March 12, 2007
March 12, 2007
To the Securities and Exchange Commission:
The intent of the Congress in the passage of the Credit Rating Agency Reform Act of 2006 was to encourage competition in and reduce the barriers to entry into the bond rating business that the Securities and Exchange Commission (SEC) had erected in 1975 with its creation of the "Nationally Recognized Statistical Rating Organization" (NRSRO) designation and had reinforced with its opaque "no action" letter approval process for new NRSROs. The SEC should keep this intent clearly in mind as it drafts the final regulations under this statute. The SEC's regulations should not unduly impair competition, inhibit entrants, nor favor incumbents; at the same time it should not discriminate against incumbents either.
I will comment on four aspects of the proposed rules:
(1) With respect to Item #6 and "reasonable fees" on the proposed NRSRO form (discussed on p. 33 of the proposed rules): The SEC should not be dictating the business model of any NRSRO, nor the fees that any NRSRO -- incumbent or entrant -- chooses to charge for its bond ratings. The market should be the judge in this respect.
(2) With respect to Exhibit #1 (credit ratings performance measurement statistics, discussed on pp. 39-41): Although there are obvious advantages to users to having performance data that are comparable across NRSROs, the danger to a detailed specification by the SEC of a set of performance data is that the SEC will thereby narrow unduly the types of data that are provided; the SEC could thereby influence the market and favor some NRSROs over others. Instead, the SEC should merely encourage the NRSROs to issue performance data as they see fit and then allow the market to make its own judgments as to which NRSROs' performances are worthy of further attention. Unless the number of NRSROs grows quite large -- in which case, the SEC can revisit this issue -- the market should be able to handle the comparability issues on its own.
(3) With respect to Exhibit #2 (procedures and methodologies used in determining credit ratings, discussed on pp. 42-45): Again, the provision of information as to the NRSROs' procedures and methodologies has obvious advantages for users. But the danger is that NRSROs may be required to reveal proprietary information as to their procedures and methodologies, which would then permit other NRSROs to free ride on that information and ultimately discourage the creation of superior procedures and methodologies. NRSROs currently reveal to the public some aspects of their procedures and methodologies. Again, this should be a market-driven process, with (at most) some modest encouragement for revelation by the SEC.
(4) With respect to proposed Rule 17g-6 -- Prohibited Unfair, Coercive, or Abusive Practices (discussed on pp. 94-105): The practices that the SEC is addressing here are best analyzed on the concept of "tying" under the antitrust laws, since an NRSRO would be conditioning the provision of one service (or the price of that service) with the client's purchasing one or more other services from the NRSRO. Modern antitrust analysis has indicated that there can be beneficial reasons for a seller to tie one good or service to another, but also that there can be deleterious, anti-competitive consequences from tying. The latter consequences can occur when the seller has market power in the tying product and the intent of the tying action is to maintain or increase the seller's market power in the tying and/or tied products -- e.g., by foreclosing or raising the costs of rivals or potential entrants ("raising rivals' costs"). This is the standard that the SEC should use in deciding what constitutes "unfair, coercive, or abusive practices", including its consideration of the practice of "notching".
If you have any questions or would like to discuss these comments further, please feel free to contact me at 212-998-0880 or through e-mail at Lwhite@stern.nyu.edu.
Lawrence J. White
Professor of Economics
Stern School of Business
New York University