With February providing the lift the S&P 500 (NYSEARCA:SPY) needed to reverse January's losses, it is time to take a look once again at the January barometer. The big fear last month was that this barometer indicated 2014 was headed for losses. I demonstrated that the January barometer is actually not useful for predicting down years. Similarly, the "ability" to predict up years is not impressive given the vast majority of years on the S&P 500 are up years: 46 of the last 63 years (since 1951), or 73%, were up years.
I hoped that February's rally could deliver more clarity on the January barometer. Unfortunately, this is not quite the case, but I think I did find some nuggets of interest that suggest some value in a more nuanced view of monthly versus yearly performance on the S&P 500.
The bubble chart below enhances the chart I showed in the last piece on the January barometer. Here, I have layered in the performance of the subsequent February using bubble size to indicate the magnitude of the change in the S&P 500 for a given February and using color to indicate the direction of the change - white for negative (an unfortunate limitation of Microsoft Excel), green for positive. Note the widespread of scenarios with February performing well and poorly across the spectrum of January performances. If you stare very intently, you should notice that if January performs anywhere from -6% to 4% AND February is a down month, then the entire year does tend to end on a downer. This year's combination of January performing with "around" a -3% change and February performing positively has occurred three times - each time resulting in a year with a 3% or so gain. I believe that is the best "January-February barometer" can do for us.
In very select scenarios, February can add some clarity to the likely direction of the year
As bad as January is at predicting down years, it turns out the remaining months fare worse, sometimes a lot worse. January is the only month in the year where the rate of false positives - a negative monthly performance where the year actually ends performing with a gain - is below 50%. Of course, by the latter part of the year, traders become more interested in how year-to-date performance might indicate likely monthly performance. Readers should also note which months have the highest rate of negative performances …
January is the best month for predicting a negative year for the S&P 500
While no pattern exists amongst the months to signal the high likelihood of a down year, down years are distinct from up years in the number of negative monthly performances. During the average down year, 62% of months deliver negative performances. During the average up year, only 33% of months deliver negative performances. This means that a string of negative performances MIGHT be a warning sign for the rest of the year - of course by the time you recognize the negative streak, it may already be too late to trade on the negative performance of the year. Regardless, I have posted below a heat map for the 17 negatively performing years since 1951. Each red square is a month with a negative performance. The far right column shows the performance for the given year. Note that 2011 was just fractionally negative.
Source for price data in all charts: Yahoo Finance
1974 sticks out as the most brutal - all but one month performed in the red. Otherwise, the pattern seems to be a streak of negative performances occurs in the beginning of the year or in the middle. Note how, remarkably, these negative years have (relatively) mild ends. October, November, and December conspicuously tend to perform positively. 1987 could have provided a particularly treacherous exception along with 2008, but 1987 managed to close with a 2% gain even with the infamous crash in October.
Overall, I think these various nuanced views add more color than the simplistic January barometer. There is of course still no magic crystal ball, but there are interesting monthly patterns to reference throughout the year.
Be careful out there!