While this month’s stock market performance hasn’t been kind to the so-called January effect, this is hardly big news. Most analysts had already recognized that other months yielded better returns of late.
Suppose there is a truly dependable and exploitable January effect, that the stock market—especially stocks of small companies—will generate extraordinary returns during the first five days of January. What will investors do? They will buy on the last day of December and sell on January 5. But then investors find that the market rallied on the last day of December, and so they will need to begin to buy on the next-to-last day of December; and because there is so much “profit taking” on January 5, investors will have to sell on January 4 to take advantage of this effect. Thus, to beat the gun, investors will have to be buying earlier and earlier in December and selling earlier and earlier in January so that eventually the pattern will self-destruct. Any truly repetitive and exploitable pattern that can be discovered in the stock market and can be arbitraged away will self-destruct. Indeed, the January effect became undependable after it received considerable publicity.
True to Malkiel’s words, January is only the 8th best performing month (5th worst) for the S&P 500 (SPY) over the past 25 years. Small cap results over the same period aren’t much different – January ranks as the 7th best month for the Russell 2000 (IWM).
But what about other calendar effects?
We can appreciate Malkiel’s skepticism, but we’re not willing to reject all connections between investment markets and the calendar. There are other theories that aren’t as widely known or easily arbitraged as the January effect, which was first popularized way back in 1942. We’ll argue that one such theory may be more important than usual this year.
Our argument begins with four observations (we’ll get to theories in a moment):
- Market sentiment often changes during the earnings reporting season – in which most of the action occurs in the first month of the quarter – and these sentiment shifts tend to persist.
- Individual investors tend to pay extra attention to their positions early in a quarter, reacting to the past quarter’s results and then looking ahead to the next performance period.
- Professional money managers often refine their strategies prior to client reviews or board meetings, which typically occur after results for the prior quarter become available.
- Investors (individuals and professionals) are even more likely to rethink strategy in January, partly because it marks a new annual reporting period but also because it tends to be a time for planning and reflection. (How are you doing on those resolutions, by the way?)
These observations are admittedly vague, but we suspect they’re relevant to stock performance. They suggest that the first month of a quarter may set the market’s tone in subsequent months. In the context of today’s markets, they tie into a few questions you may be asking about early 2014 volatility: Is January’s market drop merely noise on the way to another string of all-time highs, or is there more to it than that? For instance, doesn’t it seem a little ominous that we stumbled out of the gates this year despite sentiment being rampantly bullish? Does this tell us to be cautious going forward?
If you happen to read the Stock Trader’s Almanac, you’ll connect our questions to the “January barometer” (not to be confused with the “effect” discussed above). The Almanac’s founder, Yale Hirsch, coined the term in 1972 when he presented research showing that January’s return is a decent predictor of full-year returns. He concluded: “As January goes, so goes the year.”
We’ll take a closer look at the January barometer below, while testing two variations drawn from the observations above.
“Downsizing” the January barometer
First, we doubt that any carryover of January’s performance is likely to persist for an entire calendar year. Based on the idea that quarterly reporting cycles may have something to do with these types of anomalies, it doesn’t seem right to think that January’s events should still be relevant near the year’s end. The first month of a quarter may offer clues about the next quarter or two, but probably not three or four quarters later after investors have shifted focus to subsequent corporate earnings and investment performance reports.
In fact, even without quarterly reporting cycles, you may still question why January would continue to be a “barometer” by the third or fourth quarter. You may expect to find lower correlations of January returns with the year’s second half than with the first half, and this is exactly what we see:
Note that the 33% correlation for the “downsized” January barometer is very high for these types of relationships. It’s comfortably significant based on traditional tests (the F-stat is 8.2). By comparison, the correlation of January’s return with the 11 months from February to December is still high at 28% but less significant (the F-stat falls to 5.1).
Here’s a scatter plot and trendline for the year-by-year results:
The chart shows that 54% of the years with negative January returns included negative returns from February to June (13 of 24), while only 9% of the years with positive January returns were followed by negative February to June returns (8 of 60). In other words, the probability of a down market between February and June was six times higher after a down market in January.
Do years or quarters hold the key to the calendar?
Second, we considered whether April, July and October also qualify as barometers, based on our speculation that the January barometer is partly explained by quarterly phenomena.
In particular, we calculated correlations with subsequent returns for all 12 months to see if the beginning-of-quarter months stand out:
Needless to say, the correlations fit the hypothesis, with the four highest belonging to January, April, July and October. The odds of this happening in a purely random market are nearly 500 to 1. Call it the “JAJO effect.”
Will JAJO spell an end to the market’s MOJO?
Only time will tell if the JAJO effect is reliable.
You may take a Malkiel-like perspective and say: “Okay, let’s assume 500 researchers test 500 hypotheses – at least one of those researchers is likely to find a result with 500 to 1 odds.”
And while that may be true, we think there’s more to it. If stocks don’t recover strongly by month’s end – say, back to the S&P 500′s 2013 close of 1848.36 – the odds favor continued weakness. As January goes, so goes the first half of the year.