As January comes to a close, many market participants will be tempted to fret about the "January barometer," which holds that stock performance in January predicts the performance over the rest of the year. In a January that has seen stocks slide 3 percent, this could become a concern. But investors would probably do well to ignore this long-standing market meme.
(Read more: Stocks down 4% in January! If history repeats ...)
Proponents of the barometer point to the fact that over the past 35 years, the S&P 500 has followed January's direction 71 percent of the time. However, this statistic is skewed by the fact that it includes January in that full-year performance.
That means that in years like 1987, when the market rose 13 percent in January but finished the year just 2 percent in the green, listening to the January barometer would have yielded a loss of some 10 percent—and yet this year is still counted as a success for the barometer.
Still, even if we merely compare January's performance to the path that stocks beat over the following 11 months, January still appears to predict the S&P's path 66 percent of the time. The problem is that it is much better at "predicting" winning years than losing ones.
Going back to 1979, the S&P rose in 23 out of 35 Januarys. Over the next 11 months, the market consequently rose in 19 of those 23 years that were kicked off by winning Januarys—meaning that a positive January has successfully predicted a winning February-through-December 83 percent of the time.
But in the 12 years when the market fell in January, the market only followed along in four years. That's just a 33 percent success rate.
The reason that positive Januarys prove to be a great barometer, and negative Januarys a terrible one, is the same reason that the "January barometer" appears to exist in the first place: Stocks rise.