August 5, 2003
Jonathan G. Katz, Secretary
Re: Concept Release: Rating Agencies and the Use of Credit Ratings
Dear Mr. Katz:
Dominion Bond Rating Service Limited ("DBRS") appreciates this opportunity to comment on the above-referenced Concept Release regarding rating agencies and the use of credit ratings under the federal securities laws (the "Concept Release"). Following up on the Commission's efforts in this area in 1994 and 1997,1 the rating agency hearings it held last fall2 and its report to Congress earlier this year,3 the Concept Release seeks public input on a broad range of issues relating to the role of credit rating agencies in the U.S. capital markets in general and the SEC's role in designating Nationally Recognized Statistical Rating Organizations ("NRSROs") in particular.
DBRS is a Toronto-based, full-service credit rating agency established in 1976. Privately owned and operated without affiliation to any financial institution, DBRS is respected for its independent, third-party evaluations of corporate and government issues, spanning North America, Europe and Asia. DBRS' extensive coverage of securitizations and structured finance transactions solidifies its standing as a leading provider of comprehensive, in-depth credit analysis.
Since its inception, DBRS has worked continuously at refining its expertise in the analysis of credit quality. Investors and borrowers alike have come to rely on DBRS' ratings as the standard in the Canadian marketplace, based on the consistency, objectivity and unbiased approach taken on each rating. Due to the large volume of cross-border activity between Canada and the United States, DBRS' ratings have also gained broad market acceptance in the U.S. At this time, DBRS rates some 700 different entities and provides credit research on another 250 (mostly U.S.) companies. DBRS' ratings and research are disseminated through various means, including a proprietary subscription service that is used by more than 3,500 institutional investors, financial institutions and government bodies. In addition to being recognized by the Ontario Securities Commission ("OSC") as an "approved rating organization,"4 DBRS has also been designated by the SEC as an NRSRO.5
DBRS commends the Commission on the thoroughness and thoughtfulness of the Concept Release. As explained in more detail below, DBRS believes that the ratings and related credit analyses that the NRSROs provide are extremely beneficial to the issuers, investors and financial intermediaries who rely on them. NRSRO ratings have become so integral to the capital markets that eliminating the NRSRO concept from the Commission's rules would, we believe, cause serious problems without producing countervailing benefits. We further maintain that there is no need to establish a new regulatory regime in this area; microregulating rating agencies would impede their independence and ability to refine rating methodologies in response to market developments, and would deprive the markets of the diversity of views on which the current system thrives. The one area where we do see a need for change is the manner in which NRSROs are designated. In this regard, we recommend that the Commission abandon the current no-action letter approach in favor of a formal application process. We further recommend that this process allow for the designation of certain limited-purpose NRSROs, although we do not endorse the concept of provisional designation.
In order to evaluate the role of credit rating agencies in the capital markets, it is necessary to have a clear understanding of what a credit rating is and what it is not. A credit rating is an opinion regarding the creditworthiness of a company, security or obligation. It is not an absolute predictor of whether a particular debtor will default on a particular obligation. Credit ratings are a critical assessment tool for investors in fixed-income securities or other debt instruments, as well as for issuers seeking access to the capital markets. In addition, for more than 25 years, the SEC and other regulators have used the credit ratings issued by market-recognized credible agencies to distinguish among grades of creditworthiness of variousinstruments and to help monitor the risk of investments held by regulated entities. As the debt markets have grown more complex and more volatile, investors, issuers and regulators have grown increasingly reliant on the impartial and independent ratings and credit analyses that the NRSROs supply.
The confidence the marketplace and the regulators have placed in these rating agencies is well-deserved. Academic and industry studies uniformly show a strong correlation between credit ratings and the likelihood of default over time. We respectfully submit that recent headline-grabbing corporate failures should be seen for what they are: aberrations caused by spectacular issuer dishonesty and not signs of structural defects in the ratings industry or the regulation thereof. Indeed, the administrative and congressional scrutiny of credit rating agencies over the past eighteen months has failed to uncover any systemic flaws in the way NRSROs operate. There is no need to dismantle a system that has served the capital markets so well for so long.
With this background in mind, we turn our attention to the specific topics the Commission has raised.
Alternatives to the NRSRO Designation
The Concept Release seeks comment on whether the NRSRO designation should be eliminated from Commission rules, and if so, whether there are workable alternatives to that designation. DBRS submits that abandoning the NRSRO concept is neither advisable nor feasible.
