September Is Coming. So Are False Seasonality Fears

Pundits’ list of worries is long and growing these days, including $40 trillion in US debt, high stock valuations, rising Treasury yields and an AI-related bubble. As September approaches, many see that as an inauspicious backdrop for the calendar’s weakest month historically. But hold on. While September is stocks’ weakest month on average, this overlooks several key points about stocks’ returns and drivers. Chief among them: Seasonality doesn’t drive markets. That pundits have added this old, false worry to their list reveals sentiment isn’t uniformly euphoric, a counterintuitively bullish factor.

Scary September claims rest on a simple reflection of history: Since good data start in 1926, the S&P 500 has averaged a -0.7% decline in September, making it stocks’ weakest calendar month and the only one with a negative average return.^[i] This, plus the coincidental fact several big crashes happened in October, leads many to dub September the start of “financial hurricane season.” Thus, nearly every year, some pundits turn the calendar page and tout the summer’s unofficial end as reason to avoid stocks.

But this thinking—that markets bend to seasonal trends—is wide of the mark. First and foremost, average returns don’t tell the whole story. Several major outliers skew September’s average return downward, making things seem worse than they are. For instance, the S&P 500 fell a whopping -29.6% in September 1931—its worst month ever—amid a Great Depression-era banking meltdown and the UK’s abandoning the gold standard, which forced the Fed to raise the discount rate in an effort to attract deposits.^[ii] Just six years later, US stocks sank -13.8% in September 1937 following the Fed’s doubling of banks’ reserve requirements. Banks responded by curtailing lending and removing liquidity from the financial system.^[iii]

Or fast forward to September 1974, which featured an -11.5% drop as fears over an oil-shock-driven recession, double-digit inflation, price controls, rising unemployment and Watergate came to a head.^[iv] More recently, see -10.9% and -9.2% drops in 2002 and 2022, respectively, which both fell late in bear markets.^[v] The former was near the end of the dot-com bear market as stocks retested earlier lows after the rapid passage of Sarbanes-Oxley. The latter came near the end of the sentiment-driven bear market, as investors weighed myriad economic and geopolitical fears.

The common theme? All these monthly declines had unique, non-calendar-related causes. Stocks didn’t fall because it was September. Other factors dragged them down or had already been dragging them down when the calendar turned. Take out these epically bad Septembers, and the average return is flat.

Another way to see this is median returns, or the point at which there are equal instances of returns above and below. For September, this figure is 0.1%—positive!^[vi] Not gangbusters, but a median return in the black is pretty far off September’s lousy reputation. That hints at another reality: The S&P 500 rose in 52% of Septembers since 1925.^[vii] Stocks are up more often than not in the ninth month—another inconvenient fact for the “September Effect” crowd.

Avoiding stocks on seasonality fears can mean missing nice gains that compound later. Look no further than the last two Septembers, in which stocks rose 2.1% and 3.6%, respectively.^[viii] Two pretty solid months. Or September 2010’s 8.9%.^[ix] Or 1995 – 1998, which saw stocks rise over 4.0% in four consecutive Septembers.^[x] Or or or. The “September Effect” has oodles of counterexamples.

Beyond the skew from bad Septembers, the simple fact is that even if Septembers were strongly positive historically, it wouldn’t give you any useful sign of what to expect in the future. Past performance never predicts future performance. If it did, stocks would move in the same direction or pattern in perpetuity. They don’t. Conversely, stocks look forward, weighing the economic, sentiment and political factors most likely to influence corporate profits 3 – 30 months in the future. Monthly average returns over the past century have absolutely no bearing on where stocks are headed next.

Consider: If calendars drove returns, it stands to reason everyone would take advantage by selling out before September to front run the month’s purported weakness. This cycle would continue over time, with traders acting earlier and earlier until, eventually, folks would sell in November because September was only 10 months away. This clearly isn’t the case. Friends, take a step back and it is clear the logic doesn’t hold up here.

If anything, pundits’ rehashing “September Effect” fears today speaks to sentiment. Again, stocks look forward. Yet all of these fears (i.e., debt, valuations, Treasurys) have lingered for months, if not years at this point. At the same time, US stocks are up 13.0% this year—the bull market has marched on.^[xi] If these “risks” posed a legitimate threat, we highly doubt stocks would be performing this way. To us, markets are basically telling investors that these fears are very likely false over the next 3 – 30 months.

That many are adding in September as a worry shows people are hunting for negatives—a sign sentiment, while warmer than in previous years, isn’t uniformly euphoric. It seems there is still some skepticism lingering even in America. That shows expectations likely aren’t overly stretched, a sign this bull market still has some fuel in the tank.