9/07/2007

09-07-07 - 3

Coming out of Burning Man in the ridiculous traffic jam, I was thinking about how traffic is sort of like capitalist markets. There were 8 lanes of traffic. Each lane is sort of like an equity or something - you "select" a lane, that's where you choose to put your investment. You have some information about each lane - you can see it moving - but your information is very far from perfect, you can't tell how they merge in the future, which ones might be blocked or have real slowpokes in them. So you have to make an educated guess and pick one.

Now the interesting part comes in. In theory, if the players are all smart and fluidly move between lanes at reasonable times, then all the lanes should equalize. That is, if one is much faster, people will move into it, and it will slow down. Similarly if one lane is much slower people will leave it and it will speed up. In that ideal world, when you come into the 8 lanes it shouldn't matter which one you chose.

In reality, that's very very far from the truth, and the reason is basically that people are retarded. They do retarded things like try to box people out, or leave way too big of gaps, both of which screw up the fluid flow. They also retardedly think that it's for some reason morally superior to just stay in one lane. In reality it is far better for the health of the whole system to make a reasonable amount of calculated lane changes when the flow is better in one lane than an another.

I find that when navigating traffic, if I give people any credit for brains I tend to make sub-optimal moves. I tend to make the best moves when I just assume everyone is totally retarded and make "1st order" moves. That is, if there's a gap where I can just ahead or a much faster lane, just go for it and don't assume it's bad for some reason. Some times you see a gap that no one is taking and you think "there must be a reason why no one is taking it". Yeah, that reason is that they're retarded and if you try to go to the "2nd level" you just out-think the problem and pass up a good gap.

Of course the analogies to poker are obvious, but this also reminds me of a lesson I keep relearning over and over about the stock market. In theory markets are just like these traffic lanes - the easy obvious moves have already been done so you shouldn't gain anything by making them, but in practice that's just not the case. One recent example of course is this subprime nonsense. Anyone with half a brain knew those things were/are garbage and huge risks, and if the markets are functioning properly that risk should be built into the pricing and it shouldn't be a surprise at all when defaults start happening. Of course they weren't. You can see the exact same thing happen when a company gets a big government contract - obviously they're going to benefit from it, you would think it's too late to jump on that news, but it's almost never too late to jump on total obvious news. The only exception to all this these days is anything that can be modelled with some kind of short term automatic system; those kinds of easy moves are indeed being sucked out of markets because so many people use computer models and do automated trading that purely numeric correlation type of "easy moves" are not so obvious any more. It still remains true that if you try to go "2nd level" and think "they must have already compensated for this" - you are usually wrong, no they probably haven't compensated for the totally obvious thing that they should have compensated for.

No comments:

old rants