Tuesday, March 11, 2008

Modeling*: The Map is Not the Territory

Back in August we took exception to Ben Stein's contention that the sky was not falling in Ben Stein and the Markets (nutshell: Ben's wrong):

Mr. Stein is glossing over, at minimum, three issues that the sub-prime mess raises.

1) To the extent that sub-prime losses are held by publicly traded companies or their affiliated funds, the amount that should be subtracted from their market capitalization is not the recognized loss but the loss times their P.E. multiple (I shouldn't have to make an argument by definition, Ben knows this stuff). This gives you a much larger number.

2) The banks are reporting that they can't quantify the extent of their losses. What? Who are these guys?

They trade based on models. Here's a must read on modeling,
"Numerical Models, Integrated Circuits and Global Warming Theory", it was written in regards to climate models but if you read it to the end you will:
a) have a pretty good understanding of some of the problems of modeling in general; and
b) you will think "This guy knows what he's talking about".

Short version on modeling errors? As Alfred Korzybski* said
"The map is not the territory"...

Yesterday Eric Falkenstein Laid it out in six words:
Models are always wrong, sometimes useful.

Here are the other titles in our Ben Stein series:

Up to $4 Trillion Decline in U.S. Household Real Estate Value Predicted
Ben Stein, My Trading Floor Be-atch

Here's a bit on Korzybski from the Wikipedia article, linked above:

...The essence of Korzybski's work was the view that human beings are limited in what they know by (1) the structure of their nervous systems, and (2) the structure of their languages. Human beings cannot experience the world directly, but only through their "abstractions" (nonverbal impressions or "gleanings" derived from the nervous system, and verbal indicators expressed and derived from language). Sometimes our perceptions and our languages actually mislead us as to the "facts" with which we must deal. Our understanding of what is going on sometimes lacks similarity of structure with what is actually going on. He stressed training in awareness of abstracting, using techniques that he had derived from his study of mathematics and science. He called this awareness, this goal of his system, "consciousness of abstracting." His system included modifying the way we approach the world, e.g., with an attitude of "I don't know; let's see," to better discover or reflect its realities as shown by modern science. One of these techniques involved becoming inwardly and outwardly quiet, an experience that he called, "silence on the objective levels."

*This is another type of modeling error:

Friday night, however, I got to thinking. What if I really was God’s gift to modeling? What if I really was meant to be the next top model? What if I didn’t go and I would never know how my life could change? I mean, really, who passes up the opportunity to be famous, well-liked, and have the potential to make enough money to pay off student loans? I finally decided that “you live once” and printed out the application online.