Last week, we looked at the historical performance of years ending in "0" and found that there is some truth to the idea that the first year of a decade tends to perform poorly (and the first two quarters especially).

Now let's look at the January Barometer, the idea that however January performs, the rest of the year tends to follow.  In other words, if January shows a positive return, then the rest of the year should too, and vice versa.

The table below shows the forecasting power of each month.  What we're looking at is how often each month's return predicted the return of the following 11 months. 

Well, it turns out that January was pretty good, accurately forecasting the next 11 months nearly 70% of the time since 1928.

If January closed in positive territory, then over the next 11 months the average maximum gain in the S&P 500 was +14.8%, compared to an average maximum loss of only -6.4%.  It wasn't all hearts and flowers - there were 7 years where the S&P dropped at least -20% during the year - but at least there were 15 years where it rallied at least +20%.

If January closed in negative territory, then over the rest of the year the S&P averaged a maximum gain of +8.1% versus an average maximum loss of -12.5% (uhh, that's not good).  Only 2 of these years showed a rally of at least +20% during the next 11 months, and there were 6 of them that showed losses of at least -20%.

So is January a good barometer?  I'm sure there are some statistical rules that would suggest anything we're looking at here could just be a fluke, but empirically we have seen that when January is positive, then the rest of the year performs much (much!) better than when it is negative.

No other month really stood out from the pack...except for October.  That one was the poorest of the bunch, giving an accurate prediction only 48% of the time.  That's not a big surprise, really, given October's historical tendency for extremely volatile moves and higher-than-average tendency to see major market sell-offs and rebounds.

Jason Goepfert