OriginaveIs there a best month to buy ETFs?
Average S&P 500 returns for every calendar month, the best day of the month to invest, and whether “sell in May” survives contact with the data — from real SPY prices, measured in dollars, dividends reinvested.
Across 1993–2026, November has been the strongest calendar month for the S&P 500 as tracked by SPY: +2.45% on average, positive in 76% of years. September has been the weakest at -0.47%. That gap — 2.9 percentage points between the extremes — is the entire seasonal effect, and it comes with no promise of repeating next year.
Average return by calendar month
| Month | Average return | Positive years | Best | Worst |
|---|---|---|---|---|
| January | +0.76% | 61% | +8.01% (2019) | -8.21% (2009) |
| February | +0.00% | 53% | +6.93% (1998) | -10.74% (2009) |
| March | +0.95% | 68% | +9.69% (2000) | -12.49% (2020) |
| April | +2.00% | 74% | +12.70% (2020) | -8.78% (2022) |
| May | +1.20% | 71% | +6.32% (1997) | -7.95% (2010) |
| June | +0.44% | 65% | +6.96% (2019) | -8.36% (2008) |
| July | +1.42% | 65% | +9.21% (2022) | -7.88% (2002) |
| August | +0.06% | 65% | +6.98% (2020) | -14.12% (1998) |
| September | -0.47% | 55% | +8.96% (2010) | -10.49% (2002) |
| October | +1.68% | 64% | +10.91% (2011) | -16.52% (2008) |
| November | +2.45% | 76% | +10.88% (2020) | -7.47% (2000) |
| December | +0.97% | 70% | +6.69% (2010) | -8.80% (2018) |
403 complete calendar months of SPY, measured in dollars, dividends reinvested, 1993–2026.
Is there a best day of the month to buy ETFs?
Sort of — but not for the reason timing fans hope. Because markets drift upward over time, the earlier in the month you buy, the cheaper you get in on average. No mid-month sweet spot, no month-end discount: just drift. The honest conclusion is that the best day of the month to buy an ETF is the earliest one your payday allows.
| Always buying on… | vs the month’s average price | Cheaper than average |
|---|---|---|
| First trading day | -0.32% | 63% of months |
| The 5th (or next trading day) | -0.19% | 59% of months |
| The 10th (or next trading day) | -0.17% | 59% of months |
| The 15th (or next trading day) | +0.04% | 43% of months |
| The 20th (or next trading day) | +0.03% | 40% of months |
| The 25th (or next trading day) | +0.16% | 39% of months |
| Last trading day | +0.30% | 37% of months |
Each row: buy SPY on that day every month of 1993–2026 and compare the price paid with that month’s average price.
Does “sell in May and go away” hold up?
The famous half-year gap shows up in SPY’s data — but notice what the slogan skips: the “bad” summer half still averaged +4.16%. Selling in May meant giving that up, then paying the trading costs and taxes for the privilege.
Do other funds agree?
A seasonal pattern that only exists in one fund is noise. The cross-check: best and worst calendar months for the other big funds we track — same method, each over its own history.
The honest take
Seasonal averages describe the past; they don’t schedule the future. The average month of SPY returned +0.95% across 1993–2026 — so every month spent waiting for a “better” one had that expected cost, while the seasonal spread is a fraction of a single bad (or great) week. If a calendar rule helps you invest regularly, keep it. Just don’t expect it to beat simply starting today.
Frequently asked
What is the best month to buy the S&P 500?
November, in this data: across 1993–2026 it averaged +2.45% for SPY and finished positive in 76% of years. The seasonal edge is small and unreliable, though — being invested at all matters far more than which month you start.
What is the worst month for stocks?
September has the weakest average in this window: -0.47% per year for SPY, positive in only 55% of years. September's classic bad reputation is confirmed here.
Is there a best day of the month to buy ETFs?
Earlier beats later, but only because markets drift upward. Always buying SPY on the first trading day cost -0.32% vs the month's average price across 1993–2026; always waiting for the last trading day cost +0.30%. The best day of the month to buy an ETF is simply the earliest one your payday allows.
Does “sell in May and go away” actually work?
The gap is real in this data — November–April averaged +7.33% while May–October averaged +4.16%, and winter won 22 of 32 cycles. But the summer half still averaged a gain, so selling in May meant skipping +4.16% on average, plus trading costs and taxes.
Dig deeper
Follow one year month by month on the monthly returns dashboard, test a starting point with the what-if calculator, or see whether investing everything at once beats spreading it out under lump sum vs DCA. The numbers behind this page live on the SPY fund page and every pairing on the compare hub.