# Volatility Clustering

*Risk & Portfolio — Finicade finance glossary*

Volatility clustering is the observation that turbulent days follow turbulent days and calm follows calm. Returns themselves are nearly unpredictable, but their magnitude is strongly autocorrelated — which is the single most robust empirical fact in finance. It's the entire justification for GARCH and EWMA models, and the reason a risk estimate that weights the last month equally with the last decade will always be late to a crisis.

**Also known as:** vol clustering, persistence of volatility

**Related terms:** [GARCH](https://finicade.com/glossary/garch), [EWMA](https://finicade.com/glossary/ewma), [Historical Volatility](https://finicade.com/glossary/historical-vol), [Volatility](https://finicade.com/glossary/volatility), [Mean Reversion](https://finicade.com/glossary/mean-reversion)

**Taught in:** Risk Arena — Volatility That Moves: EWMA & GARCH

Source: https://finicade.com/glossary/volatility-clustering