by Tanta on 2/06/2008 10:08:00 AM

Moody's is throwing out some possibilities for changes to its rating system that would give more information to bond buyers, and is inviting comment from the investment community. Mr. Ritholtz over at Big Picture has a suggestion for a disclosure statement that has too many dirty words in it for me to post on my respectable blog, but here's the link.

Here are the proposed options, from Moodys' request for comment:

1. Move to a completely new rating scale for structured securities, for example, numerical rankings of 1-21. These would continue to contain ordinal rankings of expected credit risk and would probably map to corporate ratings.

2. Add a modifier to all structured ratings utilizing the existing rating scale, e.g., Aaa.sf. This would designate the issue as structured, but add no other additional information.

3. Add a suffix to the existing rating scale for structured ratings that contains additional information – for example, estimates of multi-notch rating transition risk. This could be Aaa.v1, Aaa.v2, etc. We would derive these gradations through an analytical process that would be disclosed to the market.

4. Use the existing rating scale for structured securities, and put additional analytical information in a separate scale that would exist in a separate data field. For example, an issue could have a “Aaa rating, with a ratings change risk indicator of v1”. The added field would be analogous to our existing ratings outlooks and watchlists.

5. Make no changes to the rating scale, but provide additional information and commentary through written research.

I personally like the idea of combining alpha ratings with a suffix code that indicates the assumptions built into the ratings models. For instance, a bond could be rated AAA.PONY, indicating that the assumptions built into the ratings were Prices always rise, Owners occupy all units, Nobody lied, and Your own analysts did all the due diligence.

Is that any more bizarre than giving a rating to a security with a "ratings change risk indicator"? Wasn't there a time when the rating given to a bond was supposed to be the one least likely to have to be changed?