Direct answer
Average True Range (ATR) is used in forex to quantify how much a pair typically moves, on average, over a chosen number of periods. You calculate ATR from recent price ranges, then use the resulting value to describe volatility and to set size-related distances in a way that adapts to changing market “speed.”
ATR by itself does not predict direction. It measures movement magnitude, so it is best treated as a volatility input for context and calculations rather than as a standalone signal.
Explanation: what ATR is and how to use it
ATR definition (conceptual): ATR converts raw candle ranges into an average volatility measure that considers gaps and large single-period moves.
Core inputs you choose:
- Timeframe: e.g., 5-minute, 1-hour, daily. ATR depends on the timeframe.
- Lookback period (commonly 14): the number of periods used to smooth volatility.
How ATR is computed (operation level):
- For each period, compute a true range from the current high/low and the previous close.
- Smooth those true range values over your lookback window to produce ATR.
How to use ATR in forex (practical roles):
- Volatility context: When ATR is rising, recent movements are larger; when it is falling, movements are smaller.
- Volatility filters: You can restrict certain calculations to periods where ATR is above a minimum level, to avoid acting during unusually quiet conditions.
- Distance scaling: Convert a rule like “X% of typical movement” into a price distance using ATR, so distances expand in high volatility and contract in low volatility.
Important normalization idea: ATR is expressed in the instrument’s price units. To compare ATR across pairs or timeframes, consider using a normalized ATR, such as ATR divided by price (a percentage-style view). This reduces misleading comparisons caused by different price scales.
Example and checks
Example (distance scaling concept): Suppose you use a fixed multiple of ATR (for instance, 2 × ATR) as a “typical movement distance.” If volatility increases, 2 × ATR grows, widening the distance; if volatility decreases, it tightens.
Checks you can run without needing future data:
- Visual alignment: Plot ATR on the same timeframe as price and look for whether large spikes in candles correspond to ATR rises.
- Regime observation: Compare ATR across two timeframes (e.g., daily ATR vs. 1-hour ATR) to see whether the market shows consistent or conflicting volatility regimes.
- Stability test: Recalculate ATR with slightly different lookback periods (e.g., 10 vs. 14 vs. 20). If your chosen volatility-dependent behavior changes dramatically, the method may be too sensitive.
Limitations and risks
- Direction is not included: ATR does not tell you whether price will go up or down; it only summarizes movement magnitude.
- Parameter sensitivity: The lookback period and timeframe strongly affect ATR values. Changing them changes the volatility measure.
- Lag and smoothing: Because ATR is an average, it reacts with delay to regime changes.
- Overfitting risk: Volatility rules that look good historically can fail when volatility dynamics change. Keep evaluation simple and out-of-sample when possible.
- No guarantees: Even if ATR-based calculations improve consistency of distances or filters, they cannot ensure favorable outcomes.
For verification, treat ATR as a measurable input: confirm the math on your platform, confirm units, and test whether your chosen use (filtering, scaling, or normalization) behaves reasonably across different market conditions.