Volatility
Volatility (standard deviation of returns)
Historical volatility explained: log returns, standard deviation, calm and stormy markets, and volatility alerts in Subtrigger.
3 min readUpdated
Volatility here is the statistician's measure of how jumpy the price is: the standard deviation of bar-to-bar returns, in percent. It is what options traders call historical or realized volatility, measured over a short window. Use it to tell calm markets from stormy ones, and to compare coins and stocks on the same scale.
What it measures
For each pair of consecutive closes, the log return:
Then the standard deviation of the last N returns, in percent:
Log returns are almost the same as percent changes for small moves, and they add up nicely over time. The result is the size of a typical one-bar move.
Try it
Volatility hovers low in the calm phases and spikes in the drop, then decays as the market calms down.
How Subtrigger computes it
| Crypto | US stocks and ETFs | |
|---|---|---|
| Period | 20 returns | 20 returns |
| Bar | 1 minute | 5 minutes |
| First value | right away on Binance, after 21 bars elsewhere | after 21 bars |
| Stream | ind:vol_20_1m@binance.miniticker:ETHUSDT:value | ind:vol_20_5m@nasdaq.etf:QQQ:value |
It uses 21 closes (20 returns), the forming bar included, and the population standard deviation.
How to read it
| Volatility 20 ยท 1m (BTC) | Market |
|---|---|
| below 0.04 % | very quiet, often before a breakout |
| 0.05 โ 0.12 % | normal |
| 0.15 โ 0.3 % | active: news, a trend leg |
| above 0.3 % | very volatile: liquidations, crashes, squeezes |
Smaller coins live higher on this scale; large ETFs live lower. Watch a market for a day to learn its normal range.
Alert ideas
The market waking up:
Alert idea
ETHUSDT Volatility 20, % ยท 1m ยท Binance > 0.25
A very quiet stretch on BTC, the kind that often precedes a big move:
Alert idea
BTCUSDT Volatility 20, % ยท 1m ยท Binance < 0.03
Volatility vs ATR
| Volatility | ATR | |
|---|---|---|
| Unit | percent | price |
| Based on | close-to-close returns | high, low, gaps |
| Comparable across markets | yes | no |
| Best for | regimes, filters, statistics | stops, targets, sizing |
For pros
- Volatility clusters. Calm follows calm, storms follow storms (GARCH effects). A spike rarely disappears in one bar; plan for the decay.
- Microstructure noise. At 1-minute horizons, bidโask bounce adds to measured volatility, more on less liquid exchanges. Compare like with like.
- Volatility of volatility. The change of this stream (or its z-score on your side) flags regime shifts.
Related
- ATR: range per bar in price units.
- Bollinger Bands: built on the standard deviation of prices.
- ROC: the size of a move over N bars, to compare with volatility.
FAQ
Is this the same as implied volatility?
No. Implied volatility is priced into options and looks forward; this is measured from past prices. They are on the same scale after annualizing.
Why log returns?
They are symmetric for up and down moves of the same size in log terms, they add over time, and for small moves they are almost equal to percent changes.
What volatility is "normal" for BTC?
On 1-minute bars, usually around 0.05โ0.12 %. It varies a lot between days and hours; watch it for a while before setting alerts.
Keep reading
- VolatilityATRAverage true range explained: true range, Wilder smoothing, stop distances, position sizing and ATR alerts for volatility.
- VolatilityBollinger BandsBollinger Bands and %B explained: the bands, squeezes, band walks, mean reversion and Bollinger alerts on crypto and stocks.
- StatisticsZ-scoreThe price z-score explained: how many standard deviations from the mean, ยฑ2 extremes, mean reversion and z-score alerts.