At the end of every January there is an onslaught of articles talking about the old adage, “as goes January, so goes the year.” The translation is that the performance in January is a good indicator of how the year will turn out.
With the recent holidays and lack of market volume, there is little to report on, so we thought we'd take some time out to debunk this myth.
Let's set up the basics: using S&P 500 index data going back to 1951, we can calculate that in 75% of years, the directional return — positive or negative — from the year mirrored the directional return of January. At first blush, that sounds very convincing. Unfortunately, it turns out not to be very actionable.
Why not?
We'll use an analogy. Let's pretend we're flipping a coin 12 times in a row. We start our score at 0. For every heads we add 1 to our score and for every tails we add -1. In our first flip, the coin comes up tails, making our score -1. This -1 says nothing about how the coin will turn up in the future, but what it does give us is information about what we can expect our score to be by the end of the game. Given that the coin flips are random, over the next 11 flips we are just as likely to get heads as we are tails, and so our expected score is going to be the -1 we currently have. So because the first coin flip was negative, we can expect to have a negative score at the end of the game. That is, until the end of the second coin flip, where we will learn a little more about where we stand.
So while our first coin flip gave us information about the expected final score, it told us nothing about the next 11 coin-flips themselves. Those were totally random.
That is, more or less, the same issue with the January adage. A strong January gives the full year's return a solid head start, but it says nothing about performance in the following 11 months.
What's a simple way to test this? We can look at the probability that February through December is positive given that January was positive. This is known as a conditional probability and allows us to account for the information we know at the end of January. What is this probability? Close to 70%.
If that seems high it is because the other shoe hasn't dropped. We forgot to account for the fact that markets, historically, have tended to go up. In fact, the raw probability that February through December was positive over the same time frame is 75%.
In other words, there is no extra information in January's return that helps us forecast February through December. And nor would we expect there to be from any theoretical foundation.
Corey Hoffstein is co-founder, chief investment officer and chief technology officer of Newfound Research and Justin Sibears is a managing director in Newfound's investment strategies group. This blog originally appeared on the firm's website.
