Skip to content
Menu
Finabase Learn

Glossary

Kurtosis

A measure of how often extreme outcomes occur compared with a bell curve; high kurtosis means fat tails and more frequent large moves.

On this page
  1. Markets have fat tails
  2. Why it matters
  3. Kurtosis and skewness
  4. Limits
  5. Related terms

Kurtosis measures the weight in the tails of a distribution. A normal distribution, the bell curve, has a kurtosis of 3. "Excess kurtosis" subtracts 3, so the bell curve scores zero. A positive figure means extreme results turn up more often than the bell curve predicts. Such distributions are said to have fat tails.

Markets have fat tails

Daily stock returns show high kurtosis. Most days are quieter than a bell curve would suggest, and a few are far more violent.

Why it matters

  • Risk models understate danger. A value-at-risk figure that assumes normal returns will be exceeded more often, and by more, than stated.
  • Leverage is riskier than it looks. Position sizes that survive ordinary swings can be wiped out by one extreme day.
  • Option prices reflect it. Far out-of-the-money options cost more than a bell-curve model implies.

Kurtosis and skewness

Skewness says which side the surprises lean toward. Kurtosis says how large and frequent surprises are on both sides. Stock returns over short periods tend to show negative skew and high kurtosis together.

Limits

  • The estimate is unstable. One extreme day can change it greatly, and adding or removing that day changes it back.
  • It mixes both tails. A high figure does not say whether the extremes were gains or losses.
  • Frequency matters. Kurtosis is much higher for daily returns than for annual ones.