Market Seasonality: Sell in May, the Santa Rally, and the January Effect

Calendar effects are persistent but modest statistical tilts tied to specific dates. They are not laws — they are edges that compound when treated as tilts, not rules.

The "Sell in May" Effect

The adage reflects the historical outperformance of November–April versus May–October on US and UK equities. Since 1950 (per the Stock Trader's Almanac), Nov–Apr produced the majority of the S&P 500's annual gain; May–Oct has been flat-to-negative. The effect appears in 36 of 37 developed markets studied (Jakobsen and Bouman, 2002).

Do not literally sell on May 1. Reduce gross exposure 10–20% into May–October, tilt toward defensives (utilities, healthcare, staples), and re-add risk in late October.

The Santa Rally

Equities tend to rise in the last five trading days of December plus the first two of January. Since 1969, the S&P 500 was positive in this 7-day window ~76% of the time, averaging ~1.3% (versus ~0.2% for any random 7-day window). Use a long-biased tactical tilt into the final week; do not lever aggressively. A failed Santa rally (negative 7-day return) is a mild warning for the coming year, not a forecast.

The January Effect

Small-cap outperformance in January, driven by December tax-loss selling, has weakened substantially in large-caps but persists modestly in micro-caps. The "January Barometer" has been directionally correct ~70% of the time since 1950, with notable failures (2016, 2018). Tilt toward small-caps in early January and exit by month-end if the effect fails; treat the barometer as a sentiment read, not a sizing input.

Other Tilts and Decay

Turn-of-the-month (first and last few days) and the pre-FOMC 24-hour drift (Lucca and Moench, 2015) are documented but have weakened since publication. Calendar effects erode once traded — trade the strongest, least-publicized effects and assume decay. Backtest any rule on 20 years; if the hit rate is below 60% or the average return is below the transaction-cost threshold, the effect is too weak to trade.

Action Points

1. Treat seasonality as a 10–20% exposure tilt, never an all-in/all-out switch.
2. Combine seasonal tilts with the prevailing trend; do not fight a Stage 4 markdown because "it's December."
3. Re-test each rule annually; abandon rules whose hit rate has decayed below 55%.

Seasonality is a tailwind, not an engine. Use it to lean into a position the trend already supports; never to override the trend.