Our View on Proposed Regulation of Credit Rating Agencies

By Deven Sharma†

As the U.S. Congress considers the Accountability and Transparency in Rating Agencies Act, Standard and Poor’s reiterates its commitment to working with the House and Senate to fashion a final bill that will do its part to reinvigorate the economy, create jobs, and give investors the tools they need to make wise decisions. By providing independent credit benchmarks, S&P helps create transparency -- one important contributor to the development of the credit markets.

S&P supports many of the ideas under deliberation and shares the view that through greater disclosure, oversight, and accountability, confidence in ratings can be restored. Indeed, we are making progress on all these fronts in the absence of new legislation.

Some proposals, however, are problematic and would bring unintended harm to the markets and to future economic growth.

Before discussing these proposals further, I would like briefly to address S&P’s credit rating history. We are deeply disappointed with the performance of many of our ratings on mortgage-backed instruments. Although we always expect that some portion of the debt we rate, even highly-rated debt, will ultimately default, our ratings of mortgage-backed securities issued between 2005 and 2007 have been unusually unstable and their performance has not matched our historical track record.

Why did these ratings on mortgage-backed securities perform poorly?

Put simply, our assumptions about the housing and mortgage markets in the second half of this decade did not account for the extraordinarily steep declines we have now seen. Although we did assume, based on historical data stretching back to the Great Depression, that these markets would decline to some degree, we and virtually every other market participant and observer did not expect the unprecedented events that occurred.

It should go without saying that had we anticipated fully the speed and scope of the declines in these markets at the time we issued our original ratings, many of those ratings would have been different.

It is important to note, however, that overall S&P’s ratings, including in the area of structured finance securities, have historically performed very well. We have been rating structured securities for over 30 years and have developed industry-leading processes and models for evaluating the creditworthiness of a wide array of structured transactions. Since 1978, only 1.15% of structured finance securities rated by us “AAA” have ever defaulted. For residential mortgage-backed securities (RMBS), the percentage of “AAA” ratings to default over this time period is 0.77%. Our ratings of non-mortgage asset-backed securities (“ABS”) have performed notably well, even during the recent credit crisis. As of July 2009, of the more than 6,000 “AAA” ratings we issued on non-mortgage ABS, approximately 90% have remained “AAA” and only 1% have been downgraded below investment grade. Only four “AAA” transactions in this group — or approximately 0.07% — have ever transitioned to default. Our ratings on corporate and municipal debt have also performed extremely well, reflecting a high correlation between rating levels and defaults.

Our ratings on the mortgage-backed securities were our best informed views expressed based on the best information available about these complex instruments. More importantly, we have learned from this experience and we have made major changes ourselves to restore confidence in our ratings.

S&P’s Commitment to Change

In 2008, we announced a series of initiatives aimed at improving checks and balances in our organization and promoting four broad objectives:

• ensuring the integrity of the ratings process;
• enhancing analytical quality;
• providing greater transparency to the market by disseminating relevant information about ratings, as well as information to help the market form its own view of the soundness of rating analysis; and
• more effectively educating the marketplace about ratings.

To date, we have made significant implementation progress. For example, we have:

• Established distinct groups and invested significantly in processes around our Criteria, Quality, Compliance, and Internal Audit functions. Each of these has a distinct and important role in the review of the ratings process and related controls;
• Established an Office of the Ombudsman. The Ombudsman addresses concerns related to potential conflicts of interest and analytical and governance processes that are raised by issuers, investors, employees, and other market participants across S&P’s businesses;
• Instituted a rotation system for analysts;
• Established an enterprise-wide independent Risk Assessment Oversight Committee. The Committee assesses all risks that could impact the integrity and quality of the ratings process;
• Enhanced our analyst training programs and introduced a new Analyst Certification Program in which all S&P rating analysts are required to participate;
• Enhanced and expanded our quality assurance and controls related to the development and implementation of criteria; and
• Created a separate Model Validation Group to independently analyze and validate all models, developed by S&P or provided by issuers, used in the ratings process.

In addition, we have taken steps to raise the disclosure levels on our analyses, strengthen the analytics, and educate the market.

