September Really Is the Stock Market's Worst Month — Here's What 75 Years of Data Actually Shows

The Statistic Is Real — the Panic It Usually Causes Isn't Justified

Every year around this time, some version of the same headline resurfaces: September is historically the worst month for stocks. It's not clickbait — it's an accurate description of the data. Since 1950, the S&P 500 has averaged a -0.6% return in September, with 41 negative Septembers against 34 positive ones. The NASDAQ Composite's record since 1985 is even weaker: an average -0.9% September return, with 22 losing years against 18 winning ones. No other calendar month shows a comparably consistent pattern of underperformance across both indices.

See current market data on SmartRates →

Why This Happens (and Why Nobody Fully Agrees)

There's no single, settled explanation for the September effect, which is itself informative — a well-understood, exploitable market pattern tends to get arbitraged away once enough investors try to trade around it. The leading theories include institutional portfolio rebalancing after summer, mutual funds realizing capital gains ahead of their fiscal year-end (many use an October 31 fiscal year-end, prompting September selling), and a general return of trading volume and risk-aversion after a quieter August. None of these fully explains the pattern on its own, and the honest answer is that "September is weak" has held up statistically for decades without a fully agreed-upon cause — which is a meaningfully different situation from a pattern driven by a specific, identifiable, and still-active mechanism.

Why the Long-Term Numbers Matter More Than the Seasonal Ones

Here's the context that usually gets left out of the alarming headline: over the same 1950–2024 period where September averaged -0.6%, the S&P 500's overall annualized return was about 8.2%. The NASDAQ's annualized return since 1985 is roughly 11.5%, despite its weaker Septembers. A single month averaging a modest loss, sitting inside a multi-decade record of strong annual gains, is not a reason to restructure a long-term portfolio — it's a data point that's statistically true and financially close to irrelevant for anyone investing on a multi-year horizon.

What the Data Doesn't Tell You

An average masks the range underneath it. "41 negative years vs. 34 positive" means September has still been a *positive* month for the S&P 500 in roughly 45% of the years measured — hardly a reliable one-way bet even within the pattern itself. And an average -0.6% loss includes both mild single-digit pullbacks and the handful of genuinely sharp September selloffs (2001, 2008, and 2022 among them) that pull the average down disproportionately relative to what a "typical" September actually looks like in a calm year. Treating -0.6% as what to expect this September specifically overstates how predictable any individual year actually is.

What to Actually Do With This Information

• Don't sell based on the calendar alone. Selling every August 31st and buying back in October, purely to dodge September's historical average, has to overcome trading costs and tax consequences on any gains realized — and because the effect isn't reliable enough year to year, backtests of this exact strategy don't consistently beat simply staying invested.
• If you dollar-cost average, keep doing it. A weaker average month is, if anything, a slightly better month to be a net buyer on a fixed schedule — you're no more able to predict which September you're in than anyone else, and a disciplined compound interest-style contribution schedule benefits from not trying to time individual months.
• Use any actual September dip as a rebalancing check, not a panic trigger. If a pullback pushes your portfolio meaningfully off its target allocation, that's a legitimate reason to rebalance — but rebalancing back to your plan is different from abandoning the plan because of a headline about historical averages.
• Run your own numbers rather than reacting to the aggregate. A brokerage investment return calculator can show you how a hypothetical September dip and recovery actually affects a long-term contribution schedule — for most multi-year horizons, the effect is smaller than it feels in the moment.

If You're Choosing Where to Invest, Not Just When

Seasonality aside, the more consequential decision for most long-term investors is which platform and fee structure they're using, not which month they buy in. Fidelity, Charles Schwab, and Vanguard all offer commission-free trading on stocks and ETFs with no account minimums for a standard brokerage account, which matters far more to long-run returns than trying to dodge a historically weak month.

Frequently Asked Questions

Should I sell my stocks before September to avoid the seasonal drop?

Historical averages don't reliably predict any single year, and September has still been positive in close to 45% of years measured. Trading costs, potential tax consequences, and the risk of missing a positive September make this a weak strategy compared to staying invested according to your actual plan.

Is the September effect the same as "sell in May and go away"?

They're related but distinct seasonality patterns. "Sell in May" refers to broader historical underperformance across the May–October stretch as a whole; the September effect is a narrower, specific pattern concentrated in that single month, and the two are often discussed together but measured separately.

Does the September effect apply to individual stocks, or just the index as a whole?

The data cited here is for broad indices (S&P 500, NASDAQ Composite). Individual stocks can behave very differently in any given September depending on company-specific news, earnings timing, and sector trends — index-level seasonality doesn't reliably predict a single stock's behavior.

What years were the worst Septembers on record?

2008 and 2001 stand out as the sharpest September declines in the modern data, both coinciding with broader financial crises rather than seasonality alone — a reminder that the average is pulled down by a small number of severe, event-driven years rather than reflecting a "typical" September decline.

This article discusses historical market seasonality and is not personalized investment advice; past performance does not predict future results, and any investment decision should account for your own timeline, goals, and risk tolerance — see current market data and tools before making changes to your portfolio.