# Proposed Supervisory Guidance for Internal Ratings-Based Systems for Credit Risk, Advanced Measurement Approaches for Operational Risk, and the Supervisory Review Process (Pillar 2) Related to Basel II Implementation > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/fr%3A07-811 ## Record - **Collection:** Federal Register - **Document type:** Notice - **Published:** February 28, 2007 - **Citation:** 72 FR 9084 ## Text DEPARTMENT OF THE TREASURY Office of the Comptroller of the Currency [Docket No. OCC-2007-0004] FEDERAL RESERVE SYSTEM [Docket No. OP-1277] FEDERAL DEPOSIT INSURANCE CORPORATION DEPARTMENT OF THE TREASURY Office of Thrift Supervision [No. 2007-06] Proposed Supervisory Guidance for Internal Ratings-Based Systems for Credit Risk, Advanced Measurement Approaches for Operational Risk, and the Supervisory Review Process (Pillar 2) Related to Basel II Implementation AGENCIES: Office of the Comptroller of the Currency, Treasury (OCC); Board of Governors of the Federal Reserve System (Board); Federal Deposit Insurance Corporation (FDIC); and Office of Thrift Supervision, Treasury (OTS) (collectively, the Agencies). ACTION: Proposed supervisory guidance with request for public comment. SUMMARY: The Agencies are publishing for comment three documents that set forth proposed supervisory guidance for implementing proposed revisions to the risk-based capital standards in the United States (New Advanced Capital Adequacy Framework or proposed framework). These proposed revisions, which would implement the “International Convergence of Capital Measurement and Capital Standards: A Revised Framework,” published in June 2004 by the Basel Committee on Banking Supervision (Basel II), in the United States, were published in the Federal Register on September 25, 2006 as a notice of proposed rulemaking (NPR or proposed rule). The proposed framework outlined in the NPR would require some and permit other qualifying banks to calculate their regulatory risk-based capital requirements using an internal ratings-based (IRB) approach for credit risk and the advanced measurement approaches (AMA) for operational risk (together, the advanced approaches); it also provides guidelines for the supervisory review process (Pillar 2). The proposed supervisory guidance documents provide additional detail for the advanced approaches and the supervisory review process that should help banks satisfy the qualification requirements in the NPR. DATES: Comments on the three proposed supervisory guidance documents must be submitted on or before May 29, 2007. ADDRESSES: OCC: You must include OCC and Docket Number OCC-2007-0004 in your comment. You may submit comments by any of the following methods: • Agency Web site: http://www.occ.treas.gov. Click on “Contact the OCC,” scroll down and click on “Comments on Proposed Regulations.” • E-mail address: regs.comments@occ.treas.gov. • Fax: (202) 874-4448. • Mail: Office of the Comptroller of the Currency, 250 E Street, SW., Mail Stop 1-5, Washington, DC 20219. • Hand Delivery/Courier: 250 E Street, SW., Attn: Public Information Room, Maila Stop 1-5, Washington, DC 20219. Instructions: All submissions received must include the agency name (OCC) and docket number for this proposed notice. In general, OCC will enter all comments received into the docket without change, including any business or personal information that you provide. You may review comments and other related materials by any of the following methods: • Viewing Comments Personally: You may personally inspect and photocopy comments at the OCC's Public Information Room, 250 E Street, SW., Washington, DC. You can make an appointment to inspect comments by calling (202) 874-5043. • Viewing Comments Electronically: You may request e-mail or CD-ROM copies of comments that the OCC has received by contacting the OCC's Public Information Room at: regs.comments@occ.treas.gov. • Docket: You may also request available background documents and project summaries using the methods described above. Board: You may submit comments, identified by Docket No. OP-1277, by any of the following methods: • Agency Web site: http://www.federalreserve.gov. Follow the instructions for submitting comments at http://www.federalreserve.gov/generalinfo/foia/ProposedRegs.cfm. • Federal eRulemaking Portal: http://www.regulations.gov. Follow the instructions for submitting comments. • E-mail: regs.comments@ federalreserve.gov. Include the docket number in the subject line of the message. • Fax: (202) 452-3819 or (202) 452-3102. • Mail: Jennifer J. Johnson, Secretary, Board of Governors of the Federal Reserve System, 20th Street and Constitution Avenue, NW., Washington, DC 20551. All public comments are available from the Board's Web site at http://www.federalreserve.gov/generalinfo/foia/ProposedRegs.cfm as submitted, unless modified for technical reasons. Accordingly, your comments will not be edited to remove any identifying or contact information. Public comments also may be viewed electronically or in paper form in Room MP-500 of the Board's Martin Building (20th and C Streets, NW.) between 9 a.m. and 5 p.m. on weekdays. FDIC: You may submit comments by any of the following methods: • Agency Web Site: http://www.fdic.gov/regulations/laws/federal. Follow instructions for submitting comments on the Agency Web Site. • E-mail: Comments@FDIC.gov. Include “Basel II Supervisory Guidance” in the subject line of the message. • Mail: Robert E. Feldman, Executive Secretary, Attention: Comments, Federal Deposit Insurance Corporation, 550 17th Street, NW., Washington, DC 20429. • Hand Delivery/Courier: Guard station at the rear of the 550 17th Street Building (located on F Street) on business days between 7 a.m. and 5 p.m. (EST). • Federal eRulemaking Portal: http://www.regulations.gov. Follow the instructions for submitting comments. Public Inspection: All comments received will be posted without change to http://www.fdic.gov/regulations/laws/federal including any personal information provided. Comments may be inspected and photocopied in the FDIC Public Information Center, 3501 North Fairfax Drive, Room E-1002, Arlington, VA 22226, between 9 a.m. and 5 p.m. (EST) on business days. Paper copies of public comments may be ordered from the Public Information Center by telephone at (877) 275-3342 or (703) 562-2200. OTS: You may submit comments, identified by No. 2007-06 by any of the following methods: • Federal eRulemaking Portal: http://www.regulations.gov. Follow the instructions for submitting comments. • E-mail: regs.comments@ ots.treas.gov. Please include No. 2007-06 in the subject line of the message, and include your name and telephone number in the message. • Fax: (202) 906-6518. • Mail: Regulation Comments, Chief Counsel's Office, Office of Thrift Supervision, 1700 G Street, NW., Washington, DC 20552, Attention: No. 2007-06. • Hand Delivery/Courier: Guard's Desk, East Lobby Entrance, 1700 G Street, NW., from 9 a.m. to 4 p.m. on business days, Attention: Regulation Comments, Chief Counsel's Office, Attention: No. 2007-06. Instructions: All submissions received must include the agency name and document number. All comments received will be posted without change to http://www.ots.treas.gov/pagehtml.cfm?catNumber=67&an=1, including any personal information provided. Docket: For access to the docket to read background documents or comments received, go to http://www.ots.treas.gov/pagehtml.cfm?catNumber=67&an=1. In addition, you may inspect comments at the Public Reading Room, 1700 G Street, NW., by appointment. To make an appointment for access, call (202) 906-5922, send an e-mail to public.info@ots.treas.gov, or send a facsimile transmission to (202) 906-7755. (Prior notice identifying the materials you will be requesting will assist us in serving you.) We schedule appointments on business days between 10 a.m. and 4 p.m. In most cases, appointments will be available the next business day following the date we receive a request. FOR FURTHER INFORMATION CONTACT: OCC: IRB guidance: Fred Finke, Senior Basel Policy Liaison (202-874-4468 or fred.finke@occ.treas.gov ); AMA guidance: Mark O'Dell, Deputy Comptroller for Operational Risk (202-874-4316 or mark.odell@occ.treas.gov ); or guidance on supervisory review: Akhtarur Siddique, Lead Expert (202-874-4665 or akhtarur.siddique@occ.treas.gov ); Office of the Comptroller of the Currency, 250 E Street, SW., Washington, DC 20219. Board: IRB guidance: Sabeth Siddique, Assistant Director, Credit Risk Section (202-452-3861); AMA guidance: Stacy Coleman, Assistant Director, Operational Risk Section (202-452-2934) or Connie Horsley, Senior Supervisory Financial Analyst, Operational Risk Section (202-452-5239); or guidance on supervisory review: David Palmer, Senior Supervisory Financial Analyst, Credit Risk Section (202-452-2904); Board of Governors of the Federal Reserve System, 20th Street and Constitution Avenue, NW., Washington, DC 20551. Users of Telecommunication Device for Deaf (TTD) only, call (202) 263-4869. FDIC: IRB guidance: Pete Hirsch, Chief, Large Bank Supervision (202-898-6751 or phirsch@fdic.gov ), Curtis Wong, Senior Examination Specialist, Planning and Program Development Section (202-898-7327 or cwong@fdic.gov ); AMA guidance: Mark S. Schmidt, Regional Director (678-916-2189 or maschmidt@fdic.gov ), Alfred Seivold, Senior Examination Specialist, Large Bank Supervision (415-808-8248 or aseivold@fdic.gov ); or guidance on supervisory review: Bobby Bean, Chief, Capital Markets Policy Section (202-898-3575 or bbean@fdic.gov ), Gloria Ikosi, Senior Quantitative Risk Analyst, Capital Markets Policy Section (202-898-3997 or gikosi@fdic.gov ); Federal Deposit Insurance Corporation, 550 17th Street, NW., Washington, DC 20429. OTS: IRB guidance: David Tate, Manager, Examination Quality Review (202-906-5717); AMA guidance: Eric Hirschhorn, Senior Financial Economist, Credit Policy (202-906-7350); or guidance on supervisory review: Sonja White, Senior Project Manager, Capital Policy (202-906-7857); Office of Thrift Supervision, 1700 G Street, NW., Washington, DC 20552. SUPPLEMENTARY INFORMATION: The Agencies issued an NPR on September 25, 2006, 1 which seeks comment on the New Advanced Capital Adequacy Framework that revises the existing general risk-based capital standards as applied to large, internationally active U.S. banks. 2 The public comment period on the NPR closes on March 26, 2007. 3 The proposed framework would implement Basel II in the United States. 1 See 71 FR 55830 (Sept. 25, 2006). 2 For simplicity, and unless otherwise noted, the term “banks” is used here to refer to banks, savings associations, and bank holding companies. The terms “bank holding company” and “BHC” refer only to bank holding companies regulated by the Board and do not include savings and loan holding companies regulated by the OTS. For a detailed description of the institutions covered by this notice, refer to part I, section 1, of the NPR. 3 See 71 FR 77518 (Dec. 26, 2006). As described in the NPR, Basel II sets forth a three-pillar framework encompassing regulatory risk-based capital requirements (Pillar 1); supervisory review of capital adequacy (Pillar 2); and market discipline through enhanced public disclosures (Pillar 3). The proposed framework outlined in the NPR for Pillar 1 would require some and permit other qualifying banks to calculate their regulatory risk-based capital requirements using the IRB approach for credit risk and the AMA for operational risk. 4 The NPR also requires a process for the supervisory review of capital adequacy under Pillar 2, and outlines requirements for enhanced public disclosures under Pillar 3. 5 The NPR describes the qualification process and provides qualification requirements for obtaining supervisory approval for use of the advanced approaches. 6 The qualification requirements are written broadly to accommodate the many ways a bank may design and implement robust credit and operational risk measurement and management systems, and to permit industry practice to evolve. 4 While Basel II provides several approaches for calculating regulatory risk-based capital requirements under Pillara1, only the advanced approaches are proposed for implementation in the United States. 5 Supervisory expectations pertaining to a bank's public disclosures are not part of this notice. 6 See part III, section 22 of the NPR. The proposed supervisory guidance documents are companion guidance to the September 2006 NPR and, as such, are designed to be consistent with the proposed rule and do not address any public comments since the NPR was issued. They provide additional detail that should help banks satisfy the qualification requirements in the NPR. However, the publication of these guidance documents for comment does not imply that the outcome of the NPR has already been determined. As part of the regulatory rulemaking process, the proposed guidance documents are subject to change as needed based on, among other things, the public comments on the guidance and the Agencies' decisions regarding any final rule. The Agencies believe that the proposed supervisory guidance documents are necessary to supplement the proposed framework with standards to promote safety and soundness and encourage comparability across banks. A bank's primary Federal supervisor will review the bank's framework relative to the qualification requirements in the NPR to determine whether the bank may apply the advanced approaches and has complied with the proposed rule in determining its regulatory capital requirements. In August 2003, the Agencies issued an advance notice of proposed rulemaking (ANPR), which described the proposed revisions to the existing risk-based capital framework in general terms and sought public comment. 7 The content of the ANPR was based, in large part, on the April 2003 version of the Basel II framework. 8 Contemporaneously with the ANPR, the Agencies also issued for public comment two proposed supervisory guidance documents relating to the proposed framework. 9 The first proposed 2003 guidance document described supervisory views on the credit risk measurement and management systems that should be implemented by banks that adopt the IRB approach for computing risk-based capital requirements for corporate credit risk exposures. The second proposed 2003 guidance document provided supervisory views on the operational risk measurement and management systems that should be implemented by banks that adopt the AMA for computing risk-based capital requirements for operational risk, including their operational risk management, data elements, and quantification processes. In October 2004, the Agencies also issued for public comment proposed supervisory guidance on IRB systems for retail credit risk exposures. 10 7 See 68 FR 45900 (Aug. 4, 2003). 8 See The New Basel Capital Accord (April 2003) (available at http://www.bis.org ). 