In the wake of the housing and foreclosure crisis, the United States Department of Justice (“DOJ”) and federal bank regulatory agencies have made fair lending enforcement a top priority. As a result, banks have become subject to severe examination criticism, enforcement actions and even civil actions alleging discrimination in loan pricing, particularly with respect to unsecured consumer loans. Banks have endured these harsh consequences, even where there was no evidence or even any allegation that the bank intended to discriminate. In response, many community banks have revised lending policies to eliminate virtually all loan officer discretion in the pricing of unsecured consumer loans, with the inevitable result that unsecured credit is generally less available to the very consumers that the fair lending laws were designed to protect.
In light of these unintended consequences, one would expect Washington and the bank regulatory agencies to take a step back and to rethink their aggressive and arguably unfair approach to fair lending enforcement. But that has not been the case. In fact, the Department of Housing and Urban Development’s (“HUD’s”) recent release of its hotly-debated “disparate impact” rule sends a clear signal that regulators intend to plow ahead with this same type of enforcement—only now expanding it to place more emphasis on residential mortgages. What’s more, when you factor in the Consumer Financial Protection Bureau’s (“CFPB’s”) new “qualified mortgage” regulations, it is very possible a bank attempting in good faith to comply with one agency’s regulation could find itself in inadvertent violation of another. It’s a treacherous regulatory environment, and bankers will need to step carefully if they are to successfully navigate the compliance tightrope that now stretches between the fair lending “rock” and the credit quality “hard place.”
Fair Lending Enforcement – No Intent Necessary!
An increasing number of our bank clients have come under fair lending scrutiny during recent months. The typical scenario goes something like this. During a regular compliance examination, examiners identify a sample (usually, a small sample) of unsecured consumer loans. Examiners gather data relating to the pricing of these loans and the ethnicity and/or gender of the bank’s borrowers from a combination of file review and interviews with bank officers. This data is sent to the regulators’ statisticians in Washington, who perform a complex regression analysis to determine whether there are statistically significant differences in rates charged to minority versus non-minority borrowers sufficient to justify a finding that discrimination may have occurred. When the analysis is complete, the bank is typically given a mere fifteen days to respond to the preliminary examination findings. If the bank is unable to rebut the statistical analysis, the regulators will conclude that a “pattern or practice” of discrimination has occurred. Banks should understand that the finding of discrimination as a result of this type of regulatory inquiry may be – and is often – based solely upon the results of the statistical regression analysis, without any direct evidence of discriminatory conduct.
Expanding the Fair Lending Focus to Residential Mortgage Loans
In February 2013, HUD released a final rule implementing its discriminatory effects standard in fair housing enforcement. Under the new rule, which is to be applied retroactively, HUD makes it clear that lenders can be liable for lending policies that have a disparate impact on a group of borrowers, even in the absence of any intention to discriminate.
Although much of the “discriminatory impact” scrutiny during recent bank compliance examinations has been focused on unsecured consumer loan pricing, the implication of HUD’s new rule is that this same type of scrutiny will be expanding to place more emphasis on residential mortgage loans. Given the similarity of the standards and objectives of HUD’s new rule with those under fair lending laws, it is reasonable to expect that regulatory examination and enforcement of the new rule will take a similar approach. HUD has even hinted at this, indicating in its release that lenders who are subject to challenge under the rule’s discriminatory effects standard of liability would likely be subject to a similar challenge under fair lending laws. Accordingly, banks should be prepared to face fair lending and fair housing challenges to their mortgage lending practices that are based solely upon the results of statistical analyses, rather than any discriminatory intent or overt discriminatory conduct.
Potential Impact of the Qualified Mortgage / Ability to Repay Rules
When you consider the regulators’ purely statistical approach to fair lending/fair housing enforcement within the backdrop of the CFPB’s new “ability-to-repay” and “qualified mortgage” rules, things get even scarier. It is quite possible that a bank attempting in good faith to comply with those rules could step directly into a fair lending/fair housing violation.
The CFPB’s new rules create a safe harbor from liability for residential mortgage loans that satisfy a very specific set of underwriting, or “ability to repay,” requirements. There is a significant disincentive, therefore, for banks to make loans that are not “qualified mortgages.” If banks determine to make only qualified mortgages, it is very possible that, over time, a purely statistical look at the bank’s mortgage loan portfolio will reveal disparities between loans that are made to one demographic group of borrowers versus another potentially protected group.
Consider the following example:
A bank decides to make only qualified mortgage loans. Over time, the bank makes lending and pricing decisions based solely upon the specified “ability to repay” underwriting criteria set forth in the applicable regulations with no consideration given to race, gender or any other prohibited characteristic. In short, the bank has made its lending decisions with the intent to remain within the available safe harbors from liability and not with the intent to discriminate against any customer or group of customers on a prohibited basis.
At some time in the future, the bank becomes subject to a fair lending review during its normal compliance examination. For the purposes of the bank’s fair lending analysis, regulators categorize all mortgage loans that have been made to males or jointly to males and females as “male loans” for the purposes of their regression analysis, and they categorize loans that have been made only to females as “female loans.” It turns out that, according to the regulators’ regression analysis, the bank has given favorable treatment in its lending practices to a statistically higher number of male and joint male/female applicants. Logically, this is because a higher number of male and joint male/female applicants have satisfied the debt-to-income and other requirements for a “qualified mortgage” than have female applicants, and not because the bank intended to discriminate against females. Nonetheless, as a result of the regulators’ purely statistical review, the bank’s policies and procedures will appear to have had a disparate impact on a protected class of consumers, and examiners may cite an apparent violation.
