Causation
I know, who doesn’t want to read a blog called causation? It sounds like something your logic/statistics professor was talking about while you were nursing a hangover. But there are practical applications of the concept.
In the August issue of Money Magazine, they build some conclusions from a specious relationship. And even if you didn’t pay attention in class, thanks to fluffy magazines, you can get a refresher in logic every month.
Here is their blurb (Money, August p. 24):
Maybe Older Really is Wiser
More evidence that seasoned managers do it better: A recent Standard & Poor’s study found that while the average tenure of a domestic fund head was 5.5 years, those who led funds performing in the top half for the past five years had been with their firms 8.6 years.
If you think about this statistic for a second, it makes perfect sense. Are over- or under-performing managers likely to stay around longer?
The answer, of course, is obvious. Maybe, it’s not older managers that have higher returns, it could be that the managers with higher returns keep their jobs. Not exactly a huge logical leap in an environment that rewards short-term performance.
The next time a magazine quotes some data that shows a correlation (i.e. A and B are related), and assumes that A causes B, question their logic. It could be that A causes B, B could cause A, or something else (say, C) causes A and B.
In this case, it is obvious that returns (B) would be the major cause of tenure (A) rather than the other way around.
