Kurtosis

What it measures

Kurtosis measures the "tailedness" of a return distribution — specifically, how much of the distribution's variance comes from rare, extreme outlier moves versus frequent, everyday small moves. A distribution with high kurtosis has "fat tails": extreme moves happen more often than a standard bell-curve (normal distribution) assumption would predict. This is a genuinely important risk consideration, since most standard risk models implicitly assume a normal distribution.

How readings are interpreted

High (positive) kurtosis means the security has historically been more prone to sudden, extreme price moves — both up and down — than a normal distribution would suggest, punctuated by otherwise calmer periods. Low kurtosis means returns have been more evenly and moderately distributed, without the same tendency toward sudden extreme surprises.

Conventional levels

• Kurtosis near zero (relative to a normal distribution baseline): return distribution roughly matches normal-distribution assumptions
• High positive kurtosis: fat tails — historically prone to sudden extreme moves, more risk of surprise than standard volatility measures alone would suggest
• Negative kurtosis: thinner tails than normal — historically fewer extreme surprises, more uniformly moderate moves

Where it works, and where it does not

A risk-assessment tool rather than a trading-signal tool — it doesn't tell you which way price will move, it tells you how much to trust conventional volatility-based risk estimates. Most relevant for options traders, risk managers, and anyone using standard deviation-based models who wants to sanity-check the fat-tail assumption.

Commonly read alongside

• Standard Deviation: Kurtosis is essentially a "trust check" on what standard deviation is telling you — high kurtosis means standard deviation alone is understating true tail risk
• Skew: The two are usually examined together for a fuller picture of distribution shape
• Ulcer Index: High kurtosis names are more likely to produce the kind of sharp, sudden drawdowns that Ulcer Index is designed to flag after the fact

Known limitations

• A purely statistical, backward-looking measure — high historical kurtosis doesn't guarantee future extreme moves, and low historical kurtosis doesn't guarantee their absence
• Not a directional signal at all, and easy to misuse as one
• Requires a reasonably long lookback to be statistically meaningful, so it reacts slowly to genuine regime changes

In practice

• Conventionally read as a risk-sizing input: high-kurtosis names carry more tail risk than their standard deviation alone would suggest
• Particularly relevant before earnings or other binary-outcome events, where fat-tail risk is structurally elevated
• Options traders specifically should weight high-kurtosis names' implied volatility assumptions with extra scrutiny
• Not a screening tool for entries — a background risk-context check on names you're already considering