You can find the “January Barometer” mentioned in dozens of media articles and it has been referenced a ton on CNBC so far this year, as it is every January. Here’s a recap from the Financial Post in case you have been lucky enough not to hear about it:
“Stock performance in January can say a lot about where the markets are headed for the rest of the year. At least, that’s the premise behind the January Barometer, a theory that the performance of the S&P 500 during the first month will set the tone for the rest of the year… The Stock Trader’s Almanac points out that since 1950, the Barometer has been right 76% of the time.”
Sounds harmless enough; If January is up, then the market will finish up for the year three times out of four. Good odds, right? So why does this so-called barometer drive me crazy every time I hear it? Because you need context to really determine if this indicator has any value.
Forget January entirely for a second. Would it not be helpful to know how often the market goes up in a given year regardless of any particular month? I certainly think so. In fact, since 1957 (the year the S&P 500 index was created — don’t ask me how they claim to have data from 1950-1956) there have been 56 calendar years and the S&P 500 index has risen 44 times and fallen 12 times. Why is that important? Because 44 divided by 56 equals 78%. The market goes up 78% of the time no matter what!
But if January is up then the market goes up 76% of the time. So what? Actually, that tells me that January has essentially no influence at all. In fact, we could go a step further and say that if January is up, the odds the market will rise for the full year actually go down slightly compared with the historical average. So really, January is irrelevant. It tells us nothing on its own.
It’s sort of like saying if you play blackjack in Vegas in January then the house edge is only 1%. That might sound like great odds, until you do some digging and realize that the house edge in blackjack, assuming you follow perfect basic strategy, is actually less than 1% regardless of when you play.