9 See 68 FR 45949 (Aug. 4, 2003). 10 See 69 FR 62748 (Oct. 27, 2004), and 70 FR 423 (Jan. 4, 2005) (correction). The first guidance document presented in this notice sets forth proposed supervisory guidance on IRB systems for credit risk covering the wholesale and retail exposure categories, as well as guidance on the equity and securitization exposure categories (IRB Guidance). Under the IRB framework, banks would use internal estimates of certain risk components as key inputs in the determination of their regulatory risk-based capital requirement for credit risk. As mentioned above, the Agencies previously published proposed supervisory guidance on a bank's IRB systems for corporate and retail exposures in 2003 and 2004, respectively. Since the release of those documents, the Agencies have continued to refine the proposals based on insights gained from public comment and the collective efforts of the interagency IRB working groups. The IRB Guidance updates and consolidates the previously proposed supervisory guidance on corporate and retail exposures. It also provides new guidance on systems a bank may need to differentiate the risk of other credit exposure types, such as equity and securitization exposures, as well as to recognize the benefits of financial collateral in mitigating counterparty credit risk in certain transactions or to use the double default treatment for certain wholesale exposures. The IRB Guidance is structured somewhat differently from the proposed supervisory guidance issued in 2003 and 2004. Those guidance documents contained four chapters covering corporate ratings and retail segmentation systems, quantification, data management and maintenance, and controls, with discussion of validation and stress testing contained within the rating and segmentation and quantification chapters. The structure of the IRB Guidance generally follows the key components of a bank's advanced systems for credit risk outlined in the NPR. Chapter 1 provides guidance on governance of a bank's overall advanced systems for credit risk. Chapters 2 through 5 cover the components of a bank's IRB systems for wholesale and retail exposures. Chapters 6 and 7 provide guidance on data management and maintenance and the control and validation framework. Chapter 8 provides guidance on stress testing. Chapters 9 through 11 provide guidance on the other systems a bank may need to differentiate risk in certain transactions subject to counterparty credit risk, equity exposures, and securitization exposures. The IRB Guidance supplements the NPR and provides additional context and detail to help banks meet the qualification requirements in the NPR relevant to a bank's systems and processes for credit risk. Thus, the guidance should be read alongside the NPR to obtain a full perspective of the underlying requirements in the proposed rule. The guidance does not contain additional proposed requirements that are not in the NPR. Chapters 5, 9, 10, and 11, are being issued for the first time and supplement the detailed discussion of those topics in the NPR. Similar to the previously proposed corporate and retail guidance, the IRB Guidance contains supervisory standards (designated with an “S”) that highlight important elements of a bank's advanced systems for credit risk. The supervisory standards contained in the previously proposed corporate and retail guidance documents have been consolidated and updated and new supervisory standards are proposed. The second guidance document in this notice sets forth proposed supervisory guidance on the AMA for operational risk (AMA Guidance), updating the proposed AMA Guidance published in 2003. Since the issuance of that proposed AMA Guidance, the Agencies have revised the guidance to clarify issues and simplify, wherever possible, supervisory standards. The revisions are based on insights gained from public comment and the collective efforts of the interagency AMA working group. Under the AMA framework, a bank would rely on internal estimates of its operational risk exposure to generate its regulatory risk-based capital requirement for operational risk. The AMA Guidance provides additional context and detail to help a bank meet the qualification requirements outlined in the NPR relevant to operational risk. Some of the specific revisions to the AMA Guidance include: (1) Clarifying the roles of a bank's board of directors and management in developing and overseeing the implementation of the bank's AMA framework; (2) expanding standard 5 to address the integration of the bank's operational risk management, data and assessment, and quantification processes into the bank's existing risk management decision-making processes; (3) expanding and clarifying operational risk quantification standards both to reflect the evolution of industry practices, as well as to address supervisory concerns; (4) clarifying supervisory expectations regarding the use of scenario analysis, the key elements used to support operational risk management and measurement, and eligible operational risk offsets (see standards 20, 24, and 26, respectively); (5) adding standard 25 that discusses how frequently a bank must recalculate its estimate of operational risk exposure and its risk-based capital requirement for operational risk; (6) adding standard 27 that a bank must employ a unit of measure that is appropriate for its range of business activities and the variety of operational loss events to which it is exposed; (7) expanding the discussion on dependence modeling in standard 28; and (8) adding a section that discusses a bank's use, in certain limited circumstances, of an alternative quantification system to estimate its operational risk exposure. The Agencies recognize that a bank required to adopt an AMA framework may have developed an implementation plan using the proposed supervisory standards in the 2003 proposed AMA Guidance to assess its status in meeting the requirements proposed in the ANPR and to determine additional work needed to comply with those requirements. The table below maps the current proposed supervisory standards to those in the 2003 proposed AMA Guidance. Comparison of Current Proposed AMA Supervisory Standards to the 2003 Proposed AMA Supervisory Standards Current Proposed Standard Number 2003 Proposed Standard Number 1 1 2 8 3 11 4 2 5 3 6 4 7 5 8 6 9 7 10 9, 10 11 12 12 13, 14 13 15 14 16 15 17 16 18 17 19 18 20 19 21 20 24 21 22 22 23 23 25 24 27 25 New 26 28 27 New 28 29 29 30 30 26 31 31 32 32, 33 The third document sets forth proposed supervisory guidance on the supervisory review process (Pillar 2) in the New Advanced Capital Adequacy Framework. The process of supervisory review described in this proposed guidance document reflects a continuation of the longstanding approach employed by the Agencies in their supervision of banks. However, new methods for calculating regulatory risk-based capital requirements—such as those in the proposed framework—and development of improved risk monitoring and management tools within the industry often bring changes in the relative emphasis placed on the various aspects of supervisory review. This proposed guidance document highlights aspects of existing supervisory review that are being augmented or more clearly defined to support the proposed framework. Under the framework, in determining the extent to which banks should hold capital in excess of regulatory minimums, supervisors would consider the combined implications of a bank's compliance with qualification requirements for regulatory risk-based capital standards, the quality and results of its internal capital adequacy assessment process (ICAAP), and supervisory assessment of its risk management processes, control structure, and other relevant information relating to its risk profile and capital position. The ICAAP (while not mandating the determination of economic capital) should, to the extent possible, identify and measure material risks, which may include (but should not necessarily be limited to) credit risk, market risk, operational risk, interest rate risk, and liquidity risk, and account for concentrations within and among risk types. The Agencies solicit comment on all aspects of the supervisory guidance documents. In addition, the Agencies believe an important goal for any regulatory capital system is to achieve a measure of consistency in the capital requirements assigned to exposures with similar risk profiles held by different banks. The Agencies seek comment on the extent to which this proposed supervisory guidance will promote that objective. Paperwork Reduction Act A. Request for Comment on Proposed Information Collection In accordance with the requirements of the Paperwork Reduction Act of 1995, the Agencies may not conduct or sponsor, and the respondent is not required to respond to, an information collection unless it displays a currently valid Office of Management and Budget (OMB) control number. The Agencies are requesting comment on a proposed information collection. The Agencies are also giving notice that the proposed collection of information has been submitted to OMB for review and approval. Comments are invited on: (a) Whether the collection of information is necessary for the proper performance of the Agencies' functions, including whether the information has practical utility; (b) The accuracy of the estimates of the burden of the information collection, including the validity of the methodology and assumptions used; (c) Ways to enhance the quality, utility, and clarity of the information to be collected; (d) Ways to minimize the burden of the information collection on respondents, including through the use of automated collection techniques or other forms of information technology; and (e) Estimates of capital or start up costs and costs of operation, maintenance, and purchase of services to provide information. Comments should be addressed to: OCC: Communications Division, Office of the Comptroller of the Currency, Public Information Room, Mail stop 1-5, Attention: 1557-NEW, 250 E Street, SW., Washington, DC 20219. In addition, comments may be sent by fax to (202) 874-4448, or by electronic mail to regs.comments@occ.treas.gov. You can inspect and photocopy the comments at the OCC's Public Information Room, 250 E Street, SW., Washington, DC 20219. You can make an appointment to inspect the comments by calling (202) 874-5043. Board: You may submit comments, identified by FR 4199, by any of the following methods: • Agency Web Site: http://www.federalreserve.gov. Follow the instructions for submitting comments at http://www.federalreserve.gov/generalinfo/foia/ProposedRegs.cfm. • Federal eRulemaking Portal: http://www.regulations.gov. Follow the instructions for submitting comments. • E-mail: regs.comments@ federalreserve.gov. • Fax: (202) 452-3819 or (202) 452-3102. • Mail: Jennifer J. Johnson, Secretary, Board of Governors of the Federal Reserve System, 20th Street and Constitution Avenue, NW., Washington, DC 20551. All public comments are available from the Board's Web site at http://www.federalreserve.gov/generalinfo/foia/ProposedRegs.cfm as submitted, except as necessary for technical reasons. Accordingly, your comments will not be edited to remove any identifying or contact information. Public comments may also be viewed electronically or in paper form in Room MP-500 of the Board's Martin Building (20th and C Streets, NW.) between 9 a.m. and 5 p.m. on weekdays. FDIC: You may submit comments by any of the following methods: • Agency Web Site: http://www.fdic.gov/regulations/laws/federal. Follow instructions for submitting comments on the Agency Web Site. • E-mail: Comments@FDIC.gov. Include “Basel II Supervisory Guidance” in the subject line of the message. • Mail: Robert E. Feldman, Executive Secretary, Attention: Comments, Federal Deposit Insurance Corporation, 550 17th Street, NW., Washington, DC 20429. • Hand Delivery/Courier: Guard station at the rear of the 550 17th Street Building (located on F Street) on business days between 7 a.m. and 5 p.m. (EST). • Federal eRulemaking Portal: http://www.regulations.gov. Follow the instructions for submitting comments. Public Inspection: All comments received will be posted without change to http://www.fdic.gov/regulations/laws/federal including any personal information provided. Comments may be inspected and photocopied in the FDIC Public Information Center, 3501 North Fairfax Drive, Room E-1002, Arlington, VA 22226, between 9 a.m. and 5 p.m. (EST) on business days. Paper copies of public comments may be ordered from the Public Information Center by telephone at (877) 275-3342 or (703) 562-2200. A copy of the comments may also be submitted to the OMB desk officer for the Agencies: By mail to U.S. Office of Management and Budget, 725 17th Street, NW., #10235, Washington, DC 20503 or by facsimile to 202-395-6974, Attention: Federal Banking Agency Desk Officer. OTS: Information Collection Comments, Chief Counsel's Office, Office of Thrift Supervision, 1700 G Street, NW., Washington, DC 20552; send a facsimile transmission to (202) 906-6518; or send an e-mail to infocollection.comments@ots.treas.gov. OTS will post comments and the related index on the OTS Internet site at http://www.ots.treas.gov. In addition, interested persons may inspect the comments at the Public Reading Room, 1700 G Street, NW., by appointment. To make an appointment, call (202) 906-5922, send an e-mail to public.info@ots.treas.gov, or send a facsimile transmission to (202) 906-7755. B. Proposed Information Collection Title of Information Collection: Proposed Basel II Interagency Supervisory Guidance for IRB, AMA, and the Supervisory Review Process. Frequency of Response: Event-generated. Affected Public: OCC: National banks. Board: State member banks, bank holding companies, affiliates and certain non-bank subsidiaries of bank holding companies, commercial lending companies owned or controlled by foreign banks, and Edge and agreement corporations. FDIC: Insured nonmember banks and certain subsidiaries of these entities. OTS: Savings associations and certain of their subsidiaries. Abstract: The notice sets forth three proposed supervisory guidance documents for implementing proposed revisions to the risk-based capital standards in the United States (New Advanced Capital Adequacy Framework). The proposed guidance documents concern (1) the internal ratings-based systems for credit risk (IRB), (2) the advanced measurement approaches for operational risk (AMA), and (3) the supervisory review process (Pillar II). The Agencies believe that the documentation, prior approvals, and disclosures included in the proposed IRB and AMA guidance are directly related to the information collection requirements found in the Basel II notice of proposed rulemaking (NPR) published in the Federal Register on September 25, 2006 (71 FR 55830). More specifically, the information collection aspects