Consumer loans pose a great deal of risk for financial institutions. In this post, we’ll examine that risk to lenders.
Consumer Loans & Fair Lending
Regulatory reviews typically focus on HMDA reportable loans and applications. This is because prohibited bases information (also know as Government Monitoring Information or GMI) is collected for these types of applications and not on other types of applications such as auto and unsecured. However, regulatory reviews by traditional bank regulators, such as the FDIC and the OCC, have increasingly included consumer loan products.
This expanded scope has been accelerated with the creation of the CFPB and evidenced by the enforcement actions against a number of larger indirect auto lenders.
The first question of course is how can an evaluation of fair lending, i.e. different levels of treatment based on race or gender be conducted when that type of information is not available in the loan files?
The answer is through different “proxy” methodologies in which these attributes are estimated and used in the review. We will cover these methods in detail in a future series of posts. For now, let’s evaluate the risk.
Consumer Loan Risk
Many times for a financial institution consumer loans are an afterthought when it comes to managing fair lending risk. Consumer loans, however, hold a great deal of risk for the following reasons:
- No Analysis – Most banks have NO idea what an analysis of their consumer lending may suggest. They have likely done little or no analysis.
- Loose Monitoring – Since HMDA lending is usually the focus, these loans are many times ignored and not managed well with regard to fair lending. There is usually no monitoring and very little tracking of underwriting exceptions and loan rate deviations.
- Large Datasets – As will be covered in future posts, most analysis of consumer lending will involve statistical techniques. Consumer loan pools often consist of large datasets which are quite conducive for statistical analyses. A large volume of records in a regression analysis, for example, allows for precise measurement. This means that a small difference in treatment with regard to prohibited factors can be more easily detectable than with smaller datasets.
- A Number of Different Products – There are many different types of consumer loans that can be analyzed, including auto, unsecured, HELOC and so forth. Any one can be a target. This leaves a wide array of potential targets for regulators to review.
- Variation – There tends to be much more variation in consumer lending, which again combined with large datasets means that differences between protected and non-protected groups may be more easily detectable. The rate structures, for example, for consumer loans tend to vary much more than with HMDA products which elevates risk.
In summary, consumer loans pose a great deal of risk for financial institutions. These products are increasingly within the scope of fair lending reviews, and institutions should be on notice. We will cover proxy techniques and strategies to deal with these risks in future posts.