Volatility clustering
Why do wild weeks in markets tend to come in runs, one large move following another?
In plain words
Volatility clustering is the tendency for large price moves to be followed by further large moves, and quiet periods by quiet periods. Markets go through stretches of heightened turbulence rather than distributing wild swings evenly across time.
Why it matters
Traders who size positions as if calm conditions will persist are repeatedly caught when they don't. Risk limits, stop distances and margin buffers calibrated to an average week can be overwhelmed in a cluster of turbulent weeks, even if the eventual price change is modest.
In a real week: ETH, week of 26 November 2018
ETH entered the week of 26 November 2018 after four weeks that included returns of -15.82% and -33.72%, and that week itself traded a 23.88% range with 105.0% annualised realised volatility. The worst day, 26 November, closed -6.72%, and the largest one-hour move, -3.1%, arrived on Thursday 29 November. Consecutive large moves, not isolated ones.
A common mistake
Assuming a period of calm means volatility has ended. Quiet stretches often sit between clusters, and the first large move in a new cluster is frequently mistaken for a one-off.