If you've been in the markets long enough, you've heard the whispers: “Sell in May and go away,” but the real killer month is September. I've personally watched my portfolio take a hit every fall, and it's not just bad luck—it's a decades-old pattern. After analyzing the data from the S&P 500 since 1950, I can tell you flat out: September is the worst month for the stock market, both in terms of average return and frequency of declines. But it's not the only ugly month. Let me break it down with real numbers and the why behind the slump.

The Verdict: September Claims the Crown

Based on research from Yardeni Research and CFRA, the S&P 500 has posted an average decline of about 0.5% to 1% in September since 1928. That's worse than any other month. But here's the kicker: September also has the highest percentage of negative years—about 55% of Septembers end in the red. Compare that to December, which is positive about 70% of the time.

I remember back in 2022, I was fully invested in tech stocks when September hit. The S&P 500 dropped over 9% that month alone. My gut told me to trim positions, but I ignored it. That mistake cost me thousands. Since then, I've made it a rule to reassess exposure every late August.

Below is a snapshot of average monthly returns for the S&P 500 (1950–2023) to give you a clear picture:

MonthAvg Return% Positive
January+1.0%62%
February0.0%52%
March+1.1%64%
April+1.3%68%
May+0.3%55%
June+0.2%56%
July+1.5%65%
August+0.1%52%
September-0.7%45%
October+0.5%58%
November+1.5%68%
December+1.4%72%

Sure, the averages look small, but when you factor in compound interest, a -0.7% drag every September adds up. Plus, September is often the month when major sell-offs begin—think 2008 (Lehman collapse), 2001 (9/11), and 1929 (peak before the crash).

Historical Data: The September Effect in Numbers

The “September Effect” isn't just a Wall Street myth. According to a study by the Federal Reserve Bank of New York, the effect holds true across global markets—not just the US. They found that from 1990 to 2020, September was the worst month for equity indices in 36 out of 47 countries analyzed. Japan's Nikkei 225, for instance, falls an average of 1.2% in September.

But here's a nuance most articles miss: the effect is stronger in odd-numbered years and during presidential election years. I've seen this firsthand—September 2019 (odd year) saw a 1.9% drop, while September 2020 (election year) fell 3.9%. Not a coincidence, in my opinion.

Key Stat: Since 1950, September has produced a negative return in 34 out of 71 years for the Dow Jones Industrial Average (DJIA). That's a failure rate of 48%—higher than any other month.

Why Is September So Weak? The Real Drivers

Most people point to “seasonal adjustment” or “fund managers selling losers.” That's part of it, but not the whole story. Here are the less-discussed reasons I've observed:

  • Outflows from mutual funds: Institutional investors often rebalance portfolios in September after summer vacations, leading to concentrated selling. In the first week of September, fund flows historically turn negative.
  • Tax-loss harvesting pressure: Smart money starts taking losses early to offset gains, especially if the year has been strong. This selling pressure disproportionately hits losing stocks.
  • Behavioral shift: After a strong Jan-Aug, many traders get complacent. Then a small dip triggers stop-losses, creating a cascade. I've seen this pattern in my own trades—September corrections are faster because liquidity thins out.
  • Geopolitical events cluster: September is prime time for earnings warnings, trade disputes, and political uncertainty. The government fiscal year end (Sept 30) also adds budget drama.

One often overlooked factor: seasonal affective disorder (SAD) affects traders too. Shorter days and reduced sunlight can lower risk appetite. I know it sounds flimsy, but psychological studies show a measurable dip in market optimism during autumn.

Other Contenders: October, February, and May

September is the clear winner, but it's not alone. October has a reputation for crashes (1929, 1987, 2008), but its average return is actually slightly positive thanks to recovering from September lows. February is historically flat, and May underperforms because of the “Sell in May” effect. However, none match September's consistency of failure.

Let me share a quick comparison from my decade of trading:

MonthNotable CrashesAvg Return (S&P)My Personal Rule
September9/11, Lehman, 1929 peak-0.7%Reduce stock exposure by 20% by late August
OctoberBlack Monday 1987, 2008 panic+0.5%Buy the dip if September was terrible
February2020 COVID selloff (start)0.0%Be cautious of overextended tech names
May2010 Flash Crash+0.3%Tighten stop-losses on growth stocks

How to Navigate the Worst Month (Without Panicking)

You don't have to completely exit the market every September. That would be foolish—you might miss a big rally if the pattern breaks. Instead, I use these five strategies:

  1. Sell half your winners in late August. If a stock is up 50%+, take some profits. September tends to hit high-fliers hardest. I've done this for three years straight and it's saved me from giving back gains.
  2. Buy protective puts on indices. The cost of a one-month put on the S&P 500 in September is usually cheap because volatility is low before the storm. It's like insurance. I typically set a 5% downside protection.
  3. Rotate into defensive sectors. Utilities, healthcare, and consumer staples hold up better. In September 2022, XLU (utilities ETF) lost only 0.5% vs. QQQ's 10%.
  4. Wait until mid-October to add risk. The worst of the selling is often over by October 15. I start averaging in after that using limit orders.
  5. Watch for a “September miracle.” Sometimes the Fed steps in or earnings surprise. If the S&P holds above its 200-day moving average, I don't panic—I just stay the course.
Fact-check note: Data sourced from CFRA's Stock Market Almanac and Yardeni Research's monthly return tables. My personal observations are from trading since 2014.

Frequently Asked Questions

Is September always the worst month, even in strong bull markets?
Not always—in 2013, the S&P 500 gained 3.0% in September. But in 7 out of 10 strong bull years, September still underperformed the other months. The pattern is probabilistic, not deterministic. I wouldn't bet the farm on it, but I'd definitely hedge.
Should I sell everything in August and buy back in October?
No. That market timing rarely works because you'll miss the best days. The September effect is weak enough that you're better off trimming exposure rather than going all cash. From my experience, a 20% reduction in equity exposure is the sweet spot—you cushion the blow without missing the recovery.
Does the September effect also apply to crypto or bonds?
Somewhat. Bitcoin has historically been weak in September (average -5% since 2013), but with higher volatility. Bonds actually tend to perform well in September as a flight-to-safety trade. The iShares 20+ Year Treasury ETF (TLT) has positive returns in about 55% of Septembers. So it's a good hedge.
What if the market rallies in September—should I chase it?
Statistically, a September rally is less reliable. I've been burned twice chasing early September gains, only to see them evaporate by month-end. If markets are up 3%+ by mid-September, I take partial profits. Better safe than sorry.
Are there any months that consistently beat September's badness?
February and May are close, but September wins for average decline. May's “Sell in May” effect is more about underperformance from May to October, not a single month. February's flatness comes from post-January profit-taking. If I had to rank: 1st September, 2nd February, 3rd May.