Why skewness matters
Strategies with identical average return and volatility can have very different exposure to rare losses or gains, making skewness relevant to options and carry trades.
How it is applied
Analysts calculate the standardized third central moment and inspect the distribution, downside quantiles, drawdowns, and economic sources of asymmetry. Calculate the standardized third moment of a return distribution over a stated frequency and sample, then inspect the underlying observations. Pair skewness with volatility, kurtosis, drawdown, and scenario analysis. For options and nonlinear strategies, distinguish statistical return skew from implied-volatility skew across strikes.
Portfolio example
A strategy earns small gains in most months but occasionally suffers a very large loss. The pattern typically produces negative skewness. Two strategies both average 1% monthly with similar volatility. Strategy A has frequent small losses and occasional large gains, producing positive skew. Strategy B earns small gains most months but occasionally loses 15%, producing negative skew despite an attractive average before the loss.
How to interpret it
Negative skewness can signal crash-like downside, while positive skewness can reflect lottery-like payoffs. Desirability depends on expected return, price, and portfolio role. Positive skew means the distribution has a longer or heavier right tail; negative skew indicates more severe downside tail observations. It does not mean most returns are positive or negative. Investors often value positive skew and demand compensation for strategies that sell downside protection.
Limitations and common misconceptions
Sample skewness is unstable, especially with short or smoothed histories. One outlier can dominate it, and historical asymmetry may change. Sample skewness is unstable and sensitive to outliers, frequency, stale valuations, and limited history. It provides no timing or probability for the next extreme event. Smoothing can hide negative skew. A strategy can change shape after leverage, liquidity, or market structure changes. Editorial treatment should include a payoff example and explicitly warn that average return and volatility can conceal asymmetry. When reporting a statistic, state sample dates, frequency, formula convention, and whether observations are gross or net. A histogram or scenario table can make the interpretation more concrete.
Sources and further reading
- Introduction to Risk ManagementCFA Institute