of the proposed IRB and AMA guidance tie to the following sections of the NPR: 21, 22, 44, 53, and 71. The Agencies believe that the burden estimates developed for the NPR adequately cover the additional specificity contained in the proposed IRB and AMA guidance. For the proposed Pillar II portion of the guidance, the Agencies believe that paragraphs 25, 31, 35, 37, and 42 impose new information collection requirements that were beyond the scope of the burden estimates developed for the NPR. The agencies burden estimates for these additional information collection requirements are summarized below. Note that the estimated number of respondents listed below include both institutions for which the Basel II risk-based capital requirements are mandatory and institutions that may be considering opting-in to Basel II (despite the lack of any formal commitment by most of these latter institutions). Estimated Burden: OCC Number of Respondents: 52. Estimated Burden per Respondent: 140 hours. Total Estimated Annual Burden: 7,280 hours. Board Number of Respondents: 15. Estimated Burden per Respondent: 420 hours. Total Estimated Annual Burden: 6,300 hours. FDIC Number of Respondents: 19. Estimated Burden per Respondent: 420 hours. Total Estimated Annual Burden: 7,980 hours. OTS Number of Respondents: 4. Estimated Burden per Respondent: 420 hours. Total Estimated Annual Burden: 1,680 hours. The proposed supervisory guidance documents follow: Proposed Supervisory Guidance on Internal Ratings-Based Systems for Credit Risk Table of Contents Introduction I. Purpose II. Scope of Guidance Chapter 1: Advanced Systems for Credit Risk Rule Requirements I. Overview II. Governance of Advanced Systems Chapter 2: Wholesale Risk Rating Systems Rule Requirements I. Overview II. Credit Rating Assignment Techniques A. Expert Judgment B. Models C. Constrained Judgment D. Rating Overrides III. Definition of Default IV. Independence of the Wholesale Risk Rating Process V. IRB Risk Rating System Architecture A. Two-Dimensional Risk-Rating System B. Other Considerations Chapter 3: Retail Segmentation Systems Rule Requirements I. Overview II. Definition of Default III. Retail Segmentation Architecture A. Criteria for Retail Segmentation B. Assignment of Exposures to Retail Segments Chapter 4: Quantification Rule Requirements I. Overview A. Stages of the Quantification Process B. General Standards for Sound Quantification II. Probability of Default (PD) A. Data B. Estimation C. Mapping D. Application III. Expected Loss Given Default (ELGD) and Loss Given Default (LGD) A. Data B. Estimation C. Mapping D. Application IV. Exposure at Default (EAD) A. Data B. Estimation C. Mapping D. Application V. Maturity (M) VI. Special Cases and Applications A. Loan Sales B. Multiple Legal Entities Appendix A: Illustrations of the Quantification Process for Wholesale Portfolios Appendix B: Illustrations of the Quantification Process for Retail Portfolios Chapter 5: Wholesale Credit Risk Protection Rule Requirements Chapter 6: Data Management and Maintenance Rule Requirements I. Overview II. General Data Requirements A. Life Cycle Tracking for Wholesale Exposures B. Rating Assignment Data for Wholesale Exposures C. Segmentation Data for Retail Exposures D. Outsourced Activities E. Asset Sales III. Data Applications A. Validation and Refinement B. Applying IRB System Improvements Historically C. Calculating Risk-Based Capital Ratios and Reporting to the Public D. Supporting Risk Management IV. Managing Data Quality and Integrity A. Documentation and Definitions B. Electronic Storage and Access Appendix A: Data Elements for Wholesale and Retail Exposures A. Examples of Data Elements for Wholesale Exposures B. Examples of Data Elements for Retail Exposures Appendix B: Applying Risk Rating System Improvements Historically Chapter 7: Controls and Validation Rule Requirements I. Overview II. Reviews of the IRB System III. Consistency Between IRB Systems and Risk Management Processes IV. Internal Audit V. Validation Activities A. General Validation Requirements B. Validation Activities C. Minimum Frequency of Validation Chapter 8: Stress Testing of Risk-Based Capital Requirements Rule Requirements Chapter 9: Counterparty Credit Risk Exposure Rule Requirements I. Overview II. Transactions with Counterparty Credit Risk III. Definitions IV. Netting V. Determination of Eligibility for EAD Adjustment VI. Methods for Determining EAD A. Methodologies for Repo-style Transactions and Eligible Margin Loans B. EAD for OTC Derivative Contracts C. Internal Models Methodology VII. Defaulted Counterparties Chapter 10: Risk-Weighted Assets for Equity Exposures Rule Requirements I. Overview II. Definition of Banking Book Equities III. Applying the Framework IV. Using Internal Models for Equity Exposures V. Quantification of Equity Exposures A. Reference Data B. External Data C. Estimation VI. Validation of Internal Models for Equity Exposures VII. Consistency Between Internal Models Used for Equity Exposures and Risk Management Processes Chapter 11: Securitizations Rule Requirements I. Overview II. Scope of Application III. General Principles of the Securitization Framework A. Risk Transference B. Implicit Support C. Servicer Cash Advances D. Clean-up Calls E. Maximum Capital Requirements for Securitization Exposures IV. Hierarchy of Approaches V. IRB Approaches for Securitization Exposures A. Ratings-Based Approach B. Internal Assessment Approach VI. Internal Credit Assessment Process in the IAA VII. Validation of IAA A. Supervisory Formula Approach VIII. Early Amortization Provisions IX. Data Management Requirements A. Data Elements Appendix A: Description of the Supervisory Formula Approach (SFA). Appendix B: Examples of Data Elements for Securitization Exposures Attachment A: The NPR Qualification Requirements Related to the IRB Framework Attachment B: Supervisory Standards Attachment C: Acronym List Introduction I. Purpose 1. This proposed guidance (“guidance”), published jointly by the U.S. Federal banking agencies 1 provides supervisory guidance for U.S. banks, thrifts, and bank holding companies (“banks”) that adopt the Advanced Internal Ratings-Based Approach (“IRB” or “IRB framework”) for calculating minimum regulatory risk-based capital (“risk-based capital”) requirements for credit risk under the Basel II capital regulation. 1 The Federal banking agencies are: The Board of Governors of the Federal Reserve System; the Federal Deposit Insurance Corporation; the Office of the Comptroller of the Currency; and the Office of Thrift Supervision; and will collectively be referred to as “the Agencies,” “supervisors,” or “regulators” in this guidance. 2. This guidance supplements the notice of proposed rulemaking (“NPR” or “proposed rule”) published in the Federal Register on September 25, 2006. 2 The NPR proposes a regulatory framework within which all banks subject to the proposed rule must develop their IRB systems. The NPR contains qualification requirements that each bank subject to the proposed rule must meet to the satisfaction of its primary Federal supervisor before using its IRB systems to calculate risk-based capital requirements. As stated in the preamble to the NPR, the qualification requirements for these systems are written in broad terms to accommodate the many ways a bank may design and implement a robust internal risk measurement and management system and to permit industry practice to evolve. As a supplement to the NPR, this guidance provides supervisory standards and additional detail on credit risk measurement and management systems that will assist banks in satisfying the requirements in the NPR. 2 71 FR 55830 (Sept. 25, 2006). II. Scope of Guidance 3. The focus of this guidance is on wholesale, retail, equity, and securitization exposures. A bank subject to the IRB framework for credit risk in the NPR is required to have systems for determining risk-based capital requirements for its wholesale and retail exposures. The wholesale category includes corporate exposures (for example, exposures to companies and banks, as well as commercial real estate exposures and other types of specialized lending), sovereign exposures, and other non-retail exposures. The retail category includes residential mortgage exposures, qualifying revolving exposures (QRE), and other retail exposures. 4. A bank may also need systems to differentiate the risk of other exposure types, such as equity and securitization exposures, as well as to recognize the benefits of financial collateral in mitigating counterparty credit risk in certain transactions or to use double default treatment for certain wholesale exposures. 5. In aggregation, the IRB systems and other systems for differentiating credit risk are defined in the NPR and in this guidance as a bank's “advanced systems.” This guidance covers advanced systems for all of a bank's credit-related exposure types. A bank's advanced systems also include its systems for determining risk-based capital requirements for its operational risk exposures under the proposed Advanced Measurement Approaches (“AMA”) framework, which is the subject of a separate supervisory guidance document. Certain banks subject to the proposed rule may also be required to calculate risk-based capital requirements for their market risk exposures. 6. As described in separate guidance relating to supervisory review (Pillar 2), in addition to meeting qualification requirements for regulatory risk-based capital standards, a bank must have a rigorous process for assessing its overall capital adequacy in relation to its risk profile and a comprehensive strategy for maintaining an appropriate level of capital. This process (while not mandating the determination of economic capital) should, to the extent possible, identify and measure material risks, which may include (but should not necessarily be limited to) credit risk, market risk, operational risk, interest rate risk, and liquidity risk, and account for concentrations within and among risk types. One of the main objectives of the internal capital adequacy assessment process is to identify the extent to which banks need to hold capital above regulatory minimums, in order to address risks not adequately captured by minimum regulatory capital requirements. 7. A primary objective of the IRB framework is to make the risk-based capital requirements more sensitive to credit risk. In general, the IRB framework incorporates recent developments in risk management and banking supervision. Under this framework, banks use their own internal risk rating and segmentation systems, as well as their quantification processes, to generate estimates of risk parameters that are inputs to the calculation of the risk-based capital requirements. Data that support accurate and reliable credit risk measurements, as well as rigorous management oversight and controls, including continuous monitoring and validation, are crucial to the prudent application of the IRB framework. 8. This guidance, which is written for supervisors and bankers, describes the important elements and characteristics of a bank's advanced systems for credit risk. Toward this end, this guidance designates certain of those elements as supervisory standards denoted by the prefix “S.” These supervisory standards generally implement or clarify the requirements in the NPR and, whenever possible, are principle-based to provide banks with flexibility in implementing the framework. However, when prudential concerns or the need for standardization outweigh the benefits of flexibility, the supervisory standards are specified in greater detail. Furthermore, nothing in this guidance should be interpreted as weakening, modifying, or superseding the safety and soundness principles articulated in the Agencies” existing statutes, regulations, or guidance. The standards are contained within each chapter with a full compilation of the standards provided in Attachment B. 9. Supervisors will consider this guidance in evaluating banks' advanced systems for credit risk. This guidance assumes that readers are familiar with the proposed framework for calculating risk-based capital requirements for credit risk articulated in the NPR. 10. The conceptual framework outlined in this guidance is not intended to dictate the precise manner by which banks should meet the qualification and other requirements in the NPR. Supervisors will determine compliance with the qualification requirements by evaluating, on an individual bank basis, the extent to which banks meet the substance and spirit of those requirements as they relate to each of the components of a bank's advanced systems for credit risk. However, evaluating each qualification requirement individually is not sufficient to determine a bank's overall compliance. The components of a bank's advanced systems for credit risk should complement and reinforce one another to ensure the accuracy of risk measurements. As part of the supervisory review of a bank's advanced systems, supervisors will analyze the extent to which a bank's advanced systems incorporate the substance and spirit of the standards outlined in this guidance. 11. The structure of this guidance generally follows the key components of the advanced systems for credit risk. Chapter 1 provides guidance on governance of a bank's overall advanced systems. Chapters 2 through 7 cover the components of a bank's IRB systems for wholesale and retail exposures. Chapter 8 provides guidance on stress testing. Chapters 9 through 11 provide guidance on the other systems a bank may need to differentiate risk for certain transactions subject to counterparty credit risk, equity exposures, and securitization exposures and supplements the detailed discussion of these exposure types in the NPR. The data standards and control framework provided in Chapters 6 and 7, respectively, of this guidance generally apply to these other systems as well. 12. To aid the reader, the applicable NPR qualification requirements are listed at the front of each chapter, as well as listed together in Attachment A. Also, certain NPR requirements, such as definitions, are either repeated in this guidance or paraphrased to provide context. However, readers must look to the NPR for the exact proposed rule requirements. 