Kurtosis

Kurtosis is a measure of how much data resides in the tails on each side of a distribution compared to how much is hanging out in the middle. Kurtosis is measured against a standard which is how the tails of the Normal distribution compare to the data in the middle.

Distributions that either have longer tails than the Normal curve or that have fatter tails than the Normal curve are considered to have excess kurtosis. Distributions with low kurtosis have shorter or skinnier tails than a Normal curve. In terms of practical applications, distributions with excess kurtosis are more prone to producing extremely high or extremely results. Low kurtosis curves are much less likely to have those high return or low return results.