S&P has gone to great lengths to make serious and meaningful changes in the way we go about compiling our credit ratings and we believe the market is taking notice. We believe firmly that our ratings will continue to be an important part of the information available to investors and other market participants as we move forward and the financial markets continue to improve.

An Example of Sound Regulation

Another way to restore investor confidence in ratings --- in addition to the sweeping changes I just described --- is to pursue effective regulation of credit rating agencies. Currently, Nationally Recognized Statistical Rating Organizations (NRSROS) are subject to the Credit Rating Agency Reform Act of 2006, which gave the SEC broad oversight authority and was intended to increase transparency about the ratings process and investor choice and competition among NRSROs.

As a result, there are now ten NRSROs competing for business, double what it was at the time the law was passed. As long as barriers to entry remain low, the number of NRSROs will continue to grow, providing still more options for investors.

Also, as a result of the 2006 Act, NRSROs are now required to disclose detailed performance data about their ratings.

The SEC also issued another layer of rules governing NRSROs in February 2009. Those actions by the SEC require enhanced disclosures of performance measurement statistics and procedures and methodologies; new record-keeping standards; public disclosure of ratings histories; and a comprehensive annual report from each NRSRO.

Looking Forward: A Broad Regulatory Roadmap

Although the 2006 Act has resulted in a broad and robust regulatory scheme, S&P shares the view that further regulation, appropriately crafted, can serve the goal of restoring and maintaining investor confidence.

Such regulation should follow four broad principles:

• it should be part of “beginning-to-end” regulatory solution;
• it should not interfere with analytical independence;
• it should foster greater competition in the ratings industry and not impose burdensome barriers to entry; and
• it should recognize the critical importance of international consistency.

Beginning-to-End Solutions : Any new regulatory approach regarding ratings should include “beginning-to-end” solutions, covering all aspects of the capital markets that, taken together, contribute in a systemic way to their effective and efficient functioning. In structured finance, this would include not just ratings, but appropriate regulation related to the origination and pooling of assets, the structuring and underwriting of securities, the management of collateral held by a structured vehicle, and the marketing of securities.

Analytical Independence : We believe analytical independence is a fundamental principle. At its core, a rating is an analytical determination. It results from a group of experienced professionals analyzing a set of facts and forming a judgment as to what might happen in the future. For the markets to have confidence in those ratings, they must continue to be made independently. Accordingly, we would be extremely concerned about regulatory measures that could force analysts to make judgments not based upon their own considered analysis and independent views and experience, but rather out of a desire to avoid subsequent second-guessing by regulators or others.

Fostering Competition in the Ratings Industry : A key aspect of the 2006 Act is a set of provisions designed to ease the burden of becoming an NRSRO. S&P strongly supports that legislative goal. Regulatory requirements, by their nature, are often seen by potential new entrants as burdensome; yet carefully crafted, balanced rules are necessary to establish a fair and level playing field.

International Consistency : We also believe international consistency is critical to an appropriate regulatory framework. Ratings are issued and used globally. This reflects one of their many benefits — their ability to provide a common global language for analyzing credit risk and contribute to the global flow of capital. We believe the model code of conduct published by the International Organization of Securities Commissions (IOSCO) as a possible blueprint in that regard is constructive.

Specific Proposals That Would Improve Ratings and Benefit the Markets

There have been several recent proposals for additional legislative and regulatory action which, on top of the new and far-reaching rules issued by the SEC, would provide for extensive new regulation designed to increase NRSROs’ accountability. S&P agrees with, and is prepared to support, many of the recent proposals to strengthen regulation of NRSROs by:

• increasing SEC oversight;
• protecting against conflicts of interest, which inevitably exist in connection with any credit rating business model;
• promoting the use of high quality data by NRSROs; and
• improving transparency in the ratings process.
• These recent proposals that we broadly support include, specifically:

Oversight and Accountability

• Creating a dedicated office within the SEC to oversee NRSROs and empowering the SEC to conduct frequent reviews of NRSROs to ensure that NRSROs follow their internal controls and policies for determining ratings and managing conflicts of interest;
• Providing the SEC with explicit authority to impose sanctions, including steep fines for NRSROs that fail to comply with the SEC’s rules or their own policies and procedures, and to conduct regular, annual reviews of NRSROs;
• Designating a responsible officer (or officers) within each NRSRO with authority to review compliance with procedures and methodologies, administer rating policies, and oversee compliance with the securities laws and regulations;
• Requiring NRSROs to adopt payment practices that are based in part on ratings performance and ongoing surveillance, which will help promote comprehensive and thorough analysis; and
• Mandating comprehensive analyst training programs.