13. What follows is a brief description of each chapter: Chapter 1: Advanced Systems for Credit Risk The chapter provides a discussion of the governance and system and process requirements for a bank's advanced systems for credit risk. It also outlines the key components of a bank's advanced systems for credit risk. Chapter 2: Wholesale Risk Rating Systems A key component of an IRB system for wholesale exposures is the risk rating system. This chapter describes the design and operation of wholesale risk rating systems. Banks should use the principles outlined in this chapter when designing and operating wholesale risk rating systems. Chapter 3: Retail Segmentation Systems A key component of an IRB system for retail credit exposures is the segmentation system, which groups retail exposures into segments according to risk characteristics. This segmentation is the retail portfolio analogue of assigning ratings to exposures in wholesale portfolios. This chapter describes the design and operation of an IRB segmentation system. The retail framework provides banks with substantial flexibility to use the retail segmentation that is most appropriate for their activities. Chapter 4: Quantification Another key component of an IRB system is a quantification process that assigns numerical values to the key risk parameters that are used as inputs to the IRB risk-based capital formulas. This chapter provides guidance on the quantification process for wholesale and retail exposures. These risk parameters are probability of default (“PD”), expected loss given default (“ELGD”), loss given default (“LGD”), and exposure at default (“EAD”), and for wholesale exposures only, the effective remaining maturity (“M”). The quantification of these risk parameters should be the result of a disciplined process as described in this chapter. The chapter also includes specific examples for both wholesale rating systems and retail segmentation systems in the two appendices. Chapter 5: Wholesale Credit Risk Protection This chapter supplements the detailed discussion of credit risk mitigation in the NPR by providing guidance on how banks may recognize contractual arrangements for exposure-level credit protection (eligible guarantees and eligible credit derivatives) that transfer risk to one or more third parties. Each of these forms of credit protection must meet certain specific standards of eligibility, as articulated in the NPR, for recognition of the associated risk mitigation. Chapter 6: Data Management and Maintenance A bank must have advanced data management and maintenance systems that support credible and reliable risk parameter estimates. This chapter describes how a bank should collect, maintain, and manage the data needed to support the other IRB system components for wholesale and retail exposures ( e.g. , risk rating and segmentation systems, the quantification process, and validation and other control processes), as well as the bank's broader risk management and reporting needs. Chapter 7: Controls and Validation A bank must have a system of controls that ensures that the components of the IRB system are functioning effectively. This chapter provides guidance on the important elements of an effective control environment, including independent review processes, a comprehensive validation process (evaluation of developmental evidence, ongoing monitoring, and outcomes analysis), and an internal audit review and reporting process. Chapter 8: Stress Testing of Risk-Based Capital Requirements Banks must conduct stress testing analysis of their advanced systems for credit risk as part of the risk-based capital management process. Stress testing analysis is a means of understanding how economic downturns, as described by stress scenarios, cause migration across ratings or segments and the concomitant change in required risk-based capital. This chapter discusses considerations for conducting stress testing analyses. Chapter 9: Counterparty Credit Risk Exposure For certain transactions subject to counterparty credit risk, banks may be allowed to recognize the risk mitigating effect of financial collateral through an adjustment to EAD. This chapter supplements the detailed discussion of counterparty credit risk in the NPR by describing some of the elements of counterparty credit risk mitigation, providing information to aid banks in choosing among the alternative methods to calculate EAD for these transactions, and providing some descriptions and illustrative examples of acceptable modeling practices for the estimation of EAD under the alternative methods. Chapter 10: Risk-Weighted Assets for Equity Exposures This chapter supplements the detailed discussion of equity exposures provided in the NPR. It provides guidance on determining risk-based capital requirements for equity exposures held in the banking book for banks subject to the Market Risk Rule and for all equity exposures for banks not subject to the Market Risk Rule. Chapter 11: Securitization Exposures A securitization exposure is any exposure whose credit risk reflects the tranching of risk of one or more underlying exposures. This chapter describes the concepts, eligibility, and mechanics associated with applying the three approaches for calculating risk-based capital requirements for securitization exposures. Chapter 1: Advanced Systems for Credit Risk Rule Requirements Part III, Section 22(a)(2): The systems and processes used by a bank for risk-based capital purposes [in the NPR] must be consistent with the bank's internal risk management processes and management information reporting systems. Part III, Section 22(a)(3): Each bank must have an appropriate infrastructure with risk measurement and management processes that meet the qualification requirements [in the NPR] and are appropriate given the bank's size and level of complexity. Regardless of whether the systems and models that generate the risk parameters necessary for calculating a bank's risk-based capital requirements are located at any affiliate of the bank, the bank itself must ensure that the risk parameters and reference data used to determine its risk-based capital requirements are representative of its own credit risk and operational risk exposures. Part III, Section 22(j)(1): The bank's senior management must ensure that all components of the bank's advanced systems function effectively and comply with the qualification requirements [in the NPR]. Part III, Section 22(j)(2): The bank's board of directors (or a designated committee of the board) must at least annually evaluate the effectiveness of, and approve, the bank's advanced systems. Part III, Section 22(k): Documentation. The bank must adequately document all material aspects of its advanced systems. I. Overview 1. This chapter provides a discussion of the governance and system and process requirements for a bank's advanced systems for credit risk. Board of directors and senior management oversight is critical to ensure that the design and function of the advanced systems are appropriate. Regardless of the specifics of a bank's advanced systems for credit risk, a bank should have a rigorous credit risk management infrastructure that complements these systems. 2. A bank subject to the framework for credit risk in the NPR is required to have an internal ratings-based system (“IRB system”) for determining risk-based capital requirements for its wholesale and retail exposures. S 1-1 An IRB system must have five interdependent components that enable an accurate measurement of credit risk and risk-based capital requirements. 3. The components of an IRB system are: • A risk rating and segmentation system that differentiates risk by assigning ratings to individual wholesale obligors and exposures and individual retail exposures to segments; • A quantification process that translates the risk characteristics of wholesale obligors and exposures and segments of retail exposures into numerical risk parameters that are used as inputs to the IRB risk-based capital formulas. These risk parameters are probability of default (“PD”), expected loss given default (“ELGD”), loss given default (“LGD”), and exposure at default (“EAD”), and for certain wholesale exposures only, the effective remaining maturity (“M”); • A data management and maintenance system that supports the IRB system; • Oversight and control mechanisms that ensure the IRB system is functioning effectively and producing accurate results; and • An ongoing process that validates the accuracy of the risk rating assignments, segmentations, and the risk parameters. 4. If applicable, a bank will also need systems to differentiate risk for other credit exposure types, such as for equity and securitization exposures, as well as to recognize the benefits of financial collateral in mitigating counterparty credit risk in certain transactions or to use double default treatment for certain wholesale exposures. 5. In aggregation, the IRB system and other systems for differentiating credit risk are defined in the NPR and in this guidance as a bank's “advanced systems” for credit risk. Chapters 2 through 7 of this guidance provide supplemental guidance on IRB systems for wholesale and retail exposures. Chapter 8 provides banks with guidance on conducting stress testing analyses of their advanced systems for credit risk. Chapters 9 through 11 cover additional systems a bank may need to have for other credit exposure types. II. Governance of Advanced Systems S 1-2 Senior management must ensure that all of the components of the bank's advanced systems for credit risk function effectively and comply with the qualification requirements in the NPR. 6. Senior management should provide ongoing, active oversight of the advanced systems outlined in this supervisory guidance, and articulate the expectations for the technical and operational performance of the advanced systems, including the control framework. To provide effective oversight of the advanced systems, senior management should have extensive knowledge of the advanced systems' policies, underwriting standards, lending practices, account management activities, and collection and recovery practices. Senior management should understand how these factors affect all of the components of the advanced systems. 7. The scope and depth of risk management reports should be sufficient for senior management to monitor the performance of the components of the advanced systems. Detailed reports should include, but are not limited to, the following topics: • Risk profile by rating for wholesale exposures and by segment for retail exposures; • Migration across ratings and segments with emphasis on unexpected results; • Updates to the quantification performance results; • Validation results; • Comparative analysis of risk-based and internal capital assessments; and • Control process assessments. S 1-3 The board of directors or its designated committee must at least annually evaluate the effectiveness of, and approve, the bank's advanced systems. 8. The board of directors or its designated committee should at least annually ensure that management has appropriate processes and controls in place that support effective advanced systems for credit risk. The board should be provided with information that will enable it to conclude, with reasonable assurance, that management has appropriate processes and controls in place that support effective advanced systems for credit risk. To allow for ongoing monitoring, the board should be provided with reports summarizing the design and performance of the advanced systems. The board's strategic direction and oversight is essential to effective advanced systems. S 1-4 Each bank (including each depository institution) must ensure that the risk parameters and reference data used to determine its risk-based capital requirements are representative of its own credit risk. 9. Each bank must have an appropriate infrastructure with risk measurement and management processes that meet the qualification requirements in the NPR. Each bank's advanced systems for credit risk should also incorporate the supervisory standards in this guidance. This infrastructure must be appropriate given the bank's size and level of complexity. Regardless of whether the systems and models that generate the risk parameters necessary for calculating a bank's risk-based capital requirements are located at any affiliate of the bank, the bank must ensure that the risk parameters and reference data used to determine its risk-based capital requirements are representative of the bank's credit risk profile. 10. While some organizations may conduct rating, segmentation, quantification, and validation activities on a consolidated basis, each bank subject to the capital requirements for advanced systems must determine its risk-based capital requirements for credit risk on a stand-alone basis and hold its own separate risk-based capital in proportion to the risk exposure of its portfolios. Specifically, the PD, ELGD, LGD, and EAD estimates used to determine risk-based capital levels must be applied to exposures at the exposure or segment level, and risk-based capital requirements for each relevant bank should be based on the proportionate share of each exposure or segment owned by such bank. 11. The board of directors should ensure that senior management at each bank confirm, through periodic evaluations, that risk parameters assigned to its credit exposures are appropriate on a stand-alone basis, and that the control and validation standards in Chapter 7 of this guidance are met. S 1-5 Banks should establish specific accountability for the overall performance of their advanced systems for credit risk. 12. An individual or group of individuals should be responsible for the design and operation of the overall advanced systems. This accountability includes oversight for all of the components of the advanced systems for credit risk, regardless of which organizational units perform those processes. Authority and key responsibilities should be thoroughly documented and responsible individuals should be held accountable for the performance of the advanced systems. S 1-6 A bank's advanced systems should be transparent. 13. Banks must adequately document all material aspects of their advanced systems. Adequate documentation will ensure transparency of a bank's advanced systems. A bank demonstrates the transparency of its advanced systems by comprehensively documenting all the systems” components. Transparency through documentation is important so that third parties, such as a bank's supervisors and auditors, are able to understand, evaluate, and assess the effectiveness of the bank's advanced systems. 