Glossary
Exponentially weighted moving average
An average that weights recent observations most, with weights that fall geometrically with age. Applied to squared returns it estimates volatility.
For variance the update is var_t = lambda x var_(t-1) + (1 - lambda) x r_(t-1)^2. The RiskMetrics system popularised lambda = 0.94 for daily data. A weight halves every ln(0.5) / ln(lambda) periods.
EWMA reacts to new shocks and forgets old ones smoothly. It has no long-run level, so multi-day forecasts stay flat.
Example: yesterday's variance is 0.0001 and the return was -2%. The new variance is 0.94 x 0.0001 + 0.06 x 0.0004 = 0.000118, a volatility of 1.086% a day.