Kurtosis is a statistic that describes how much of a distribution’s probability mass lies in its tails (extreme values) relative to its center. In practice, kurtosis helps you understand whether a dataset (for example, a series of asset returns) is prone to more frequent extreme outcomes than a normal (bell‑shaped) distribution.
Key intuition
– High kurtosis (heavy or “fat” tails): more observations far from the mean → greater chance of extreme gains or losses.
– Low kurtosis (light tails): fewer extreme observations than a normal distribution.
– Normal (reference) distribution has kurtosis 3 (or excess kurtosis 0). Excess kurtosis = kurtosis − 3.
Types of kurtosis
– Mesokurtic: kurtosis ≈ 3 (excess ≈ 0). Example: Normal distribution.
– Leptokurtic: kurtosis > 3 (excess > 0). Heavy tails; more extreme events.
– Platykurtic: kurtosis < 3 (excess 0 → leptokurtic (fat tails).
– Excess kurtosis 0 = fat tails; <0 = thin tails.
– Kurtosis is sensitive to outliers and small-sample noise.
Source
Material adapted and summarized from Investopedia
Editor’s note: The following topics are reserved for upcoming updates and will be expanded with detailed examples and datasets.