Thursday, August 27, 2009

Investment heuristics

Note to self:

The concept is that when faced with something new and unfamiliar, people try and understand it by relating it to something they have previously encountered and understand. They then apply this model to it and perhaps run a few checks to see if it is similar, and tweak their understanding of the new thing according to the results of the checks. This is a lazy and efficient way to understand things, if you assume some basic knowledge in the domain -- you don't need to reinvent the wheel.

However, this system works best when you already have a good 'library' of models to draw from, and know what sort of questions to ask. If you can ask good questions, you will be able to isolate and identify key differences which make something new very different from something old.

Example:

Deer --> has 4 legs --> like a horse? --> Eats grass --> ok i guess it's safe to hang around.
Lion --> has 4 legs --> like a horse? --> eats deer --> shit!

Good questions.

Lion --> has 4 legs --> like a horse? --> has lots of hair --> must be a hairy version of a horse. Let's try and ride it, it must be more comfortable...

Bad questions.


Same with investment products.
Hmm, pays me interest every year? --> like a deposit? --> did i buy it at the same bank i put deposits with? --> yes, and it has a bank name in the product's name. --> probably a deposit. should be safe.

Bad questions!


So to empower people who use heuristics to understand financial products better, you have to increase the library of products that they understand. Easier said than done of course, since it is very effortful to build something from the ground up and understand it that way. And with a whole universe of things out there, how do you make sense of it?

Of course, luckily blur naive people can hope to rely on the collective wisdom of teachers around us. If you don't know what a lion is and whether it is safe, you could always ask someone who knows about wild animals. As long as he does not have an interest in feeding you to the lion so he can eat the deer, this is fine.

But perhaps if we can break up and organise the universe into small discrete manageable chunks based on basic ideas, one can try and understand it in small basic pieces. Then one just needs to be able to ask the right questions to figure out how complicated things differ from basic things. hmm.