Fair Lending Compliance Management: Developing Strategies for Emerging Challenges

Fair Lending Compliance Management: Developing Strategies for Emerging Challenges August 20, 2014 2014 Crowe Horwath LLP 1

Agenda: The principal concepts of fair lending Current trends in fair lending enforcement Indirect automotive lending enforcement and the use of BISG for consumer loan analysis Proposed HMDA reporting requirements Key factors in fair lending compliance management 2014 Crowe Horwath LLP 2

Principal Concepts of Fair Lending What is fair lending? Fair lending represents a body of federal laws and regulations relating to the prohibition of discriminatory practices in lending. The regulations that fall under fair lending include the Equal Credit Opportunity Act (ECOA), the Home Mortgage Disclosure Act (HMDA), the Community Reinvestment Act (CRA) and the Fair Housing Act (FHA). Who faces enforcement risk factors? Although fair lending enforcement may impact entities such as rental associations, municipalities, and insurers, the vast majority of fair lending risk is borne by institutional lenders, including depository institutions, retail mortgage lenders, and captive financiers. 2014 Crowe Horwath LLP 4

Principal Concepts of Fair Lending (cont d) Fair lending forbids discrimination based on: Race Religion Ethnicity National origin Gender Marital status Age Disability Receipt of income from public assistance sources The applicant s exercise of rights under the Consumer Credit Protection Act (CCPA) 2014 Crowe Horwath LLP 5

Principal Concepts of Fair Lending (cont d) What are some types of activity being scrutinized by regulators? Underwriting risk Disparities in underwriting outcomes based upon a prohibited basis Inequitable application of exceptions to policy based upon a prohibited basis factor Pricing risk Statistically significant differences in interest rates, fees, or other characteristics on a prohibited basis Inadequacy of policy and/or procedures in governing outcomes Steering Disproportionate and/or deliberate referral of prohibited basis group applicants to less favorable loan products Redlining Insufficient application and origination activity to individuals in predominantly minority areas within assessment area 2014 Crowe Horwath LLP 6

Current Trends in Fair Lending Enforcement The Unique Alternative to the Big Four Why is fair lending a hot issue? Although the underlying regulations have been in place for decades, the creation of the CFPB, combined with vigorous enforcement by the U.S. Department of Justice (DOJ), has resulted in substantially increased regulatory risk for mortgage and automotive lenders. Opaque methodology and aggressive enforcement have created an atmosphere of uncertainty throughout the financial services industry. To address the lack of identifiable applicant race or ethnicity data, the CFPB has begun to use an advanced probabilistic modeling technique known as Bayesian Improved Surname Geocoding (BISG), which assumes applicant characteristics and helps to identify violations. Use of third parties for various lending services e.g., auto dealers, credit cards, broker relationships has increased. Consumer complaints throughout the industry are trending upward. 2014 Crowe Horwath LLP 7

Poll Question #1 The Unique Alternative to the Big Four Which of the following characteristics represent areas that are considered to be prohibited basis factors under ECOA? a) Age b) Race c) Marital status d) Disability e) All of the above f) Unsure/don t know 2014 Crowe Horwath LLP 8

Recent Enforcement Activity (cont d) The Unique Alternative to the Big Four PNC Financial December 2013 Successor to action against National City Bank Ordered to pay $35 million in compensation for pricing discrimination against Hispanic and African-American borrowers Plaza Mortgage Company September 2013 Ordered to pay $3 million in compensation for pricing discrimination against Hispanic and African-American borrowers Capital One October 2013 Successor to action against Chevy Chase Bank Ordered to pay $2.85 million in compensation for pricing discrimination against Hispanic and African-American borrowers The consent orders can all be found at www.justice.gov/opa/pr/2013 2014 Crowe Horwath LLP 10

Recent Enforcement Activity (cont d) The Unique Alternative to the Big Four Texas Champion Bank February 2013 (www.justice.gov/opa/pr/2013) Asset size of $367 million Ordered to pay $700,000 in restitution for mortgage pricing discrimination against Hispanic borrowers Southport Bank October 2013 (www.justice.gov/opa/pr/2013) Asset size of $256 million Ordered to pay $687,000 for mortgage loan pricing discrimination Community State Bank of St. Charles January 2013 Asset size of $189 million Ordered to invest $165,000 for lending to African American neighborhoods in Saginaw Fort Davis State Bank December 2013 Asset size of $69 million Ordered to pay $159,000 for unsecured consumer loan pricing discrimination The consent orders can all be found at www.justice.gov/opa/pr/2013 2014 Crowe Horwath LLP 11

