volatility

Mean-reversion trading signals are generated by computing rolling historical volatility over a configurable window, ranking it percentiles over a lookback period, and outputting integer signals (1, -1, 0) to trigger long, short, or neutral positions.

Yes, pandas and numpy compute rolling volatility and percentile rankings for cryptocurrency OHLCV data, allowing you to apply adjustable annualization factors and window lengths to generate trading signals for assets like BTC-USDT.

The best way to identify volatility regimes is ranking historical volatility into percentiles over a 120-day lookback, using configurable low and high percentile thresholds to systematically trigger long positions in low-volatility regimes and short positions in high-volatility regimes.

You configure annualization factors and window lengths by adjusting the parameters for the rolling historical volatility calculation, applying distinct annualization settings for equities versus crypto to accurately percentile-rank OHLCV time series.

A volatility percentile signal returns 0 when current historical volatility falls between your configured low and high percentile thresholds, indicating a neutral market regime where no mean-reversion long or short position is triggered.

Yes, the volatility signal approach works without relying on price direction alone by exploiting volatility mean-reversion, systematically identifying low-volatility regimes to go long and high-volatility regimes to short based solely on historical volatility percentiles.