Update On Fair Lending Exams

At ABA's National Graduate School for Compliance Management, Bobbie Jean Norris, FDIC's Section Chief - Fair Lending and Outreach Branch, updated the >
Reviewing Denials
Early fair lending examinations involved a review of denials to determine that they looked fair and reasonable. Fair lending examinations then moved to a comparison of denials to approvals. Now, the FDIC examiners compare approved files to approved files, controlling for a specific prohibited basis and lending practice.

It can still be useful to review denials. Norris described how examiners can quickly review denials to identify marginally qualified applicants and discard clearly unqualified applications. Using a similar technique among approvals, the examiner can produce a set of applications from marginally qualified applicants to compare treatment.

Looking at Approvals
Norris stated that the examination emphasis has moved away from denials and underwriting into specific lending practices. The FDIC has referred 35 cases to the Department of Justice ("DOJ"). Of these cases, none were based on an analysis of denials. Most of these cases involved discriminatory rates and terms, and were developed based on comparison of approved files to approved files, controlling for prohibited bases.

Norris noted that, in spite of these agency findings, the DOJ is still looking for a case involving underwriting decisions based on denials. Norris suggested that banks should continue to monitor their denial ratios and audit denied files to identify any problem trends to take corrective or preventative action before an examination.

Comparative File Analysis
The most important part of the fair lending examination is comparative file analysis. Selecting files to compare is not a straightforward procedure. The examination procedure begins with a set of minority files and an equal number of "control" files.

The number of control files used in the examination depends on the nature of the investigation and what type of practices are being reviewed. It may be useful to pull up to four times as many control files. Beginning with this large a sample increases the likelihood of finding similarly situated files. It also strengthens the conclusion if no similarly situated files are found.

Norris recommends using four times as many control files as minority files. She finds this technique to be more efficient and effective than gradually expanding the sample. She stated that, if after reviewing four times as many control files as minority files, no discrimination is revealed, then the examination should be concluded.

The referrals
Of the 35 cases that FDIC has referred to DOJ, 14 were based on spousal signatures. The remaining 21 involve a combination of race, gender, marital status, ethnicity, and source of income.

Almost all of the referralsinvolve consumer loans, not mortgage loans. Norris believes that the industry is placing emphasis on mortgage lending discrimination at the expense of giving adequate attention to consumer lending.

Most of these recent referrals were based on pricing. Norris shared several examples showing single women being charged several points more than men for similar loans.

Indirect lending
FDIC is currently investigating a case involving indirect loans. In this case, the examiner identified a practice of a dealer charging higher rates to single female customers. The bank's defense that only the dealer, not the bank, benefits from the higher rate does not relieve the bank of liability. The bank is responsible, Norris stated.

This investigation also clarifies the that the bank has an obligation to observe patterns from dealer-originated loans. Even though indirect loans arrive one at a time, the bank must observe the pattern and as the pattern emerges, becomes liable for the discrimination if it continues to purchase loans from that dealer.

Policies
All banks should review their loan policies frequently. The FDIC has examined banks that have not reviewed their policies in five years or more. Norris recommends that policies be reviewed every year or two. Policies should be updated to reflect any changes in bank products, the community, and interpretation of fair lending laws.

Most important
Asked what is the most important element of a fair lending program, Norris responded: "believe that it can happen to you." Too many bankers believe that it can't happen in their bank. It can.

ACTION STEPS

Review your lending policies - all of them - for fair lending elements and set a schedule for getting revised policies to the board.
Schedule an audit of consumer loans. Use the techniques discussed in this article to identify files for review and comparison.
If you purchase loans from dealers, analyze files separately for each dealer. Then compare the types of deals made by dealers. Look for any dealer charging higher rates or different terms based on gender, age, or ethnic surname.
Send a fair lending memo to each dealer reminding them of their obligations under the law. State that your bank will not purchase loans from a dealer that may be discriminating.