What each one measures
Average Daily Range (ADR) is the average distance between each day's high and low over a set number of days.
daily range = high − low · ADR = average of the last N daily ranges
Average True Range (ATR), introduced by J. Welles Wilder in 1978, starts from the same high-to-low range but replaces it with a larger figure whenever the market opened away from the previous close. The true range of a day is the largest of three distances:
- today's high minus today's low;
- today's high minus yesterday's close;
- yesterday's close minus today's low.
ATR is the average of those true ranges. Wilder smoothed it with a running average rather than a simple one, which makes it react a little more slowly, but the idea is the same. Both are measured in price units: dollars on a stock, pips on a currency pair, points or ticks on a future.
The one difference: gaps
If a market opens where it closed, the second and third distances can never beat the first, so true range equals daily range and ATR equals ADR. They only split apart when price jumps between one session and the next.
That makes the gap between the two numbers a property of the market you trade:
| Market | How often it gaps | ADR vs ATR |
|---|---|---|
| Forex | Trades around the clock on weekdays; gaps mostly at the weekend open. | Nearly identical most weeks. |
| Index and commodity futures | Nearly 24-hour sessions with a short daily break. | Close; ATR a little higher after big news. |
| Stocks and ETFs | Closed overnight, so every open can gap, and earnings gaps can be large. | ATR often clearly higher. |
| Crypto | Never closes, so there is no gap between days. | Effectively the same. |
So on EUR/USD the ADR-or-ATR debate barely matters. On a stock with an earnings date in the lookback window, it decides whether your stop survives the next open.
Worked example: same five days, two numbers
A stock closed at $100.00. Over the next five sessions it did this, with an earnings gap on day three:
| Day | High | Low | Close | Daily range | True range |
|---|---|---|---|---|---|
| 1 | 102.00 | 99.00 | 101.00 | 3.00 | 3.00 |
| 2 | 103.00 | 100.50 | 102.50 | 2.50 | 2.50 |
| 3 (gap up) | 109.00 | 107.00 | 108.00 | 2.00 | 6.50 |
| 4 | 109.50 | 106.50 | 107.00 | 3.00 | 3.00 |
| 5 | 108.00 | 105.50 | 107.50 | 2.50 | 2.50 |
| Average | ADR 2.60 | ATR 3.50 |
Day three's own range was only $2.00, the quietest of the five. But even its low was $4.50 above the previous close, and ATR counts that jump: 109.00 − 102.50 = 6.50. ADR ignores it. One gap makes ATR 35% larger than ADR.
Now size a trade with a $50,000 account risking 1% ($500), using one full unit of each as the stop distance:
- ADR stop ($2.60): $500 ÷ 2.60 = 192 shares.
- ATR stop ($3.50): $500 ÷ 3.50 = 142 shares.
Both risk the same $500 if the stop fills at its level. The ADR version is 50 shares larger and has a stop that a single repeat of day three's gap would jump straight over.
Which to use for your trading style
| You trade | Better fit | Why |
|---|---|---|
| Intraday, flat by the close | ADR | You never hold through a gap, so the range inside the session is the risk you actually face. ADR also tells you how much of a normal day is left. |
| Swing trades held overnight | ATR | Your stop has to survive the open as well as the session, and only ATR includes that move. |
| Forex, crypto, near-24h futures | Either | With few gaps the two numbers are close, so use whichever your platform shows. |
| Stocks through earnings | ATR, and wider | An earnings gap can exceed any recent average. Size smaller or stand aside rather than trust either figure. |
Setting a stop from ADR
Because ADR describes one day, intraday traders usually work in fractions of it. A stop of a third to a half of ADR sits outside the ordinary back-and-forth of a session without risking a whole day's movement on one idea.
ADR also answers a second question: how much room is left today? If a pair's ADR is 80 pips and it has already travelled 75 pips since the open, a new trade in the same direction needs the day to be unusually large to reach its target. That is a reason for caution, not a rule; big days happen, often on news.
This is how RiskDesk works: it averages the high-to-low range of the last 14 daily bars, sets the stop at a chosen fraction of ADR, and sizes the position so a stop-out costs the risk you picked. For a gentler introduction to the idea, see what ADR is and how it sets a stop.
Setting a stop from ATR
Swing traders usually work in multiples of ATR, placed below entry for a long and above it for a short:
stop = entry − (ATR × multiple)
Multiples of 1.5 to 3 are common. A tighter multiple means a bigger position and more stop-outs from noise; a wider one means a smaller position and a larger move before the trade pays. Where to place a stop loss covers that trade-off, and why a stop below a real chart level usually beats any formula.
In the example above, a 1.5× ATR stop is $5.25 away, which cuts the position to 95 shares for the same $500 risk. That looks timid next to 192 shares until you remember the $6.50 true range on day three.
How many days to average
- 5 days reacts fastest. It picks up a change in conditions within a week, but a single wild day distorts it for the whole week.
- 14 days is Wilder's original ATR setting and the most common default for both measures. It is the usual balance of reactive and stable.
- 20 days is roughly a trading month. It is smoother and slower, and suits positions held for weeks.
Whatever you pick, keep it fixed. Results produced with a 14-day stop say nothing about a 5-day stop, and switching lookbacks after a run of losses is curve-fitting in real time.
What neither measure can tell you
- They are averages, not limits. Plenty of days are larger than average, and news days can be several times larger.
- They look backwards. A market that was quiet for 14 days has a small ADR and ATR right up to the day it isn't.
- They say nothing about direction. A large range can be a trend or a whipsaw; volatility measures cannot tell the two apart.
- A stop is not a guaranteed price. A gap through your level fills at the next available price, however carefully the distance was chosen.
Once the distance is set, position sizing turns it into a size, and risk of ruin covers how much to risk per trade in the first place.