S&P 500 Seasonality: 75 Years of Patterns

The baseline: what an average year looks like

Before any seasonality pattern is useful, you need the baseline. Across 77years of S&P 500 history, here is what the average year has done.

Roughly three out of every four years is positive. Roughly one in four is negative. The negative years are not bunched, but they are real. Any seasonality pattern below has to be read against this baseline.

Month of year: when does the market actually go up?

Average monthly return for each calendar month, compounded year by year and averaged across 76 years of history. Read top to bottom: which months have done the heavy lifting, which months have dragged.

September has been weak with remarkable consistency. November has been strong. The reasons usually invoked include tax-loss selling in October giving way to a year-end rally, plus the return of institutional flows after summer. The why matters less than the consistency. Seven decades of data is enough sample size to take the pattern seriously without treating it as a guarantee.

"Sell in May, buy in November" — does it actually work?

The proverb claims the market does most of its work in the winter months, and the summer is dead weight. We ran the test across 77 years.

The data supports the proverb. November through April has historically returned +7.01% on average, while May through October has returned +2.11%. The winter half does the majority of the work.

That does not mean an investor should sell in May. The summer half is still positive on average, and the friction cost of jumping in and out twice a year (taxes, spreads, the risk of missing the recovery) typically eats the apparent advantage. The pattern is most useful as context for why some years feel stronger than others, not as a trading signal.

Day of week: a Monday-versus-Friday tale

Average daily return on each trading day of the week. The differences are small (per-day), but compound across thousands of trading sessions.

Wednesday has been the strongest day on average; Monday the weakest. The Monday effect (Mondays underperforming) is well-documented and persists across most modern data. The usual hypothesis: weekend news flow tends to be negative on net, and digestion happens on Monday. Wednesday and Friday tend to be the strongest days.

First half of the month versus second half

We split every calendar month into the 1st-15th and the 16th through month-end. Average daily return in each window:

The first half of the month has historically been stronger. One contributing factor is the inflow of fresh capital from month-beginning paycheck contributions and 401(k) allocations, which lands in markets in the first few trading days. The effect is real but small per day.

What the data says, in one screen

• From 1950 to 2026, the S&P 500 has finished positive in 74.0% of calendar years.
• The average S&P 500 calendar year over the last 77 years has returned 9.58%.
• November has been the strongest month on average (1.86%); September has been the weakest (-0.63%).
• Wednesday has been the strongest day of the week on average; Monday the weakest.
• The first half of the month (1st-15th) has historically averaged 0.05% per trading day, vs 0.02% in the second half.
• 'Sell in May, buy in November' tested: November to April has averaged 7.01% per year over 77 years, vs 2.11% for May through October. The full year has averaged 9.58%.