They found that January has the highest average return (1.5%) of all the months. It also has the lowest standard deviation (2.9%), which makes the outperformance even better on a risk-adjusted basis. However, CXO Advisory warns that the January uptick appears to be on a downtrend. Indeed, the current decade from 2001 to 2008 is the only decade that the January Effect has been negative.

But eyeballing their data, one could arrive at a more encouraging interpretation. The 1930s and 1970s saw noticeably above-average January gains of 2% and 3%, respectively. As the stock market in the current decade is turning out to be similarly volatile, perhaps the Januaries to come in 2009 and 2010 could see strong rallies after all.

The thesis could be that there is unusually heavy tax-loss selling during the down years in turbulent decades. They are then followed by unusually heavy buying to re-instate positions in the stocks sold for tax reasons. For more, see Big January Effect this year?

On average, says the Stock Trader’s Almanac for 2009, buying stocks trading at 52-week lows in the week before Christmas and holding them for a 4- to 6-week period has generated average gains of 13% since 1974, versus 3.4% in the market. This “Free Lunch” strategy performs best, says the Almanac, after market corrections and when there are a lot of new lows. That would seem to fit the description for 2008.

In the interests of disclosure, I should mention that I purchased a small-cap exchange-traded fund in November (iShares CDN Small Cap Index Fund). It’s intended as a long-term investment, but a nice uptick in January would help improve the morale.

By the way, for anyone still contemplating tax-loss selling, there is some academic research that suggests it may be better to do it in January, believe it or not. Apparently, you would be further ahead because the price gain from the January effect would on average offset the lower tax-loss claim.