Historical vs. Implied Volatility
There are two types. Historical Volatility (HV) is computed from past price data - typically the annualized standard deviation of 14-day or 30-day log returns. Implied Volatility (IV) is derived from option prices and represents the market's forward expectation; the DVOL index from Deribit BTC options is the primary benchmark. HV reflects realized facts while IV captures market expectations. The gap between them (the volatility risk premium) provides a potential edge for trading strategies.
Volatility Clustering
Crypto price movements exhibit 'volatility clustering': high-volatility days tend to follow other high-volatility days, and periods of calm tend to persist. This autocorrelation is captured by GARCH-family models and used for next-day volatility forecasting. Systematic strategies exploit this property through vol-targeting: expanding positions during low-vol regimes and contracting during high-vol regimes to maintain constant portfolio risk.
Volatility and Position Sizing
ATR (Average True Range) based sizing is a standard method for incorporating volatility into position decisions. By measuring each asset's 14-day ATR and sizing positions so that a 1-ATR move equals a fixed percentage of account value (e.g., 1%), risk is equalized across assets of differing volatility. Without ATR sizing, a portfolio holding both crypto (5-10% daily moves) and traditional assets (sub-1% moves) would have massively skewed risk exposure.
Volatility Regimes
Crypto markets exhibit cyclical volatility regimes. Common patterns include elevated volatility around Bitcoin halving events, altcoin volatility spikes during DeFi summers or meme-coin booms, and event-driven volatility from regulatory announcements. Detecting regime transitions and adapting strategy parameters accordingly is a key design challenge in systematic crypto trading.