![]() |
![]() |
![]() |
![]() |
![]() |
Home > Regulation & Examinations > Bank Examinations > Side by Side: A Guide to Fair Lending |
![]() |
|
![]() |
![]() |
Side by Side: A Guide to Fair Lending |
COMPARING THE FILES
Creating a Comparison Spreadsheet Identifying Exceptions to Policy or Standard Practice Identifying Unequal Treatment Several basic steps can facilitate an orderly analysis of the sample of applications created. These include:
Creating a Comparison Spreadsheet After creating the sample of loan applications, place them on a spreadsheet. Appendix D to this guide contains an example of one such spreadsheet - "Mortgage Lending Decision Analysis (MLDA) Worksheet." Using a spreadsheet will help to determine if the reasons for an applicant's denial are consistent with the institution's lending policies and practices and whether policies and practices are consistently applied without regard to any prohibited bases. Appendix D lists additional items that a lender might include on the spreadsheet to more adequately determine unfavorable treatment of target group applicants. Other factors can be added as a lender may deem necessary. In particular, a lender may want to include such items as assistance in completing an application or cross-selling more suitable loan products or other services. These and other factors could indicate whether the quality of assistance offered to target applicants and control applicants differ in any way. Identifying Exceptions to Policy or Standard Practice Review all applicant profiles to identify any exceptions to loan policies for each item on the sample spreadsheet such as minimum and maximum mortgage loan amounts, analytical ratios and other decision factors used by the institution. In addition, any exceptions made to articulated policy standards or common practices regarding the nature and length of employment, credit history, underwriting and appraisal standards, or any other loan decision criteria should be identified. It also should be determined whether there is any pattern to the exceptions such that the exceptions may have become a standard practice or policy. The likely effect this may have on protected groups should be evaluated. The control group should be compared with the target group to establish equal or different treatment in the terms and conditions of loans (e.g., the length of the amortization periods, interest rates, special fees, or higher fees on smaller loans) and the quality of assistance offered to an applicant. Terms and Conditions. Each key ratio should be compared for target applicants and control applicants to discern individual differences in treatment and possible patterns. Specific geographic areas should be reviewed for differential treatment as this may indicate "redlining". When comparing target group applicants to control group applicants, it is not necessary for all the applicant characteristics to be similar. The focus should be on similar "flaws" in applicant characteristics that may have resulted in the denial of a target group applicant but not a control group applicant. For example, if target group applicants appear to be denied credit based on debt-to-income ratios that exceed the institution's normal underwriting standards, you should look for approved control group applicants with debt-to-income ratios similar to those of denied applicants. Other inconsistencies may include:
Quality of Assistance. It is important to remember that the risk of disparate treatment is higher in the treatment of applicants who are either "marginally" qualified or "marginally" unqualified. Here there is more room for lender discretion and "valid" reasons for denying loans to minority applicants may not have been applied in the cases of non-minority applicants with similar characteristics. In addition, offsetting reasons for approving a loan to a marginally qualified non-minority applicant could involve information the lender may usually fail to consider for minority applicants. For example, the lender may request an explanation for derogatory credit report information from a white applicant but fail to do so for a minority applicant, whose qualifications otherwise fall within the institution's underwriting guidelines, with derogatory credit information. A pattern of inconsistencies should raise questions that require a review of lending policies and discussions with loan staff. For example, if "insufficient income" is cited often as the reason for denial of applicants from a prohibited basis group, one might ask for the institution's definition of income; whether all of an applicant's income is considered; whether the loan policies are too restrictive; or whether more restrictive practices exist in the market area. Another example would involve whether there is a possible pattern of "under-appraisals" for members of the target group. On the spreadsheet include a category for "Loan-to-Selling Price Ratio" to help determine if the loan-to-price ratio is consistently less than the loan-to-value ratios for members of the target group or for properties in some of the institution's lending areas. These ratios can indicate possible under-appraisals, which may be a problem for protected groups. An institution should then reconsider its appraisal review process and discuss the problem with staff or fee appraisers. In addition, foreclosure files, collection files, and institution owned real estate files should be reviewed individually or, if there is a large enough number, a sample may be reviewed. Exceptions found in the sample review process should be discussed with loan department personnel to establish any additional information that may have a bearing on a particular application's decision. Each file where an inconsistency is present should be reviewed to determine whether any practices have a discriminatory effect on minorities, women, or other protected groups. If no reasonable explanation for the inconsistencies can be determined, then a violation may exist. Where denials or more restrictive terms correlate with particular census tracts or neighborhoods and, again, no reasonable explanation for the inconsistencies can be determined, then unlawful redlining also may be a possibility. |
Last Updated 07/28/1999 | supervision@fdic.gov |