The Myth of September's Siren

Summer's unofficial end brings students back to school and, according to the punditry, money managers back to markets. Upon their return, we're told, they don't like what they see, frown and sell-repositioning assets to coincide, we guess, with expectations the leaves will change.^[i]Hence, claim some, September is bad for stocks, and you should prepare accordingly! But in our view, investors shouldn't fret September's historically weak average-this is coincidence without causality, and either way, past performance is not indicative of future returns.

It's true, since 1928, September has the highest frequency of negative months-46, above the average of 34 for the other 11 months.^[ii] September also has the worst historic average return of any month (Exhibit 1).

Exhibit 1: S&P 500 Price Index Monthly Averages

Source: FactSet, as of 09/03/2014. Average monthly price returns from 12/31/1928 - 12/31/2013.

What about September gives stocks the willies? Nothing-this is coincidence without causality, in our view, a cardinal sin in analyzing markets. Of the 46 negative Septembers, 20 occurred during bear markets. And this doesn't mean bears start in September-only one of the 13 since 1928 has: 1929's. Stocks move in cycles, and if the month you're referring to happens to fall amid more cyclical downturns, it'll likely have worse average returns.

Now, that doesn't totally eviscerate the September-is-bad argument. During bull markets, September also happens to have the highest frequency of negativity (26 times, exceeding the average of 20). But September's average return in bull markets is positive-and it's not even the weakest month, showing the impact of a couple really bad Septembers, like 1931's -30% drop (Exhibit 2).

Exhibit 2: S&P 500 Price Index Monthly Average During Bull Markets

Source: FactSet, as of 09/03/2014. Average monthly price returns from 12/31/1928 - 12/31/2013 during the S&P 500's 13 bull markets.^[iii]

Beyond the numbers, recent history pokes holes into the "September Swoon" theory. In the 10 years between 2004 and 2013, September was negative a whopping two times.^[iv] So was December, the most frequently positive month since 1928.^[v] One of the two of those negative Septembers came in 2008, 11 months into a bear market. We are pretty sure markets didn't peak in October 2007 in anticipation of the calendar turning to September in less than a year. Our skepticism here stems from the fact we are fairly confident market participants know September follows August and precedes October. The Roman calendar after all, wasn't built in a day or yesterday.

Past September performance is no more predictive of future results than regular past performance. Now, to be fair, most folks seem to get this, but they blend September fears and factors like a Fed rate hike announcement or geopolitics, suggesting these factors plus the month's sketchy history mean the long-awaited correction-a short-term, sentiment-driven market pullback of -10% or more-is coming. While corrections can start for any (or no) reason, just because we haven't had one since 2012 doesn't mean we're "overdue"-September or not. (Nor are geopoliticaljitters or Fed rate hikes big bull market risks, but those are discussions for another day.)

The hesitant September chatter suggests false fears persist-a bullish sign. Even optimistic "experts" are hedging a bit with correction warnings. For investors, the presence of false fears indicates we're far from the euphoria typical at a bull market's end. This, combined with the backdrop of underappreciated economic growth, midterm-election-generated gridlock and more indicates the bull market has legs to continue. Flopping in and out of stocks because of the flip of a calendar page is likely fruitless.