Home > News & Events > Conferences & Events > FDIC Teleconference on Fair Lending Issues
FDIC Teleconference on Fair Lending Issues
November 16, 2010
1. Are we required to gross up non-taxable income. For example, a person on social security who only gets $12,000 a year, would we need to gross this income up by 1.20% to ensure fair lending?
While it is not required that an institution gross up non-taxable income, Fannie, Freddie and FHA all allow a gross up factor of 25% on this type of income. (Or alternatively you may use the current federal and State income tax withholding tables.) The decision to use this factor is discretionary. However, it is important that if you decide to gross up income, in order to avoid any appearance of a fair lending violation, ALL similar types of non-taxable income should be grossed up and in a similar manner.
2. I have three questions:
1) Examiners recognize that the economic climate has caused many institutions to decrease their lending or number of loan products. This is not a fair lending concern on its face, provided that the loans the institution does make are made in a manner that does not treat members of protected classes differently or made under policies that have a disparate impact on protected classes.
2 & 3 - The FDIC does not endorse any particular tool, product or training that is available; however, trade associations are a good resource to find out what trainings are available. Also, in addition to the FFIEC Interagency Fair Lending Procedures, make sure you are aware of any guidance issued by your regulator.
3. I noticed the “Hot Topics” section of the presentation did not include any specific loan servicing or default area. Are there specific areas (Collections, Loss Mitigation, and Customer Service) that we should be concerned with or be proactively monitoring?
All of these areas should always be monitored to ensure fair and equitable treatment. Please refer to the FFIEC Fair Lending Examination Procedures for more detail.
4. Our bank is entering into a relationship with a mortgage broker whereby the broker will be processing the majority of our approved requests and the bank will be responsible for declining requests that do not meet the broker’s criteria. The result will be that the vast majority of our LAR will be non-originations. Is there advice you can offer as to how to present this to our examiners?
It is important to first identify who is the creditor (under HMDA) in this new business arrangement. You can use the “HMDA Reporting - Getting It Right!” Guide to make sure you are reporting these applications properly on the LAR. Once you verify that the LARs are correct, there should not be a problem explaining the business arrangement to the examiners.
5. Only one product is originated in both our Secondary Market Channel and In-House Channels. If the customer expresses an interest in a 15 or 30 year fixed rate product they would have to go to the Secondary Market channel. However, our In-house Channel offers a similar product with a 5 year balloon. Does the bank need to underwrite the In-House product as an option for the customer, although the customer did not express an interest in the 5-year balloon?
It would not be necessary to underwrite the balloon loan for the customer unless the customer expresses an interest in such product. However, to ensure all applicants have equal access to credit, it would be advisable to make sure the customer is aware of the balloon option and then document the fact that the customer chose the secondary market product over the balloon loan.
6. What is the Ethnicity and Race for Iranian borrowers?
According to the US Census Bureau, there is no wrong or right answer to identifying a person’s race and ethnicity. You can research this on the US Census website. Census data relies strictly on self identification, according to the race or races with which they most closely identify. Persons who report themselves as Hispanic can be of any race. Self-identification of a person’s race and ethnicity does not need to conform to any biological, anthropological, or genetic criteria. For HMDA purposes, you should report the ethnicity and race as each applicant identifies themselves in the Government Monitoring Information on the application. If you need to use visual observation to report race and ethnicity for Regulations B & C, then, the identification of the applicant’s race and ethnicity would be left to the interviewer’s observation of the borrower.
7. We use the Appraisal Date as the Application Date -- Can we be off by a day or two in reporting Application Dates? Is it OK to use the Date of the Note as the Action Date or do we have to use the funding date?
We recommend that you get a copy of “A Guide to HMDA Reporting - Getting It Right!” for 2010 HMDA reporting. You can download this guide from the FFIEC website. As long as you follow the guidance in HMDA Getting It Right!, you should not have any problems explaining to examiners how or why you reported HMDA data as you did.
Specifically, the Guide states in appendix A (I.A.2.a) the application date should be reported as the date the loan application was received by your institution or the date shown on the application. Therefore, if you received the completed application on the date of the appraisal, then you could use the appraisal date as the application date.
