whole sample came in October 2017; the move that ends a quiet spell can cross several such stops in one bar. The ATR described the past accurately the whole time and had no way to see the change coming.
The second failure is asking it for direction. It contains none.
A third is comparing readings across instruments. They are in different units.
A fourth is a stop inside one ATR. Within ten sessions, one ATR was reached 58.82% of the time.
And a fifth is widening a stop on an open position. That raises a risk already agreed.
Related
ATR covers the calculation and why true range includes the gap. Stop loss covers where the distance from step three is used. And position sizing covers the division in step four that turns the distance into a number of shares.
I keep this one on the chart when almost everything else comes off, because it answers a question I actually have. Not which way, and not when, just how far this thing typically travels, so a stop can sit outside that. Its output feeds an arithmetic step rather than a judgment.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.