« The Wisdom of Bill Gross | Main | The Winners of “JK Lasser’s Your Income Tax 2008″ »
What Does a Negative January Mean for the Rest of the Year?
By JLP | January 20, 2008
As of Friday’s market close, the S&P 500 Index is down 9.75% in 2008. I posted that I hoped that this January’s performance wasn’t an ominous sign for the rest of year. That p0st got me to thinking about whether or not a negative January meant a negative return for the year. So, armed with the monthly returns for the S&P 500 Index going back 1926, I ran some numbers and here is what I found out:
Since 1926, there have been 29 years in which January had a negative return (not including 2008). Of those 29 years, 15 of them had a negative return for the entire year following the negative January. In other words, there appears to be a 50-50 chance of a negative return for the year following a negative January.
If there’s anything to be concerned about it’s the fact that should this year’s -9.75% start hold for the entire month of January, it will mark the worst January in the S&P index’s history:
What all this means for the here and now is anybody’s guess. My advice is to not let it worry you. Just make sure you’re properly diversified and let the market do what it’s going to do.
Topics: Investing |
January 20th, 2008 at 5:04 pm
You know I love this stuff, JLP. What about the return for the rest of the year without January included? It looks like only 13 of the 29 years had negative returns in the 11 months following a negative January. (Unless I’m doing something wrong…)
January 20th, 2008 at 5:30 pm
Cool Information. A chart you won’t see distributed by any mutual fund wholesalers, I’ll tell you that.
Based on the above info. looks like the best investors could ‘hope’ for is to make their money back or a little better.
January 20th, 2008 at 6:11 pm
Ciaran, not distributing information like this might be one of the few investor-friendly moves from mutual fund wholesalers. Not that they can hide what’s happening, but trying to predict the next 11 months from the market performance in January is a weak technical indicator (it doesn’t seem like there’s been enough history yet to make any real statistical conclusions). Anyone who’s supposed to be investing for the long-term certainly doesn’t need more encouragement to think they can predict the future. There’s enough news already to drive the people who sell low and buy high.
January 20th, 2008 at 9:57 pm
[...] In case you haven’t noticed, the stock marketing has taken a beating so far in 2008. Never on to be afraid of the numbers, JLP digs into the data and shows us that, historically, there’s a 50/50 chance that the market will end up positive (or negative) for the year given the start we’ve had thus far. So you tell me… Is the glass half full or half empty? [...]
January 21st, 2008 at 1:48 am
JLP,
While it is true that a glance at the January returns tells you there is about a 50/50 chance of the market being negative as a whole this year, I suspect this would true of positive or negative returns in any particular month of the year and is therefore not a terribly useful conclusion. However, a slight reorganization of the data can provide some a bit more insight.
I binned your data into 4 groups: 0 to -2, -2 to -4, -4 to -6, and -6 to -8. Each return bin had an average of about 0 for the yearly return. As you might expect, though, the variation in the yearly returns increases dramatically as the January returns approach 0. The 0 to -2 binning has a standard deviation of 23.76 in its yearly return but the -6 to -8 binning has a standard deviation of only 2.96.
Obviously the sample sizes are too small to draw conclusions with any confidence, but with that caveat I would say that while you can’t say much about whether the S&P will be up for 2008 as a whole you can make a credible case that it will be up from where it is right now.
January 21st, 2008 at 8:51 am
BizLinks | 1.21.08
Unimpressed by U.S. stimulus plan, world’s markets dive Oil drops below $90 a barrel Questioning the Bonus System — after the year Wall Street’s had, are any bonuses justified? Chinese banks face writedowns from U.S. subprime mess What Does…
January 21st, 2008 at 2:09 pm
Around the investment blogging world
It has been a while since our last edition of around the investment blogging world and I thought now was a great time to see what is being said about the current market situation. Around the investment blogging world gives…
January 21st, 2008 at 5:12 pm
The interesting fact is that if you simply bought SPY on tuesday, you would outperform the market in 2008 by 10%!
January 21st, 2008 at 7:41 pm
Hold on to your hats, S&P futures are falling like a brick out of a New York skyscraper.
February 6th, 2008 at 11:51 am
Hi, this is a good analysis. The ‘january’ effect was first discussed by lakonishok and Smidt, in a paper dated of 1988.
In their study 90 years of data was observed and they concluded that each year return is higly dependent on teh january return.
After this ugly 2008 January I’ve made an similar analysis ans i found that the january return explains 30% of the anual return, of each year. Period observed 1990/2007. I also noticed that this model explains better the bull years than the bear years.
So, i’m not sure id my model will say much more than your 50/50 conclusion… but from the estimated regression one shall wait for a negative year. In fact, one shall expect the market to be lower in the end of the year than it was in the end of january… ouch!
February 24th, 2008 at 10:44 pm
There are two papers that you should read to get a handle on this.
Firstly, Cooper, M.J., McConnell, J.J., Ovtchinnikov, A.V., 2006. The other January effect. Journal of Financial Economics 82, 315-341. discusses the existence of the January effect in the US (ie the predictive power of January) arguing that it does indeed exist.
Then there is a second paper Easton, S. and Pinder, S. 2007. “A Refutation of the Existence of the Other January Effect”, International Review of Finance, 3-4 (7): pp. 89-104. which very clearly demonstrates that this effect is not really there either in the US or OS. indeed, the paper shows that five of the remaining 11 months are better predictors than January using raw returns.