The Psy-Fi Pages

Wednesday, 9 July 2014

Q is for Quantification Fallacy

The Quantification Fallacy is a logic error where the premises don't justify the conclusion. It's very common in investing lore where the tendency to omit just small amounts of qualifying information can often lead to strong conclusions being drawn from frankly dubious underlying data.

Example

The classic example uses an appeal to authority: experts agree that investors should buy stocks.  Hurrah, you might think. Unfortunately the quantifier "some" has been omitted: some experts agree that investors should buy stocks. Not quite so hurrah, now we have to think for ourselves.

It also occurs in more subtle ways: 15 stock portfolios offer investors all the diversification that they need to protect against systemic risk. Want to guess what's missing? Yep, actually this should state: some 15 stock portfolios offer investors all the diversification that they need to protect against systemic risk. Unfortunately there's no way of telling which ones will work until after the event, which makes this a less useful "rule" than one otherwise might want. We have to think for ourselves, again.

Causes

The quantification fallacy is an error in logic of a type often used by people trying to manipulate situations to their own ends. You'll find compliance practitioners attempting to frame situations in ways that are favorable to them, for their own reasons; reasons that usually involve extracting money from us. We'll often fall prey to this type of problem in situations of technical difficulty - the 15 stock diversification issue is a classic problem. It's a very simple rule bandied about by people who don't really understand how it was derived.

Mitigation

Try adding the word "some" to the front of any very precise statement and see if it still makes sense. This doesn't always work, of course, but I've found it a useful rule of thumb for detecting the quantification fallacy at work. More generally the use of logical fallacies to confuse, bamboozle and generally discombobulate us can only really be conquered by gaining an understanding of the things. Try printing off this useful poster and sticking it to your bedroom ceiling. It'll give you something to think about while you're staring at it. Some of the time.

1 comment:

  1. First time at your block, I think is very good & valuable. I really enjoy the A to Z of Behavioral Bias. It is a must read. As an investor I tend to lucubrate strategies, they are usually wrong, always victim of a new bias. So, I decided to create a quantitative model to defend me from myself. Oh! Much better! I am tempted to tell you the model works, but still not a millionaire, just afloat. Turn to you in 2 years.

    ReplyDelete