First, NRSRO ratings have become so firmly embedded in the U.S. capital markets that eliminating the NRSRO designation would be enormously disruptive. Investors, issuers and financial intermediaries rely on these rating agencies to provide independent, objective and credible opinions on creditworthiness. As one of the buy-side participants in the SEC's Rating Agency Hearings said of NRSRO ratings: "They are the very structure of the marketplace . . . . They are the risk language that we all speak and rely on."6
Moreover, because the reliability of NRSRO ratings has been proven over time, and because -- despite recent SEC and Congressional scrutiny -- no systemic flaws in the NRSRO system have been identified, there is nothing to be gained by up-ending the NRSRO structure. Eliminating the NRSRO system would cause a profound market disruption without any countervailing benefit.
A third reason not to eliminate the NRSRO designation is that there is no viable alternative. Performing reliable credit analysis and evaluation is a very difficult and capital-intensive process. Over the years, each of the NRSROs has devoted considerable resources to assembling the staff, developing the expertise and refining the methodologies necessary to produce credible ratings and analyses. Moreover, NRSROs' commitment to evaluating credit risk does not end with the issuance of a rating or report; rather they monitor creditworthiness on a day-to-day basis and adjust their ratings over time as needed. In order to ensure the objectivity of their ratings, DBRS and the other NRSROs have also developed extensive internal controls to manage potential conflicts of interest. DBRS, for example, engages in no business other than producing credit ratings and related research. DBRS further prohibits its employees from (and itself refrains from) purchasing any security issued by companies that it rates or otherwise follows. All ratings are decided on a committee basis, thus neutralizing any potential positive or negative bias on the part of a particular analyst. And the company does not tie analyst compensation to any particular rating or the amount of revenue generated from issuers within the analysts' respective areas.
Replicating the expertise, experience, commitment and objectivity of the NRSROs would be difficult if not impossible to achieve. In particular, DBRS submits that allowing broker-dealers to use internally-generated credit ratings for net capital purposes is a poor substitute for the objective credit assessments supplied by the NRSROs. Because credit analysis is such a capital- and labor-intensive undertaking, only a handful of broker-dealers would be capable of assuming this task, and it is unlikely that these few firms could accommodate the needs of the rest of the broker-dealer community in a timely and cost-effective manner.
Allowing broker-dealers to assess the creditworthiness of their own securities for net capital purposes would also introduce substantial conflicts of interest into what has heretofore been a relatively conflict-free process. In addition to the inevitable pressures imposed by the net capital rule's treatment of non-investment grade securities, a broker-dealer's internal credit analysis could also be influenced by its investment banking or market making activities. Because NRSROs do not have these conflicts, the regulatory structure used to oversee rating agencies today would be of little use in overseeing the credit-rating activities of broker-dealers; instead, a whole new regulatory structure would need to be developed. DBRS respectfully submits that in light of recent issues relating to investment analyst conflicts of interest and the spate of as-yet untested regulations that have been adopted to address those conflicts, this is a particularly inopportune time to impose a new layer of conflicts and related controls on the broker-dealer industry.
Allowing broker-dealers to use credit spreads instead of NRSRO ratings for net capital purposes is problematic as well. Credit ratings like those issued by DBRS are produced by a "through-the-cycle" methodology that emphasizes stability. Spreads on the other hand, reflect the last trade in the marketplace and not a company's underlying fundamentals. As such, credit spreads tend to be volatile and are not always an accurate assessment of risk. They areparticularly unsuited to evaluating new or private securities issues or thinly traded securities. For these reasons, DBRS does not believe credit spreads present a viable alternative to the use of NRSRO ratings.
DBRS also believes that investors would be ill-served by eliminating the objective component under Investment Company Act rule 2a-7. As it stands now, the credit analysis performed by NRSROs and money market fund advisers complement each other. Fund investors benefit from this "two-view" approach which adds a stream of independent, reliable information to the investment manager's internal decision-making process. It is hard to see how eliminating this independent information would improve the safety or soundness of fund management.
Finally, with regard to the question of oversight, DBRS submits that the SEC, with its twenty-eight years of experience, is by far the most appropriate body to designate and monitor NRSROs. Transferring responsibility for credit rating agencies to the SROs or other federal regulators who have no experience in this field would be inefficient and would be of no benefit to investors, issuers or financial intermediaries.
