December (Not So) Super Duper Bullish
One bit of follow up (inspired by Ed Mamula) to my previous post December Super Duper Bullish.
That post showed that December has historically been bullish for the stock market. Ed asked whether this bullish seasonality was affected by the market’s performance in the preceding 11 months, Jan – Nov (which would make sense given “tax loss selling” and all that jazz).
In short, yes it is. To illustrate…
The table above shows December results when Jan-Nov was either up (left column) or down (right).
From this 30,000 foot view, December returns in years when Jan-Nov was up have been hugely bullish, and when Jan-Nov was down, flat.
How consistent has this observation been?
In the graph below I’ve assumed an investor was long the S&P 500 in December when Jan-Nov had been up (green) or down (red) since 1930.
I’d call that pretty darn consistent.
We’re dealing with a very small number of observations here (so be extra wary that this is all just noise) but the stark contrast between the two is hard to argue with.
By my calculation we stand at about a -3.5% loss YTD, so we’re right on the cusp of the cutoff point. In any case, the important takeaway is that the much touted bullish December seasonality might not be so strong this year.
Note: I’m sure that none of this is new and other folks have blogged about this observation before. But it’s all new to me, so meh. P.S. Thanks Ed!
Happy Trading,
ms
Geek note: returns are dividend-adjusted. Dividends interpolated from quarterly data before 1988. S&P 500 TR index used 1988 and beyond.
. . . . .
To stay up to date with what’s happening at the MarketSci Blog, we recommend subscribing to our RSS Feed or Email Feed.
Filed under: Time-based | 6 Comments
One of your more important tenets is the view that trading need not be binary (all in or all out) but rather should be continuous. For this study, instead of a binary view of up year-to-date vs. down YTD, see how the magnitude of YTD change might influence December returns. This could be done either by grouping YTD returns into quintiles (16 samples per bucket is not too bad), or make a scatter chart for a visual check of a relationship.
Thanks for your good work!
Can you please run an analysis of December performance in pre-election years?
Since the January effect is much stronger on smaller companies, I would guess that you would also see a stronger relationship here if you used a small company index instead of the S&P 500 in your sample
Hello Adreno – great comment – unfortunately the only months that IMHO have been consistently strong for large or small caps are January and October. Everything else has been inconsistent.
Read more:
http://marketsci.wordpress.com/2010/12/05/large-vs-small-cap-seasonality-strategy/
http://marketsci.wordpress.com/2010/12/03/the-tax-loss-selling-effect-and-large-vs-small-cap-performance-by-month/
michael
I am not able to open the seasonality map or the large vs small cap. relative strength variation of it for Dec. ’11. Have I missed something?