Saturday, May 15, 2010

How to implement the Franken & LeMieux NRSRO Amendments?

Interesting post on the Baseline Scenario on how to implement the Franken and LeMieux credit rating agency-related amendments to the Senate financial sector reform bill:
The two (perhaps contradictory) amendments each try to implement a proposed solution that runs into some of the critiques. The Franken amendment has rating agencies assigned to debt issues by a neutral arbiter; critics maintain that lack of competition may reduce the quality of analysis. The LeMieux amendment removes legal mandates to obtain a NRSRO rating and the preferential treatment those issues currently receive. However, it leaves out details about whose advice agencies and public trusts should seek out instead.

This is not such a difficult problem. We already have an example of a successful private rating agency, whose imprimatur is desired or in some cases required by law, that is paid for by fees on the seller, and has been operating since 1894: Underwriters Laboratory. The UL publishes safety standards for almost 20,000 different types of products, many of which are adopted by other standard-setting organizations like ANSI (American National Standards Institute) and Canada’s IRC (Institute for Research In Construction). Although generally not actually required by federal law, the sale of many types of products in the US would be difficult without UL listing. Also, many local jurisdictions responsible for building and fire codes mandate the use of UL approved products. In all cases, the manufacturer must submit samples and pay fees to UL in order to win approval.

The comparison to NRSROs is apt. In both cases, a third party sets standards based on theory, models, and best practices. In both cases, the issue is the assessment of risk by experts in that type of risk. In both cases, approval is desired by the market or required by local ordinance or rules. And in both cases, the seller pays the fees; so the third party might be led to relax their standards in order to capture some extra fee income. Yet in the case of fire safety the model has been functioning well for over 100 years, but in financial safety there has been a rash of fires as one rated product after another has blown up. Why?

There are a few key differences. Until 2007, UL was completely non-profit, so as long as user fees covered their costs there was little incentive to chase extra revenue by relaxing standards. There is no real competition in the US market for UL (although Europe has its own standard-setting body that manages the ), so manufacturers have little leverage to push for easier standards. The LeMieux amendment could allow for the creation of a not-for-profit entity to take the place of NRSROs, while the Franken amendment would reduce competition, limiting it to delivering a better, more reliable rating rather than adjusting standards to capture fees.

Critics of the amendments, including those who support a buyer-pays model, need to address the question of why the UL model for risk assessment has worked well, and why it can’t be applied to debt rating. Is the model broken? If so, I expect a rash of building fires any day. Is rating of financial safety fundamentally different than, say, electrical safety?
Some of the comments are kind of interesting too:
The analogy between UL and the NRSROs doesn’t take one very far. The products that the organizations rate are just too different, and the chief difference is simply that UL-rated products are tangible. As a result, it doesn’t generally pay for manufacturers to game the approval process by adding complexity that hides fundamental safety deficiencies. In the world of tangible products, increased complexity almost always equals increased manufacturing costs and lower expected profit. By contrast, in the securities world, increased complexity has a negligible impact on the cost of “manufacturing” the security and can result in a higher expected profit to the issuer and underwriter by hiding features that accrue to their benefit....

When a fire happens, it can generally be traced to a single root cause, or maybe two (e.g., somebody was smoking in bed + the sprinkler system failed).

Put another way, the conditions that cause physical failures do not change from year to year, because the laws of physics do not change.

As a result, if UL failed to do their job properly, it would be pretty easy to tell. It would not even take an expert.

In the financial sector, the top experts always disagree about what is safe and what is dangerous. That is the nature of the sector.

The financial sector is different because its “laws” change daily. The cause of the crisis was not ratings agency failure. The cause was a universal desire to get something for nothing combined with low interest rates. Had the ratings agencies tried to be careful, they would simply have been ignored, because nobody wants to hear about risks when they see their neighbors are getting rich doing nothing year after year after year.

No comments: