Seal of the Board of Governors of the Federal Reserve System BOARD OF GOVERNORS

WASHINGTON, D. C.  20551
SR 02-19
October 29, 2002


SUBJECT:   Use of Statistical Sampling in the Review of Commercial and Industrial Loans and Commercial Real Estate Loans during On-Site Safety and Soundness Examinations of Community Banks

                      This letter introduces a statistical sampling method to improve the comprehensiveness and effectiveness of credit reviews for community bank examinations.  By supplementing the traditional credit review of problem, watch, insider and other loans with a statistical sampling technique, examiners may gain a broader and better understanding of a bank's credit quality and risk management practices.  This in turn may provide examiners with a basis for determining whether a community bank's internal classifications of commercial and industrial and commercial real estate credits can be relied upon in determining the extent and severity of regulatory classifications in a safety and soundness examination. 

                      The sampling methodology is designed to assess the accuracy of the bank's internal classifications.  If the findings of the statistical sampling reveal no material differences between a bank's internal classifications and those of the examiner, the bank's internal classifications, as augmented by examination findings, may be considered in determining regulatory classifications.  If the findings of the statistical sampling result in material differences between the classifications assigned by the bank and examiner or other concerns are noted such as control weaknesses, then further credit review and transaction testing should be performed to evaluate the full extent of problem assets.  That expanded review should be consistent with the minimum loan coverage required for banks posing supervisory concerns as set forth in SR letter 94-13, entitled Loan Review Requirements for On-site Examinations.

                      This letter also clarifies the Federal Reserve's policy on loan review coverage in community banks.  Going forward, loan reviews conducted as part of full-scope Federal Reserve examinations are expected to comply with the 1994 guidance or the guidance on the loan sampling process outlined in the attachment to this letter.  Credit reviews are a primary means for examiners to evaluate both the quality of a bank's loan portfolio and the effectiveness of credit risk management practices, both of which must be considered in assigning an asset quality rating.  The level, distribution, severity, and trend of regulatory classifications for both on- and off-balance sheet exposures are important factors, as are the adequacy of underwriting standards, internal controls, the soundness of credit administration practices, the reliability and integrity of risk identification and review practices, the adequacy of the allowance for loan and lease losses, and the extent of asset concentrations.  (See SR letter 96-38, Uniform Financial Institution Rating System, for further detail and discussion of the factors to be considered in assigning an asset quality rating.)

                      The sampling-based procedures, outlined in the attached guidance, are generally appropriate for state member community banks currently rated CAMELS composite and asset quality "1" or "2" and with assets of $1 billion or less.  The statistical sampling approach is not recommended for use at de novo banks or other banks with unusually high or low capital ratios. Reserve Banks wishing to experiment with a sampling approach at de novo banks or at larger or troubled institutions are instructed to contact Board staff.

                      After completion of an examination using the statistical loan sampling process, Reserve Banks should:  (i) record the bank's classification information into the loan sampling software and forward to Board staff a feedback form summarizing the results; and (ii) record in the confidential section of the examination report the percent of commercial loans and leases reviewed, the number of sampled borrowers that were incorrectly classified by the bank, the weighted classified asset ratio, and the extrapolated weighted classified asset ratio.  This information will help in the continued development and refinement of loan sampling in the credit review process. 

                      Each Reserve Bank is asked to distribute this guidance to supervisory staff responsible for community bank supervision.  In support of this process, each Reserve Bank has designated an individual to become familiar with the loan sampling software and technical requirements and to assist other examiners in its use.

                      Questions regarding sampling procedures should be directed to Robert Walker, Senior Supervisory Financial Analyst, at 202-452-3429 or David Elkes, Senior Supervisory Financial Analyst, at 202-452-5218.

Richard Spillenkothen


Attachment (746 KB PDF)

SR letters 94-13 and 96-38
Commercial Bank Examination Manual, Section 2080.1

SR letters | 2002