14. Documentation should encompass, but is not limited to, the internal risk rating and segmentation systems, risk parameter quantification processes, data collection and maintenance processes, and model design, assumptions, and validation results. The guiding principle governing documentation is that it should support the requirements for the quantification, validation, and control and oversight mechanisms as well as the bank's broader credit risk management and reporting needs. Documentation is critical to the supervisory oversight process. Chapter 2: Wholesale Risk Rating Systems Rule Requirements Part III, Section 22(b)(1): A bank must have an internal risk rating and segmentation system that accurately and reliably differentiates among degrees of credit risk for the bank's wholesale and retail exposures. Part III, Section 22(b)(2): For wholesale exposures, a bank must have an internal risk rating system that accurately and reliably assigns each obligor to a single rating grade (reflecting the obligor's likelihood of default). The bank's wholesale obligor rating system must have at least seven discrete rating grades for non-defaulted obligors and at least one rating grade for defaulted obligors. Unless the bank has chosen to directly assign ELGD and LGD estimates to each wholesale exposure, the bank must have an internal risk rating system that accurately and reliably assigns each wholesale exposure to loss severity rating grades (reflecting the bank's estimate of the ELGD and LGD of the exposure). A bank employing loss severity rating grades must have a sufficiently granular loss severity grading system to avoid grouping together exposures with widely ranging ELGDs or LGDs. Part III, Section 22(b)(4): The bank's internal risk rating policy for wholesale exposures must describe the bank's rating philosophy (that is, must describe how wholesale obligor rating assignments are affected by the bank's choice of the range of economic, business, and industry conditions that are considered in the obligor rating process). Part III, Section 22(b)(5): The bank's internal risk rating system for wholesale exposures must provide for the review and update (as appropriate) of each obligor rating and (if applicable) each loss severity rating whenever the bank receives new material information, but no less frequently than annually. I. Overview 1. This chapter describes the design and operation of IRB risk rating systems for wholesale exposures. Banks will have latitude in designing and operating wholesale risk rating systems, subject to four broad principles: Two-dimensional risk rating system—Banks must be able to make meaningful and consistent differentiations among credit exposures along two dimensions—obligor default risk and loss severity in the event of a default. Rank order risks—Banks must rank obligors by their likelihood of default, and wholesale exposures ( e.g. , loans, facilities) by the loss severity expected in the event of default. Quantification—The risk rating system must be designed to facilitate quantification of obligor ratings in terms of PD and loss severity in terms of ELGD and LGD. Accuracy—The risk rating system must be designed to ensure that ratings are accurate, so that obligors within a rating grade have similar default risk and wholesale exposures within a loss severity rating grade have similar risk of loss in the event of default. II. Credit Rating Assignment Techniques 2. In general, a credit rating is a summary indicator of the relative risk of a credit exposure. Credit ratings can take many forms. Regardless of the form, meaningful credit ratings share two characteristics: • They group exposures to discriminate among possible outcomes. • They rank the perceived level of credit risk. 3. Banks have used credit ratings of various types for a variety of purposes. Some ratings are intended to rank obligors by risk of default and some are intended to rank wholesale exposures by expected loss, which incorporates risk of default and loss severity. Only risk rating systems that distinguish probability of default from loss given default meet the two-dimensional requirements for the IRB framework. 4. Banks use different techniques, such as expert judgment and models, to assign credit risk ratings. How ratings are assigned is important because different techniques will require different validation processes and control mechanisms to ensure the integrity of the rating system. Validation and controls are discussed in Chapter 7 of this guidance. Some rating assignment techniques are described below; any of these techniques—expert judgment, models, constrained judgment, or a combination thereof—could be acceptable in an IRB system, provided the bank meets the qualification requirements in the NPR and the substance and spirit of the standards outlined in this guidance. A. Expert Judgment 5. Historically, banks have used expert judgment to assign ratings to wholesale exposures. With this technique, an individual weighs relevant information and reaches a conclusion about the appropriate risk rating. The rater makes informed judgments based on knowledge gained through experience and training. 6. The key feature of expert-judgment systems is flexibility. The prevalence of judgmental rating systems reflects the view that the determinants of default are too complicated to be captured by a single quantitative model. The quality of management is often cited as an example of a risk determinant that is difficult to assess using a quantitative model. In order to foster internal consistency, banks employing expert judgment rating systems should provide narrative guidelines that set out specific quantitative and qualitative rating criteria for each rating grade. However, the expert should decide how much weight to give to each of these criteria in assigning a risk rating grade to an obligor. 7. The flexibility possible in the assignment of judgmental ratings has implications for how the accuracy of the ratings is reviewed. One goal of the ratings review validation process is to confirm that raters followed policy. However, two individuals exercising judgment can use the same information to support different ratings. Thus, individuals reviewing an expert judgment rating system should have sufficient credit expertise and a thorough knowledge of how the bank's rating methodology and policies should be applied. B. Models 8. In recent years, models have been developed to assign ratings to wholesale exposures. In a model-based approach, inputs are numeric and provide quantitative and qualitative information about an obligor. The inputs are combined using mathematical equations to produce a number that is translated into a categorical rating. An important feature of models is that the rating is perfectly replicable by another party, given the same inputs. 9. Models to assign wholesale ratings typically are statistically derived or based on expert-judgment techniques. 10. Some models are the result of statistical optimization, in which well-defined mathematical criteria are used to choose the model that has the closest fit to the observed data. Numerous techniques can be used to build statistical models; regression is one widely recognized example. Such models are often referred to as scoring models or scorecards, because they produce a single number, or “score,” as an output that may be related, for example, to the estimated probability of default of each individual obligor in a portfolio. Regardless of the specific statistical technique used, a knowledgeable independent reviewer should exercise judgment in evaluating the reasonableness of a model's development, including its underlying logic, and the methods used to handle the data. 11. In other cases, banks have built rating models by asking their experts to decide what weights to assign to critical variables in the models. Drawing on their experience, the experts first identify the observable variables that affect the likelihood of default. They then reach agreement on the weights to be assigned to each of the variables. Unlike statistical optimization, the experts are not necessarily using clear, consistent criteria to select the weights attached to the variables. Indeed, expert-judgment model building is often a practical choice when there is not enough data to support a statistical model building. Despite its dependence on expert judgment, this method can be called model-based as long as the resulting equation, most likely with linear weights, is used to rate the credits. Once the equation is set, the model can be replicated, a feature shared with statistically derived models. However, while some banks refer to these types of expert-derived models as “scorecards,” they are not scoring models in the conventional use of the term. The term scoring model or scorecard is customarily reserved for a rating model derived using strictly statistical techniques, as described in the preceding paragraph. Generally, independent credit experts use judgment to evaluate the reasonableness of the development of these expert-derived models. C. Constrained Judgment 12. The alternatives described above present the extremes; in practice, banks use risk rating systems that combine models with judgment. Two approaches are common. Judgmental systems with quantitative guidelines or model results as inputs. Individuals exercise judgment about risks subject to policy guidelines containing quantitative criteria such as minimum values for particular financial ratios. Banks develop quantitative criteria to guide individuals in assigning ratings, but the criteria may need to be augmented with additional information. One version of this constrained judgment approach features a model output as one among several criteria that an individual may consider when assigning ratings. The individual assigning the rating is responsible for prioritizing the criteria, reconciling conflicts between criteria, and, if warranted, overriding some criteria. Even if individuals incorporate model results as one of the factors in their ratings, they will exercise judgment in deciding what weight to attach to the model result. The appeal of this approach is that the model combines many pieces of information into a single output, which simplifies analysis, while the rater retains flexibility regarding the use of the model output. Model-based ratings with judgmental overrides. When banks use rating models, individuals are permitted to override the results under certain conditions and within tolerance levels for frequency. Credit-rating systems in which individuals can override models raise many of the same issues presented separately by pure judgment and model-based systems. If overrides are rare, the system can be evaluated largely as if it is a model-based system. If, however, overrides are prevalent, the system will be evaluated more like a judgmental system. D. Rating Overrides 13. Regardless of the rating assignment technique in use, banks should define, within their IRB rating system documentation, what constitutes a ratings override. A judgmental override occurs when judgment is used to reject a rating suggested by an objective rating process, such as a model or scorecard. A policy override occurs whenever a rating is assigned in a manner that deviates from the bank's approved rating policy and procedures. Overrides should be specifically identified, monitored, and analyzed to evaluate their impact on the bank's IRB rating system. III. Definition of Default S 2-1 Banks must identify obligor defaults in accordance with the IRB definition of default. 14. The consistent identification of defaults is fundamental to any IRB risk rating system. For IRB purposes, a bank's wholesale obligor is in default if, for any wholesale exposure of the bank to the obligor, the bank has: • Placed the exposure on non-accrual status consistent with the Call Report Instructions or the Thrift Financial Report (“TFR”) and the TFR Instruction Manual; • Taken a full or partial charge-off or write-down on the exposure due to the distressed financial condition of the obligor; or • Incurred a credit-related loss of 5 percent or more of the exposure's initial carrying value in connection with the sale of the exposure or the transfer of the exposure to the held-for-sale, available-for-sale, trading account, or other reporting category. 15. Partial charge-offs or write-downs for reasons not related to the distressed financial condition of the obligor do not trigger the default definition. For example, taking a write-down or charge-off to reflect forgiveness of a minor fee for relationship purposes unrelated to financial distress does not trigger the default definition. 16. An obligor in default remains in default until the bank has reasonable assurance of repayment and performance for all contractual principal and interest payments on all exposures of the bank to the obligor (other than exposures that have been fully written-down or charged-off). IV. Independence of the Wholesale Risk Rating Process S 2-2 Banks should demonstrate that their wholesale risk rating processes are sufficiently independent to produce objective ratings. 17. Independence in the rating process helps to ensure the integrity of ratings. Banks can promote more independence by implementing a variety of controls and reporting structures. For example, a bank could structure its organizational reporting lines so that the credit approval and the rating assignment decisions are separate from each other. Banks that separate the credit approval process from the rating assignment/review functions are often better able to manage the conflicts that arise between loan volume and credit quality goals. Banks should be aware of the full range of potential conflicts and should develop effective controls to mitigate any conflicts that might arise. 18. However, banks that choose to maintain less separation in organizational reporting lines between credit approval and rating assignment should strengthen controls and consider conducting a post-closing review process. A post-closing review provides an independent review of a rating that has been assigned by those who are not fully independent of the approval process. Any post-closing review, which serves to ensure that the initial rating is appropriate, should be conducted shortly after a credit is originated. The less independent the rating process is, the more rigorous the post-closing review should be. 19. Whether ratings integrity is achieved by creating structural independence in reporting lines or through a combination of other control processes, a bank should demonstrate that its rating processes ensure integrity in ratings throughout the economic cycle. V. IRB Risk Rating System Architecture A. Two-Dimensional Risk-Rating System S 2-3 IRB risk rating systems must have two dimensions obligor default and loss severity corresponding to PD (obligor default), and ELGD and LGD (loss severity). 20. Regardless of the type of rating system(s) used by a bank, the IRB framework imposes some specific requirements. The first requirement is that an IRB risk rating system must be two-dimensional. Banks will assign obligor ratings, which will be associated with a PD. They will also assign either a loss severity rating(s), which will be associated with ELGD and LGD estimates, or ELGD and LGD estimates directly to each wholesale exposure. 21. The process of assigning the obligor rating and either loss severity ratings or ELGD/LGD values—hereafter referred to as the rating system—is discussed below, and the process of quantifying the PD, ELGD and LGD risk parameters is discussed in Chapter 4. Obligor Ratings S 2-4 Banks must assign discrete obligor rating grades. 