Sexy dividends

In this age of massive stock repurchases, dividends have a distinctly stodgy, old-fashioned air about them, despite the fact that over the past few years they've been taxed at the same rate as capital gains. Yet, according this paper (PDF) published by fund manager Tweedy, Browne Company (ed. man, that's about the stodgiest, WASPiest name I've come accross in a long while).
(Hat tip: Abnormal Returns)

It concludes that there's plenty of evidence that high dividend yield stocks outperform market averages. (Not surprisingly, Tweedy, Browne recently launched a fund with this investmen strategy).

Needless to say, this is interesting. I must confess that I'm rather partial to dividends, as they're the most transparent way to return cash to investors. However, the paper solely emphasizes correlation and does not delve into causation, which is very unsatisfactory.

There are many possible explanations to account for the apparent success of a high dividend yield strategy. Maybe it's a sector effect. That is, perhaps firms in successful sectors happen to pay dividends. Or it could be that dividends are indicative of some positive trait, such as high capital spending discipline. I'm sure the academic literature has many other theories.

Also, the paper neglects to present studies that fail to support this conclusion (such as this this one).

As always, beware of research with deep conflicts of interest, no matter how academic it may look.