The Fall 2016 edition of the Consumer Finance Protection Bureau’s Supervisory Highlights devoted several pages to the topic of redlining.
Some of what the CFPB had to say about fair lending was not new—for example, that “redlining is a priority area in the Bureau’s supervisory work” or that “redlining is a form of unlawful lending discrimination under ECOA.” Really! Obviously, these statements are not new news.
However, the article did include some interesting comments regarding how the CFPB determines an institution’s “peers.”
According to the CFPB, when they do their redlining analyses and choose peers, they do it two main ways:
- First, they consider all the banks within a designated geography, e.g. an MSA (Metropolitan Statistical Area) for example, to be your peers.
- Second, they analyze the distribution of the number of applications or originations for this all-bank MSA aggregate and compare that distribution to your bank’s distribution by, say, minority percent (0-10%, 20-50%, etc.). Pretty standard stuff.
But the regulators go one step further and compare your bank’s odds of having applications or originations in majority minority areas as other lending institutions. That is, they are comparing odds ratios between banks.
Comparing odds ratios between banks can be misleading since odds ratios are not unique. For example, an odds ratio of .5 can be calculated either by 5/10 or 10/20. Nonetheless that is what the CFPB does (insert link to section of Supervisory Highlights).
Refined Institution Groups
Another interesting part is what comes next when the CFPB does an analysis using refined groups of institutions. As opposed to the aggregate analysis discussed above, the regulator tries to tailor your comparators to more accurately reflect the characteristics of your bank. This tailoring can take on various dimensions.
Most characteristically asset size is used. Thus you are going to be compared to another bank in the same MSA that is of similar asset size. Certainly, tailoring the comparisons by asset size may make sense, but the comparison may be further refined along additional dimensions such as number of loans or originations, loan type, purpose and lien status.
The Supervisory Highlights article states that past redlining investigations have relied on multiple refined bank comparisons. But the article also states that examiners have taken suggestions from banks as to who are the appropriate peers in any give market.
Suggesting Your Own Institution Groups
The opportunity to make suggestions to the regulators as to what is the appropriate peer set for each of your markets should not be missed. After all, who knows your business better than you do?
Based on experience, when compiling this list of peers asking your business leader who their competitors are should not proceed by putting a blank sheet of paper in front of them and asking, “Who are your peers?” While it’s necessary to get their buy-in, the result of that exercise typically is not very productive. The end result is too many “peers”—many of which are irrelevant for your proposed refined peer list.
As a compliance officer, you should provide a beginning proposed list of peers to your business head as a starting point for compiling the peer list that the regulators will see. Your starting point for the refined peer list can be developed by combining your HMDA data and the institution data set that is part of the HMDA dataset.