Conflicts of Interest

• Requiring disclosure of a uniform body of underlying information by issuers, according to requirements appropriate to every class of securities to all NRSROs. This would lead to increased transparency and promote the issuance of unsolicited ratings, thus improving competition in the rating industry;
• Requiring analyst rotations so that analysts do not consistently analyze the same issuer or security, further reducing the potential for conflicts of interest;
• Requiring “look-back“ reviews whenever former NRSRO employees leave to work for an issuer, thus providing a further check on the integrity of the rating process;
• Requiring that fees for structured transactions are paid at defined milestones in the process, rather than when the transaction closes, which would help avoid potential conflicts of interest and protect against so-called “ratings shopping”; and
• Requiring issuers to make disclosures when they seek preliminary ratings from NRSROs, which will further protect against ratings shopping.

Data Quality

• Requiring disclosures about the use of any third party due diligence services; and
• Requiring NRSROs to receive due diligence certifications from independent, third party firms in connection with credit ratings on structured finance securities backed by residential mortgages.

Transparency

• Requiring frequent dissemination of robust and comparable data about the historical performance of ratings, including default and transition studies, which will permit investors and other market participants to compare NRSROs and improve competition in the industry;
• Requiring broad dissemination of ratings methodologies, analytical assumptions, and practices related to data quality;
• Requiring disclosures about the use of “servicer reports“ in connection with structured finance transactions; and
• Improving documentation and broadening public disclosure of internal controls, including conflicts of interest policies and codes of conduct. On this last category, I will add that while we believe strongly in promoting transparency and a well-informed market, the key is to promote relevant transparency; we should seek to avoid the risk of overloading the market with too much information that has no genuine use.

Proposals That Raise Concerns For NRSROs and the Markets Generally

There are several proposals that raise serious concerns, not only for the effect the proposals would have on S&P and other NRSROs, but also for the broader implications they would have for the United States and global financial markets as a whole.

Amendments to Pleading Standards

One proposal that has been discussed in recent months could have the effect — at least as interpreted by aggressive plaintiffs‘ lawyers not always seeking to advance the broader interests of the market — of lowering the threshold legal requirements for bringing a securities fraud claim against NRSROs and only NRSROs.

The courts have indeed affirmed that credit ratings are opinions that are matters of public concern protected by the First Amendment. However, the First Amendment provides no exemption from liability to any company, including a rating agency that intentionally misleads or defrauds investors. Indeed, the First Amendment provides no protection in a securities fraud action under the Securities Exchange Act — the very statute that this proposal would seek to amend.

If the proposal under consideration were to become law, plaintiffs‘ lawyers would use it to argue that NRSROs may be sued for securities fraud under the Securities Exchange Act whenever they act “unreasonably.” This would differ materially from the legal standard applicable to every other defendant — including auditors, equity analysts, issuers, underwriters, and others — who must be found to have acted intentionally with bad faith (or in legal terms with “scienter“) before they can be found liable. Such a distinction is inappropriate and unfair. We are not suggesting that NRSROs should receive special treatment in such cases, but rather that they should be subject to the same pleading standards as other defendants.

It is important to understand that the very nature of credit ratings makes NRSROs uniquely susceptible to potential harm from the creation of a new lower liability standard. Credit ratings are not statements of existing fact but rather opinions about the future.

If S&P or other NRSROs could potentially be liable under the securities laws even where they act in good faith, plaintiffs‘lawyers would inevitably file suit against them any time rated securities default, or even when ratings are simply downgraded. There would be limitless opportunities for such second-guessing because, at any moment, S&P has well over 1 million ratings outstanding and rates more than $32 trillion of debt.

