I took the CSI’s Technical Analysis Course a while ago which is mostly based on trying to identify patterns that can be used as predictors for stock market movements. The statistic mentioned in the title, however, did not come from that course. There is no shortage of stats and predictive models out there that people use, ranging from sophisticated computer analyses to blind rules of thumb like, “buy the stock that just fell the most”.
According to Brooke Thackray in his book Thackray’s 2009 Investor’s Guide, between (and including) the years 1950 to 2007, if January’s performance was positive then the REST of the year’s performance tended to be positive, and vice versa. This worked out 74% of the time. (All references in this post are for the S&P 500.)
Note: I highlighted the word REST in both statements. This is because if you include January’s performance in the entire calendar year’s performance you would slightly skew the results and get a higher predictive rate.
Further, it is noted in the book that a positive January was a better predictor for positive performance for the rest of the year (89% accuracy) then negative January performance was a good predictor for negative performance for the rest of the year (48% accuracy). Since the markets have historically spent more time going up than down, this makes sense.
For all you numerologists out there, here’s to a strong January! :)