Skip to main content
aifinhub
Risk & Portfolio Construction Formula

Skewness Formula

Skewness is the third standardized moment of a return distribution: the average cubed deviation from the mean, divided by the cube of the standard deviation. It measures asymmetry. Negative skew means a long left tail (occasional large losses), which most strategies exhibit and which deflates risk-adjusted scores that assume symmetry.

By AI Fin Hub Research · AI Fin Hub Team
Best Next MoveCalculators

Returns Distribution Analyzer

Paste a returns CSV. Histogram, normal-overlay, QQ plot, skewness, excess kurtosis, Jarque-Bera test, tail-weight index. See why Sharpe alone misleads.

CalculatorOpen ->

On This Page

Formula

Copy the exact expression or work through it step by step below.

Skew = ( (1/n) x sum_i (R_i - mean_R)^3 ) / sigma^3 where sigma = population standard deviation (divisor n)

Variables

R_i

Period return

Individual return observation in the sample.

mean_R

Mean return

Average of the returns, the center from which deviations are measured.

(R_i - mean_R)^3

Cubed deviation

Deviation from the mean raised to the third power. Cubing preserves sign, so large positive deviations and large negative deviations push the result in opposite directions; the imbalance between them is the skew.

sigma^3

Cube of standard deviation

Standardizes the third moment so skewness is unitless and comparable across assets. Using the population standard deviation (divisor n) matches the population skewness definition.

Step By Step

  1. 1

    Compute the mean and the population standard deviation of the returns.

    Returns -10%, +2%, +3%, +5% have mean 0% and standard deviation computed below.

  2. 2

    For each return, cube its deviation from the mean and average the cubes.

    Deviations -0.10, 0.02, 0.03, 0.05; cubes -0.001, 0.000008, 0.000027, 0.000125; mean = -0.00021000.

  3. 3

    Compute the cube of the population standard deviation.

    Variance = mean of squared deviations = (0.01 + 0.0004 + 0.0009 + 0.0025)/4 = 0.00345; sigma = 0.05874; sigma^3 = 0.0002027.

  4. 4

    Divide the mean cubed deviation by sigma cubed.

    -0.00021000 / 0.0002027 = -1.036.

Worked Example

Detecting left-tail risk in a four-return sample

Returns

-10%, +2%, +3%, +5%

Mean

0%

Deviations equal the returns (mean is 0). Cubes: (-0.10)^3 = -0.001, (0.02)^3 = 0.000008, (0.03)^3 = 0.000027, (0.05)^3 = 0.000125. Sum = -0.00084; mean = -0.00021. Variance = (0.01 + 0.0004 + 0.0009 + 0.0025)/4 = 0.01380/4 = 0.00345; sigma = 0.058737; sigma^3 = 0.00020266. Skewness = -0.00021 / 0.00020266 = -1.036.

Skewness of about -1.04, a clear negative skew. The single -10% return drives the long left tail: most outcomes are small positives, but the rare loss is large. Strategies with this profile (selling options, carry trades) look attractive on mean and Sharpe yet hide concentrated downside, which is exactly why the deflated Sharpe ratio penalizes negative skew.

Common Variations

Sample (adjusted) skewness: applies a bias correction factor of n^2 / ((n-1)(n-2)) for small-sample estimates, as used by spreadsheet SKEW functions.
Fisher-Pearson standardized moment: the population form shown here, the convention in most statistical libraries by default.
Nonparametric skew: (mean - median) / standard deviation, a quick robust proxy that avoids the third moment entirely.

Try These Tools

Run the numbers next

Sources & References

Related Content

Keep the topic connected

Planning estimates only — not financial, tax, or investment advice.