None of this would matter, of course, except that the ratings are used for capital ratio requirements. But I think we need to do away with the ratings, as they are just another way to game the system.

If we need to be able to rate bonds in order to allow higher leverage, but can’t find a way to rate bonds, then we need to reduce the allowable leverage.

A false sense of security is very dangerous.

Like stock prices, aren’t forecasts almost random walks? ;)

Seriously, there are many ways that a forecast can come about. There is a cloud of uncertainty. The larger the cloud, the less certain the forecast, and the closer the forecast approaches the mean, in general. (If there is no cloud of uncertainty, there is no “regression towards the mean”.) As the time of the forecast approaches, the cloud becomes smaller. This is the case, even if the current data is close to previous expectations. It is not too surprising, then, if a new forecast incorporating new data is further from the mean than previous forecasts, even if the new data is not surprising.

To be fair, however, if the rating agencies actually had spines and amended ratings when they thought something would change versus when they could demonstrate that something had changed, they would probably be criticized for that too (especially if what they forecast did not materialize).

I’ve started a Blog with a small collection of short stories called Baseline Fables. The stories are inspired by Baseline Scenario.

The Blog address is: http://www.tippygolden.wordpress.com

Hope it’s OK to post this here.

There is no pricing mechanism (ie no market) that would make them ‘price in’ anything.

What probably happened is that housing prices fell farther, and unemployment went higher, than they expected since the last time they did projections. They wouldn’t be the only ones who underestimated these things.