Are September and October Bad for Stocks?

Despite the S&P 500 finishing August down just -1.6% for the month, the rocky path there seemingly has many investors on edge.^[i] Now the calendar turns to September and October—times of allegedly woeful returns and heightened risk of a market crash. In our view, this popular myth is about as useful to investors now as it ever was—which is to say, not at all. Regardless of how this September and October play out, they are still just months with nothing inherently different or more bearish than the other 10.

September and October’s subpar reputations mostly stem from their lackluster historical returns. Since the end of 1925 (when reliable S&P 500 data begin), September is the only month with a negative average return (-0.6%).^[ii] October’s historical average return is positive, but not by much—0.6%.^[iii] This places it third from the bottom, edging out May’s 0.24%.^[iv] But those fearing October don’t point to the average—rather, they cite past October crashes. In October 1929, the S&P 500 dropped -19.7% during the 1929 stock market crash.^[v] In October 1987, the nexus of that year’s brief bear market, US stocks fell -21.5%.^[vi] Most recently, they sunk -16.8% in October 2008—the heart of the financial crisis.^[vii] Given this history and recent volatility, sitting out the next couple of months may seem sensible.

However, we don’t see a sound basis for this. Correlation isn’t causation, and the calendar isn’t a timing tool.^[viii] October’s crashes are pure coincidence. Each described in the prior paragraph occurred during a bear market with fundamental causes that had nothing to do with the month. Focusing on them is also quite selective. What about October 1974’s gangbusters 16.8% return—or October 1982’s stellar 11.5%?^[ix]

Turning to September: Despite its worst-in-show title, it has still been positive more than half the time. Squeamishness about a month that has helped investors more often than not seems odd to us. Moreover, as with October, just a few hugely negative Septembers—including 1931 (-29.6%), 1937 (-13.8%) and 1930 (-12.7%)—skew the average.^[x] But the median—which removes outliers’ skew—is 0.1%—positive![xi] For evidence, look at the complete tally of Septembers since good stock market data begin in 1926:

Exhibit 1: Historical September Returns

Source: Global Financial Data, Inc., as of 8/29/2019. S&P 500 Total Return Index percentage change in September, 1926 - 2018.

Even if there were something special^[xii] about September and October, it wouldn’t be actionable, in our view. Markets are efficient in part because investors constantly seek an edge. If stocks reliably performed worse in September, investors would immediately act on this information by selling beforehand. This would move the dip up to late August. Investors would notice! Next year, selling would likely start around mid-August. Rinse and repeat until not a shred of seasonal pattern remains.

This brings us to some lessons that apply all year: Alleged seasonal stock trends trick investors—especially myopic ones. Statistical patterns that lack a fundamental cause aren’t worth heeding, in our view. Volatile times can make stock market myths seem more plausible—making it the perfect time to refocus on your financial goals. Two months is a blip for investors whose long-term goals require long-term equity-like returns. We understand it probably doesn’t feel like a blip in the moment. But questioning popular narratives surrounding alleged short-term stock movements may provide context and help you stay cool when headlines urge reaction.