Is the October effect real?

Only as folklore. October’s average and median returns are unremarkable. What actually sets the month apart is variance: October has one of the highest standard deviations of any month, so the rare bad Octobers are extremely bad while most are forgettable.

What is the worst month for the stock market historically?

September. It is the only month with a slightly negative average S&P 500 return going back to 1928. October gets the fear and September does the damage, which is a decent reminder that reputation and statistics do not always agree.

Why do stock market crashes seem to happen in October?

Nobody has proven a cause. The usual explanations are Q3 earnings season, mutual funds selling ahead of an October fiscal year end, tax-loss harvesting, and reflexivity, where investors who expect October to be bad hedge in ways that amplify declines. Three events across a century is not enough data to settle it.

Should I sell my stocks before October?

Almost certainly not. You would exit dozens of normal or positive Octobers to avoid a handful of bad ones, and you would need to know in advance which is which. The people who made money on October crashes were holding cash and buying, not timing the calendar.

The October Effect: Real Pattern or Folklore

The October effect is the belief that stocks tend to crash in October. The folklore is real, the average return is not unusually bad, and both facts coexist because of how human memory works. October has hosted some of the worst single days in stock market history. It has also had plenty of forgettable, slightly-positive months. The reputation is built on a few catastrophic outliers, not a typical experience.

The way I think about it, October is the month where if something is going to break, it tends to break visibly. The crashes that defined the modern market memory all clustered there: Black Thursday 1929, Black Monday 1929, Black Tuesday 1929, Black Monday 1987, the 2008 cascade. Five of the ten worst single-day percentage drops in S&P 500 history happened in October. So the fear isn't crazy. It just isn't most Octobers.

Plain English

“Seasonality” in markets means the tendency for returns to differ by calendar month. The October effect is one of the more famous claimed seasonal patterns, alongside the “Sell in May and go away” idea and the January effect.

What the Average October Looks Like

Run the numbers on the S&P 500 since 1928 and October's average monthly return is roughly 0.5%, slightly positive, slightly worse than the all-month average of about 0.6%. Median is higher than average. October's standard deviation, though, is among the highest of any month. Translation: most Octobers are unremarkable, but the bad ones are very bad. The mean is dragged around by a few outlier years.

Compare to September, which is statistically the worst month with an average return slightly negative across the same window. September is the actually-worst month by average. October just has the famous crashes. Reputation and statistics don't always agree.

The Crashes That Built the Reputation

Three October events did most of the work:

1929: Black Thursday (October 24, the Dow down about 11% intraday before a late recovery), Black Monday (October 28, the Dow down 12.8%), Black Tuesday (October 29, down another 11.7%). The week that started the Great Depression.
1987: Black Monday (October 19), the S&P down 20.5% in a single day, still the worst one-day percentage drop in the index's history.
2008: The week of October 6, S&P 500 down 18%, the worst week of the financial crisis. Followed by extreme volatility through the rest of the month.

Three crashes across nearly a century is not a pattern in any statistical sense. But each was so severe that the month became permanently associated with collapse. Mark Twain's line lands here: “October. This is one of the peculiarly dangerous months to speculate in stocks. The others are July, January, September, April, November, May, March, June, December, August, and February.”

Why People Argue Crashes Cluster in October

Several theories, none of them airtight:

Earnings season pressure. Q3 earnings come out in October. Bad guidance, especially after a long bull, can trigger waves of selling.
Fiscal year-end positioning. Many mutual funds have an October fiscal year end and rebalance ahead of it.
Tax-loss harvesting. Selling losers before year end has historically been concentrated in late fall, with October as a starting line.
Memory and reflexivity. Investors expect October to be bad, become hypersensitive in October, and small declines turn into bigger ones because of preexisting fear.

The reflexivity argument is the most interesting and the hardest to test. Markets that expect volatility tend to produce volatility, because hedging activity itself creates flow. October's reputation may now be partly self-sustaining.

What This Means for Actually Investing

Almost nothing. The base-rate October return is positive. The historical worst-cases were dramatic but rare. Selling in October to avoid crashes means missing dozens of normal Octobers to dodge a few bad ones, and you'd need to know in advance which Octobers were going to be the bad ones. Nobody does.

For long-term holders, calendar-based timing has consistently underperformed buy-and-hold across every reasonable backtest. The October crash narrative is a great campfire story and a terrible investment thesis. The ones who actually benefited from October crashes are the people who already had cash and bought into them, not the people who tried to predict them by month.

Takeaway

October hosts famous crashes but produces normal returns most years. The reputation is built on three outlier events across nearly a century. Don't time the calendar. Do remember that the worst weeks in market history happened in months that didn't feel especially scary going in.

The Take

The interesting October isn't the bad one. It's the average one, where nothing happens and the long-term holder accumulates while the people who sold in September wait for re-entry. That's the pattern that builds wealth. The October effect is what newspapers write about. It is rarely what the market is actually doing.