22. While banks may use models to estimate probabilities of default for individual obligors, the IRB framework requires banks to group the obligors into discrete rating grades. Each obligor rating grade, in turn, must be associated with a single PD. S 2-5 The obligor rating system must rank obligors by likelihood of default. 23. For example, if a bank uses a rating system based on a 10-point scale, with 1 representing obligors of highest financial strength and 10 representing defaulted obligors, rating grades 2 through 9 should represent groups of ever-increasing risk. In a rating system in which risk increases with the rating grade, an obligor with a rating grade 4 is riskier than an obligor with a rating grade 2, but need not be twice as risky. S 2-6 Banks must assign an obligor to only one rating grade. 24. As noted above, the IRB framework requires that the obligor rating be distinct from the loss severity rating, which is assigned to the wholesale exposure. The obligor rating should focus on the obligor's ability and willingness to service any obligation and to follow through on any commitments it has with the bank to avoid default. For example, in a 1-to-10 rating system, where risk increases with the number rating grade, an otherwise defaulted obligor with a fully cash-secured transaction should be rated 10—defaulted—regardless of the remote expectation of loss on a specific exposure. Conversely, a nondefaulted obligor whose financial condition warrants the highest investment grade rating should be rated 1, even if the bank's transactions are subordinate to other creditors and unsecured. Since the obligor rating is assigned to the obligor and not to its individual exposures, the bank must ensure that all the exposures to the same obligor bear the obligor's rating grade. 25. At the bottom of any IRB rating scale is at least one default rating grade. Once an obligor is in default on any exposure to the subject bank, the obligor rating grade associated with all of its exposures to that bank will be the default rating grade—even for those exposures of the obligor that have not triggered any element of the definition of default. Ratings Philosophy and Expected Ratings Migration S 2-7 A bank's rating policy must describe its ratings philosophy and how quickly obligors are expected to migrate from one rating grade to another in response to economic cycles. S 2-8 In assigning an obligor to a rating grade, a bank should assess the risk of obligor default over a period of at least one year taking into account the possibility of adverse economic conditions. 26. The term rating philosophy is used to describe how obligor rating assignments are affected by a bank's choice of the range of economic, business, and industry conditions that are considered in the rating process. It establishes the bank's philosophy on the manner in which it rates credits and the scenarios under which ratings would be expected to change. In assigning an obligor rating grade, banks must consider both the current risk characteristics of the obligor and the impact that adverse economic, business, and industry conditions could have on the obligor's ability to repay; however, nothing in this guidance requires any specific rating philosophy be employed. 27. Rating grades should group obligors that are expected to share similar default frequencies. The rating assignment for an obligor may be based upon a combination of obligor-specific (idiosyncratic) risk characteristics and the general economic, business, and industry (systematic) risk characteristics or conditions that obligors in the rating may experience. 28. The time horizon used for the assignment of obligors to rating grades should be one year or longer. The obligor rating should reflect the obligor's ability as evidenced by its financial capacity, as well as its willingness to service any obligation and to follow through on any commitments it has with the bank to avoid default. The time horizon chosen for the rating assignment process should be appropriate to the business line or geography for which the respective obligor rating system will be used. 29. That general description, however, still leaves open different possible implementations, depending upon what range of future systematic risk conditions the bank considers when making a rating assignment and the weight given to those conditions. In practice, it appears that most banks have adopted a rating philosophy where an obligor's rating would have some sensitivity to changes in economic conditions. Regardless of the approach taken, banks should document their choice of economic, business, and industry conditions considered in each risk rating system and the expected frequency of rating changes over economic cycles. Such differences have important implications for validation and other aspects of the operation of rating systems, and therefore should be clearly articulated and well understood. A bank should also understand the effects of ratings migration on its risk-based capital requirements and ensure that sufficient capital is maintained during all phases of the economic cycle. 30. A bank's ratings philosophy can be empirically demonstrated through an analysis of how its obligors migrate across rating grades as economic and industry conditions change. While individual obligor ratings may change due to changes in obligor-specific risk characteristics, the average migration observed through time is likely to reveal how sensitive rating assignments are to systematic risk changes. Rating systems in which obligor ratings are more closely linked at a given point in time to particular economic conditions are more likely to be associated with higher overall average rates of rating migration than are other systems. Ratings that respond primarily to obligor-specific (idiosyncratic) changes may be less sensitive to changes in economic and industry conditions, and be more stable throughout the economic cycle. Obligor-Rating Granularity S 2-9 Banks must have at least seven discrete obligor rating grades for non-defaulted obligors and at least one rating grade for defaulted obligors. 31. A risk rating system's grades should be sufficiently numerous to ensure that management can meaningfully differentiate risk in the portfolio, without being so numerous that they limit the system's practical use. To determine the appropriate number of rating grades beyond the minimum seven non-default rating grades, each bank should perform its own internal analysis. S 2-10 Banks should justify the number of obligor rating grades used in its risk rating system and the distribution of obligors across those grades. 32. Some portfolios may have a majority of obligors assigned to only a few of the available rating grades. The mere existence of a concentration of exposures in a rating grade (or rating grades) does not, by itself, reflect weakness in a rating system. For example, banks focused on a particular type of lending, such as asset-based lending, may lend to obligors having similar default risk. Banks with focused lending activities may use the minimum number of obligor rating grades, while banks with a broad range of lending activities should have more rating grades. However, banks with a high concentration of obligors in a particular rating grade should perform a thorough analysis that supports such a concentration. 33. A concentration of obligors in a rating grade is inappropriate when the financial strength of those obligors varies considerably. If such is the case, the following questions should be answered: • Are the criteria for each rating grade clear? Are rating criteria too vague to allow raters to make clear distinctions? Ambiguity may be an issue throughout the rating scale or it may be limited to the most commonly used ratings. • How diverse are the obligors? Is the bank targeting a narrow segment of obligors with homogeneous risk characteristics? • Are the bank's internal rating categories considerably broader than those of other lenders? Recognition of Implied Support S 2-11 Banks may recognize implied support as a rating criterion subject to specific supervisory considerations; however, banks should not rely upon the possibility of U.S. government financial assistance, except for the financial assistance that the U.S. government has legally committed to provide. 34. Implied support is support from a third party that is less than a legally enforceable guarantee. Banks that use implied support as a ratings criterion typically rely on a wide range of policies and procedures for its use. As the impact of implied support arrangements has typically been difficult to quantify, the circumstances under which banks use such arrangements as a ratings criterion should be limited. 35. Supervisors will assess the appropriateness of a bank's usage of implied support as a ratings criterion. A bank should recognize implied support only if the following are true: • The support is from a parent corporation or sovereign; however, banks should not rely upon the possibility of U.S. government financial assistance, except for the financial assistance that the U.S. government has legally committed to provide; • The implied support provider is rated investment grade by an NRSRO; • The implied support is a factor only in assigning an obligor rating, not a loss severity rating; • The final rating assigned to the obligor reflects greater credit risk than the rating assigned to the implied support provider (the parent corporation or sovereign); • The bank has considered the magnitude of the rating benefit accorded from the recognition of implied support and the bank has performed and documented comprehensive due diligence to assess the parent corporation or sovereign's willingness and capacity to support the obligor. To assess the willingness to support the obligor, a bank may consider prior situations where the support provider has supported the obligor or other obligors under similar circumstances, extended credit to the obligor at beneficial rates, or made large scale investments of cash or resources in the obligor. To assess capacity, a bank should conduct a thorough analysis of the financial position of the support provider and its ability to provide support including during periods of financial stress; • There is broad market recognition of the implied support. This can be evidenced through a number of market indicators including situations where the external ratings of the parent corporation and subsidiary are closely linked or the ratings of the parent or sovereign reflect an expectation of support. It could also include evidence derived from traded credit spreads of the parent and subsidiary; • For a bank whose rating system design incorporates external ratings as a tool in assigning an internal rating, the internal rating does not additionally incorporate implied support when there is evidence that the external rating has already benefited from the assumption of support; • The bank has established a stand-alone rating for the obligor and continues to monitor the stand-alone rating throughout the term of the exposure; • The bank's internal tracking processes monitor the dollar volume of credit exposures where implied support is a material consideration in the rating assignment; and • The provision of significant implied support to a subsidiary or subsidiaries is incorporated into the parent corporation's obligor rating. Loss Severity Ratings S 2-12 Banks must have a loss severity rating system that is able to assign loss severity estimates (ELGD and LGD) to each wholesale exposure. 36. The term loss severity rating system refers to the method by which a bank assigns loss severity estimates to wholesale exposures. This assignment can be accomplished through a loss severity rating process or via direct assignment to each wholesale exposure. A wholesale exposure's ELGD and LGD estimates are expressed as a percentage of the estimated EAD of the exposure. Both the ELGD and the LGD are required inputs into the IRB risk-based capital formulas. S 2-13 Banks should have empirical support for their loss severity rating system and the rating system should be capable of supporting the quantification of ELGD estimates (and LGD estimates if approved for internal estimates). 37. ELGD and LGD analysis is in the early stages of development compared to default risk modeling. Over time, banks' methodologies are expected to evolve. Longstanding banking experience and existing research on ELGD and LGD, while preliminary, suggests that type of collateral (in terms of liquidity and marketability), collateral values, seniority, industry position and whether an exposure is secured or unsecured are the most commonly used predictors of loss severity. 38. Whether a bank assigns ELGD and LGD values directly or, alternatively, rates wholesale exposures and then quantifies ELGD and LGD for the rating grades, the bank should conscientiously identify characteristics that influence ELGD and LGD. Each of the loss severity rating categories should be associated with empirically supported ELGD and LGD estimates. (Even though the grouped exposures have common characteristics and a common expected ELGD and LGD, realized loss severity for individual exposures may vary). Loss Severity Rating/LGD Granularity S 2-14 Banks must have a sufficiently granular loss severity rating system to group exposures with similar estimated loss severities or a process that assigns estimated ELGDs and LGDs to individual exposures. 39. While there is no stated minimum number of loss severity ratings, the systems that provide ELGD and LGD estimates must be granular enough to separate wholesale exposures with significantly varying estimated LGDs. For example, a bank using a loss severity rating-scale approach that has credit products with a variety of collateral packages or financing structures should have more ELGD and LGD rating grades than those banks with fewer options in their credit products. 40. Like obligor rating grades, the mere existence of an exposure concentration in an ELGD or LGD rating grade (or rating grades) does not, by itself, signify a rating system's weakness. However, banks with a high concentration within ELGD and LGD rating grades should perform a thorough analysis that supports such a concentration. B. Other Considerations Rating Criteria S 2-15 Rating criteria should be written, clear, consistently applied, and include the specific qualitative and quantitative factors used in assigning ratings. 