Recent Enforcement Activity (cont d) The Unique Alternative to the Big Four Since the DOJ s Fair Lending Unit was established in February 2010, it has filed or resolved 34 lending matters under the FHA, the ECOA, or the Servicemember s Civil Relief Act. The settlements in these matters provide for more than $1 billion in monetary relief for affected communities and individual borrowers. 2014 Crowe Horwath LLP 12

Ally Financial Consent Order The Unique Alternative to the Big Four In December 2013, the CFPB and DOJ placed Ally Financial under a consent order, levying a total of $98 million in restitution and fines for discriminatory pricing outcomes among its population of indirect dealer auto loans. Ally, which conducts indirect loans for more than 12,000 automotive dealers, was cited for the following: Pricing disparities affecting black, Latino, and Asian applicants (20-30 basis point range difference from white non-latino applicants on average) Failure to conduct ongoing monitoring or analysis of possible disparities in outcomes Failure to offer fair lending training to its network of auto dealers Ally Financial was effectively provided two choices by the CFPB, either: 1. Eliminate the existing dealer compensation system in which dealers have discretion to offer pricing exceeding the lender s buy price ; or 2. Implement a quarterly and annual analysis of possible pricing disparities at the dealer and institutional level, using the CFPB s methodologies. Ally Financial chose the latter option, to enhance its monitoring program while maintaining existing pricing practices. 2014 Crowe Horwath LLP 14

Implications of the CFPB s Automotive Fair Lending Enforcement Which choices are automotive lenders currently facing? The CFPB has effectively given indirect automotive lenders the choice between (1) ceasing to offer dealer discretion over pricing or (2) implementing comprehensive monitoring and compliance programs related to fair lending. How have market peers reacted so far? In response, lenders such as Wells Fargo and Ally have announced that they will not switch to a flat-fee system and will instead work to augment their compliance and monitoring programs to satisfy regulatory expectations. What are the broader implications for industry participants? If deprived of pricing discretion, dealers face adverse effects relating to the ability to generate revenue as well as to compete with offers from rivals. The adoption of a flat-fee system could endanger existing dealer-lender relationships, and this is what makes an industry-wide shift highly unlikely. As a result, to comply with regulatory expectations, lenders will need to take proactive measures to enhance their compliance and monitoring programs. 2014 Crowe Horwath LLP 15

Challenges of Meeting Regulatory Expectations The Unique Alternative to the Big Four What challenges confront auto lenders in complying with the CFPB mandate? Compliance Program Readiness Although fair lending laws are not new, the CFPB s regulation of automotive finance companies raises the expectation on program readiness. Automotive lenders, in confronting the ongoing operational challenges of their industry, might be required to implement or augment compliance oversight and training programs that may exceed existing capabilities. Resource Constraints of Nonbank Lenders Fair lending compliance presents challenges related to current lending practices. While the CFPB s mandate may conform to the capabilities of the largest retail banks, many lenders face significant resource constraints making sufficient oversight of both dealer and institutional activity challenging. Complexity of Analytical and Technical Requirements The CFPB appears to be aggressively enforcing fair lending expectations of automotive lenders. Lenders are expected to use BISG as a proxy for ethnicity and race. This requires significant technical resources and using large, complex data sets from both lender and public data sources. 2014 Crowe Horwath LLP 16

Challenges of Meeting Regulatory Expectations (cont d) The Unique Alternative to the Big Four How dealers seek to approach compliance program readiness: The National Automotive Dealer Association (NADA) has published its Fair Credit Compliance Policy and Program, which offers guidance to dealers seeking to enhance their fair lending controls. NADA s Fair Credit Compliance Policy and Program was significantly influenced by the remediation steps implemented following the consent orders issued in 2007 in United States v. Pacifico Ford, Inc. (E.D.Pa.) and United States v. Springfield Ford, Inc. (E.D. Pa.). The program steps in place include: Creation of a fair credit policy Establishment of a fair credit compliance program Appointment of a program coordinator tasked with oversight of the compliance program Program components, including: Establishment of a standard dealer participation rate Definition of specific allowable reasons for deviation Development of a review process to ensure appropriateness of the deviation Development of training, oversight, and reporting processes related to fair credit 2014 Crowe Horwath LLP 17

