On-Balance Volume (OBV)

I. What is OBV

On-Balance Volume (OBV) is a technical analysis indicator first introduced by American investment analyst Joseph Granville in 1963 in his book Granville’s New Key to Stock Market Profits. OBV was one of the earliest indicators to systematically incorporate volume into technical analysis and holds landmark significance in the history of technical analysis.

OBV belongs to the Cumulative Volume category of indicators and has no parameters to configure. Its core philosophy originates from Granville’s classic assertion:

“Volume precedes price” — changes in volume often appear before changes in price.

The calculation logic of OBV is extremely simple: each day, based on whether the closing price is higher or lower than the previous day’s close, the current day’s volume is added to or subtracted from a running cumulative total. Volume on up days is treated as “buying power,” and volume on down days is treated as “selling power.” By observing the trend of the OBV curve and its relationship with price, traders can determine the direction of capital inflows and outflows.

---

II. Mathematical Principles and Calculation

2.1 Core Formula

Let $C_t$ be the closing price and $V_t$ be the volume on day $t$. The recursive formula for OBV is:

$$OBV_t = \begin{cases} OBV_{t-1} + V_t & \text{if } C_t > C_{t-1} \\ OBV_{t-1} - V_t & \text{if } C_t < C_{t-1} \\ OBV_{t-1} & \text{if } C_t = C_{t-1} \end{cases}$$

The initial value is typically set to $OBV_0 = 0$ or $OBV_0 = V_0$.

2.2 Intuitive Understanding

The essence of OBV is a signed cumulative volume:

• When price rises, the entire day’s volume is considered buying power and added to OBV
• When price falls, the entire day’s volume is considered selling power and subtracted from OBV
• When price is unchanged, OBV remains the same

2.3 Calculation Steps

1. Set the initial OBV value (usually 0)
2. Starting from day 2, compare each day’s closing price with the previous day’s closing price
3. If up: add the day’s volume to OBV; if down: subtract the day’s volume from OBV; if unchanged: OBV stays the same
4. This produces a continuous OBV curve

---

III. Python Implementation

import numpy as np
import pandas as pd


def on_balance_volume(close: pd.Series, volume: pd.Series) -> pd.Series:
    """
    Calculate On-Balance Volume (OBV).

    Parameters:
        close  : Series of close prices
        volume : Series of volume

    Returns:
        Series of OBV values (cumulative volume)
    """
    # Calculate price change direction: up +1, down -1, unchanged 0
    direction = np.sign(close.diff())
    # Set the first value to 0 (no comparison possible)
    direction.iloc[0] = 0

    # OBV = cumulative sum of directional volume
    obv = (direction * volume).cumsum()

    return obv


def obv_with_ema(close: pd.Series, volume: pd.Series,
                 signal_period: int = 20) -> pd.DataFrame:
    """
    Calculate OBV and its EMA signal line.

    Parameters:
        close         : Series of close prices
        volume        : Series of volume
        signal_period : EMA period for the signal line, default 20

    Returns:
        DataFrame containing obv and obv_ema columns
    """
    obv = on_balance_volume(close, volume)
    obv_ema = obv.ewm(span=signal_period, adjust=False).mean()

    return pd.DataFrame({
        'obv': obv,
        'obv_ema': obv_ema
    })


# ============ 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(100000, 1000000, n_days).astype(float)
    })

    # Calculate OBV and signal line
    obv_df = obv_with_ema(df['close'], df['volume'], signal_period=20)
    result = pd.concat([df[['close', 'volume']], obv_df], axis=1)

    print("=== OBV Calculation Results (last 15 rows) ===")
    print(result.tail(15).to_string())

    # Determine OBV trend
    result['obv_trend'] = np.where(result['obv'] > result['obv_ema'], 'Bullish', 'Bearish')
    print("\n=== OBV Trend Statistics ===")
    print(result['obv_trend'].value_counts())

    # Detect OBV-price divergence (simple example)
    price_higher = df['close'].iloc[-1] > df['close'].iloc[-20]
    obv_higher = obv_df['obv'].iloc[-1] > obv_df['obv'].iloc[-20]
    if price_higher and not obv_higher:
        print("\nWarning: Bearish divergence detected -- price rising but OBV falling")
    elif not price_higher and obv_higher:
        print("\nWarning: Bullish divergence detected -- price falling but OBV rising")
    else:
        print("\nPrice and OBV are moving in the same direction, no divergence")

---

IV. Problems the Indicator Solves

4.1 Trend Confirmation

The most important use of OBV is confirming the validity of price trends:

• Price rising + OBV rising: The uptrend is supported by volume, the trend is reliable
• Price falling + OBV falling: The downtrend is confirmed by volume, the trend is reliable
• Price rising + OBV flat/falling: The rally lacks volume support, the trend may reverse
• Price falling + OBV flat/rising: Selling pressure is weakening, a rebound may be brewing

4.2 Divergence Signals — “Volume Precedes Price”

OBV divergence is the core application of Granville’s theory:

• Bearish Divergence: Price makes a new high but OBV does not — buying power is exhausted, warning of a decline
• Bullish Divergence: Price makes a new low but OBV does not — selling power is exhausted, warning of a rally

4.3 Breakout Confirmation

When price breaks through a key support or resistance level, observe whether OBV simultaneously breaks through its corresponding support or resistance level. If it does, the breakout is more reliable; if OBV fails to confirm, it may be a false breakout.

4.4 Typical Trading Strategies

1. OBV Trend Following: Draw trendlines on OBV and use OBV trendline breakouts as trading signals
2. OBV + Moving Average: Go long when OBV crosses above its EMA; go short when it crosses below
3. Divergence Strategy: Take contrarian positions when divergences appear
4. OBV Breakout Strategy: When OBV makes a new high/low, expect price to follow

---

V. Advantages, Disadvantages, and Use Cases

Advantages

Advantage	Description
Extremely simple	No parameters required, calculation logic is concise and intuitive
Leading nature	”Volume precedes price” — OBV changes typically appear before price changes
Effective trend confirmation	Price and volume moving in the same direction is an important sign of trend health
Long history, well-validated	Has undergone over half a century of market testing since 1963

Disadvantages

Disadvantage	Description
All volume attributed to one side	Simplistically assigns the entire day’s volume to buyers or sellers, overly crude
Sensitive to gaps	Large gaps can cause misleading jumps in OBV
Absolute value meaningless	OBV values depend on the starting point, making cross-instrument or cross-period comparisons impossible
Noisy in range-bound markets	In trendless markets, OBV fluctuates frequently, reducing signal readability

Use Cases

• Best suited for: Confirming trend strength and sustainability in trending markets
• Moderately suited for: Breakout confirmation alongside price pattern analysis
• Not suited for: Narrow range-bound markets (OBV will flip frequently); markets lacking reliable volume data

Comparison with Similar Indicators

Indicator	Calculation Method	Characteristics
OBV	Full volume add/subtract	Simplest, intuitive signals, but crude
A/D Line	Weighted by close position	Considers intraday price position, more refined
CMF	Weighted period average	Bounded indicator, suitable for short-term buying/selling pressure
MFI	Volume-weighted oscillator	Bounded 0-100, similar to volume-weighted RSI