Glossary
Skewness
A measure of how lopsided a distribution of returns is: whether extreme losses or extreme gains are the more likely surprise.
Skewness describes the asymmetry of a set of returns. Standard deviation says how spread out returns are. Skewness says whether the spread is even on both sides of the average.
- Zero skew: symmetric, like the bell curve.
- Negative skew: many small gains and a few large losses. The long tail is on the left.
- Positive skew: many small losses and a few large gains. The long tail is on the right.
Examples in markets
- Negative: stock indexes over short periods, which fall faster than they rise. Selling options, lending to risky borrowers, and carry trades earn steady income and occasionally lose a great deal.
- Positive: buying options, lottery tickets, and venture capital, where most bets lose a little and a few pay many times over.
Why it matters
Measures built on standard deviation, such as the Sharpe ratio, treat both sides alike. A negatively skewed strategy can show an excellent Sharpe ratio for years before its large loss arrives. Many investors accept lower average returns for positive skew, which is one explanation offered for the poor average results of lottery-like stocks.
A naming trap
Volatility skew in options is a different thing: it is the pattern of implied volatility across strikes.
Limits
Skewness is estimated from past data, and rare events dominate the estimate. A few years with no crash will show little skew for a strategy that has plenty.
