Speech by SEC Commissioner:
Proposed Rules for Nationally Recognized Statistical Rating Organizations
Commissioner Kathleen L. Casey
U.S. Securities and Exchange Commission
SEC Open Meeting
May 18, 2011
Thank you, Chairman Schapiro. It almost goes without saying how much appreciation is due the staff working on today’s release. But it is particularly noteworthy and worthwhile to recognize the work of Randall Roy. For many who have waded into the credit rating reform debate following the crisis, they may have missed the fact that the SEC and, in particular great folks like Randall, have been engaged in rating agency reforms, pretty much non-stop, over the past 5 years since the passage of the 2006 Rating Agency Reform Act. He has done so while at the same time coordinating on the consideration of many of these issues internationally as well. It would be difficult to overstate his tremendous work and effort on these issues. So, I truly want to commend him today.
Today caps another in a series of rule proposals we have undertaken to address regulatory concerns and weaknesses in the rating agency space. This effort to craft a coherent, rational, effective framework has not always been easy. Challenges are posed to the degree some reforms proposed or contemplated conflict and/or are animated by competing philosophies of how best to address concerns about the role and reliability of credit ratings and the regulation of credit rating agencies.
While the stated objectives of both the 2006 Credit Rating Agency Reform Act and the Dodd Frank Act seek to achieve greater accountability and integrity in ratings through transparency and competition and restoring market discipline, there are some provisions of the new law, including some of the rules we are proposing today, that if not considered and implemented carefully, could result in undermining these objectives.
As the release before us notes, many of the new Dodd Frank requirements are self-executing. Where this is the case, the proposed rules, in many respects, mirror the language of the Act in order to give effect to the law quite literally.
The internal controls requirements of the new law are one such example. And, while the Commission has the discretionary authority to prescribe factors the NRSRO must take into consideration with respect to its internal control structure under the law, should the Commission consider in the future exercising that authority, it must do so judiciously. The 404 control model is particularly burdensome for smaller entities, and in the rating space, could be life-threatening. The consequence could actually be to raise the barriers to entry in the rating space and undermining competition.
Some of the new requirements, build on or expand on existing requirements already adopted by the Commission, and are largely consonant with the broader objectives noted above.
Others, however, threaten to cross the line into regulating the substance of credit ratings. While the release notes a balancing effort to ensure that our rules do not have the effect of essentially regulating the substance of credit ratings or the procedures and methodologies by which an NRSRO determines credit ratings which is prohibited by the law, I am concerned that this is such a slippery slope that we must avoid stepping so closely to that line that we easily, although unintentionally, cross it.
There are two areas in particular where this arises I would encourage commenters to focus on – the look-back provisions requiring a credit watch for ratings determined to be influenced by a conflict and the disclosure matrices that provide standardized definitions for such terms as transition and default rates.
With respect to the look back provisions, I am also interested in commenters’ views on how practical its operation would be and whether our proposed approach would inject undue market uncertainty around a rating to the potential detriment of issuers and investors.
With respect to the disclosure matrices, I appreciate the interest in attempting to provide comparability to investors. Nonetheless, I am concerned that we are prescribing standardized definitions and approaches for transition and default rates despite the different meanings of NRSRO credit ratings and their definitions of default, thereby hardwiring standardized, static procedures and definitions in our rules that do not reflect market practice.
I would like to make one last point, and it should not be viewed as diminishing Randall’s good work here or the work of the Division of RiskFin. But, I wish to highlight an ongoing concern I have about how we are engaging in cost/benefit analysis with respect to the significant number of rulemakings that have flowed – and keep flowing – from Dodd-Frank.
Specifically, the Commission has been operating under an articulated view that cost-benefit analysis is only required and relevant to those rules – or those aspects of our rules – that we implement through the exercise of our discretionary authority. As a result, the Commission has not engaged in a cost-benefit analysis of the rulemakings that were essentially dictated by the law. This seems to be a narrow and limiting approach to the requirements and objectives of cost-benefit analysis.
I believe we are severely limited in our ability to act consistently with congressional interest and intent if we fail to understand the full impact of the new law and the regulations flowing from it. By limiting our cost-benefit analysis to those measures over which the Commission has full discretion, we fail to consider all the costs and benefits that will result from a particular regulatory action, whether or not that action was undertaken at the direct command of Congress, or through the exercise of our own judgment.
Indeed, I question whether it is even possible to adequately weigh the costs and benefits of discretionary regulatory actions without knowing the full, cumulative impact of the regulatory structure that ultimately will be erected via the twin actions of Congress and the Commission.
Moreover, it will be difficult, if not impossible, to assess effectively the ultimate impact of our rules and their interoperability if we are deliberately ignorant of all the resultant costs and benefits of those rules. Our rules do not exist in a vacuum – particularly in the context of the Dodd-Frank rulemakings, our rules intersect and interact. When we are choosing amongst regulatory policies, the total costs of other, related rules could have a significant impact on which approach is adopted. But without thorough and inclusive cost-benefit analyses, we are regulating blind.
This is not to say that we should adopt an approach to cost-benefit analysis that is inappropriately rigid in the opposite direction. Of course there will be instances in which it may be unnecessary to evaluate the costs and benefits of a regulation’s statutory underpinning. For example, it may not be useful to consider certain costs and benefits when congressional action simply creates a new baseline for our regulations. But by prematurely curtailing our cost-benefit analyses in all instances, we hobble ourselves as we make the sweeping regulatory changes mandated by Dodd-Frank.
Again, I wish to thank the staff for their hard work and I have no questions.