Chapter 29 · Volatility Indicators
Volatility indicators
True range, ATR, Bollinger Bands, Keltner and Donchian — and how ATR turns a stop from a guess into a measurement.
Volatility indicators answer how far, not which way. They are the family that most directly changes how you trade, because position size and stop distance both depend on how much a stock ordinarily moves — and neither depends at all on whether you are bullish.
True range, and why it exists
True range is the greatest of three distances:
- 1High − Low
- 2|High − Previous close|
- 3|Previous close − Low|
Worked: high 431, low 405, previous close 402.
- High − Low = 26
- |431 − 402| = 29
- |402 − 405| = 3
- True range = 29, not 26.
The overnight gap is counted. That is the entire reason true range exists rather than simply High − Low: a stock that opens Rs 3 above yesterday's close and then travels Rs 26 has moved 29 rupees away from where holders last stood, and a stop has to survive all of it.
ATR
ATR is Wilder's smoothed average of true range, conventionally over 14 periods:
| Formula | |
|---|---|
| First value | Simple average of the first 14 true ranges |
| Afterwards | ATR = (Previous ATR × 13 + Current TR) ÷ 14 |
Why a stop should be sized in ATR
Worked: sizing a stop and a position from ATR
Capital Rs 5,00,000. Risk on any one trade 1%, so Rs 5,000. The stock trades at Rs 428 with an ATR of Rs 12.
| Step | Calculation | Result |
|---|---|---|
| Stop distance | 2 × ATR = 2 × 12 | Rs 24 |
| Stop price | 428 − 24 | Rs 404 |
| Shares to buy | Rs 5,000 risk ÷ Rs 24 per share | 208 shares |
| Position value | 208 × 428 | Rs 89,024 |
| As a share of capital | 89,024 ÷ 5,00,000 | 17.8% |
Notice what happened: the position size fell out of the stock's volatility, not out of how much you liked the idea. A quieter stock with an ATR of Rs 5 would have given a Rs 10 stop and 500 shares — a larger position for the same rupee risk. This is the entire mechanism by which risk is held constant across different stocks.
Bollinger Bands and the squeeze
Bollinger Bands
| Component | Formula |
|---|---|
| Middle band | SMA(20) |
| Upper band | SMA(20) + (2 × standard deviation of the last 20 closes) |
| Lower band | SMA(20) − (2 × standard deviation) |
| %B | (Price − Lower) ÷ (Upper − Lower) |
| Bandwidth | (Upper − Lower) ÷ Middle |
Worked: SMA(20) = 420, standard deviation = 8. Upper = 420 + 16 = 436; lower = 420 − 16 = 404. Bandwidth = 32 ÷ 420 = 7.6%. With price at 428, %B = (428 − 404) ÷ 32 = 0.75 — three quarters of the way up the band.
The squeeze is the useful part. Bandwidth at a multi-month low means the stock has gone quiet, and quiet resolves. The bands say nothing about which direction it resolves in — that is a different family's job.
Keltner and Donchian
| Channel | Formula | Difference |
|---|---|---|
| Keltner | EMA(20) ± (2 × ATR(10)) | Built on ATR, so gaps count; smoother than Bollinger |
| Donchian | Highest high and lowest low of the last n periods | No averaging at all — it is a record of extremes |
| Bollinger squeeze test | Bollinger bands sitting inside the Keltner channel | A standard, objective definition of 'unusually quiet' |
Volatility is not direction
A stock at the upper Bollinger band is not overbought; it is far from its own average, which is exactly what a strong stock does. In a powerful trend price can ride the upper band for weeks. Read the bands as a measure of distance, and take direction from somewhere else.
Saved in this browser only — there is no account to create. Clearing your browser data clears your progress.