Sell in May / Statistics
SELL IN MAY STATISTICS
Which numbers actually tell you whether Sell in May is robust?
Averages alone are not enough. A useful seasonal test looks at consistency, downside, sample size and whether the result survives different historical lookbacks.
Five statistics worth checking
Average return
Useful summary, but vulnerable to outliers.
Median return
Shows the middle outcome and reduces the impact of extremes.
Win rate
Measures how often the seasonal window ended positive.
Worst year
Highlights downside hidden by attractive averages.
Sample size
Patterns based on too few observations deserve skepticism.
Lookback stability
Compare 10, 20 and 25+ year histories to detect fragile results.
Test the exact window yourself
Compare May–October with November–April on the symbol you care about.