Monday, September 17, 2007

Economist discuss Iraq data

The Freakonomics guys link to the following report by economist Michael Greenstone. He takes a pretty detailed look at the recent data out of Iraq to try and come to a definitive answer regarding the 'surge' (the increase in US troops since January).

Firstly, it's good to see someone with a lot of experience in information assessment try to tackle this question. Secondly, it's good to see someone who (I assume) has no political axe to grind entering the debate.

I haven't read the whole article yet, so might get to a longer post about it later.

The first thing that struck me as a bit strange though was Greenstone's faith in the predicitive power of Iraqi government bonds. He states:

I examine the price of Iraqi state bonds, which the Iraqi government is currently servicing, on world financial markets. After the Surge, there is a sharp decline in the price of those bonds, relative to alternative bonds. The decline signaled a 40% increase in the market's expectation that Iraq will default. This finding suggests that to date the Surge is failing to pave the way toward a stable Iraq and may in fact be undermining it.
The logic makes sense - bonds represent future income so their price can be indicative of the likelihood of that income eventuating. However, why do the people trading these bonds have any better sense of if or when Iraq will fall/fail? The don't. They watch the same news broadcasts and read the same information you and I do. Their opinions suffer from the same biases their sources do. So do ours.

I guess there might be some merit in the averaging effects of a large market, but that just means you come to an average estimate, not a better one.

You just have to look at oil futures pre Iraq war to see that no one in that market was predicting what would come.

I'll read the paper and see what else he says. I think he makes some other very useful points. I guess I'm just not a fan of prediction markets. They are basically gambles. You may as well say the odds of a horse winning are best represented by the win and place money. You could bet that way, but you're never going to make much. I have 15 years of watching my father bet on horse races to back me up!


Digg this
Sphere: Related Content

No comments: