Since 1994, the Federal Reserve has used two econometric models jointly known as SEER or the System to Estimate Examination Ratings to monitor the condition and performance of state member banks. These surveillance models have played an important role in the Federal Reserve's supervisory process, supporting risk-focused examinations by identifying changes in the financial condition of banks that warrant examiner review. In the time since the System implemented these models, however, new banking laws and regulations, financial innovation, and changing business practices have substantially affected the activities of U.S. banks. In addition, supervisory practices have evolved as the Federal Reserve has refined its risk-focused examination approach and increased its emphasis on the evaluation of risk identification and control processes at banking organizations.
Accordingly, in 2004 the Federal Reserve formed a working group to develop and recommend options for enhancing the econometric models used in the System's off-site surveillance programs. The group's primary objective was to develop a modeling framework that improved upon the performance of the SEER models in identifying weak and potentially weak banks and that considered the effect of nontraditional activities on banking risk. Other important areas of focus included designing supporting reports to clearly link model results to the primary areas of examination focus and developing training materials to ensure the most effective possible use of off-site model results. Based on the recommendations of this group, the Federal Reserve is now replacing the SEER models with a new econometric framework referred to as the Supervision and Regulation Statistical Assessment of Bank Risk model or SR-SABR. This change is being implemented using December 31, 2005 Call Report data.
This SR letter briefly describes the new model, details the screening thresholds for SR-SABR within the State Member Bank Watch List program, and updates Watch List follow-up procedures. Most of the procedures adopted when the Watch List Program was implemented in 2000 will remain in place and are restated below. Some Watch List procedures have been clarified to reflect experience gained with the Watch List process since 2000.
The SR-SABR model assigns a two-component surveillance rating to each bank. The first component is the current composite CAMELS rating assigned to the bank. The second component is a letter (A, B, C, D, or F), reflecting the model's assessment of the relative strength or weakness of a bank compared to other institutions within the same CAMELS rating category1. An SR-SABR rating that includes an "A" denotes a bank with particularly strong financial and supervisory indicators compared to other banks within its CAMELS rating category. An SR-SABR rating including an "F" indicates that a bank is reporting poor financial results or showing other signs of significant weakness compared to similarly rated banks. For example, a 1A rating signifies a 1-rated bank that reports strong financial and supervisory indicators when compared to all 1 and 2 rated banks, while a 1F indicates that, while the bank currently maintains the strongest possible composite CAMELS rating, its financial or other supervisory indicators place it among the weakest of the banks currently rated either 1 or 2. SR-SABR ratings that include a "B" generally correspond to banks with financial and supervisory measures that are comparable to most banks in the CAMELS rating category. Those with a "C" have weaker measures than those of most other banks in their CAMELS rating category and those with a "D" have significantly weaker financial or supervisory measures compared to other banks in their rating category.
Three separate econometric models contribute to SR-SABR surveillance ratings. Two of the models estimate the probability of an adverse supervisory rating change for a bank if it was examined within the next quarter. The first estimates the probability of an adverse rating change for banks currently rated CAMELS 1 or 2. The second estimates the probability of an adverse rating change for banks currently rated 3, 4, or 52. Together these models are used to assign an "adverse change" rating. They utilize seven financial variables computed using Call Report data and seven supervisory variables that have been statistically significant in explaining adverse ratings assigned over the past three years. The third model is retained from the SEER framework and estimates the probability that a bank will fail or become critically undercapitalized within the next two years. This model is referred to as the "viability" model and includes 11 financial variables computed using Call Report data. It was estimated based on financial results for the large group of banks that failed in the late 1980s and early 1990s.
For further information on SR-SABR, examiners and analysts should contact the designated surveillance contact at their respective Reserve Bank. The Surveillance Resources page located on the Board's Division of Banking Supervision and Regulation intranet website also includes more detailed background information on SR-SABR.
The Watch List program applies to all state member banks and includes both state member banks with known weaknesses and those with characteristics that could affect supervisory assessments of the quality of bank management or of the overall safety and soundness of a bank. It helps to ensure that weaknesses existing at supervised banks are being addressed appropriately and that potential emerging problems can be promptly identified in between regularly scheduled on-site safety and soundness examinations. With adoption of the SR-SABR model, state member banks are included on a Watch List and require quarterly written analyses when they meet any of the following criteria:
Reserve Banks and Board staff may also add state member banks to the Watch List for reasons other than those listed above. For example, they may elect to include selected de novo banks, banks reporting rapid asset or loan growth or significant changes in business mix, and other institutions with financial characteristics that suggest the need for heightened off-site monitoring in between on-site examinations.
Board staff will distribute a preliminary quarterly Watch List to Surveillance Contacts at each Reserve Bank upon the finalization of quarterly Call Report processing. To assist examiners and analysts in interpreting SR-SABR model results, Board staff will also distribute SR-SABR Schedule of Risk Factors Reports (SRFs). The SRFs highlight financial ratios which cause the model to flag a bank as particularly strong or weak. These reports also include peer statistics to highlight the relative position of a bank compared to other institutions with similar CAMELS composite ratings. In addition, supplemental monitoring screens will be distributed to assist in analyzing Watch List banks and in identifying other banks that may require additional supervisory attention.
Upon notification from Board staff that quarterly surveillance materials are ready for review, Reserve Banks should perform the following procedures:
Questions about these procedures should be directed to Kevin Bertsch, Assistant Director, Surveillance, Financial Trends, and Analysis Section, at (202) 452-5265.