Kurtosis

Last updated:

The fat-tailedness of your returns. Why "100-year events" keep happening every decade.

Quick Answer

What is Kurtosis?

Kurtosis measures the "fat-tailedness" of a return distribution. Normal distributions have kurtosis = 3 (excess kurtosis = 0). Equity returns typically have excess kurtosis of 3–10, meaning extreme moves happen far more often than a normal distribution would predict. Higher kurtosis = more black-swan tail risk.

Kurt = E[(R − μ)⁴] / σ⁴

Formula

Kurt = E[(R − μ)4] / σ4
Fourth standardized moment. "Excess kurtosis" = kurtosis − 3, where 3 is the kurtosis of a normal distribution.

Raise each deviation from mean to the fourth power, average, divide by σ⁴. Larger numbers indicate fatter tails.

Intuition — what is this number telling you?

Higher kurtosis = more "black swan" tail risk. The 1987 crash, 2008 collapse, and March 2020 sell-off were all 5+ sigma events under a normal distribution, meaning they should happen roughly once every 10,000+ years. Real markets produce them every decade.

Practical implication: any risk model assuming normal distributions (basic VaR, basic Sharpe statistical tests) underestimates real-world tail risk. Use PSR, fat-tailed VaR, and CVaR to compensate.

Worked example

Step-by-step

S&P 500 daily returns since 1926 have excess kurtosis of approximately +15. Highly fat-tailed.

BTC daily returns have excess kurtosis around +25. Even fatter tails — extreme moves are common.

What's a good Kurtosis value?

Excess kurtosis levels by asset:

AssetExcess Kurtosis
Normal distribution0 (reference)
Diversified equity3–10
Single stocks5–20
BTC15–30
Altcoins30–100+
Short volatility strategies50–200+ (extreme)

Related metrics

Skewness  ·  Value at Risk (VaR)  ·  CVaR (Conditional Value at Risk)  ·  Standard Deviation

Frequently asked questions about Kurtosis

What does high kurtosis mean for my portfolio?

Extreme moves (both up and down) happen more often than a normal-distribution model would predict. Tail risk is higher than naive VaR suggests.

Is excess kurtosis or kurtosis reported?

Foliolytic shows excess kurtosis (kurtosis − 3). 0 means normal-distribution-like; positive means fatter tails.

How does kurtosis affect Sharpe ratio?

High kurtosis means Sharpe statistics are noisy. Use Probabilistic Sharpe Ratio, which adjusts for kurtosis.

Does Foliolytic compute kurtosis?

Yes — excess kurtosis in the tail risk section.

See Kurtosis on your real portfolio

Upload your brokerage CSV — Foliolytic computes Kurtosis plus 70+ other metrics using real historical prices, real Treasury yields, and real CPI data. Free, no signup, your data stays in your browser.

Analyze your portfolio free →