This Report brings to the fore many of the issues that commentators have been citing for a long time about problems inherent in the ratings market. This (long) post summarizes the Report and discusses briefly the findings, but future posts will most likely draw more on this Report and what it is likely to mean for US and EU regulation in the future.
Staff of the SEC have issued a report on the ten Nationally Recognized Statistical Rating Organizations (NRSROs) registered in the United States. These include the "big three" of Standard and Poor's, Moody's and Fitch, and the report - required by the Dodd-Frank Act - makes for some interesting reading. The Dodd-Frank Wall Street Reform and Consumer Protection Act 2010 requires the SEC to examine each NRSRO annually and report on findings. The aim of the increased oversight was to ensure compliance with the new reporting, disclosure and examination requirements as set out in Dodd-Frank. The CRAs, which were criticized for issues of inaccurate ratings relating to mortgage-backed securities leading up to the financial crisis, still have "apparent failures" in following their own methodologies and procedures, the report found.
One CRA, which is not identified (more on this later), apparently allowed some "dissemination" of a pending rating before it's public release. The report also notes apparent failures in some instances to make timely and accurate disclosures, to establish effective internal control structures for the rating process and to adequately manage conflicts of interest. The Report itself splits the ten CRAs into two groups, the first comprising the "big three" (Standard and Poor's, Moody's and Fitch), and the seven smaller CRAs operating as NRSROs. The Report notes that the larger CRAs are mainly financed along the lines of the issuer-pay model. While some of the smaller agencies have traditionally kept to subscriber-pay models, these are now beginning to explore more issue-pay models, in particular with respect to the rating of asset-backed securities.
The report also notes that since the 2008 Public Report, the larger agencies have taken steps to address the concerns raised. That report had examined operations relating to ratings given to certain Residential Mortgage-Backed Securities (RMBS) and Collateralized Debt Obligations (CDOs) and had highlighted concerns about the sufficiency of resources devoted to ratings and surveillance, the adequacy of disclosures, documentation of the rating process and management of conflicts of interest. The 2011 report states: "Each of the three larger NRSROs appears to have devoted notable resources and effort to responding to the concerns and recommendations outlined in the 2008 Public Report". Further detail on this point is notably absent, although conclusions can be drawn from the number and breadth of concerns raised in the Report.
The Report's essential findings also show that one larger CRA failed to follow its methodology for certain asset-backed securities, applying the wrong methodology to one class of securities which have now been put on review. This was reported to the Staff during the period of the review, causing concerns to be raised about the speed with which the error had come to light, been disclosed, and then remedied. This, according to the Report, led the Staff to express concerns about the extent to which lack of resources - a criticism of the 2008 Report - remains a factor, making the lack of detail about the extra resources put in place by the "big three" even more frustrating. The CRA in question apparently has greater resources devoted to the surveillance of ratings of asset-backed securities, yet the error still took months to come to light. The Report questions if this means that the CRA is following its own published criteria for the review of asset-backed securities ratings. Interestingly, the Report cites market share as one of the possible factors that led to a delay in discovering and disclosing the error and further delays in taking remedial action.
The Report further found that one of the smaller NRSROs was slow to disclose changes to its rating methodology for certain asset-backed securities and how those changes would apply to its ratings, and slow to apply those changes to outstanding ratings of affected asset-backed securities.
It is the issues highlighted by the Report in relation to management of conflicts of interest that is perhaps the most illuminating. The Staff identified "troubling weaknesses" in two of the smaller NRSROs with regards to their securities ownership policies and procedures, and the implementation of these. It was found that a lack of adequate documentation made it impossible to verify that the procedures had been adhered to, and that one key employee in particular had violated the policy by failing to adhere to securities trading pre-approval requirements. Moreover, it would appear that this same key employee held securities related to a business sector for which he participated in criteria development, in clear violation of the policy. In other of the smaller NRSROs, the securities and ownership policies, including monitoring and enforcement were found to be "haphazard and inconsistent" and one instance was identified where a key analyst may have directly owned a security of a company that was subject to a rating action in which the analyst participated, contrary to Rule 17g-5(c)(2) Dodd-Frank. Even more surprising is that the Report states that the Staff made recommendations that each enhance its policies and procedures, including documentation and enforcement, and presumably hope to find progress in the next annual report. The Report does stress that employee securities ownership policies play a crucial role in preventing the misuse of material non-public information, before stating that this area will comprise a focus of future reviews.
