Speech by SEC Commissioner:
Statement before the Open Meeting regarding NRSRO Proposing Release
Paul S. Atkins
U.S. Securities and Exchange Commission
March 3, 2005
The SEC first addressed the lack of transparency in the NRSRO no-action process when I first worked at the Commission under Chairman Breeden. Even in those days, we saw that our jurisdiction was rather debatable. Then, in 1994, the SEC released a concept release outlining formal procedures for designating and monitoring NRSRO activities. Since that time, market participants and the Congress have, on more than one occasion, sought guidance, held hearings and requested Commission action. In 1997, the SEC defined the term "NRSRO" in a rule proposal but failed to take further action. The collapse of Enron and Worldcom highlighted this inaction as the media spotlight turned once again to credit rating agencies.
Today, the Commission is proposing a different definition for the term "NRSRO". Credit rating agencies meeting the proposed three-pronged test would qualify as NRSROs under SEC rules.
The limited approach taken in the current proposal reflects the SEC's lack of statutory authority over NRSROs. This informal designation for net capital rule purposes took on a life and meaning all its own over the years. Unfortunately, for far too long, investors and industry participants believed the SEC monitored and set standards for the NRSROs. The Washington Post recently called the designation a US Government "stamp of approval." I hope that this open meeting and proposal will draw public awareness to the SEC's limited legal authority over credit rating agencies.
Today, credit rating agencies obtain formal NRSRO status by requesting a no-action letter from the staff. Broker-dealers rely on the NRSRO no-action letter as a security blanket for compliance with federal securities laws. The no-action approach raises basic questions of transparency, fairness, and the appropriate use of administrative powers, not to mention "informal rulemaking". Although this proposal does not alleviate my reservations about the no-action process, it does provide some guidance to credit rating agencies seeking the NRSRO designation.
The proposed NRSRO definition remains vague and the interpretive guidance contained in the release text is broad. However, this may be the first tentative step in reinstituting fairness and transparency into the current process. Should we eventually adopt this formulation, I will expect that the staff will keep the Commission completely apprised of the credit rating agencies seeking NRSRO status. I will also expect that the staff adhere to the time parameters established for staff action.
Not only do I support the goal of increased transparency in the no-action process, but I also support increasing competition among NRSROs. Ideally, I would like to see a real reduction in the regulatory barriers to entry in the credit rating industry.
First, I am concerned about granting NRSRO status only to those credit rating agencies that the staff deems to be "nationally recognized". The proposed rule does not provide objective thresholds for meeting the national recognition prong. The subjective nature of the national recognition standard is best illustrated by the following descriptive phrases from the proposing release: "reasonably relied upon in the marketplace"; "wide acceptance in the marketplace"; "generally accepted in the financial markets as an issuer of credible and reliable ratings by the predominate users of securities ratings". These general statements do not provide appropriate objective guidance.
I hope that commenters will address whether the SEC places too much emphasis on this subjective nationally recognized prong. The proposal justifies this standard in order to ensure the credibility and reliability of credit ratings used to monitor the financial soundness of broker-dealers. But is it important that a rating agency be widely recognizable if its ratings are credible and reliable? The proposal states that the nationally recognized standard reflects "a level of assurance [of] those relying on the ratings with regard to the dependability and consistency of the ratings…"
Even if true, the "nationally recognized" standard creates a dilemma for credit rating agencies seeking national recognition without the NRSRO status. This barrier to entry unnecessarily limits competition. In fact, the Department of Justice submitted a comment letter on the SEC's 1997 proposal noting that the "nationally recognized" test "is likely to create a nearly insurmountable barrier to new entrants in to the market for NRSRO services."
What value is the SEC creating by limiting competitive pressures? One problem with the proposed approach may be that it centers around the fluid concept of "national recognition" -- which is necessarily determined on a case-by-case basis - as a proxy for determining the credibility and reliability of ratings. Events over the past thirty years, including Enron and Worldcom, cause me to question whether this anti-competitive standard serves investors the best.
The continued use of the national recognition standard, versus some objective measure of credibility, will not prevent future failures to predict troubles at rated companies. We have to acknowledge that there will always be, unfortunately, financial and other fraud that leads to corporate failures. In 2002, the staff to the Senate Committee on Government Affairs submitted a report to the SEC noting that the NRSROs exhibited "a disappointing lack of diligence in their coverage" of Enron and concluded that ratings published by NRSROs must be based on diligent, careful work.
How do we achieve this increased diligence or efficiency in the rating process? It can only result from increased competition as agencies strive to meet the obvious market demand for timely, accurate information. Or, an unattractive alternative would be merit regulation.
The SEC's proposal explicitly chooses national recognition over competition and innovation. Assuming that Congress grants the SEC further authority to oversee NRSROs, an effective oversight program may ultimately involve an invasive and highly complicated regulatory regime. What sort of merit regulation could be utilized in an effort to control the diligence of the rating process? Well, we could outline the number of issues rated per analyst, their educational requirements, the appropriate methodology for credit ratings, and the level and type of fees to charge. Would such an approach actually have any chance of success? Would it add any value for the investor?
By taking this approach, we risk creating a misplaced investor belief that the SEC's extensive oversight would prevent any and all future market failures by the entrenched players.
Furthermore, if a Congressional grant of regulatory oversight for credit rating agencies is compromised by 1st Amendment protections, I am concerned that the imperfect regulatory structure will place the SEC in the same situation that it is in today. By that I mean, the market may overestimate the amount of protection afforded by the Commission's designation of the NRSRO status and corresponding regulatory oversight. After years of our reviewing this issue, I am not aware if Congress or the Commission has studied the effectiveness of such a regulatory structure on the reliability of credit ratings.
Another solution would be to increase the competitive landscape. Competitive pressures and innovation may be the best cure for concerns about stale ratings and high prices. I would like to see the SEC consider provisional status, admitting foreign credit rating agencies, and/or removing the nationally recognized prong of the definition. I support the publishing of this proposed definition for comment and look forward to hearing suggestions as to how it can be improved to increase competition within this industry.