Classify the current volatility regime of each asset
Describes how turbulent each asset is. It reads up to 200 daily candles and computes ATR(14) as a percentage of price and Bollinger Band width percentage. Thresholds on those two metrics classify each asset into a low, medium, high, or explosive volatility regime, and the run is summarized to a dominant or mixed regime. It describes where volatility is elevated or compressed; it does not recommend a position size.
Two backward-looking movement measures and a threshold rule. Neither is the conventional standard deviation of returns, and the difference is the lesson.
ATR(14), J. Welles Wilder Jr. (1978). True range per bar is TR_t = max(H_t - L_t, |H_t - C_{t-1}|, |L_t - C_{t-1}|). Including the previous close is what makes it gap-aware: a market opening far from its last close registers movement a simple high-minus-low would miss. For the first bar, with no previous close, TR is taken as H - L. The seed is ATR_14 = (1/14) * sum(TR_1..TR_14), then smoothed recursively as ATR_t = (13*ATR_{t-1} + TR_t)/14. Reported here as a percentage, 100 * ATR_t / C_t, so assets at different price levels can be compared. ATR% is a normalised RANGE-BASED measure of movement.
Bollinger BandWidth = (upper - lower) / middle. With bands at +/- 2 standard deviations of the close around a 20-period SMA, that is 4 * sd(close) / SMA20. It is a normalised measure of the dispersion of price LEVELS - not of returns, not annualised, and not the volatility used in option pricing or a Sharpe ratio. Whether the standard deviation uses divisor N or N-1 is an implementation choice. Dividing by the middle band makes it scale-free; it does not by itself make two assets statistically comparable.
The low/medium/high/explosive labels come from thresholds on these two numbers. The cut-offs and what happens when the two measures disagree belong to the run's published classification rule; the concepts here explain the measurements, not a substitute threshold set. A threshold is a discretisation of a continuum: assets either side of one are barely different, and labels are comparable only within a single threshold set.
Both inputs are computed from history and describe what has happened. Volatility clustering - large moves tending to follow large moves - is a well-documented empirical regularity, which is why recent movement carries some information about near-term movement. That is a statement about magnitude, not direction, and it is a property of return volatility, so it transfers to these two proxies only approximately.
Published reports are readable without an account. The newest is Ether (ETH) and XRP extend a month-long slide as Bitcoin falls below $75,000: trend, momentum and news (Nov 9, 2026). Create an account to run research on the stocks and crypto you choose.