41. Each obligor and loss severity rating (including ratings with modifiers such as + or −) should be defined. The definitions should describe all significant quantitative and qualitative ratings criteria used to promote consistent application of risk ratings. The ratings should be sufficiently transparent to allow replication by a third party. This is particularly important in expert-judgment rating systems where establishing the transparency of rating assignments is more challenging. Without clearly defined rating criteria, expert-judgment rating systems are not sufficiently transparent. A risk rating system with vague criteria or one defined only by PDs, ELGDs, or LGDs is neither replicable nor transparent. Transparent criteria promote accurate and consistent ratings within and across business lines and geographies, and permit the rating process to be refined over time. Use of External Rating Tools 42. Banks may use results from external rating tools, such as vendor default models or agency ratings, as inputs into their internal rating processes for obligors and wholesale exposures. The validation standards in this guidance apply to a bank's use of external rating tools as well as internal ones. Therefore, banks should apply the same level of rigor to their external tools as to their internal tools. In addition, any external rating tool employed should be consistent with the architecture of the bank's IRB rating systems. To verify this consistency, a bank should analyze and understand: • The predictive ability of the external rating tool; • The factors and criteria used by the external rating tools to assign ratings; and • The expected effect of using the external rating tool on the migration of internal ratings. 43. Sole reliance on external rating tools is not appropriate. Every rating tool has limitations, and banks should have a process to ensure that accurate ratings are assigned despite such limitations. How much additional analysis is required will depend on the exposure's rating, relative size and complexity. Banks should maintain data on the critical factors underpinning an external rating tool's obligor or loss severity ratings (as the banks would for any rating assignment process). Timeliness of Ratings S 2-16 Risk ratings must be updated whenever new material information is received, but in no instance less than annually. 44. A bank should have a policy that ensures that obligor and loss severity ratings reflect current information. That policy should also specify minimum financial reporting and collateral valuation requirements. When loss severity ratings or estimates depend on collateral values or other factors that change periodically, that policy should take into account the need to update these factors. 45. Banks' policies may include an alternative timetable for updating ratings of exposures below a de minimis amount that the bank determines has no material impact on risk-based capital levels. For example, some banks use triggering events to prompt them to update their ratings on de minimis exposures rather than adhering to a specific timetable. Multiple Ratings Systems 46. A bank's complexity and sophistication, as well as the size and range of products offered, will affect the types and number of rating systems employed. However, each risk rating system should conform to the standards in this guidance, must be validated for accuracy and consistency, and should be used consistently. Validation exercises should produce evidence that the ratings have been applied consistently. Chapter 3: Retail Segmentation Systems Rule Requirements Part III, Section 22(b)(1): A bank must have an internal risk rating and segmentation system that accurately and reliably differentiates among degrees of credit risk for the bank's wholesale and retail exposures. Part III, Section 22(b)(3): For retail exposures, a bank must have a system that groups exposures into segments with homogeneous risk characteristics and assigns accurate and reliable PD, ELGD, and LGD estimates for each segment on a consistent basis. The bank's system must group retail exposures into the appropriate retail exposure subcategory and must group the retail exposures in each retail exposure subcategory into separate segments. The bank's system must identify all defaulted retail exposures and group them in segments by subcategories separate from non-defaulted retail exposures. Part III, Section 22(b)(5): The bank's retail exposure segmentation system must provide for the review and update (as appropriate) of assignments of retail exposures to segments whenever the bank receives new material information, but no less frequently than quarterly. I. Overview 1. This chapter describes the design and operation of an IRB retail segmentation system. An IRB retail segmentation system groups retail exposures into segments with homogeneous risk characteristics within each of the three retail exposure subcategories (residential mortgage exposures, qualifying revolving exposures (QRE), other retail exposures). Examples of segmentation techniques include the use of obligor (such as income and past credit performance) and exposure (such as product type and loan-to-value) characteristics; or grouping loans by similar estimated default rates and estimated loss severities. The segmentation system used for IRB will often differ from segmentation used for other purposes, such as for marketing and scorecards. The retail risk parameter estimates that determine risk-based capital requirements are assigned at the segment level. 2. The retail IRB framework provides banks substantial flexibility to use the retail segmentation that is most appropriate for their activities, subject to the following broad principles: • Differentiation of risk—Segmentation should provide meaningful differentiation of risk. Accordingly, in developing the segmentation system, banks should select risk drivers that separate risk distinctly and consistently over time. • Reliable risk characteristics—Segmentation uses borrower risk characteristics and loan-related risk characteristics that reliably differentiate a segment's risk from that of other segments and that perform consistently over time. • Consistency—The risk drivers used to segment exposures must be consistent with the predominant risk characteristics the bank uses to measure and manage credit risk. • Accuracy—The segmentation process should generate segments that separate exposures by realized performance. It should be designed so that actual long-run outcomes closely approximate the retail risk parameters estimated by the bank. 3. Defaulted retail exposures must be segmented separately from non-defaulted exposures. In addition, retail segments should not cross national jurisdictions unless the bank can demonstrate that the exposures in the different jurisdictions have homogeneous risk characteristics. II. Definition of Default S 3-1 Banks must use the IRB definition of default when identifying defaulted retail exposures. 4. For retail exposures, banks must use the following definition of default for its IRB system: A retail exposure of a bank is in default if: • The exposure is 180 days past due, in the case of a residential mortgage exposure or revolving exposure; • The exposure is 120 days past due, in the case of all other retail exposures; or • The bank has taken a full or partial charge-off or write-down of principal on the exposure for credit related reasons. 5. The exposure remains in default until the bank has reasonable assurance of repayment and performance for all contractual principal and interest payments on the exposure. 6. For retail exposures, the definition of default is applied to a particular exposure rather than to the obligor. That is, default by an obligor on one obligation would not require a bank to consider all other obligations of the same obligor in default. III. Retail Segmentation Architecture A. Criteria for Retail Segmentation S 3-2 Banks must first place exposures into one of the three retail exposure subcategories (residential mortgage, QRE, and other retail). Banks must then separate exposures into segments with homogeneous risk characteristics. S 3-3 A retail segmentation system must produce segments that accurately and reliably differentiate risk and produce accurate and reliable estimates of the risk parameters. 7. While banks have considerable flexibility in determining retail segments, they should consider factors affecting the risk characteristics of both borrowers and loans when determining segmentation criteria. Statistical modeling, expert judgment, or some combination of the two may determine the most relevant risk drivers. 8. Examples of acceptable approaches to segmentation include: • Segmenting exposures by common risk drivers that are relevant and material in determining the loss characteristics of a particular retail product. For example, a bank may segment mortgage loans by LTV band, age from origination, geography, and/or origination channel. • Segmenting exposures by common risk drivers that are relevant and material in determining the loss characteristics of a particular borrower population. For example, a bank may segment by credit bureau score bands, behavior score bands, and/or delinquency status. In the case of mortgage products, more borrower information may be available and a bank could include the debt-to-income ratio, current income, and/or years at present location. • Segmenting by grouping exposures with similar estimated loss characteristics, such as expected average loss rates, expected default rates, or expected loss severity rates. Some banks have developed models that rank order default risk or generate an estimated default rate, loss severity, and/or exposure at default for individual exposures. A bank could use such estimates as criteria in their segmentation system. 9. Each retail segment will have an estimated PD, ELGD, LGD, and EAD. In some cases, it may be reasonable to use the same risk parameter estimates for multiple segments. This may occur more frequently for bank estimates of ELGD and LGD as banks may have less robust historical data for estimating these IRB risk parameters. In such cases, the bank should demonstrate that there are no material differences in ELGD or LGD among those segments. Over time, supervisors expect banks to develop more precise data and methodologies for determining ELGD and LGD. 10. Data for certain retail loans are sometimes missing or incomplete, such as data for purchased loans or loans originated with policy exceptions. The overall segmentation system should adequately capture the risk associated with these loans based on the data available. In some cases, missing or incomplete data itself may be a significant risk factor used for segmentation purposes. 11. A bank should substantiate the degree of granularity in its segmentation system and the distribution of exposures across segments. (Here, “granularity” is how finely the portfolio is segmented.) 12. Banks have flexibility in determining the granularity of their segmentation system. Each bank should perform internal analysis to determine how granular segments must be to group homogeneous exposures. For example, a bank using credit score ranges to segment its portfolio should provide the rationale for the ranges chosen. 13. A concentration of exposures in a segment (or segments) does not, by itself, reflect a deficiency in the segmentation system. For example, a bank may lend within a narrow risk range and, therefore, have a smaller number of segments than a bank that lends across a wider spectrum of risk. However, a bank with a high concentration of exposures in a particular segment will be expected to show that the bank's segmentation criteria are carefully delineated and well-documented. The bank should be able to demonstrate that there is little risk differentiation among the exposures within the segment, and that the segmentation method produces reliable estimates for each of the risk parameters. A bank should not artificially group exposures into segments specifically to avoid the 10 percent LGD floor for mortgage products. A bank should use consistent risk drivers to determine its retail exposure segmentations and not artificially segment low LGD loans with higher LGD loans to avoid the floor. S 3-4 Banks should clearly define and document the criteria for assigning an exposure to a particular retail segment. 14. Banks should choose risk drivers that accurately reflect an exposure's risk. Risk drivers selected must be consistent with risk measures used for credit risk management. 15. The method of segmentation will help determine the risk parameters, as well as which techniques should be used for validation and which control mechanisms will best ensure the integrity of the segmentation system. Described below are some techniques for determining whether the segmentation was done appropriately: • Statistical Models—Banks may incorporate results of statistical underwriting models or scoring models directly into their segmentation process. For example, a bank may use a custom or bureau credit score as a segmenting criterion. In that case, the bank should support the choice of the score, and should demonstrate that it has adequate controls for the credit scoring system. • Inputs to Models—Banks may incorporate the variables from a statistical model into their segmentation processes. For example, a bank that uses a statistical model to predict losses for its mortgage portfolio could select some or all of the major inputs to that model, such as debt-to-income and LTV, as segmentation criteria. As part of its validation and controls for the segmentation system, the bank should provide an appropriate rationale and empirical evidence for its choice of the particular set of risk drivers from the loss prediction model. • Expert Judgment—Banks may combine expert judgment with statistical analysis in determining segmentation criteria. However, expert judgment must be well-documented and supported by empirical evidence demonstrating that the chosen risk factors are reliable predictors of risk. 16. A bank should be able to demonstrate a strong relationship between IRB risk drivers and comparable measures used for credit risk management. Specifically, a bank should demonstrate that the segmentation system differentiates credit risk across the portfolio and captures changes in the level and direction of credit risk using measures that are similar to those used in credit risk management. For example, even if a bank uses custom scores for underwriting or account management, generic bureau scores may be used for IRB segmentation purposes if the bank can demonstrate a relationship between these measures. 17. Banks should have clear policies to define the criteria for modifying the segmentation system. Changes in the segmentation system should be documented and supported to ensure consistency and historically comparable measurements. B. Assignment of Exposures to Retail Segments S 3-5 Banks should develop and document their policies to ensure that risk-driver information is sufficiently accurate and timely to track changes in underlying credit quality and that the updated information is used to assign exposures to appropriate segments. 