The January Effect

“The January Effect” purports to be a market phenomenon in which share prices rise in the calendar month of January.  An alleged effect like this should prompt the simple question, “Why?” A US-centric explanation suggests that investors seek to create tax losses in December to offset capital gains earned during the preceding 11 months; in January, these transient sale transactions are reversed, leading to rises in share prices.

Were the January effect to be a consistent phenomenon, it would point to a level of inefficiency in the market because any predictable price movements should be eroded to insignificance by arbitrage. For example, if one group of investors (Group A) knew that another group of investors (Group B) were going to be selling shares in December, Group A would buy as Group B induced prices to fall by selling. By the same token, knowing that Group B would be buying shares back in January, Group A would be selling at a profit. Absent other catalysts, price indifference should result.

This begs the question whether or not any such phenomenon exists in the broader South African equity market. The average monthly return on the JSE since 1960 is 1.5%. Although the average return for the 52 Januaries in the sample is 2.4%, statistical tests tell us that the 0.9% difference is not significant. In fact, no month has an average return that is statistically significantly different from the overall average – except December. 

The South African equity market appears to have a “December Effect” with the average monthly return in December being 4.6% (which is shown to be very highly significantly different from the overall market average). Historically, the monthly return in December has been higher than the overall market average 69% of the time. 

Furthermore, the statistics also show that December’s returns are the most positively skewed of all calendar months, having the lowest minimum return and the highest maximum return too. 

Is this a case of market “peace on earth and goodwill towards men”? Does the South African equity market become so imbued with festive holiday spirit that equity returns are significantly higher than average? The tax argument might have some plausibility, with individual tax years ending in February in South Africa. Although without any statistical significance, the average monthly return in February is only 0.8%, some 50% of the overall market average. It is conceivable that December and January are a sort of ramp up to a sell-off in February for tax purposes.

Although interesting, the phenomenon of a “December Effect” points to an element of inefficiency in the market. It is also a short-term performance phenomenon. A preferable outcome would be to not have any one month being significantly different from the average. Instead, returns should be earned more consistently, regardless of the month. An analysis ofFoord’s equity returns over the past 18 years reveals that no single month has a return that is statistically significantly different from the overall monthly average of 2.02%. While Foord’s average returns in December follow (and, indeed are better than) the market’s, there is little to differentiate the December returns from those of the other months.

Are there logical inferences from this analysis? Indeed. First, it appears that Foord’s returns are not influenced by some aberrant and seemingly arbitrary behavioural phenomenon. More importantly, though, Foord’s average returns in the other months are consistently greater than the market, so much so as to erode to insignificance the difference between those other returns and the returns in December. The result is consistent outperformance over time, and it is this outcome that investors can and should be seeking.

Mike Soekoe, with statistical input from Darron West of the University of Cape Town