Sell in May and the Halloween Indicator

"Sell in May and go away" is a decades-old trading adage claiming that equity returns from May through October are meaningfully weaker than returns from November through April — the "winter" half of the year. Skeptics long dismissed it as folklore, but academic studies covering dozens of countries and over a century of data have found the effect is real and statistically significant more often than chance alone would explain, earning it the more formal name the "Halloween indicator" (since the winter half roughly starts around Halloween).

No single cause is fully agreed on. Candidate explanations include summer vacation reducing trading activity and volume, seasonal affective effects on risk appetite, and institutional patterns like year-end portfolio rebalancing concentrating buying in the winter months. None of these fully explains why the effect appears so consistently across countries with very different holiday calendars and cultures, which is part of why it remains a puzzle rather than a settled mechanism.

As a concrete illustration, one long-sample study found average winter-half returns roughly several percentage points higher than average summer-half returns for a broad market index — a gap far too large and too consistent to write off as pure noise, though considerably smaller in more recent decades than in the earlier part of the sample.