Calendars Help Track Time, Not Time Stocks

This shouldn't guide your investment choices. Photo by in-future/iStock by Getty Images.

Every year, pundits trot out the fable September is the "Worst Month for Stocks"™ and October another minefield. Packaged with current fears like North Korea, central bank decisions and Brexit, fall's onset looks scary-leading some to think it's prudent to take a breather from stocks. However, in our view, there is nothing special for markets about September and October-and presuming otherwise could prove damaging for investors.

Those touting these months' perils point to their subpar average returns: Since 1925, the S&P 500 has averaged -0.9% in September (in price returns). The Global Financial Data World Ex-US Index's monthly average (also price returns) was better but still negative (-0.5%).^[i] For both indexes, September is the only month with a negative average return.

This may be so-but the S&P 500 has risen in 45.7% of Septembers, compared to a 59.1% positive frequency for all months. (Exhibit 1) September isn't leading the pack, but it's far from a guaranteed negative-and we don't see the logic in skipping a month that's positive almost half the time.

Exhibit 1: S&P 500 September Returns Still Frequently Positive

Source: FactSet, as of 9/6/2017. S&P 500 monthly price returns, January 1925 - August 2017

Historical October monthly returns are also below-average for both US (0.4% in October vs. 0.6% for all months) and non-US stocks (-0.1% in October vs. 0.4% for all months). But just a few outliers-during some of the worst economic downturns in history, no less-skew monthly averages. For US stocks, for example, all three of the worst offending Septembers since 1925 came during deep economic contractions and bear markets in the 1930s-down -13% in 1930, -30% in 1931 and -14.2% in 1937. Similarly, Octobers in 1929 (Black Tuesday), 1987 (Black Monday) and 2008 (the Global Financial Crisis) pulled down the month's US average. Removing these three raises October's US average to 1.1%. Do the same for non-US stocks, and October's average rises to 0.6%. Likewise, average US September returns improve to -0.3% without the worst-in-show trio, while non-US Septembers erase their losses. The lesson: Three extreme outliers in a pool of 92 years shouldn't color your perception of these months going forward.

Recent history shows what September and October skeptics would have missed by sitting out. As Exhibit 2 shows, of 10 September - October periods from 2007 to 2016, 7 have been positive for global stocks-some by a lot. This doesn't exactly scream, "run for the hills!"

Exhibit 2: Monthly MSCI World Returns-Last Decade

Source: FactSet, as of 9/12/2017. MSCI World Index monthly return with net dividends for September and October, 2007 - 2016.

How would you feel missing global markets' 7.9% gains in October 2015? Or their 9.3% returns in September 2010? Sitting out the fall might feel great if you avoid bad months, but in our view, it isn't worth the risk of missing significant upside. Now, if you have a reason to believe a bear market (a lasting, fundamentally driven decline exceeding -20%) has begun, moving out of stocks can make sense. But that is very different than expecting a weak month based on trivial past returns. And even then, avoiding bear markets isn't necessary to achieve long-term equity-like returns. Participating in bull markets is.

Adopting such a myopic perspective is feckless anyway. Over short periods like a month or two, stocks can gyrate for any or no reason as sentiment swings (sometimes wildly) in response to shifting hopes and fears. To paraphrase the great Ben Graham, markets are like a voting machine in the short run and a weighing machine in the long run. No one can say how investors will "vote" with their money in the next couple months, no matter when those months fall during the year.

September and October's inferior histories aren't telling, either. Statistically, one month has to be worst^[ii]-just like days of the week or times of the day. Historical patterns lacking a fundamental rationale-whether based on the calendar or something else-are just noise, and past prices don't predict the future. Consider this: If the calendar were actually predictive, what would you do? If you answered, "sell before the month stocks always fall, then buy again when the all-clear sounds," you're in good company-with basically every investor on earth. Markets are efficient, you see-they incorporate all widely available information. This includes theories-however hair-brained-about when alleged bargains are available. When folks trade on such theories, any "advantage" they may have offered disappears.

Lastly, while some point to a bevy of threats making the next few weeks uniquely dangerous for stocks, we aren't buying it. Today's fears aren't specific to September or October. They've been around for months, if not years, and we've addressed them all-North Korea here, central bank worries here and the UK's health here.^[iii] Whether or not any (or something else) trigger volatility in the coming weeks, we don't believe they're a valid reason to exit the market.

The calendar's flip toward autumn doesn't mean a fall is near. Though volatility could roil markets at any time, growth in the US and elsewhere appears to be broadly outpacing expectations today, making it a poor time to move out of stocks. Our advice: Don't try to time short-term market moves-not in September, October, or any other time of year. Hop, skip and jumping your way through the calendar is no way to reach your long-term financial goals.