Kurtosis
How fat the tails of the return distribution are — how often extreme days happen.
Formula
Excess kurtosis = mean((r − mean)⁴) ÷ σ⁴ − 3
How to read it
Shown as excess kurtosis, where a normal distribution reads 0. Above 0 means extreme days are more common than a bell curve predicts.
How to use it
A warning that volatility and VaR understate the risk.
Limitations
Unstable over short samples.
Also called: kurtosis, excess kurtosis.