We all agree – please tell me we all agree – that correlation does not mean causation. That buying a Ferrari does not make you rich. (And, for very similar reasons, that a Harvard education does not increase one’s lifetime earning potential by whatever-insane-figure-the-latest-study-claims.1
We know it in our heads, at least. But jumping from one to the other is a darned tricky habit to break. Further evidence of just how hard is provided by a recent blog on Forbes.com from David Leinweber. This is the guy who first launched on the world the undisputed alpha dog of the meaningless correlation pack: the uncannily strong link between butter production in Bangladesh and returns on the U.S. stock market. His analysis was first carried out in 1994 and has been very widely cited ever since then despite not being published until his 2009 book Nerds on Wall St. It is telling that David still gets occasional requests for the butter stats. (Not only do people miss his point, they are too lazy to get their own data.)
He cites other examples, including a paper that explores the correlation between the number of nine year-olds and the stock market, and another that tests GDP growth against what he refers to as “the economic indicator in your pants”. The less said the better. (Although the author of the latter paper does note that his result “yields little in terms of feasible policy recommendations.” Which is probably a good thing.)
Returning to investment world, the point here is just that we are all prone to confusing correlation with causation, and even the knowledge that we are prone to it does not stop us. (I suspect most of you had never thought about the Harvard thing before, for example.) It’s just part of the price of being human.
1In case it’s not clear: most individuals who go to Harvard already have above-average expected lifetime earnings before they even put in a college application. While, in this particular example, it is reasonable to think that attending Harvard does add to their earnings potential, it is not reasonable to give the school credit for the whole of the difference between what this group earns and what the rest of the population earns.