2010-32280. Credit Rating Standardization Study

AGENCY:

Securities and Exchange Commission.

ACTION:

Request for comment.

SUMMARY:

The Securities and Exchange Commission is requesting public comment to help inform its study pursuant to Section 939(h) of the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 on the feasibility and desirability of: Standardizing credit ratings terminology, so that all credit rating agencies issue credit ratings using identical terms; standardizing the market stress conditions under which ratings are evaluated; requiring a quantitative correspondence between credit ratings and a range of default probabilities and loss expectations under standardized conditions of economic stress; and standardizing credit rating terminology across asset classes, so that named ratings correspond to a standard range of default probabilities and expected losses independent of asset class and issuing entity.

DATES:

The Commission will accept comments regarding issues related to the study on or before February 7, 2011.

ADDRESSES:

Comments may be submitted by any of the following methods:

Electronic Comments

• Use the Commission's Internet comment form (http://www.sec.gov/​rules/​other.shtml); or
• Send an e-mail to [e-mail]. Please include File Number 4-622 on the subject line.

Paper Comments

• Send paper comments in triplicate to Elizabeth M. Murphy, Secretary, Securities and Exchange Commission, 100 F Street, NE., Washington, DC 20549-1090.

All submissions should refer to File Number 4-622. This file number should be included on the subject line if e-mail is used. To help us process and review your comments more efficiently, please use only one method. The Commission will post all comments on the Commission's Internet Web site (http://www.sec.gov). Comments are also available for Web site viewing and printing in the Commission's Public Reference Room, 100 F Street, NE., Washington, DC 20549, on official business days between the hours of 10 a.m. and 3 p.m. All comments received will be posted without change; we do not edit personal identifying information from submissions. You should submit only information that you wish to make available publicly.

FOR FURTHER INFORMATION CONTACT:

Randall W. Roy, Assistant Director, Division of Trading and Markets, at [phone]; Alan A. Dunetz, Branch Chief, Division of Trading and Markets, at [phone]; Kevin S. Davey, Securities Compliance Examiner, at [phone]; Kristin A. Devitto, Securities Compliance Examiner, at [phone]; Mark M. Attar, Branch Chief, Division of Trading and Markets, at [phone]; or Raymond A. Lombardo, Branch Chief, Division of Trading and Markets, at [phone], Securities and Exchange Commission, 100 F Street, NE., Washington, DC 20549-7010.

Discussion:

On July 21, 2010, President Obama signed the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 (the “Dodd-Frank Act”) into law. Under Section 939(h) of the Dodd-Frank Act, the Securities and Exchange Commission (the “Commission”) is required to study the feasibility and desirability of: (A) Standardizing credit ratings terminology, so that all credit rating agencies issue credit ratings using identical terms; (B) standardizing the market stress conditions under which ratings are evaluated; (C) requiring a quantitative correspondence between credit ratings and a range of default probabilities and loss expectations under standardized conditions of economic stress; and (D) standardizing credit rating terminology across asset classes, so that named ratings correspond to a standard range of default probabilities and expected losses independent of asset class and issuing entity. Not later than one year after the date of enactment of the Dodd-Frank Act, the Commission must submit to Congress a report containing the findings of the study and the recommendations, if any, of the Commission with respect to the study.

Request for Comment:

The Commission believes that submissions by interested parties with a wide range of views, including those of investors who use credit ratings, portfolio managers, credit rating agencies, investment firms, underwriters, issuers, regulators and the academic community, will provide valuable information as it conducts the study required by Section 939(h) of the Dodd-Frank Act. Accordingly, the Commission requests commenters' views on each of the topics to be addressed in the Commission's study under Section 939(h) of the Dodd-Frank Act. In particular, the Commission seeks commenters' views in response to the following questions:

(1) Is it feasible and desirable to standardize credit ratings terminology, so that all credit rating agencies issue credit ratings using identical terms?

a. Do commenters agree that the term “credit ratings terminology” as used in Section 939(h) of the Dodd-Frank Act refers to the symbols and numbers credit rating agencies use to denote credit ratings and the definitions and meanings they promulgate for those symbols and numbers? If not, what other (or additional) credit rating terminology should this study focus on? Commenters who identify other terminologies should indicate for all subsequent questions whether they are discussing the other terminologies or ratings symbols and numbers and their corresponding definitions and meanings.

