Average True Range, or ATR, measures recent price-range behavior. It includes gaps from the previous close, so it can describe variation that a candle's high-minus-low range misses. ATR is expressed in the instrument's price units. A rising ATR does not choose a direction: sharp advances and sharp declines can both increase it.
Traders sometimes use a multiple of ATR to define a stop distance or compare market conditions. The calculation makes that distance responsive to recent observations, but it does not establish a profitable exit or guarantee that a future move stays within the average. The fictional examples here explain the arithmetic, not a recommended multiplier or position size.
True range takes the largest of three distances
True range = maximum of current high minus current low, absolute current high minus previous close, and absolute current low minus previous close. In a fictional bar with previous close 100, high 105 and low 102, those distances are 3, 5 and 2. True range is therefore 5, not the visible intrabar range of 3.
For a bar with previous close 100, high 99 and low 95, the distances are 4, 1 and 5, again producing 5. Absolute values prevent the gap terms becoming negative. These examples demonstrate upward and downward gaps without implying that the next bar will continue in either direction.
Declare how the ranges are averaged
A Wilder-style 14-period update is ATR = (prior ATR × 13 + current true range) ÷ 14, after an initial seed commonly taken as the mean of 14 ranges. With a fictional prior ATR of 4 and current range of 11, the update is (52 + 11) ÷ 14 = 4.5. A simple rolling average would use a different update because it removes a specific old observation.
Fidelity documents the Wilder-style update; MetaQuotes describes ATR as an average of true ranges. Check the actual platform or indicator's smoothing and seed history instead of assuming every ATR14 implementation matches. Intraday ATR14 uses 14 bars of that timeframe, not automatically 14 days. Compare matching settings and enough prior data before investigating a value mismatch.
Convert a distance into position risk
Assume a linear XAUUSD contract of 100 ounces per lot and a USD account. With hypothetical ATR of $8 and a study stop of 1.5 ATR, the price distance is $12. At 0.05 lots, planned gross distance risk is 12 × 100 × 0.05 = $60. A stop filling beyond its requested level, commission or financing can increase the realized loss.
If the study's gross distance-risk budget is $48, theoretical volume is 48 ÷ (12 × 100) = 0.04 lots. Round down to an allowed volume step, then recalculate rather than rounding upward silently. If minimum volume exceeds the budget, the proposed trade does not fit those assumptions. Costs must be budgeted separately; $48 is an illustrative amount, not advice on account risk.
| ATR | 1.5 ATR distance | Theoretical lots before volume rules |
|---|---|---|
| $8 | $12 | 0.0400 |
| $12 | $18 | 0.0267 |
| $16 | $24 | 0.0200 |
A wider stop does not preserve risk by itself
When volatility rises, an ATR-based stop may become wider. Keeping the same lot size increases its cash distance risk. A strategy can instead resize, skip the setup or follow another declared rule, but each choice changes behavior and must be tested. Do not move an existing stop farther away and describe the position as having its original risk.
State whether ATR is frozen at entry or recalculated later. A trailing exit also needs a direction rule so an expanding ATR does not unintentionally loosen the stop. Chart-derived levels must comply with tick size and broker distance restrictions. No volatility formula removes the possibility of a gap or unavailable fill.
Compare instruments with a normalized measure
ATR ÷ current positive price × 100 gives a price-normalized range measure. A hypothetical ATR of 20 at price 2,000 is 1%, as is ATR of 0.30 at price 30. Equal percentages do not mean equal lot risk: quantities, tick values and account-currency conversion still differ. Use a consistent close or other declared price denominator.
Review ATR across calm and stressed periods with the same settings. It summarizes a trailing sample and responds after new ranges arrive, so it may understate a sudden change at the beginning of a shock. An economic calendar and a spread observation answer different questions and cannot be replaced by ATR alone.
Checklist and frequently asked questions
Is ATR a directional signal? No. Does two ATR guarantee enough room for a trade? No multiplier guarantees that. Why does my ATR differ from another chart? Check timeframe, feed, averaging method, seed history and incomplete bars before assuming the arithmetic is wrong.
- Record timeframe, period, smoothing method, seed history and the price units displayed.
- Use the value known at the decision time and state whether the current bar is completed.
- Specify multiplier, entry, invalidation and cash-risk conversion together.
- Include costs, volume rounding, gaps and execution differences in the test.