Average True Range (ATR)

I. What is ATR

The Average True Range (ATR) is a volatility indicator first introduced by American technical analysis pioneer J. Welles Wilder Jr. in 1978 in his classic book New Concepts in Technical Trading Systems.

ATR belongs to the Volatility class of indicators. It measures the magnitude of price fluctuations over a given period without indicating direction. The default period is $n = 14$.

Unlike simply using High - Low to measure volatility, ATR introduces the concept of “True Range” (TR), which incorporates the previous bar’s closing price into the calculation, thereby accurately capturing the additional volatility caused by price gaps.

ATR is a core component of many other technical indicators, including SuperTrend, Keltner Channels, ADX, and various volatility-based stop-loss methods.

---

II. Mathematical Principles and Calculation

2.1 True Range (TR)

For the $t$-th bar, the True Range is defined as the maximum of the following three values:

$$\text{TR}_t = \max\bigl(H_t - L_t,\; |H_t - C_{t-1}|,\; |L_t - C_{t-1}|\bigr)$$

Where:

• $H_t$ = current high price
• $L_t$ = current low price
• $C_{t-1}$ = previous bar’s closing price

The three components represent:

1. $H_t - L_t$: the current bar’s range
2. $|H_t - C_{t-1}|$: the distance between the current high and the previous close (captures upward gaps)
3. $|L_t - C_{t-1}|$: the distance between the current low and the previous close (captures downward gaps)

2.2 ATR Calculation

ATR uses Wilder’s Smoothing Method (SMMA) to smooth the TR series:

First ATR value (at the $n$-th bar):

$$\text{ATR}_n = \frac{1}{n} \sum_{i=1}^{n} \text{TR}_i$$

This is the simple moving average of the first $n$ TR values.

Subsequent ATR values ($t > n$):

$$\text{ATR}_t = \frac{\text{ATR}_{t-1} \times (n - 1) + \text{TR}_t}{n}$$

When $n = 14$:

$$\text{ATR}_t = \frac{\text{ATR}_{t-1} \times 13 + \text{TR}_t}{14}$$

2.3 Calculation Steps Summary

1. Calculate the True Range for each bar
2. Take the simple average of the first 14 TR values as the initial ATR
3. Apply the recursive formula: ATR = (previous ATR x 13 + current TR) / 14
4. Repeat step 3 until the end of the series

---

III. Python Implementation

import numpy as np
import pandas as pd

def atr(high: pd.Series, low: pd.Series, close: pd.Series,
        period: int = 14) -> pd.Series:
    """
    Calculate the Average True Range (ATR).

    Parameters:
        high   : Series of high prices
        low    : Series of low prices
        close  : Series of closing prices
        period : Smoothing period, default 14

    Returns:
        ATR series (Wilder's Smoothing)
    """
    # Calculate True Range
    prev_close = close.shift(1)
    tr1 = high - low
    tr2 = (high - prev_close).abs()
    tr3 = (low - prev_close).abs()
    tr = pd.concat([tr1, tr2, tr3], axis=1).max(axis=1)

    # Apply Wilder's Smoothing (equivalent to EMA with alpha=1/period)
    atr_values = pd.Series(np.nan, index=close.index)

    # First ATR = SMA of the first 'period' TR values
    first_atr = tr.iloc[1:period + 1].mean()  # TR is valid from the 2nd bar
    atr_values.iloc[period] = first_atr

    # Recursively calculate subsequent ATR values
    for i in range(period + 1, len(close)):
        atr_values.iloc[i] = (atr_values.iloc[i - 1] * (period - 1) + tr.iloc[i]) / period

    return atr_values


# ============ Usage Example ============
if __name__ == '__main__':
    np.random.seed(42)
    n_days = 100

    # Generate simulated OHLCV data
    base_price = 100 + np.cumsum(np.random.randn(n_days) * 0.5)
    df = pd.DataFrame({
        'open':   base_price + np.random.randn(n_days) * 0.3,
        'high':   base_price + np.abs(np.random.randn(n_days) * 0.8),
        'low':    base_price - np.abs(np.random.randn(n_days) * 0.8),
        'close':  base_price,
        'volume': np.random.randint(1000, 10000, n_days)
    })