The SEC staff currently designates NRSROs through a no-action letter process. In granting such a designation, the staff considers a range of factors, the most important of which is whether the organization is widely accepted in the United States as an issuer of credible and reliable ratings by users of securities ratings. Among the other factors that the staff now considers are:
Because DBRS believes that the marketplace is the best judge of what constitutes a reliable credit rating, DBRS also believes that market acceptance is the most important test for determining whether a rating agency should be designated as an NRSRO. This market acceptance could be demonstrated in a number of ways, including proof of substantial use of an agency's ratings by major users of credit ratings in the U.S.; proof that the agency has been retained to rate securities issued by a broad group of well-capitalized firms; or even proof thatthe agency's ratings are widely cited in the U.S. financial and mainstream press. As the Concept Release notes, emphasizing market acceptance in designating NRSROs has been criticized for imposing unjustified barriers to entry into the business of acting as a credit rating agency. However, as the most recently designated NRSRO and as one headquartered outside the United States, DBRS is proof that it is possible to become nationally recognized by active participants in the U.S. capital markets without the NRSRO designation.
In addition to market acceptance, there are other factors that the Commission could continue to consider before designating a rating agency as an NRSRO. For example, DBRS believes it is appropriate to examine whether an agency seeking NRSRO designation maintains policies and procedures reasonably designed to avoid conflicts of interest and to prevent the misuse of material, nonpublic information, and to evaluate whether a rating agency has adequate resources or other safeguards to maintain its independence from the issuers it rates. It would also be appropriate, in DBRS' view, for the Commission to evaluate an agency's commitment to transparency, that is, the degree to which it makes its ratings publicly available and discloses the reasons for its ratings.
What the Commission should not do, in DBRS' opinion, is interpose itself in the process by which a credit analysis is performed or a rating is issued. Whether credit opinions are produced though traditional methods or statistical models, the Commission should neither dictate how a rating is done nor define how the quality of a rating should be evaluated. Credible, reliable rating agencies may utilize different methodologies, adopt varying outlooks and reach different conclusions regarding the creditworthiness of an issuer or obligation. This richness of opinion contributes to the safety and soundness of the markets and would be lost if every NRSRO were obliged to follow the same script. Indeed, ratings diversity increases the "watchdog" function credit ratings agencies play, and their ability to function independently helps to disperse their power. Furthermore, mandatory standardization of the ratings process would ossify credit risk practice and theory, thereby impeding ratings agencies' ability to evolve with the natural evolution of the marketplace. Credit ratings are under constant scrutiny by market participants; the Commission should allow the market to determine whether or not an agency's credit opinions have value.
In addition to requesting comment on the types of information the Commission should evaluate in granting an NRSRO designation, the Concept Release also seeks comment on the process by which NRSROs are designated. Based on its recent experience, DBRS respectfully submits that the current no-action letter process is unnecessarily cumbersome and insufficiently transparent. DBRS therefore recommends that the Commission adopt a formal application process that provides clearly articulated standards and allows for notice and the opportunity for public comment. Establishing a time period within which the Commission will act on an application will add certainty to the process, although the period selected should be long enough to permit a thorough review of the application. Applicants who are not granted NRSRO designation should be notified of the reasons for their rejection so that they may improve theiroperations in the specified areas and increase their chances of submitting a successful application in the future. DBRS believes that these measures will greatly increase the transparency of the designation process and enhance investor confidence.
Finally, in order to promote healthy competition in the credit rating industry, DBRS supports the idea of granting limited-purpose NRSRO designations to credible agencies whose ratings have achieved broad market acceptance in a limited sector of the debt market or with respect to a limited geographic region. However, DBRS does not support the granting of provisional designations. It is not clear what the utility of a "provisional" rating would be, and maintaining two classes of NRSROs in a given situation would likely cause confusion in the marketplace.
Examination and Oversight of NRSROs
Many of the same principles that DBRS articulated in its discussion of recognition criteria apply with equal force to the question of the Commission's oversight of NRSROs. In particular, for the reasons explained above, DBRS believes that it would be inappropriate for the SEC to involve itself in the credit ratings process, and that the marketplace should be the primary monitor of ratings quality on an ongoing basis.
That being said, DBRS is aware that the Commission cannot simply designate the NRSROs and walk away. DBRS submits, however, that any ongoing regulatory oversight should be limited to ensuring that a rating agency continues to meet the minimum standards that led to its NRSRO designation in the first place. As it stands today, the SEC's staff reserves the right to withdraw its no-action relief from any rating agency that ceases to comply with the representations it made in securing an NRSRO designation. Should the Commission adopt a formalized application process for designating NRSROs, a formalized process for revoking that designation (with rights of appeal) should be instituted as well.