Appendix D reads as follows regarding the action taken date:
8. In the past, banks have allowed discretionary pricing on consumer loans, including mortgages, based on relationship or to match a competitor’s rate or fees. Should this practice be discontinued? What advice can you give to help us stay in compliance from a fair lending perspective, but still be able to take care of customers who have been loyal to us?
Allowing discretion in pricing is not a prohibited practice, but it does expose an institution to increased fair lending risk. The general principle is that more discretion equals greater risk. If an institution allows discretionary pricing, it raises a “red flag” for examiners, and it will always be closely reviewed during examinations.
An institution that allows discretionary pricing should establish monitoring mechanisms to ensure that pricing is not unfairly impacting any particular prohibited basis group. For example, this monitoring could occur by conducting comparative file reviews or by closely tracking pricing deviations through an exception report. Furthermore, maintaining documentation of the reason for the pricing deviation is pertinent in the event that questions are raised. Additionally, an institution should have regular training for its employees so everyone has a clear understanding of not only the institution’s policies and procedures, but also the fair lending laws and regulations.
9. I saw in the recent FIL 47-2010 regarding overdraft legislation that “inconsistent application of waivers of overdraft fees will be evaluated in light of fair lending statutes and legislations.” Our understanding in the past was that overdraft was not a Fair Lending concern as these programs qualified under the “incidental credit” exclusion (Reg. B). Does this mean that regulators are going to have a broader interpretation of Fair Lending than they have had in the past? If so, what other areas would you expect that Fair Lending will be expanded to include?
The FDIC’s position that overdraft programs are subject to ECOA is neither a broader interpretation of fair lending nor is it new. FIL-47-2010 merely reiterated the same position that had been previously stated in prior supervisory issuance by the Federal Reserve, the FDIC, and other federal regulatory agencies.
FIL-47-2010 provides in relevant part:
The statement merely reiterates previous statements of the same principle made in multiple regulatory issuances including but not limited to:
With that said, not all overdraft protection plans qualify as “incidental credit.”For example, plans where bank customers opt into the plan in writing do not constitute “incidental credit.” See 12 C.F.R. §226.4(c).
10. I am an independent compliance consultant that works with various banks helping them with Fair Lending reviews and risk assessments. Do you see any problem with me using information from borrowers’ drivers licenses that are stored in the bank’s Customer Information System to obtain demographic information for fair lending analysis such as age, sex, and possibly race? As you know, the government monitoring information is only available for HMDA-covered loans. If I am trying to conduct fair lending analyses on other types of loans, such as consumer loans or vehicle-secured loans, there isn’t much to go on other than making assumptions about race via surnames and gender via first names. If I could use information from the drivers license that is imaged and in the bank’s system, I could get demographic information without having to make assumptions.
Do you see any problem with doing that? I am independent from the bank and this is done after the loan decision has been made (these are existing loans). To go one step further, would this be OK for a bank’s compliance officer to do too, if that person is independent from the lending function and it’s done after the loan decision has been made?
Regulation B does not prohibit the practice of photo copying or digitally storing photo identification, but it does prohibit the collection of certain demographic information. Given that photo identification usually contains this prohibited information, it is a best practice not to retain a copy of the photo identification. However, if a bank chooses to retain a copy of photo identification, it is best if it is segregated from loan files. Furthermore, the photo identification should only be requested to comply with the USA Patriot Act and not be used as part of the underwriting process.
Keeping the above in mind, photo identification that is available in a customer information system could be used to obtain demographic information to assist in a fair lending analysis. As stated in the question, it would be a more accurate way to conduct the analysis than making assumptions based on factors such as surnames.
11. We are a small community bank and we contract with a third party for independent review of our Fair Lending and have recently had a Fair Lending Risk Assessment completed. To my knowledge, our bank has not been cited for Fair Lending infractions.
Each FDIC-supervised bank is expected to have a compliance management system (CMS) that appropriately manages all compliance risks, including fair lending risks. There is not a “one-size fits all” approach for a CMS to be effective because every institution is unique in size, demographics, product offerings, etc. Two components of an effective CMS are monitoring and audits of all business areas of a bank to ensure compliance with all laws and regulations, including fair lending.