I am not suggesting that S&P receive special legal treatment in lawsuits for securities fraud under the Exchange Act. Rather, I am raising probable, unintended consequences not just for NRSROs but, more importantly, for issuers who must access the debt markets and for investors who want and need additional information about the debt. For example, increased liability standards would discourage new entrants into the ratings industry, reducing competition and limiting the diversity of opinions available to investors. In addition, increased exposure to lawsuits could make it more difficult for new businesses to access the capital markets. If faced with a dramatic increase in suits, rating agencies would likely rate only those entities that are least likely to default or be downgraded or that have a long history. As a result, issuers who present greater credit risk, such as newer businesses – for example, green or technology companies that represent emerging drivers of economic growth – may have a hard time getting rated and therefore, greater difficulty accessing capital.

I am simply saying that we should be subject to the same legal standards as everyone else in such cases. The discussion draft suggests the possibility of new language explicitly stating that Congress would not be amending existing pleading requirements under the federal securities laws, and that NRSROs would be subject to the same legal standards that apply to issuers and underwriters. While this provision would certainly help to alleviate the significant concerns I have raised, the discussion draft still contains much of the problematic language I have discussed and, therefore, should be made clearer in order to avoid any possibility of these very serious potential harms.

Collective Liability

Another proposal that has given us great cause for concern is one that would subject each NRSRO to “collective liability“ for judgments against any NRSRO. Thus, if an analyst at one NRSRO commits securities fraud, S&P could be required under some circumstances to compensate the victim of the fraud even if S&P had nothing whatsoever to do with the other NRSRO’s fraudulent activity.

No NRSRO should be required to act as an insurer and compensate plaintiffs for harm caused by employees of its competitors.

This proposal is particularly alarming because it contrasts so sharply with the primary goal of the 2006 Act — to increase competition in the credit rating industry and lower barriers to becoming an NRSRO. Indeed, it is difficult to imagine a greater deterrent to entering this industry than the knowledge that one may be required to act as an insurer and held financially responsible for the fraudulent actions of and harm caused by its competitors and its competitors‘employees.

Sharing and Verification of Information

One other troubling provision being considered would mandate that NRSROs disclose to one another all information they receive from issuers. On its face, the proposal would also require NRSROs to disclose— to competitors no less — all of their own internal work papers, meeting notes, and any other analytical support for their rating opinions.

This would constitute an unprecedented intrusion into competitive businesses and fundamentally subvert intellectual property rights in a manner that would undoubtedly chill robust analysis by NRSROs and otherwise restrict development and innovation in the ratings industry. In addition to such sweeping disclosures, the proposal would also require NRSROs to review and “verify“ any information similarly disclosed to them by their competitors. This would constitute a seemingly insurmountable burden, considering that S&P alone has well over 1 million ratings outstanding right now and currently rates more than $32 trillion of debt.

The proposal raises a number of serious questions: How would an NRSRO go about “verifying“ all of this information, which presumably would include financial statements, management projections, and the like? Would NRSROs be required to verify such information even if they did not rate the relevant issuer or security? Would NRSROs be expected to establish entire departments dedicated to verifying the reams of information flowing in from their competitors?

Rather than NRSROs, these roles are properly held by those involved in the securitization process who are responsible for the review and verification of such information. Whatever the answers to these and other questions raised by this provision, it is clear that this concept is unprecedented and would inevitably create extraordinary barriers to entry, particularly for start-up firms who would not conceivably be equipped to take on the massive burden of verifying information on virtually all of the major securities issuances in the U.S., as well other issuances abroad.

Potential Interference with Analytical Independence

A core provision in the 2006 Act prohibits the SEC from regulating “the substance of credit ratings or the procedures and methodologies by which any [NRSRO] determines credit ratings.” This important limit on SEC authority was designed to protect NRSROs‘ analytical independence and to ensure that they are able to issue rating opinions free from government-mandated analytics, which would stifle innovation, lead to less robust ratings, and have the appearance to investors that the government is somehow “sanctioning“ or “endorsing“ a particular rating or NRSRO. Some recent proposals, however, cross that threshold and suggest a substantive role for regulators in determining how an NRSRO forms its rating opinions. One example is a provision in the discussion draft that could be read to strictly define the meaning of a credit rating, limiting it to an assessment of risk that investors “may not receive payment in accordance with the terms of issuance“ — in other words, the likelihood of default.

While likelihood of default is the single most important factor in our analysis of creditworthiness, investors have repeatedly cited other factors as useful to them in their investment decisions. Our analysis therefore addresses secondary factors, including the payment priority of an obligation following default. For example, when a corporation issues both senior and subordinated debt, we typically assign a lower rating to the subordinate debt. We also consider the projected recovery that an investor would expect to receive if an obligation defaults.