b. Are there credit rating terminologies used by different credit rating agencies that are currently comparable? If so, please identify and explain how they are comparable.

c. Identify differences in the credit rating terminologies used by credit rating agencies. What is the significance of these differences?

d. What issues do commenters encounter when they seek to compare ratings from different credit rating agencies?

e. Some credit rating agencies employ multiple credit rating scales designed to distinguish between different types of issues and/or issuers. For example, a credit rating agency may employ different credit rating symbols for ratings of long term securities, short term securities, money market funds, claims paying abilities of insurance companies, and issues and/or issuers in different jurisdictions. Do commenters believe that some types of credit rating symbols used by credit rating agencies are more or less suitable to standardization? Is it feasible or desirable to use a single credit rating scale for all types of issues and issuances? Should a standardized credit rating scale include separate symbols for different types of credit ratings? If so, what separate credit symbols should be included in the standardized credit rating terminology? Alternatively, should credit rating terminologies for some types of issues or issuers not be standardized? If so, for which types of issuers or issuances?

f. The credit ratings of some credit rating agencies address probability of default while the ratings of other credit rating agencies address expected loss. Other rating scales may address other metrics such as, for example, distance to distress (e.g., with respect to the public finance ratings of some credit rating agencies). Do commenters believe that it is more or less desirable to have credit ratings of different credit rating agencies address different risks? Why?

g. Some credit rating agencies employ credit rating modifiers including, for example, “credit watch” and “rating outlook” to indicate a view as to the likelihood that a credit rating may change. Do commenters believe that it is feasible or desirable to include such credit rating modifiers in a standardized credit rating terminology? Why?

h. If commenters believe that standardizing credit ratings terminology is desirable and feasible:

i. What level of detail should be included in the standardized credit rating terminology?

ii. What mix of quantitative and qualitative factors should be referenced in each rating definition?

iii. Should a standardized credit rating terminology address likelihood of default, expected loss, or some other metric?

iv. Some credit rating agencies issue a number of broad categories of credit ratings that can be further delineated using identifiers (e.g., pluses and minuses) to allow additional gradations of ratings. How many gradations of credit quality should be included in a standardized terminology for credit ratings?

v. Should a standardized credit rating terminology employ a separate terminology for certain asset classes (e.g., for structured finance ratings)? Are there asset classes or types of ratings, such as short term or financial strength ratings, where a separate terminology should be considered?

vi. What organizations or combination of organizations should be responsible for developing and administering the standardized credit rating terminology? For example, should the Commission develop and administer the standardized terminology? Should an independent board or organization be formed to develop and administer the standardized terminology?

vii. What time period should be allowed for credit rating agencies to map their existing ratings to a new credit rating terminology, or for private contracts and investment management agreements that reference credit ratings to be changed to refer to the standardized terminology?

viii. Do commenters believe that it would be more desirable for credit rating agencies to retain their existing credit rating terminologies and make publicly available detailed information on how each credit rating agency's ratings can be mapped to a standardized terminology? Or would it be more desirable if the credit rating agency used only the standardized terminology?

(2) Is it feasible and desirable to standardize the market stress conditions under which credit ratings are evaluated?

a. Under what market stress conditions are credit ratings currently evaluated?

b. To what degree do commenters believe that credit rating agencies currently identify the market stress conditions under which credit ratings are evaluated? To the extent these market stress conditions are identified by credit rating agencies, do commenters believe that the market stress conditions used by different credit rating agencies at comparable credit rating levels are similar? If so, how are they similar? If not, how do they differ?

c. Do commenters believe that market stress conditions can be defined in a consistent manner across different industry sectors and geographic regions?

d. Do commenters believe that standardized market stress conditions are equally relevant to the evaluation of all asset classes or issuers? For example, are there some asset classes or issuers where the relative degree of idiosyncratic risk versus systemic risk differs? If so, are market stress conditions less relevant, for example, to asset classes and issuers where there is a higher level of idiosyncratic risk?

e. If commenters believe that it is feasible and desirable to standardize the market stress conditions under which credit ratings are evaluated:

i. What parameters should be defined in these market stress conditions? For example, unemployment rates, declines in GDP and financial market declines are widely referenced indicators of market stress. What other parameters do commenters believe should be defined?

ii. How should market stress conditions differ across different industry sectors and geographic regions?

iii. Should these stress conditions reference specific historical market stresses such as, for example, the Great Depression or the 2008 financial crisis?

iv. Should each credit rating level have its own specifically defined stress conditions?