Bayesian Improved Surname Geocoding (BISG) The Unique Alternative to the Big Four How do examiners impute race and ethnicity for indirect automotive loans? The CFPB/DOJ is using the BISG method, which generates demographic data based on two characteristics, the borrower s surname and the borrower s address. Does BISG require the collection of additional data from the applicant? No. Under Regulation C, lenders are permitted to collect only the government monitoring information (GMI) identifying race, ethnicity, and gender in accordance with requirements to report activity for HMDA. BISG uses characteristics derived from the applicant s surname and residential address to generate probabilities of representation among demographic groups. What groups does the BISG process identify? There are six groups in total: White Black Hispanic Asian/Pacific Islander American Indian/Alaska Native Multiracial 2014 Crowe Horwath LLP 18

Bayesian Improved Surname Geocoding (BISG) (cont d) How does BISG work? The Unique Alternative to the Big Four BISG calculates each applicant s probability of being in a given racial/ethnic category. This is done by calculating demographics that are based on the applicant s address data cross-referenced at the census block group level with the 2010 census data set and surname (using recorded demographics of all surnames reported at least 100 times in the 2000 census). Then, using Bayesian probability, the surname and geographic values are used to calculate the probabilities for each applicant. Here is an excerpt of the formula: Source: The formula is excerpted from the study establishing the BISG methodology, Using the Census Bureau s surname list to improve estimates of race/ethnicity and associated disparities, by Marc Elliott at the Rand Corporation. 2014 Crowe Horwath LLP 19

Bayesian Improved Surname Geocoding (BISG) (cont d) The Unique Alternative to the Big Four How should an institution use the imputed race and ethnicity to monitor its indirect automotive lending portfolio? Lenders are expected to conduct periodic, ongoing reviews of possible rate spread disparities at both an institutional and an individual dealer level. Institutions can use outcomes based on imputed BISG probabilities to identify disparities among populations that are possibly statistically significant. Review for statistical significance: 20-30bp rate spread disparity cited at Ally Financial 10bp cutoff referred to by the Philadelphia Fed during its 2013 Interagency Fair Lending Hot Topics webinar in October 2013 Address dealer-level and/or institutional risk factors, as applicable: If dealer disparities are identified, an institution should seek to confirm awareness of ECOA requirements by dealer and may require further steps limiting rate spread discretion and/or termination of relationship, depending on the extent and duration of noncompliance. Factors of particular importance include the adequacy of fair lending training, institutional monitoring, and, most significantly, the adequacy of institutional policies in limiting discretion in establishing rate spread levels. Use BISG to determine exposure and remuneration, if applicable: Since no individual is conclusively categorized in one group or another, the calculation of remuneration or rate relief would necessitate using discrete thresholds in identifying affected borrowers. 2014 Crowe Horwath LLP 20

Poll Question #2 The Unique Alternative to the Big Four To analyze risks of disparate impact for consumer loans, the BISG process uses which of the following: a) Borrower address b) Surname c) Income level d) Employment status e) Borrower address and surname f) Unsure/don t know 2014 Crowe Horwath LLP 21

Proposed HMDA Reporting Requirements The Unique Alternative to the Big Four What are the proposed changes to HMDA? Section 1094 of the Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd- Frank) requires the expansion of required fields to be reported under HMDA. On July 24, 2014, the CFPB issued its proposed rule for the expansion of requirements. What is the regulators intended use for the data? The CFPB has emphasized its determination to use HMDA data to increase cognizance of the housing market and, more broadly, the availability of credit. The CFPB seeks to use qualitative data to significantly enhance the use of econometric modeling to identify targets for regulatory scrutiny. What are the most significant changes to HMDA? Mandatory reporting of home equity lines of credit (HELOCs) and reverse mortgages Quarterly reporting for large institutions Changes to reporting thresholds; 25 loan minimum for depository institutions Inclusion of an additional 37 data fields, among them numerous qualitative factors, expanded borrower data, and items related to qualified mortgage and ability-to-pay rules 2014 Crowe Horwath LLP 23

Proposed HMDA Reporting Requirements (cont d) The Unique Alternative to the Big Four 2014 Crowe Horwath LLP 24

Proposed HMDA Reporting Requirements (cont d) The Unique Alternative to the Big Four What are some of the operational challenges you should begin to consider to prepare for the new HMDA requirements? How does your institution currently collect HMDA data? Can existing personnel collect and record the required data values? If your institution uses a mortgage application and/or underwriting system, what steps is it taking to prepare for the change? Do the individuals responsible for what might be newly covered areas, such as HELOCs and reverse mortgages, have sufficient experience with HMDA? Has your institution conducted data reviews to see that HMDA data is recorded accurately? What steps is your institution taking to prepare for the potential implications of the new data disclosure? Regulators are not the only potential analysts of HMDA public data; consumer rights organizations, advocacy groups, competitors, and others will also have access. 2014 Crowe Horwath LLP 25

