Two Easy Steps Toward Rating Agency Reform

|
 |  Includes: MCO, SPGI
by: Mark Alexander
The failure of Nationally Recognized Statistical Rating Organizations (NRSROs) to accurately rate mortgage backed securities during the last several years has fueled a growing consensus that the rating agency system in the United States needs to be reformed. Policymakers have introduced a number of potential reforms to our rating agency system, but unfortunately, many ideas that have been presented could take years to debate and mold into law. In the interim, there are a couple of reforms that should be relatively uncontroversial, would be relatively easy to implement, and could do a lot to improve the rating system used in the United States.

The first would be to establish a well publicized ranking system for the NRSROs. Rankings could be established for major industry groups. Structured finance might be considered one large group or rankings could be established for separate asset classes. If such a ranking system is established, it should begin with initial rankings that are established based on performance during some fixed window of time (perhaps the past five years) prior to implementing the ranking system.

These rankings should be based on common sense factors. A few examples follow.

  1. Ratings should correlate with defaults and credit losses. For example, AAA rated securities should experience much lower default rates and credit losses, on average, than below investment grade securities.
  2. Ratings should be reasonably responsive. A rating agency should receive a more favorable ranking for making appropriate rating revisions sooner than competitors.
  3. Consistent ratings that do not change dramatically are desirable.

Different evaluation methods might be needed for different rating agencies depending on each NRSRO’s guidelines. However, a small team of motivated analysts should, within a few months, be able to establish a ranking framework and initial rankings that differentiate reasonably well between fair ratings and rating failures.

A reasonably well devised program would foster healthy competition among rating agencies. It would also encourage issuers and other market participants to rely on ratings from agencies whose ratings have proven most accurate. Such a plan might also encourage issuers and market participants to utilize smaller firms with specialized expertise and established track records.

A second change that would improve the rating agency framework would be to establish a single rating scale based on expected credit losses to be used by all NRSROs. To the extent possible, this framework should be reasonably consistent with the existing guidelines currently used by the largest NRSROs. An expected loss based rating scale would be desirable because an investor’s likely loss, which is affected by the default probability and the likely loss in the event of default, is generally more important than the probability of default alone.

Disclosure: No positions