Oct

25

Of course, we always say correlation is not causation, but it goes the other way around as well. Let me explain. Years ago, when I lived in Monterey and Carmel, California, a friend of mine sold radio advertising for a country-western station. Monterey and Carmel, even back then, were upscale areas. My friend had a lock on all of the car dealers' radio advertising. I asked him why. He explained it.

When I would go give them my pitch on advertising on my station, they'd say, "Our people don't listen to country western music." He would say, "Well, go out in your car lot, look at all those cars that have been turned in (the trade-in cars, the used ones), and see where the car dials are set. Then you know what radio stations you should advertise on."

The owner would go out and check, and sure enough, almost all the dials had been set at or close to his country western station. My friend never touched the cars, didn't do anything with them. The owners bought advertising time from him.

What he had noticed, however… and this is interesting… is that when the cars were cleaned (remember, back then we didn't have punch buttons and screens to set radio stations, we had dials…) after the cars were detailed, people that were detailing them would usually spin the left dial to the left. That's where his radio station was, 550 or 600, something like that, so it was always close, simply because of the way the car was detailed.

So there was causation. Dials were set there for a reason, but there was no correlation to what people were really listening to.

And of course, we all know somebody who never went to classes in college and got straight A's. There's supposed to be a correlation with going to classes, taking notes, participating, sitting in the front row, etc., to grades. True for some people, but causation does not always mean correlation.

Adam Grimes writes:

This is a good illustration of the measurement problem–sometimes we're measuring something that is not connected to the causal mechanism we think it is. I think this is a great example of correlation not showing causation–dials turned to stations (observed) not caused by people listening to stations! The thing they were measuring was a poor proxy for what they thought they were measuring.

Nils Poertner adds:

Medicine suffers from this dilemma as well. eg, eyesight has a lot to do with habits (not just genetics or age), but if we buy into the idea of genetics /age (only), it is true for us. either way we are right.

Stefan Jovanovich writes:

The Family Office follows the LW rule. We have given up believing we know where the spectrum of time and events shifts from correlation to causation and back again. We don't think we have any idea about the connections between price and "value" in common stocks. We know, from observation and study, that the traders with skill do have the ability to make money at the casino; and, at the same time, we are absolutely certain that Eugene Fama is right - the discount mechanism that produces bids and asks has no special bubble territories.

We don't have anyone who can trade. So, our algorithm chooses companies based on their histories of success and assumes that their success as investments will depend on the fates and the probability that they will do better at surviving the random walk than the other ones we could have bought.

The algorithm says we should own NVDA common stock and the ADRs of two banks in Argentina.

Full disclosure: The FO has 4 adults who "manage" the money: an Eagle Scout EE PhD whom Nvidia pays to "research" the taping out of semiconductor circuitry, two lawyers - one of whom went to class, and a doctor who sells her time doing EMGs.

Humbert Q. offers:

There’s a wonderfully folksy aphorism in the car business that feels relevant here. “Half of every advertising dollar you spend is wasted, the problem is you don’t know which half”.

Stefan Jovanovich provides some history:

The wisecrack OJ notes was first recorded in an interview with Wanamaker, the department store magnate. He is quoted as saying to the reporter that he knew that half the money he spent on full page ads in the Bulletin was an utter waste of money. But he could never figure out whether it was the ads or the things he put in them that were the source of the loss.


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search