It might also be appropriate to institute some form of periodic self-certification whereby the NRSROs attest that their ratings continue to enjoy widespread market acceptance; that they continue to maintain internal policies and procedures reasonably designed to protect material, non-public information and to prevent conflicts of interest; and that they continue to broadly disseminate both their credit ratings and the reasons therefor.7 Additional measures may also be warranted provided that they do not deprive NRSROs of their independence or eliminate the diversity of opinion that has served the markets so well. Care also must be taken to ensure thatany new regulation does not have a chilling effect on the development of new credit analysis techniques and practices.
Conflicts of Interest
Like the other NRSROs, DBRS derives most of its revenue from fees charged to issuers, and also receives fees from investors who subscribe to its credit analyses and reports. Questions have been raised as to whether this fee structure compromises the objectivity of credit ratings; in particular, whether the receipt of fees from issuers presents the potential for rating inflation.
In exploring this topic, it is important to note that the current industry fee structure is the result of the complexity of the debt markets and the desire to have credit ratings broadly disseminated to the investing public. As the witnesses at the Rating Agency Hearings explained, performing high-quality credit analysis is a costly process, and although the public wants access to credit ratings, they do not necessarily want to pay for it.8 The only way rating agencies can afford to provide initial valuations and ongoing credit monitoring to the public is to charge the issuers whose securities they rate.
It is also important to recognize that eliminating fees from issuers would not necessarily eliminate rating agency conflicts of interest. Potential conflicts can arise from any number of relationships, including those with government bodies, regulators, investors, prospects and financial institutions. For example, accepting fees only from investors might still compromise the objectivity of rating agencies since investors have a strong interest in maintaining high ratings on the securities in their portfolios. Moving to an exclusively subscriber-funded business model would also diminish the fairness of the markets, since only those who pay for credit ratings would have access to them. And eliminating public dissemination of ratings could cause market confusion by exposing investors to rumors of rating actions and the like.
DBRS further notes that although the current industry fee structure has been in place for decades, there is no evidence that it has had a deleterious effect on the quality of credit ratings. There are a number of reasons why this is so. Perhaps most important is the fact that rating agencies live and die by the quality of their ratings and their reputation for objectivity. The fact that NRSROs derive substantial fees from issuers is widely known. If an NRSRO were seen to appease any issuer by supplying an inflated rating, the marketplace would discount that agency's opinions across its ratings universe. Such a discount would be an economic catastrophe for the rating agency, and would lead to a loss of national recognition. Moreover, a rating agency cannot avoid the reputational impact of any conflict of interest by concealing the reasons for itsratings, since ratings have to be transparent in order to be deemed valuable by market participants.
To safeguard their reputations and ensure the objectivity of their ratings, DBRS and the other NRSROs have developed a range of internal controls to manage potential conflicts of interest. As noted previously, DBRS engages in no business other than producing credit ratings and related research, and no one issuer accounts for a significant percentage of the company's total revenues. Furthermore, ratings at DBRS are determined by a committee comprised of the firm's most senior staff with input from analyst teams that produce initial rating opinions and the rationales therefor. This collaborative process effectively neutralizes any positive or negative bias on the part of any one individual. In order to further eliminate an analyst's or rating committee member's individual interest in a credit analysis or valuation, DBRS prohibits its employees from purchasing any security issued by companies that it rates or otherwise follows. The company likewise refrains from buying such securities for its own account. Finally, DBRS does not compensate its analysts on the basis of any particular ratings or the amount of revenue generated from issuers within the analysts' respective areas. Rather, analyst compensation depends on the experience, skill and quality of the analyst's work, as well as on the company's general revenues. DBRS believes that these internal policies effectively address the potential conflicts posed by the current NRSRO fee structure.
In addition to the potential conflict of interest posed by the NRSROs' receipt of fees from issuers, the Concept Release also seeks comment on whether subscribers may learn of potential rating actions or other nonpublic information by virtue of their preferential access to rating analysts.9 In part, this concern stems from the fact that NRSROs, like all credit rating agencies, may receive material nonpublic information from issuers pursuant to an exemption from Regulation FD.
The issue of NRSRO-subscriber contact was thoroughly vetted at the SEC's Rating Agency Hearings last fall. Two consistent themes emerged from this discussion. The first is that investors derive substantial benefits from their ability to speak to credit rating agency analysts.10 The second is that there is absolutely no evidence that credit rating agencies selectively disclose nonpublic information to their subscribers.11 For these reasons, DBRSbelieves prohibiting NRSROs from discussing ratings actions with subscribers would do more harm than good.
