Monday, December 16, 2013

'Tis the Season(ality)

As we embark upon the holiday season, I wanted to take a moment to update a data set which relates to the seasonal performance of the S&P 500 industry groups.  The history runs monthly from January 1990 through November 2013 - nearly 24 years and 23 industry groupings.  

While using seasonal tendencies alone is not a robust investment strategy, it does add an interesting aspect to to the discussion.  I believe such analysis has the most relevancy when combined with relative valuation work (i.e. an industry group entering a seasonally strong/weak period is more likely to show its tendency if also supported by relative under/over valuation measures).

I also believe that many analysts make an easy mistake when discussing seasonality by choosing to comment on average relative performance.  The danger of using an average is that one or two very strong, or very weak observations can skew the general tendency.  For example, although the average relative performance in April of the Autos & Components sector is 4.91% better than that of the index (one of the strongest monthly outcomes), it has only outperformed in 13 of the 24 Aprils - just one observation away from a coin toss.

Therefore, to help overcome the tendency of averages to skew results, I prefer to rely on median performances and the absolute number of periods in which a sector has done better than the market.  Below I display the monthly median relative performance (blue) followed by the number of months in which the Auto & Components sector outperformed (yellow) the S&P 500.  There is very little seasonal tendency in this sector, save some weakness in May.















After examining 276 industry months (23 industries across 12 months), only 14 industries show performance consistency of 75% or greater in a given month.  In five of these instances we find industries which have outperformed, and another nine which have underperformed similarly.
















The most seasonal sector would appear to be Capital Goods which appears three times (consistently strong in November and December and consistently weak in October).  The strongest seasonals though, would appear to be Software & Services in June (outperform 75% of the time, by a median of 2.19%) and Transportation in October (75% and 2.34%).  The weakest seasonal would be Food & Staples Retailers which tend to underperform in April 75% of the time by a median of 2.17%.

Seasonality is just a beginning and not an investment strategy in and of itself.  For those who would like to see the complete data file and graph set to combine with their own valuation work, please drop me an email and I will be happy to provide the full results.

Best Wishes to all for a Very Merry Christmas and a Happy New Year.