Higher Moments: Skewness and Kurtosis
What You'll Learn
You'll discover how the third and fourth moments of a distribution capture its asymmetry and tail heaviness—properties that mean and variance alone cannot reveal.
Beyond Mean and Variance
You already know that expectation (the first moment) tells you the center of a distribution, and variance (related to the second central moment) tells you its spread. But distributions can differ in other important ways:
- Skewness measures asymmetry: Does the distribution lean more to one side? Are there longer tails on the left or right?
- Kurtosis measures tail behavior: Does the distribution have heavy tails (prone to extreme values) or light tails?
The Third Moment: Skewness
Skewness is based on the third central moment:
where μ = E[X] and σ is the standard deviation. The cubing preserves the sign of deviations, so:
- Skew > 0: Right-skewed (long right tail, mean > median)
- Skew = 0: Symmetric (like the normal distribution)
- Skew < 0: Left-skewed (long left tail, mean < median)
Analogy: Imagine a hiking trail. Positive skew means a gradual uphill with a steep drop-off on the right; negative skew is a steep climb then a gentle descent.
The Fourth Moment: Kurtosis
Kurtosis is based on the fourth central moment:
Higher kurtosis means more probability mass in the tails and at the peak—indicating greater risk of extreme outcomes. Often we use excess kurtosis = Kurt - 3, where 3 is the kurtosis of a normal distribution.
Key Takeaway: Skewness (third moment) reveals asymmetry in a distribution, while kurtosis (fourth moment) measures the heaviness of its tails—both essential for understanding risk and shape beyond just center and spread.