Of course, the bank would have the ability to respond to examination findings and show that its decisions were based upon legitimate, non-discriminatory factors. In the event of a fair housing challenge, the bank would have the ability under HUD’s new burden-shifting standard to show that its lending policies were “necessary to achieve one or more substantial, legitimate, nondiscriminatory interests.” But the reality is that even responding to an alleged fair lending/fair housing challenge is costly and time-consuming. To respond to the regulators regression analysis, the bank would likely have to hire counsel and qualified statisticians simply to understand the basis of the charges against it. If the matter is referred to the DOJ or another consumer-protection agency, the bank would be required to comply with voluminous document production requests in a litigation-type setting. Requests for production of loan files and pricing information going back as far as six years are not uncommon. Even if the bank was successful in its defense, the costs of that defense and the related document production would be substantial, and the bank would be subject to reputational damage.
Walking the Compliance Tightrope – How to Defend Against A Fair Lending/Fair Housing Challenge
In the current regulatory environment, bankers need to be aware of the potential landmines and take steps to defend against a potential fair lending/fair housing challenge. The best defense against such a challenge can be summarized in the following important steps:
- Understand why examination information is being collected and how it is likely to be used.
- Stay involved in the regulatory process and be proactive in responding to the regulators’ initial findings and regression analysis.
- Consider engaging outside statistical expertise at the earliest indication of an adverse finding.
Step 1. Understand why examination information is being collected and how it is likely to be used.
As mentioned above, in the typical examination scenario, examiners will identify a sample of loans and then gather data relating to those loans from a combination of file review and interviews with bank officers. Examiners may also ask bank officers to review a list of borrowers and confirm the ethnicity or gender of those borrowers (often simply from a list of borrower names or surnames). From this information, the regulators’ statistical experts in Washington will prepare a complex regression analysis, the results of which will determine whether the bank has apparently discriminated against its customers.
Because most banks typically have a logical explanation for loan pricing other than discrimination, it is critical that information provided to examiners for use in the regression analysis be complete, accurate and reflective of the bank’s actual pricing policies. The examiner’s job is simply to ask questions (typically in a “yes/no” or “check-the-box” context) and to pass on the information to the statisticians for analysis. Consequently, bankers must be proactive to ensure that there is no disconnect between the bank’s actual pricing practices and the practices that will be reflected by the data that is collected during the examination. A flawed statistical analysis may have dire regulatory and reputational consequences for the bank. The results of a regression analysis are only as accurate as its underlying data inputs. With this in mind, bankers should ensure that information – especially information relating to the factors considered by loan officers in making and pricing loans – is accurately and completely conveyed to examiners, and further, that the examiners are accurately and completely conveying that information to the analysts who will be preparing any regression models.
Step 2. Stay involved and be proactive during and after the regulatory examination.
A bank’s ability to respond to and defend alleged fair lending/fair housing violations diminishes the further along the regulators go in the investigation and enforcement process. Moreover, once the regulators have made a referral to the DOJ or other agency, the bank will typically no longer have an avenue of appeal within the applicable federal banking agency and some form of enforcement action will become virtually unavoidable.
It is important to understand the process and be aware of what is going on from the beginning. Bankers should stay in regular communication with examiners during and after the compliance examination and inquire as to the progress of the examination and the status of any findings. Generally, a bank will have several months between the time that examiners collect information and the results of the regression analysis are finalized. During this period, bank management should proactively follow-up with examiners, requesting status updates as to the progress and results of the regression analysis. It is also common for examiners to contact the bank periodically during this analysis period, requesting confirmation of certain information or following up on questions posed by the statisticians. The bank should use any such request as an opportunity to ask for a status update and, as with the initial information request, the bank should make sure that it fully understands how the information that it is providing will be used so that it may respond in a manner that will accurately reflect the bank’s actual lending practices.
Step 3. Consider bringing in outside expertise at the earliest indication of adverse findings.
The regression models used by the regulators to evidence fair lending/fair housing violations are complex and are prepared by statistical experts. Most banks don’t have the in-house statistical expertise to fully analyze the regulators’ regression models or to create their own, competing analysis. Accordingly, at the first indication of an adverse finding, bank management should consider consulting with experienced statistical experts to assist the bank in identifying any potential flaws or disparities in the regulators’ regression analysis and in heading off adverse examination findings.
In the unfortunate event that the bank receives a formal written notice of adverse findings (typically permitting the bank a mere 15 days to respond), immediate consultation with statistical experts and regulatory counsel becomes imperative. This “15-day letter” almost always signals the bank’s last meaningful chance to request more time or to respond before the regulators’ findings become “final,” triggering the significant and costly consequences discussed above. In our experience, the regulators have been willing to provide extensions of the 15-day response period and to consider competing regression analysis at this stage. Any successful response, however, must be carefully targeted to address the quantitative data included in, and the results of, the regulators’ regression model.
Be Proactive to Prevent Fair Lending/Fair Housing Criticism
Even in advance of a compliance examination, all banks should take steps to review the effectiveness of their compliance management control systems as those controls relate to fair lending and fair housing. Lending policies and pricing procedures addressing all types of loans should be in place and should be reviewed regularly. These policies and procedures should prominently express the bank’s policies regarding non-discrimination against all protected classes of consumers. Any variations or exceptions from the bank’s pricing policies should be infrequent and should be approved by committee to ensure that those exceptions are being made consistently. For example, one loan officer should not make a 2% pricing exception based upon a specific factor while another loan officer makes a 1% exception based upon the same factor. All exceptions should be thoroughly documented. Finally, all loan officers should receive regular and documented training on the bank’s compliance management systems, the Equal Credit Opportunity Act, the Fair Housing Act and related regulations.
Careful adherence to the above recommendations may not ensure that justice is served in every case, but it will discourage the regulators from misapplying an exceedingly blunt enforcement instrument at the expense of your bank and its good reputation.