Poll Question #3 The Unique Alternative to the Big Four Which of the following is not a newly proposed HMDA data field? a) Credit score b) Borrower age c) Bankruptcy history d) Introductory rate period e) None of the above f) Unsure/don t know 2014 Crowe Horwath LLP 26

Key Factors in Fair Lending Compliance Management Statistical Analysis Institutional Fair Lending Analysis The adoption of BISG and the intent to expand public mortgage disclosure data to include more qualitative data points both reflect a broader trend toward the use of advanced statistical techniques to drive the examination process. As a result, institutions are increasingly expected to monitor fair lending risk factors, which often requires the use of complex statistical metrics to identify risks of noncompliance. Performing statistical analysis assists in identifying risk factors that require more detailed review. The Office of the Comptroller of the Currency (OCC) and the Federal Reserve have emphasized that the identification of statistical measures of disparity may not be reflective of a substantive disparity and require additional analysis and/or technical review. 2014 Crowe Horwath LLP 28

Key Factors in Fair Lending Compliance Management Model Governance What is model governance? Supervisory Guidance on Model Risk Management (SR Letter 11-7) represents interagency guidance on the development, use, and validation of statistical models employed for financial decision-making. How does model governance relate to fair lending? Models such as those employed for credit scoring, automated decision-making, or fair lending analysis must be managed to prevent flaws that could amplify institutional risk factors. The presumption of the reliability of a model is no longer sufficient; the reliance on statistical modeling now requires a controlled and empirically driven process. 2014 Crowe Horwath LLP 29

Key Factors in Fair Lending Compliance Management Model Governance (cont d) Key steps in model governance: Assign responsibility and oversight for models. Take inventory of model variables and formulas. Confirm statistical adequacy. Avoid discriminatory factors. Reflect design and policy in model use. Follow change management protocols. Limit exceptions or overrides. Periodically test models for accuracy and management of risk of disparate impact. 2014 Crowe Horwath LLP 30

Key Factors in Fair Lending Compliance Management Model Governance (cont d) 2014 Crowe Horwath LLP 31

Key Factors in Fair Lending Compliance Management Other Emerging Factors What are additional areas facing increased regulatory scrutiny? Complaint resolution Unfair, deceptive, or abusive act or practice (UDAAP) Credit card programs Third-party risk management 2014 Crowe Horwath LLP 32

Poll Question #4 The Unique Alternative to the Big Four Your financial institution has decided that it doesn t want to deal with mortgage loans under $80,000. What type of discrimination could occur? a) Overt discrimination b) Disparate treatment c) Disparate impact d) Unfair and deceptive e) None of the above f) Unsure/don t know 2014 Crowe Horwath LLP 33

Upcoming events: Please join us at the Auto Finance Summit in Las Vegas on Oct 6-8, 2014. For further information and conference details, please visit: http://autofinancesummit.com/ 2014 Crowe Horwath LLP 34

Contact Information Niall Twomey, CRCM Niall.Twomey@crowehorwath.com 708.935.1139 Paul Osborne, CPA Paul.Osborne@crowehorwath.com 317.432.6274 Reid Simon, CRCM, CPA Reid.Simon@crowehorwath.com 415.312.1989 Eric Durham, CRCM Eric.durham@crowehorwath.com 630.4505514 Mark Blosser Automotive Industry SME Mark.Blosser@crowehorwath.com 574.286.8390 Crowe Horwath LLP is an independent member of Crowe Horwath International, a Swiss verein. Each member firm of Crowe Horwath International is a separate and independent legal entity. Crowe Horwath LLP and its affiliates are not responsible or liable for any acts or omissions of Crowe Horwath International or any other member of Crowe Horwath International and specifically disclaim any and all responsibility or liability for acts or omissions of Crowe Horwath International or any other Crowe Horwath International member. Accountancy services in Kansas and North Carolina are rendered by Crowe Chizek LLP, which is not a member of Crowe Horwath International. This material is for informational purposes only and should not be construed as financial or legal advice. Please seek guidance specific to your organization from qualified advisers in your jurisdiction. 2014 Crowe Horwath LLP 2014 Crowe Horwath LLP 35