Issues raised with regulation of conflict of interest continue with a focus on the larger NRSROs, two of which were found not to have specific policies or procedures in place for managing potential conflict of rating issuers that may be significant shareholders in the NRSRO itself or a parent company, a situation the Staff found could give rise to a conflict of interest. Indeed, the Report states that two of the larger agencies issue ratings for companies that "may be" significant shareholders of the NRSRO or the NRSRO parent company. The Report recommends that the agencies establish procedures for managing these conflicts of interest where they arise, and this may include complete separation of such functions within the firm, although the Report makes no suggestions. This conflict of interest parallels that of the subscriber-pay business model, which the Report highlights as providing potential for conflicts of interest in three smaller NRSRO which appear to have weaknesses with regard to their policies and procedures to manage this. According to the Staff, the subscriber-pay model is more likely to see the subscriber exert pressure on the NRSRO in order to obtain a more desirable outcome.
Two of the smaller NRSROs were also found to have weaknesses in policies and procedures concerning the disclosure and management of conflicts of interest associated with certain ancillary business. This latter is defined in the Report as "any services which an NRSRO may provide which are not core credit rating services". Again, detail here is scant, and the Report does not specify the services referred to, but it is likely that some of these were product design and financial advice. The financial crash highlighted the practice of CRAs effectively advising on the market, assisting in financial product design, and then assigning these a rating. While separation of such functions was present prior to the crash, standards have been tightened since.
The Report goes on to examine internal supervisory controls, and found that in three of the smaller NRSROs these are "weak", including one firm lacking any internal audit function and having limited compliance personnel. The Report also found that of three of the smaller and one of the larger NRSROs, "the procedures for disseminating a pending rating action appeared to allow for limited dissemination of a pending rating action in some instances prior to public dissemination".
The Report notes in Part C that disclosure by one smaller NRSRO regarding certain rating methodologies appears "weak", while public descriptions of its committee process and details about the number of analysts working on credit ratings appear to be "misleading". Overall disclosure could be improved, according to the Report. However, as no NRSRO is named in the Report, it is difficult to see how the market can react to these findings. Admittedly, with only three larger players, and a total of ten NRSROs in the US market, it may be unrealistic to call for complete transparency in the hopes that the markets will respond by punishing those CRAs with poor documentation procedures. Indeed, transparency is not the focus of Dodd-Frank. But the trend in the Report does appear to be to call for increased disclosure from the NRSROs themselves. It seems a missed opportunity for this to be the same in the Staff Reports.
Finally, the Report gives no guidance on how to solve the interminable, and by now infamous, question of conflicts of interest in the ratings industry. Whether the issuer is a shareholder in the NRSRO or not, whether the rating is funded on the issuer-pay or subscriber-pay model, and whether the NRSRO employee holds securities in the issuer (or pays into a pension which holds them) all give rise to potentially material conflicts of interest. The latter is perhaps the simplest to address, but answers to all three remain tricky - and the subject of a future post.
The report notes that the staff made various recommendations to the NRSROs to address the staff’s concerns and that in some cases the NRSROs have already taken steps to address such concern, but the SEC believes that working with the CRAs without disclosing bad practices is the best way for these issues to be resolved. The report stresses that the "essential findings" set out are those of staff members who conducted the examinations, and do not constitute any expression of the Commission per se, nor do they constitute any "material regulatory deficiency", although they may form the base for such in the future.