Sell in May and go away — everyone knows the saying. But do seasonal strategies actually work? The data says: Yes, with limitations.

The Halloween Effect: November to April

The best-known calendar anomaly: Stocks perform significantly better from November to April than from May to October. S&P 500 data since 1950:

November-April: Average +7.1% per half-year
May-October: Average +1.7% per half-year
Difference: 5.4 percentage points per year

The strategy: Invest in stocks from November to April, switch to bonds or cash from May to October. Historically, this simple rule beat buy-and-hold — at significantly lower risk.

The January Effect

Small-cap stocks rise disproportionately in January. Reasons: Tax-loss selling in December depresses prices, buyers return in January. The effect has weakened in recent years.

Monthly and Weekday Effects

Turn-of-the-month: Last 3 days and first 2 days of each month are the strongest trading days
Monday weakness: Mondays are historically the weakest weekday
Friday strength: Fridays tend positive — traders close short positions before the weekend

Earnings Seasonality

Certain months are more volatile:

January, April, July, October: Earnings Season = high volatility
August, December: Lowest volume = sudden movements with little liquidity

Santa Claus Rally

Last 5 trading days of the year + first 2 of the new year. Historically positive in 75% of cases.

AI and Seasonal Patterns

Our AI integrates seasonal factors as one of many inputs — but never relies on them alone. Context matters: Sell in May does not work every year. The AI recognizes when seasonal patterns apply and when they do not.

Bottom Line: Seasonal strategies are real and measurable — but they are not guarantees. Use them as filters or weighting factors, not as standalone strategies.