    # Calculate ATR
    df['ATR'] = atr(df['high'], df['low'], df['close'], period=14)

    print("=== ATR Results (Last 10 Rows) ===")
    print(df[['close', 'high', 'low', 'ATR']].tail(10).to_string())

    # ATR-based stop-loss example
    df['stop_loss_long'] = df['close'] - 2.0 * df['ATR']   # Long stop-loss
    df['stop_loss_short'] = df['close'] + 2.0 * df['ATR']  # Short stop-loss

    print("\n=== ATR Stop-Loss Levels (Last 5 Rows) ===")
    print(df[['close', 'ATR', 'stop_loss_long', 'stop_loss_short']].tail(5).to_string())

---

IV. Problems the Indicator Solves

4.1 Volatility Measurement

ATR is the most direct and pure volatility measurement tool. It unifies “bar body + gaps” into a single consideration, accurately reflecting how far market participants are willing to push the price within a given time frame.

• Rising ATR: market volatility is increasing, uncertainty is growing
• Falling ATR: the market is calming down, possibly in a consolidation phase

4.2 Stop-Loss Placement

ATR is widely used for adaptive stop-loss calculation. A common approach is to set the stop-loss at the entry price +/- $k \times \text{ATR}$.

Multiplier $k$	Characteristics
1.0	Tight stop, easily triggered, suitable for short-term trading
1.5	Moderate stop, balances protection and tolerance
2.0	Loose stop, gives price sufficient breathing room
3.0	Very loose, suitable for long-term trend following

4.3 Position Sizing

ATR can be used to normalize volatility across different instruments, enabling equal-risk position allocation:

$$\text{Position Size} = \frac{\text{Risk Per Trade}}{k \times \text{ATR} \times \text{Contract Multiplier}}$$

This is the core position sizing philosophy of the Turtle Trading rules.

4.4 Volatility Breakout Strategy

When the price breaks above or below $\text{Close}_{t-1} \pm k \times \text{ATR}$, it is considered a move beyond the normal volatility range and may signal the start of a new trend.

4.5 Applications in Other Indicators

ATR is a core component of the following indicators:

• SuperTrend: a dynamic trend-following stop-loss line built on ATR
• Keltner Channels: channels constructed using ATR instead of standard deviation
• ADX: uses ATR to normalize directional movement
• Chandelier Exit: ATR-based trailing stop-loss

---

V. Advantages, Disadvantages, and Use Cases

Advantages

Advantage	Description
Accurately captures gaps	True Range incorporates gaps into volatility, more accurate than High - Low
Adaptive	Automatically adjusts with market conditions, no manual parameter changes needed
Wide range of applications	Stop-loss, position sizing, volatility filtering, building block for other indicators
Cross-instrument comparability	Standardized volatility measurement for horizontal comparison across markets
Simple and stable calculation	Wilder’s recursive smoothing is computationally efficient and numerically stable

Disadvantages

Disadvantage	Description
No directional information	ATR cannot determine trend direction; must be combined with directional indicators
Lagging	As a moving average of historical data, it reacts with a delay to sudden volatility changes
Absolute value affected by price level	High-priced assets have larger absolute ATR; use ATR% for cross-instrument comparison
Does not distinguish up vs. down volatility	Both sharp rallies and sharp declines cause ATR to rise

Use Cases

• Best suited for: stop-loss placement, position sizing, volatility filters
• Well suited for: combining with trend indicators to gauge trend “energy”
• Not suited for: use as a standalone buy/sell signal generator

Comparison with Similar Indicators

Indicator	Volatility Measurement Method	Characteristics
ATR	Wilder’s smoothing of True Range	Accounts for gaps, smooth and stable
Standard Deviation (sigma)	Degree of closing price deviation from the mean	Sensitive to extreme values
Historical Volatility	Standard deviation of log returns	Annualized and comparable, suitable for options pricing
Parkinson Volatility	Uses only High/Low	More efficient than close-based methods, but ignores gaps