(3) Is it feasible and desirable to require a quantitative correspondence between credit ratings and a range of default probabilities and loss expectations under standardized conditions of economic stress?

a. To what extent do credit rating agencies or others assign a quantitative correspondence between credit ratings and a range of default probabilities and loss expectations?

i. To what extent do commenters believe that the correspondence is similar for comparable ratings from different credit rating agencies?

ii. To what extent do commenters believe that the correspondence is similar across industry sectors and geographical regions?

iii. To what extent do commenters believe that the correspondence is constant throughout the economic cycle?

iv. To what extent do commenters believe that the correspondence has been constant over time? For example, do commenters believe that the range of default probabilities and loss expectations corresponding to the credit ratings of different credit rating agencies have become more or less conservative over time?

b. Does the ability to assign a correspondence between credit ratings and a range of default probabilities and loss expectations in a sector vary depending on the degree to which a rating methodology for that sector is more or less quantitative in nature? Are there other factors, such as the quality or amount of historical performance data or structural complexity that may make it more or less difficult to assign a correspondence between credit ratings and a range of default probabilities and loss expectations?

c. Does the likelihood of rating transitions for similarly rated assets vary among asset classes? If so, how should variation in the likelihood of rating transitions be addressed when a quantitative correspondence is assigned between credit ratings and a range of default probabilities and loss expectations?

d. Is there a role for market-based measures such as credit spreads or option-based approaches (i.e., Merton-type models which provide a distance to default measure based on equity prices) in determining a correspondence between credit ratings and a range of default probabilities and loss expectations?

e. If commenters believe that requiring a quantitative correspondence between credit ratings and a range of default probabilities and loss expectations under standardized conditions of economic stress is feasible and desirable:

i. What factors should be considered in determining the range of default probabilities and loss expectations associated with each rating? Should specific time horizons be specified for each default probability and loss expectation range? If so, how many different time horizons should be specified for each credit rating, and what are appropriate time horizons?

ii. The ratings of some credit rating agencies primarily address probability of default while others address expected loss. Should credit rating agencies be allowed to choose whether their ratings address one or the other? Should a single rating address both probability of default and loss expectation or should default probabilities and loss severity be addressed separately?

iii. What are the views of commenters on how the accuracy of the quantitative correspondence assigned by a given credit rating agency between its credit ratings and a range of default probabilities and loss expectations should be measured?

(4) Is it feasible and desirable to standardize credit rating terminology across asset classes, so that named credit ratings correspond to a standard range of default probabilities and expected losses independent of asset class and issuing entity?

a. To what degree do commenters believe that credit ratings are currently comparable across asset classes? For example, do commenters believe that credit ratings of structured finance products or municipal securities are comparable to credit ratings in other sectors?

b. In cases where credit rating agencies currently use the same credit rating terminology for multiple asset classes, what is the view of commenters on the adequacy and transparency of the procedures credit rating agencies use to achieve comparability?

c. What mix of quantitative and qualitative factors should be considered when standardizing credit rating terminology across asset classes, so that named credit ratings correspond to a standard range of default probabilities and expected losses?

i. To what degree should standardization be based on quantitative factors such as, for example, historical performance metrics including rating transition and default studies? What other quantitative factors should be considered?

ii. To what degree should standardization be based on qualitative factors such as, for example, analyst judgment regarding the comparability of credits from different sectors? What other qualitative factors should be considered?

d. Are there asset classes where the risk characteristics of the asset class, limitations on the quality of data, structural complexity, limitations on historical performance data, or other factors make it more difficult to apply to that asset class a standardized credit rating terminology which applies to other asset classes and issuers so that named ratings correspond to a standard range of default probabilities and expected losses?

All interested parties are invited to submit their views, in writing, on these questions.

By the Commission.

Dated: December 17, 2010.

Elizabeth M. Murphy,

Secretary.