The force index multiplies the session’s price change by its volume, so direction, size and participation land in one number. Raw values are noisy and scale with price and volume; smoothed versions are read against their own zero line.
What the force index measures
The force index combines three things in one number: the direction of a session’s move, its size and the volume behind it. Alexander Elder introduced it in his 1993 book Trading for a Living. A large gain on heavy volume produces a large positive value, and a small loss on light volume a small negative one.
Its units are dollars times shares. Because share prices and volume differ enormously between stocks, the absolute numbers only make sense against the same stock’s history.
Force index price and volume calculation
The raw, one-session force index is the change in the closing price from the previous session multiplied by the session’s volume. Smoothed versions apply an exponential moving average to the raw values.
Force index (1) = (Close - Previous close) × Volume
Force index (n) = EMA of Force index (1) over n sessions
EMA(t) = EMA(t-1) + (2 / (n + 1)) × (Force index (1) at t - EMA(t-1))Close, Previous close- Closing prices of the session and the session before, in dollars.
Volume- Shares traded in the session.
n- The smoothing length; this guide seeds each EMA with the simple average of its first n raw values.
The first raw value needs two sessions, because it uses the previous close. An unchanged close gives exactly 0, whatever the volume.
Force index smoothing periods
Elder smoothed the raw values with exponential moving averages: a 2-session EMA for short-term timing and a 13-session EMA for the intermediate trend. The 2-session version still swings sharply from one session to the next, while the 13-session version changes sign far less often. Longer settings, such as 50 or 100 sessions, are sometimes used to describe a broader balance of buying and selling.
Force index calculation: a worked example
The table computes the raw force index and both smoothed versions for a hypothetical stock. EMA(2) starts at session 3 and EMA(13) at session 14, when enough raw values exist for their seeds. Smoothed values are rounded to whole units for display.
| Session | Close | Change | Volume | Force index (1) | EMA(2) | EMA(13) |
|---|---|---|---|---|---|---|
| 1 | 45.00 | 1,000,000 | ||||
| 2 | 45.40 | +0.40 | 1,200,000 | 480,000 | ||
| 3 | 45.10 | -0.30 | 900,000 | -270,000 | 105,000 | |
| 4 | 45.80 | +0.70 | 1,500,000 | 1,050,000 | 735,000 | |
| 5 | 46.00 | +0.20 | 1,100,000 | 220,000 | 391,667 | |
| 6 | 45.70 | -0.30 | 800,000 | -240,000 | -29,444 | |
| 7 | 46.30 | +0.60 | 1,400,000 | 840,000 | 550,185 | |
| 8 | 46.50 | +0.20 | 1,000,000 | 200,000 | 316,728 | |
| 9 | 46.20 | -0.30 | 1,300,000 | -390,000 | -154,424 | |
| 10 | 46.60 | +0.40 | 1,100,000 | 440,000 | 241,859 | |
| 11 | 46.90 | +0.30 | 900,000 | 270,000 | 260,620 | |
| 12 | 46.40 | -0.50 | 1,600,000 | -800,000 | -446,460 | |
| 13 | 46.70 | +0.30 | 700,000 | 210,000 | -8,820 | |
| 14 | 47.00 | +0.30 | 1,000,000 | 300,000 | 197,060 | 177,692 |
| 15 | 46.20 | -0.80 | 2,000,000 | -1,600,000 | -1,000,980 | -76,264 |
| 16 | 46.00 | -0.20 | 1,700,000 | -340,000 | -560,327 | -113,940 |
Session 15 carries the largest raw value in the table: a $0.80 decline on 2,000,000 shares gives -1,600,000. That one session pulls EMA(13) from +177,692 to -76,264, below zero.
Force index zero line explained
Zero separates periods in which closes rose from those in which they fell, weighted by volume. Above zero, volume-weighted gains have outweighed losses over the smoothing window; below zero, losses have. A cross of the 13-session version through zero is a slower, rarer event than a cross of the 2-session version, which in the example changes sign several times.
Force index divergence example
A divergence appears when price and the smoothed force index disagree. In a typical bearish case, price sets a new high but the 13-session force index peaks lower than at the previous high, because the second advance came with smaller daily gains or lighter volume.
The illustration in this guide shows that hypothetical case: the second advance climbs more slowly on lighter volume, so the force index computed from those bars stays well below its first peak and turns negative once the decline begins. As with any divergence, it compares two past swings; it can persist while price keeps rising and does not say when, or whether, a reversal happens.
How to read the force index on a stock chart
Plot the 13-session force index below price with a zero line, and add the 2-session version for a short-term view. Read the side of zero and the trend of the line first. Then look for extremes compared with the stock’s own history, which mark sessions where large moves met heavy volume, and inspect those sessions on the chart; raw spikes often coincide with gaps or news-driven sessions.
Limitations of the force index
The force index is easy to compute, but its scale and its reliance on single sessions need care. Keep these limits in mind.
- Values grow with share price and volume, so they cannot be compared across stocks or over periods in which typical volume changed a lot.
- It uses close-to-close changes only: a session that swings widely but closes unchanged scores 0.
- Single sessions dominate the raw values, as session 15 does in the example.
- Smoothing adds lag, and the EMA seed changes the early values.
- Stock splits change both price changes and volume, so use split-adjusted data consistently.
- It describes past sessions and does not forecast the next one.
Data cadence and what CleverScan can express
On daily bars the force index updates once per session; on weekly bars the change is measured week over week. TradeClever’s Stock Charts do not plot the force index, but they show the daily closes and volume pane it is built from. CleverScan can express the raw one-session value, because change is the latest price change in dollars and volume is the latest session’s shares.
change * volume > 0 and volumeRatio > 1.5Stocks whose latest one-session force index is positive and whose volume is more than 1.5 times the 20-session average. CleverScan does not compute the 2- or 13-session smoothed force index.
Continue with
Sources
- Alexander Elder, Trading for a Living (Wiley, 1993)
Read our data and methodology notes for calculation context. This guide explains research tools and observations. It does not recommend a trade or investment.