With regard to the other questions posed in the Concept Release, DBRS believes that it would be appropriate for the SEC to condition NRSRO designation on a rating agency's adoption of formal procedures to address potential conflicts of interest, including those arising from potential issuer influence and selective disclosure to subscribers. These might include procedures reasonably designed to minimize the rating agency's dependence on any one issuer; procedures reasonably designed to minimize analysts' and rating committee members' personal interest in the outcome of any credit analysis; and procedures reasonably designed to prevent the misuse of material, nonpublic information.12 However, just as there is more than one valid rating methodology, DBRS submits that there also is more than one set of procedures that could address rating agencies' potential conflicts of interest. Thus, DBRS does not support the imposition of any specific conflict-related conditions, such as net asset or revenue requirements, caps on per-issuer revenues, prohibiting NRSROs from offering consulting or other advisory services, or prohibiting analysts and committee members from participating in the solicitation of new business.13
In addition to requiring NRSROs to adopt reasonable internal policies and procedures, DBRS also supports mandatory disclosure by NRSROs of the existing and potential conflicts discussed herein. Each of these measures would codify current NRSRO practices which have served the market place very well.
Alleged Anticompetitive, Abusive and Unfair Practices
DBRS supports the goal of full and fair competition among credit rating agencies and believes that for the most part, the relationship among credit agencies, rated entities and subscribers serves the securities markets well. As discussed above, DBRS feels that market acceptance is the most important test for determining whether a rating agency should be designated as an NRSRO, and that the Commission should avoid imposing standards for NRSRO recognition that would involve the Commission in the credit rating process.
We also believe that certain rating agency practices mentioned in the Concept Release, such as unsolicited ratings, actually contribute to full and fair competition among the agencies, as long as they are not carried out in an abusive manner. Unsolicited ratings clearly should not be issued solely for the purpose of inducing the payment of fees, without issuer coöperation or investor support. DBRS has never engaged in such a practice, and believes it should be discouraged. On the other hand, unsolicited ratings which are initiated based upon investor demand, or for the purpose of gaining a better picture of an entire industry group, can provide great value to the market by supplying an additional, independent source of information concerning an issuer. Such ratings do not raise regulatory concerns at all.
With respect to the practice of "notching"14 which is described in the Concept Release and further discussed in the comments of Fitch, Inc.,15 DBRS believes that this may be one industry practice that does interfere with an open and competitive market. Accordingly, this practice may warrant further Commission review.
DBRS considers transparency to be a key factor in the ratings process; indeed, we believe that transparency has been a historic strength of our company. In order to ensure that those who use our ratings understand the bases for our opinions, we back up each of our ratings with detailed reports on individual companies and industries. These reports openly convey our views on both current ratings and the direction of ratings. We also hold regular seminars, investor meetings and conference calls, all of which allow for an open and informative dialogue with the investment community. Although DBRS feels that it is possible to accurately assess an issuer's creditworthiness using only public information, it is our practice to identify with a "p" any reports produced without issuer involvement, in order to provide context to subscribers and the public. Where we have ceased to rate or follow an issuer, we disclose the fact that our ratings are not current.
While DBRS is committed to publicly disseminating its ratings and concise explanations of its reasons and methodologies, we also believe that NRSROs should be entitled to provide more in-depth coverage and analysis to investors on a subscription basis, if that is their business model. As explained previously, DBRS and the other NRSROs have established effective controls to prevent the selective disclosure of ratings, ratings actions or other non-public information to such subscribers.
DBRS further believes that while the Commission could properly consider a rating agency's overall commitment to transparency as part of the NRSRO recognition process, it is unnecessary to establish specific disclosure requirements for the credit ratings business, either through direct government intervention or the imposition of mandatory industry standards. For the reasons articulated above, we maintain that the marketplace already exerts adequate control over this industry. Where transparency is concerned, a rating agency simply will not be "nationally recognized" unless the market understands the meaning of and reasons for its ratings.
DBRS believes that NRSROs have served the markets well in the past and continue to do so today. While the Concept Release raises some important questions about the role of these credit rating agencies under the federal securities laws, DBRS believes on balance that radical changes -- either the elimination of the NRSRO designation or the institution of a new regulatory regime for NRSROs -- are not necessary. The one area in which we do see a need for improvement is in the process by which NRSROs are designated. In this regard, we support the adoption of a formal application process along the lines described above.
We appreciate the opportunity to comment on these important issues, and would be happy to supply additional information to the Commission or the staff if you so desire.
cc: (By Hand)
Hon. William H. Donaldson