In the question it states that the subject bank is a small institution that has contracted with a third party for an independent fair lending review. It also states that a fair lending risk assessment has been performed. If the scope of the third-party fair lending review and risk assessment are comprehensive and no major exceptions were identified as a result, an additional internal fair lending audit is likely not warranted. Otherwise if risk factors or exceptions are identified, examiners would likely expect the bank to perform some type of internal review to investigate any concerns and take appropriate action. The best course of action will be dependent on the results of a third-party review and risk assessment.
For the second question, the Interagency Fair Lending Procedures outline the process examiners use when reviewing an institution’s fair lending program. Becoming familiar with these procedures would serve as a good starting point for designing an internal fair lending audit program which should be tailored to the business activities and risks of the institution.
12. Where would we find the information to determine if we have minority census tracts in our market area?
Demographic information is available on the Federal Financial Institutions Examination Council (FFIEC) website located at www.ffiec.gov. On the main page of the website, click on the link for “Census Reports” under the Consumer Compliance heading. Through this link you can produce reports for specific states either at the MSA or county level. The reports provide demographic information on the minority population for each census tract in the chosen area as well as for other factors such as median family income.
13. From a statistical point of view, isn’t it the case that the more finely a set is divided, the more likely that one subset will represent a deviation from the norm? In other words, if the average (or median) for 100 items is, say, 75 then the odds of one of two subsets composed of, say, 40 and 60 items’ deviating from the average or median by 25% (pick your percentage) is much lower than the odds of one of 25 subsets so deviating? Applied to Fair Lending, the odds of deviating from the norm are greater the more protected classes there are. Do the regulators take this into account, or doesn’t it matter in their analysis?
This question assumes we are talking about something that naturally has variation, such as the height of the population. If we were to draw a random sample of 1/10 of the population versus a random sample of 1/100 of the population, there is indeed a greater likelihood that the smaller sample will deviate from the averages of the full population.
For fair lending and pricing in particular, we are comparing the adjusted mean interest rate (adjusted for all the pricing control factors) for the control group to the adjusted mean interest rate of the target group. After all the adjustments, which should reflect all the pricing factors used by the bank, we expect that the adjusted mean interest rates should be the same for both control and target group (i.e. there is no statistically significant difference). It does not matter how many target groups there are because there should be no statistically significant differences between the means of ANY target group and the control group once all the pricing factors have been controlled for.
14. On average, once the data gathering phase is completed, and all data required is provided to the FDIC, how long does it take the FDIC to issue the 15-day letter?
Once the FDIC has all of the data it needs to complete its review, it typically takes 60-90 days to issue a 15-day letter, if one is required.
15. Please provide information on how many outliers are reviewed each year v. the number of HMDA reporting institutions v. the number that are referred to DOJ for discrimination?
The FDIC has identified approximately 200 outlier institutions since pricing data was first reported in 2004 and in the past five years, has made 110 referrals to the Department of Justice based on findings of discrimination (29 in 2006, 15 in 2007, 12 in 2008, 21 in 2009 and 33 in 2010). For information about the number of HMDA reporting institutions, please visit the FFIEC’s website.
16. Do the other regulatory authorities use the same practices in identifying outliers and conducting outlier reviews?
Each agency has its own screening process and outlier review process.
17. Is there a matrix of some kind that can be used to list all loans with the required items that need monitoring?
There is no such matrix but please refer to the FFIEC Interagency Fair Lending Procedures for guidance regarding what areas of risk examiners will review during a fair lending review.
18. We want to break down our HMDA information so that we can report it to the Board of Directors. Is there a program or software that can help us with this?
There are various programs that are available that assist banks in analyzing their HMDA data. The FDIC does not endorse any particular program but does encourage its banks to monitor its HMDA data.
19. The institution I work for is located in a low/mod income community. Many of the mortgage requests are for loans under $50,000. Can we set a minimum loan amount on the mortgages we offer without violating fair lending laws?
Setting any sort of limits on lending should always be reviewed for possible fair lending concerns. In the situation you describe above, we would advise the institution to determine whether such a practice has a disparate impact on a protected class and if so, whether there is a legitimate business reason for such practice. This will be the same sort of analysis that our examiners will conduct during an examination and thus the institution would be wise to conduct this analysis beforehand to ensure there are no concerns.
|Last Updated firstname.lastname@example.org|