Memorandum

To: All UBPR Users	Date: November 18 , 2003
From: Task Force on Surveillance Systems	Subject: Changes to UBPR Peer Groups Effective with March 31, 2004 UBPR

Summary:

The FFIEC Task Force on Surveillance Systems will implement two changes to the peer group structure used in the UBPR effective with the March 31, 2004 version of the report. The first combines several asset-based commercial bank peer groups. The second replaces the current de novo bank peer group with three "class-of" peer groups.

Background:

Peer group definitions used in the UBPR have evolved over time. Originally, 25 asset-based peer groups were used for all banks. As the financial services industry changed, the need for greater definition in the peer group structure became apparent and further business line groupings were added.

The current peer group structure is multi-tiered. Financial institutions are first segregated into one of four broad groups. These groups include commercial banks, FDIC-insured savings banks, credit card specialty banks, and bankers banks. Within the first three groups, institutions are then subdivided by asset size. The commercial bank group is further divided based on location (urban or rural determined by whether the bank is located in a Metropolitan Statistical Area) and number of branches. For the most part, the asset based peer groups have produced reasonable benchmarks to gauge bank performance. In addition, analysis conducted by the task force shows that key financial measures for overhead, margins, capitalization, and profitability continue to differ depending on bank location (urban or rural) and number of branches.

However, the asset ranges used to define commercial peer groups have remained substantially unchanged since their inception 20 years ago. Additionally, the number of banking institutions has declined significantly. As a result, some current peer groups now contain less than 100 institutions. In some instances the resulting peer group statistics have proven to be less reliable because of the impact a few institutions can have on averaged ratios. The planned consolidation of peer groups is designed to address these distortions.

The planned changes will only affect commercial bank peer groups. The number of commercial bank asset-based peer groups will decline from 24 to 15. For example, banks in current peer groups 1 (assets > $10B) and 2 (assets between $3B and $10B) will be combined into new peer group 1 (assets > $3B). Current peer groups 15 through 24, which have a variety of asset ranges below $50M, will be combined into four new peer groups that divide banks with less than $50M in assets based on their location and number of branches. Changes to commercial bank asset-based peer groups will be made retroactively for all quarters.

Additionally, a more fundamental change to the way peer group statistics are calculated for de novo banks will be implemented. Previously all newly chartered banks were placed in peer group 25 for three years or until they reached $50M assets. A "class-of" structure for de-novo banks that would group banks by the year of their opening. De novo banks would be compared to their "class-of peers" for five years, or until attaining $750M assets. After 5 years, these banks would be moved to the appropriate asset based peer group. Initially the class-of structure will span three years, i.e. banks opened in 2001, 2002 and 2003. Additional classes will be added until five separate de novo peer groups are attained.

See the spread sheet for a comparison of current and revised peer groups.

Impact on Bank Analysis:

For many groups of banks, the task force does not expect the changes to have a significant effect. For example, Net Income to Average Assets (ROA) for banks in current peer group 15 is .76% while comparable new peer group 12 will have an average ROA of .71%. For other groups, the changes will be more noticeable but will generally improve the value of computed benchmarks. For example, the peer group average for Non Interest Income to Average Assets for banks in current peer group 21 is 3.71% while banks in comparable new peer group 14 will be compared to a more reasonable peer average of 1.43%. It should be noted that the task force will continue producing peer group distribution reports to assist in evaluating any significant changes in peer ratios and bank rankings that arise from the new peer groups (available on the FFIEC UBPR website).

For de novo banks, the task force believes that the proposed changes will provide a marked improvement in usefulness of the benchmark statistics. Under the current structure, the peer group averages reflect a static composite bank that is 1.5 years old, with loss operations and a high capital ratio. The new structure will compare banks that are at similar maturity levels. Basing the comparisons of banks within the de novo categories on the degree to which they have become "seasoned" and overcome the initial expenses of opening should strengthen the UBPR's identification of outliers.

Questions or comments regarding the upcoming changes may be addressed to John Smullen by email at jsmullen@fdic.gov or by phone at 703 516-5732.