ATR (Average True Range): Measuring Volatility for Position Sizing

Origins of the ATR

The Average True Range (ATR) was developed by J. Welles Wilder Jr. and introduced in his 1978 book New Concepts in Technical Trading Systems -- the same book that gave us the Relative Strength Index (RSI), the Average Directional Index (ADX), and the Parabolic SAR. Wilder originally designed the ATR for commodity markets, where overnight gaps between the close and the next open were common and significant. However, the indicator has since become a standard tool across all asset classes, including equities, forex, and futures.

Unlike most technical indicators, the ATR is not a directional indicator. It does not tell you whether the price is going up or down. It measures only one thing: the degree of price movement -- volatility. This makes it fundamentally different from tools like the RSI or moving averages, and it serves a different purpose in a trading system. The ATR is primarily a risk management tool, not a signal generator.

True Range: The Foundation

Before computing the ATR, you need to understand True Range. Wilder defined True Range as the greatest of three values:

The first component -- High minus Low -- is simply the intraday range for the current period. If there were no overnight gaps, this would be sufficient. But stocks frequently open at a different price than the previous close. A stock might close at $50 and open the next day at $52 due to overnight news. The intraday range that day might be $52 to $53 (just $1), but the true range of movement from the prior close is $3 ($50 to $53).

This is exactly what components (2) and (3) capture. By measuring the distance from the previous close to the current high and current low, the True Range accounts for gap moves that the simple High-Low range misses entirely.

Why Gaps Matter

Consider two scenarios for a stock that closed at $100 yesterday:

Scenario A: Today opens at $100, trades between $98 and $103. High - Low = $5. |High - Prev Close| = $3. |Low - Prev Close| = $2. True Range = $5.

Scenario B: Today opens at $105 (gap up on earnings), trades between $104 and $107. High - Low = $3. |High - Prev Close| = $7. |Low - Prev Close| = $4. True Range = $7.

In Scenario B, the simple intraday range is only $3, which drastically understates the actual volatility experienced by anyone holding the stock overnight. The True Range correctly captures the $7 of total movement. This distinction is critical for risk management -- your stop-loss needs to account for gap risk, and the True Range incorporates it.

Computing the ATR

The ATR is simply the N-period average of the True Range. Wilder's original recommendation was N = 14 periods, which remains the default on most charting platforms.

Initial ATR Value

The first ATR value is a simple arithmetic mean of the first 14 True Range values:

Subsequent ATR Values (Smoothing)

After the initial value, Wilder used the same smoothing technique as in the RSI:

This is equivalent to an exponential moving average with a smoothing factor of 1/14. The smoothing means that the ATR changes gradually rather than jumping around with each new True Range reading. Older True Range values have diminishing but non-zero influence on the current ATR.

ATR-Based Position Sizing

This is arguably the most important practical application of the ATR, and it is the one that distinguishes professional risk management from amateur approaches.

The principle is straightforward: risk a fixed dollar amount (or percentage of account equity) per trade, and let the ATR determine how many shares you can buy. More volatile stocks get smaller positions; less volatile stocks get larger positions. The risk per trade remains constant regardless of the stock's volatility.

The Formula

Where:

• Dollar Risk per Trade = Account Equity x Risk Percentage. For example, a $100,000 account risking 1% per trade would risk $1,000.
• ATR = Current 14-period ATR of the stock.
• Multiplier = A constant (typically 1.5 to 3) that determines how many ATRs your stop-loss is away from the entry price.

Worked Example

Suppose you have a $100,000 account and risk 1% per trade ($1,000). You want to buy a stock trading at $50 with a 14-day ATR of $2.00. You plan to place your stop-loss at 2x ATR below your entry price.

Stop distance = 2 x $2.00 = $4.00 per share.

Position size = $1,000 / $4.00 = 250 shares.

Total position value = 250 x $50 = $12,500 (12.5% of account).

Now consider a more volatile stock trading at $50 with a 14-day ATR of $5.00:

Stop distance = 2 x $5.00 = $10.00 per share.

