MAIX8 Research
← All lessons
Core

Is this quiet, or does it only feel quiet?

A coin shows 30% volatility. Is that high or low? Compared to what?

The idea

Volatility numbers are meaningless as absolutes. 30% is sleepy for a small cap and wild for an index. The fix is to rank today's reading against the same asset's own history and quote a percentile. Compression matters because volatility clusters and mean-reverts, so a reading near the bottom of its own distribution raises the odds of expansion — but it tells you the size of the next move, never the direction.

The formula

realized vol = stdev(daily log returns over 30d) x sqrt(365)
compute it for every day in the history, then:
percentile = count(past readings < today) / count(all readings)

Worked example — BTC

Measured on Binance spot daily candles at . Re-run the formula on current data and you should reproduce the method, not necessarily these figures.

30-day realized volatility29.7%
Percentile vs own history10.9th
Sample size970 readings
Median of that history44.2%
now 29.7%21.1983.88
BTC 30-day realized volatility: full history, and where today sits

29.7% sounds unremarkable until it is ranked: it is the 10.9th percentile of 970 readings. The number only became information after it was compared to its own past.

Common mistake

Treating compression as a direction signal. It forecasts magnitude. Anyone telling you a squeeze must resolve downward is adding an opinion to a measurement.

Do it yourself

Export daily candles for any pair, apply the formula above, and compare your number to the one on this page. If they disagree, one of us is wrong and it is worth finding out which.