18. Under the IRB framework, a bank initially assigns retail exposures to segments based on the risk-driver information available at the time of origination or acquisition. The bank should then continue to monitor the risk characteristics of the exposures and assign exposures to appropriate segments based on refreshed information gathered by the bank as part of its monitoring process. 19. In accordance with industry practices in retail credit risk management, a bank should have a well-documented policy on monitoring and updating information about exposure risk characteristics. The policy should specify the risk characteristics to be updated and the frequency of updates for each product type or sub-portfolio within its retail portfolio. Updating of relevant information on these risk drivers should be consistent with sound risk management. S 3-6 The bank's retail exposure segmentation system must provide for the review and update (as appropriate) of assignments of retail exposures to segments whenever the bank receives new material information, but no less frequently than quarterly. 20. Decisions regarding the frequency of obtaining refreshed information should reflect the specific risk characteristics of individual segments and/or the potential impact on risk-based capital levels. The frequency of updates will generally vary for different risk drivers and for different products. The underlying principle is that, in every estimation period, retail exposures are assigned to segments that accurately reflect their risk profile and produce accurate risk parameters. 21. Banks should assess their approach to updating information and migrating exposures when validating the segmentation process. Chapter 4: Quantification Rule Requirements Part III, Section 22(c)(1): The bank must have a comprehensive risk parameter quantification process that produces accurate, timely, and reliable estimates of the risk parameters for the bank's wholesale and retail exposures. Part III, Section 22(c)(2): Data used to estimate the risk parameters must be relevant to the bank's actual wholesale and retail exposures, and of sufficient quality to support the determination of risk-based capital requirements for the exposures. Part III, Section 22(c)(3): The bank's risk parameter quantification process must produce conservative risk parameter estimates where the bank has limited relevant data, and any adjustments that are part of the quantification process must not result in a pattern of bias toward lower risk parameter estimates. Part III, Section 22(c)(4): PD estimates for wholesale and retail exposures must be based on at least 5 years of default data. ELGD and LGD estimates for wholesale exposures must be based on at least 7 years of loss severity data, and ELGD and LGD estimates for retail exposures must be based on at least 5áyears of loss severity data. EAD estimates for wholesale exposures must be based on at least 7 years of exposure amount data, and EAD estimates for retail exposures must be based on at least 5 years of exposure amount data. Part III, Section 22(c)(5): Default, loss severity, and exposure amount data must include periods of economic downturn conditions, or the bank must adjust its estimates of risk parameters to compensate for the lack of data from periods of economic downturn conditions. Part III, Section 22(c)(6): The bank's PD, ELGD, LGD, and EAD estimates must be based on the definition of default [in the NPR]. Part III, Section 22(c)(7): The bank must review and update (as appropriate) its risk parameters and its risk parameter quantification process at least annually. Part III, Section 22(c)(8): The bank must at least annually conduct a comprehensive review and analysis of reference data to determine relevance of reference data to bank exposures, quality of reference data to support PD, ELGD, LGD, and EAD estimates, and consistency of reference data to the definition of default contained [in the NPR]. I. Overview 1. Quantification is the process of assigning numerical values to the key risk parameters that are used as inputs to the IRB risk-based capital formulas. This chapter provides guidance on the quantification process for wholesale and retail exposures. For both wholesale and retail portfolios these risk parameters are the probability of default (“PD”), expected loss given default (“ELGD”), loss given default (“LGD”), and exposure at default (“EAD”). Wholesale exposures also require determination of the exposure's maturity (“M”). Risk parameters are assigned to each exposure for wholesale portfolios and to each segment for retail portfolios. Specific quantification issues related to counterparty credit risk transactions, equity exposures, and securitization exposures are described in Chapters 9, 10, and 11, respectively. 2. In any discussions of the IRB system, the risk rating or segmentation system design and the quantification process should be considered together. This chapter focuses on quantification given an existing risk rating or segmentation system design, as covered in Chapters 2 and 3, respectively. 3. Section I establishes an organizing framework for considering quantification and develops general standards that apply to the entire process. Sections II, III, and IV cover specific supervisory standards that apply to PD, ELGD and LGD, and EAD respectively. The maturity risk parameter receives somewhat different treatment in section V, since it is much less dependent on statistical estimates from historical data. Special cases and applications for quantification are covered in section VI. A. Stages of the Quantification Process 4. For each risk parameter, quantification may be broken down into four stages: obtaining historical reference data; estimating the relationship between risk characteristics and the risk parameters in the reference data; mapping the correspondence between risk characteristics in the reference data and those in the existing portfolio; and applying the relationship between risk characteristics and risk parameters to the existing portfolio. An evaluation of a bank's quantification process focuses on the overall adequacy of the bank's approach, including an understanding of how the bank breaks down the quantification process where applicable into the four stages. 5. Banks are not required to separate the quantification process into four stages. The four stages are a conceptual framework, and may serve as a useful analytical and implementation guide. Readers may find it helpful to refer to the appendices to this chapter, which illustrate how this four-stage framework can be applied to quantification approaches in practice. The four stages of quantification are described below. Data—First, the bank constructs a reference data set, or source of data, from which risk parameters can be estimated. A “reference data set” consists of a set of exposures and their associated identifying information and risk characteristics. Reference data sets may include internal data, external data, or pooled data from different internal and external sources. Internal data refers to any data on exposures held in a bank's existing or historical portfolios, including data elements or information provided by third parties (e.g., data from a credit bureau about one's own customers would be considered internal data). External data refers to information on exposures held outside the bank's portfolio, including aggregate industry trends or economic data. The reference data is described using a set of observed characteristics; consequently, the data set contains variables that can be used for this characterization. For example, risk characteristics for wholesale exposures include obligor and exposure characteristics related to the risk parameters, such as agency debt ratings, risk ratings, financial measures, geographic regions, and the economic environment and industry/sector trends during the time period of the reference data. Risk characteristics for retail exposures include borrower and loan characteristics, such as loan terms, loan-to-value, credit score, income, debt-to-income, or payment history. A bank may use more than one reference data set to improve the robustness or accuracy of the risk parameter estimates. Estimation—Second, the bank applies statistical techniques to the reference data to determine the relationship between risk characteristics and the estimated risk parameter. The result of this step is a model that ties descriptive risk characteristics, or drivers, to the risk parameter estimates. In this context, the term “model” is used in the most general sense; a model may be a simple calculation of historical averages or a more sophisticated approach based on advanced statistical techniques (e.g., regression). This step may include adjustments for differences between the IRB definition of default and the default definition in the reference data set, as well as adjustments for data limitations. More than one estimation technique may be used to generate estimates of the risk parameters, especially if there are multiple sets of reference data or multiple sample periods. If multiple estimates are generated, the bank should have a clear and consistent policy for reconciling and combining them into a single estimate at the application stage. Mapping—Third, the bank creates a link between its portfolio data and the reference data based on corresponding characteristics. Variables or characteristics used in the estimation model are mapped, or linked, to the variables that are available for the existing portfolio. In order to map effectively, a bank should have reference data characteristics that allow the construction of rating and segmentation criteria that are consistent with those used on the bank's portfolio. An important element of mapping is making adjustments for differences between reference data sets and the bank's exposures. The bank should map each reference data set and each combination of risk characteristics used in any estimation model. Application—Fourth, the bank applies the relationship estimated for the reference data to the actual portfolio data. The ultimate aim of quantification is to attribute a PD, ELGD, LGD, and EAD to each exposure within the wholesale portfolio and to each segment of exposures in the retail portfolio. If multiple data sets or estimation methods are used, the bank should adopt a means of combining the various estimates at this stage. For wholesale portfolios, this step may include adjustments to default rates or loss rates to “smooth” the final risk parameter estimates. If the estimates are applied to individual transactions, the bank must in some way aggregate the estimates at the rating level. For retail portfolios, the bank may simply apply the risk parameter estimates derived for each segment to the corresponding segment in the existing portfolio. However the application stage could be more complex if multiple data sets or estimation methods were used or if the mapping stage required adjustments. 6. The four-stage quantification process described above outlines a framework that a bank may use for assigning numerical values to the IRB key risk parameters. Whether the quantification process explicitly delineates each aspect of the four stages of quantification for PD, ELGD, LGD, and EAD, or the quantification process is more integrated, each aspect of the quantification process for the key risk parameters should be justified, documented, and subject to monitoring and follow-up. 7. A number of examples are given in this chapter to aid exposition and interpretation of specific quantification issues. None of the examples is sufficiently detailed to incorporate all of the considerations discussed in this chapter. Moreover, technical progress in the area of quantification is rapid. Thus, banks should not interpret a specific example that is consistent with the standard being discussed, and that resembles the bank's current practice, as being a “safe harbor.” Banks should consider this guidance in its entirety when determining whether systems and practices are adequate. B. General Standards for Sound Quantification 8. Several core principles apply to the overall quantification process of risk rating and segmentation systems. Those principles and the general standards that reflect them are discussed in this introductory section. Other supervisory standards specific to particular stages or risk parameters are discussed in later sections. 9. The risk parameters should be estimated in a manner consistent with sound credit risk management practices and the IRB standards. In addition, a bank should have processes to ensure that these estimates are independently and thoroughly validated and the results reported to senior management. 10. Supervisory evaluation of the quantification process requires consideration of all the standards in this chapter, both general and specific. Particular practical approaches to quantification may be highly consistent with some standards, and less so with others. In assessing a bank's approach, supervisors will weigh the approach's strengths and weaknesses using all the supervisory standards in this chapter as a guide. S 4-1 Banks should have a fully specified process covering all aspects of quantification (reference data, estimation, mapping, and application). The quantification process should be fully documented. 11. A fully specified quantification process should describe how all four stages (data, estimation, mapping, and application) are addressed for each parameter. The linkages between the bank's quantification and validation processes should also be explicit. 12. An important aspect of the quantification process is the appropriate capture and analysis of developmental evidence in support of techniques applied by the bank. A few examples of such developmental evidence are: • For reference data—a discussion of how the best available data are chosen from various sources so that the data include periods of economic downturn conditions and the portfolio in the reference data is comparable to the existing portfolio; • For estimation—discussions of why the bank uses various averaging methods on historical data, how it specifies downturn estimates, or how it develops predictive models; • For mapping—discussions of how risk characteristics in the reference data compare with those in the existing portfolio; and • For application—a discussion of the combination of multiple estimates, aggregations of estimates across exposures, or any judgmental adjustments. 13. Major decisions in the design and implementation of the quantification process should be justified and fully documented. Documentation promotes consistency and allows third parties to review and replicate the entire process. S 4-2 Risk parameter estimates must be based on the IRB definition of [Text truncated at 120,000 characters. The full text is on the page linked above.] --- Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/fr%3A07-811. Public record. Not legal advice.