Position size = $1,000 / $10.00 = 100 shares.

Total position value = 100 x $50 = $5,000 (5% of account).

Both trades risk exactly $1,000. But the more volatile stock receives a smaller position because its expected range of movement is larger. This is the essence of volatility-adjusted position sizing -- equalizing risk across trades regardless of each stock's inherent volatility.

ATR for Stop-Loss Placement

Placing stop-loss orders is one of the most difficult practical problems in trading. Set the stop too tight, and normal price fluctuations will trigger it before the trade has a chance to work. Set it too wide, and losses on failed trades become unacceptably large.

The ATR provides a principled framework for stop-loss distance. The idea is to set the stop far enough away from the entry price that normal volatility will not trigger it, but close enough that the stop will fire before losses become catastrophic.

Common ATR Stop-Loss Multipliers

• 1.5x ATR: Tight stop. Used for short-term trades where quick exits on adverse moves are acceptable. Higher probability of being stopped out by noise.
• 2x ATR: Standard stop. Provides room for normal fluctuation while limiting losses. The most commonly used multiplier.
• 3x ATR: Wide stop. Used for longer-term trend-following positions that need significant room to breathe. Lower probability of noise stops, but larger loss per failed trade.

For a stock with an ATR of $3.00, a 2x ATR stop-loss placed below the entry would be at entry price minus $6.00. This means the stock can move $6.00 against you before the stop triggers -- roughly two average daily ranges.

The Chandelier Exit

The Chandelier Exit is a trailing stop technique developed by Charles Le Beau that uses the ATR. It is named for its resemblance to a chandelier hanging from the ceiling -- the stop hangs down from the highest point reached by the stock.

The standard parameters are a 22-period lookback for the highest high and a 3x ATR multiplier. As the stock makes new highs, the exit level ratchets up. If the stock falls more than 3 ATRs from its highest recent high, the position is closed.

The Chandelier Exit has the desirable property of adapting to both the stock's trend and its volatility. In a strong, low-volatility uptrend, the exit stays relatively close to the price. In a volatile uptrend, the exit gives more room. When volatility expands (as it often does at trend reversals), the exit adjusts automatically.

Trailing Stop with ATR

A simpler version of the same idea is a pure ATR trailing stop. After entering a long position, you set the initial stop at entry minus N x ATR. As the stock moves in your favor, you move the stop up (but never down) to remain N x ATR below the highest close since entry. This locks in profits while still giving the position room to fluctuate within normal volatility.

Keltner Channels

Keltner Channels are a volatility envelope indicator that uses the ATR to set channel width around a moving average. The original version was created by Chester Keltner in the 1960s, but the modern version -- popularized by Linda Bradford Raschke -- uses the following construction:

The specific parameters vary by practitioner. Some use ATR(14) with a 2x multiplier, others use ATR(10) with 1.5x. The concept is the same: the channels expand when volatility increases and contract when volatility decreases.

Keltner Channels are sometimes compared to Bollinger Bands, which use standard deviation instead of ATR to set channel width. The practical difference is that ATR-based channels tend to be smoother because the ATR changes more gradually than standard deviation, which can spike dramatically on a single extreme day.

Normalizing ATR: The ATR Percentage

One limitation of the raw ATR is that it is expressed in price units. A stock trading at $200 with an ATR of $4.00 and a stock trading at $20 with an ATR of $4.00 have the same absolute ATR but very different relative volatilities.

To compare volatility across stocks with different price levels, normalize the ATR by dividing it by the stock price:

The $200 stock has an ATR% of 2.0%, while the $20 stock has an ATR% of 20.0%. The second stock is far more volatile in relative terms. ATR% allows you to compare volatility across different securities, identify the most and least volatile stocks in a universe, and screen for volatility characteristics.

ATR vs. Standard Deviation: Why Traders Prefer ATR

Both ATR and standard deviation measure volatility, but they do so in different ways, and each has advantages depending on the use case.