Another factor is credit stability, which accounts for the expectation that some types of issuers and obligations are prone to displaying a period of gradual decay before they default, while others may be more vulnerable to sudden deterioration or default. Accordingly, in 2008, we introduced an explicit “stability“ measure into our ratings criteria, which addresses whether, in our view, an issuer or security has a high likelihood of experiencing unusually large adverse changes in credit quality under conditions of moderate stress.

Any government mandate that arguably would prohibit an NRSRO from considering any secondary credit factors that are relevant to our opinion of creditworthiness would lead not only to homogenized ratings, thus depriving investors of the full breadth and diversity of NRSROs‘ opinions, but could also result in undue investor reliance on rating opinions and a misperception that Congress or the SEC has endorsed NRSROs‘ methodologies and their ultimate rating opinions — again, precisely the opposite of Congress‘ goal in pursuing proposals to remove the NRSRO designation from existing laws and rules.

Removing Protections for Forward-Looking Statements

S&P also has serious concerns about proposals providing that rating opinions shall not be deemed “forward-looking statements” under the federal securities laws. These proposals ignore that the very essence of a rating is that it is forward looking — it speaks to the likelihood that a particular obligor will pay back principal and interest in the future. Unlike statements that speak to an entity’s current financial condition, ratings expressly relate to what may likely happen on a going-forward basis. If ratings were not forward-looking, but instead simply reported on existing facts about an issuer or security, they would serve very little, if any, purpose to the markets.

Corporate Governance Restrictions

Another proposal would regulate and restrict the corporate governance practices at NRSROs or their parent companies. Among other things, the proposal would dictate the composition of boards of directors, restrict the activities of board members and control their compensation structure and length of term. In addition, board members would be required to oversee analytical processes, including the development of ratings criteria and methodologies — subjects that would ordinarily fall far outside the knowledge and experience of corporate board members, who are not trained analytical staff or management of NRSROs.

Such interference with NRSROs‘ corporate governance structure would treat NRSROs differently and much more harshly than all other market participants, including, for example, auditors and equity analysts. Members of Congress and critics of NRSROs have frequently observed that NRSROs should be treated “the same as“ auditors and equity analysts; yet, the effect of this proposal represents the opposite approach.

Prohibitions on the Activities of NRSRO Affiliates

Finally, another provision under discussion would impose restrictions on the affiliates of NRSROs, prohibiting them from engaging in an array of services unrelated to any credit rating service. S&P has further strengthened its policies in this area since 2007 and strongly supports the goal of protecting against potential conflicts of interest. These new proposals, however, would sweep too broadly and would increase barriers to entry. By way of example, S&P is owned by The McGraw-Hill Companies, Inc., which itself operates a number of entities and business units that have nothing to do with — and are segregated both structurally and substantively from — S&P’s credit rating activities. McGraw- Hill’s education and information and media segments, for example, have nothing to do with S&P’s ratings, yet would be covered by this sweeping prohibition. Imposing restrictions on the ability of units like these to do business with issuers on matters totally unrelated to ratings would be unfair and could well prevent new businesses that currently offer similar services from considering entrance into the ratings business as an NRSRO.

This problem could be solved rather easily by creating a safe- harbor provision for those NRSROs, such as S&P, that maintain policies and procedures that establish firewalls to insulate ratings-related activities from other business activities under the same corporate umbrella.

Although S&P agrees completely with the goal of improving the quality and transparency of credit rating analysis, we urge caution in the crafting of proposals that would ultimately result in less comprehensive ratings, covering a narrower scope of world markets to the detriment of investors and business enterprises large and small.

Any regulatory scheme that effectively scales back the availability of robust, independent rating opinions will result, inevitably, in a reduction to the flow of capital in global markets, stalling innovation and growth in emerging sectors and beyond.

Throughout our history, S&P’s consistent approach has been to learn from experience and to improve and strengthen our analytics, criteria, and review processes when appropriate. We will continue this approach going forward, working with Congress and governments, legislatures and policymakers worldwide as they strive to develop productive solutions that restore stability in the global capital markets.