Standard Deviation

Standard deviation is the academically standard measure of volatility. It measures the dispersion of returns around their mean. Historical volatility (as used in options pricing and academic finance) is typically calculated as the annualized standard deviation of log returns.

Why ATR Is Often Preferred for Trading

Gaps are captured: This is the primary advantage. Standard deviation of close-to-close returns does capture gaps (since the return from close to close includes any overnight gap). However, the standard deviation of intraday ranges (High - Low) does not. The True Range explicitly accounts for gaps, making the ATR a more complete measure of the price movement a trader actually experiences.

Intuitiveness: The ATR is expressed in the same units as the stock price (dollars per share), making it immediately interpretable. "The stock has an ATR of $3" means it typically moves about $3 per day. Standard deviation of returns is expressed as a percentage, which is less intuitive for setting stop-loss levels and position sizes in dollar terms.

Stability: Because the ATR uses Wilder's smoothing technique, it changes more gradually than standard deviation, which can spike dramatically on a single extreme day. This stability is desirable for position sizing and risk management, where you want your risk parameters to adjust smoothly rather than whipsawing.

No directional assumption: The ATR measures total price movement regardless of direction. Standard deviation measures dispersion around a mean return, which implicitly assumes a trend component. For the purpose of setting stop-losses and sizing positions, total movement is more relevant than dispersion around a trend.

ATR in Different Market Environments

The ATR is not constant -- it expands and contracts with market conditions. Understanding these dynamics adds a layer of information beyond the raw number.

Low ATR (contracting volatility): Periods of low volatility often precede significant price moves. When the ATR reaches unusually low levels relative to its recent history, it may indicate that a breakout (in either direction) is building. Volume dry-up combined with ATR contraction is a classic pre-breakout pattern.

High ATR (expanding volatility): Volatility expansion often occurs at major turning points -- the beginning of trends, panic selling events, and blow-off tops. A sharp spike in ATR during a downtrend can indicate capitulation, where sellers exhaust themselves and the stock may be approaching a bottom.

Rising ATR in an uptrend: This typically occurs during the middle or late stage of a strong trend, when price moves are accelerating. It can be a sign of momentum, but also a warning that the trend is becoming overextended.

Falling ATR in an uptrend: A healthy, sustainable uptrend often shows a stable or slightly declining ATR. The stock is making steady progress without wild swings. This is generally a positive sign for trend continuation.

Practical Implementation Tips

1. Always use ATR for position sizing in systematic strategies. Fixed share counts or fixed dollar amounts ignore volatility differences between stocks and create uneven risk exposure. ATR-based sizing is the minimum standard for serious risk management.
2. Recalculate regularly. The ATR changes over time. A stock that had an ATR of $2 last month might have an ATR of $4 now. Use the current ATR, not a stale value, for sizing and stop placement.
3. Use ATR% for cross-security comparisons. The raw ATR is meaningless for comparing stocks at different price levels. Normalize by price.
4. Do not use ATR as a directional signal. It tells you how much the stock is moving, not which direction. High ATR does not mean bearish, and low ATR does not mean bullish.
5. Combine ATR with directional analysis. Use the RSI, moving averages, or insider trading signals for direction. Use the ATR for sizing and risk management. These are complementary tools.

ATR in Quantitative Signal Generation

In quantitative trading systems, the ATR plays a critical role in both the signal scoring and the risk management layers. It serves as the volatility anchor that connects signal strength to position size.

A high-conviction signal at a high-ATR stock should receive a smaller position than the same conviction signal at a low-ATR stock. The conviction determines whether to trade; the ATR determines how much to trade. This separation of signal quality from risk sizing is a hallmark of well-designed trading systems.

Alpha Suite uses the ATR extensively in its signal generation pipeline. The barrier model for computing take-profit and stop-loss levels is anchored to the ATR, ensuring that targets and stops are calibrated to each security's actual volatility. Position sizing uses ATR-based risk normalization, and the system adjusts for excess kurtosis (fat tails) to account for the fact that real returns have more extreme moves than a normal distribution would predict.