Consumer Data Industry Association Fair Lending Teleseminar

Transcription

1 Consumer Data Industry Association Fair Lending Teleseminar May 10, 2016 D. Jean Veta, Covington & Burling LLP Michael Nonaka, Covington & Burling LLP Marsha J. Courchane, Charles River Associates

2 Agenda The CFPB s Increasing Role in Fair Lending CFPB 2015 Fair Lending Report (April 2016) New HMDA Rules Supreme Court s Inclusive Communities Decision Fair Lending and its Intersection with UDAAP Indirect Auto Lending Small Business Lending 2

3 The CFPB s Role in Fair Lending CFPB Examination and Enforcement Jurisdiction: Insured depository institutions and insured credit unions with total assets of more than $10 billion Certain nondepository institutions that are covered persons, including those who the CFPB has reasonable cause to determine are engaging or have engaged in conduct that poses risks to consumers with regard to the offering or provision of consumer financial products or services and those who are larger participants in certain markets for consumer financial services. Credit bureaus; Auto finance lenders; Debt collectors ECOA vs. FHA 3

4 The CFPB s Role in Fair Lending The CFPB s Office of Fair Lending and Equal Opportunity Patrice Alexander Ficklin, Director From April 2014 April 2015, fair lending enforcement actions required institutions to provide approximately $224 million in remediation to about 303,000 consumers Number of fair lending enforcement actions by year: 2013 (two); 2014 (one); 2015 (four); 2016 (one to date) Making policy through enforcement actions E.g., Indirect Auto Lending Leaked Memos Use of Bayesian Improved Surname Geocoding ( BISG ) Non-English Speaking Customers 4

5 The CFPB s Role in Fair Lending: Individual Liability Broad scope both individual liability and failure to admit wrongdoing Congress DOJ the Yates Memo; new FCPA pilot program SEC under increasing scrutiny for failing to assess individual liability CFPB individual liability cases to date Small entities; owner/operators Clearly unlawful conduct Potential sanctions Civil money penalties Sancho: ban from the financial products industry Potential application in fair lending cases 5

6 The CFPB s Role in Fair Lending: CFPB v. PHH PHH s appeal of Director Cordray s ruling in captive reinsurance case has potentially broad consequences for industry and the Bureau The facts ALJ found that PHH violated RESPA through a mortgage reinsurance kickback scheme, and required disgorgement of $6.5 million On administrative appeal, Director Cordray ordered a massive increase in disgorgement to $109 million. The increase was due to: Violation of RESPA every time PHH accepted a payment from a mortgage reinsurer Total gross revenue from premiums vs. profits Did Cordray correctly interpret RESPA? If so, did the industry have fair notice of his interpretation? Contrary to HUD guidance Contrary to widespread industry practice Is there no statute of limitations applicable in CFPB administrative hearings? Is the Bureau constitutional? Single Director structure Director can be fired only for cause 6

7 The CFPB s Role in Fair Lending: The Future Fintech Coming under increasing scrutiny Relationship between fintech lenders and banks Innovations in Underwriting Big Data Non-traditional data New Small Business Initiative Early stages Fair lending risk 7

8 Changes to HMDA Reportable Data 80 Fed. Reg (October 28, 2015) Applicant/Borrower Characteristics Applicant/Borrower Age Co-Applicant/Co-Borrowers Collateral Characteristics Detailed Property Type 1-Unit, 2-4 Units, 5+ Units Detailed Occupancy Status Differentiate b/w Investor Property & Second Home Underwriting/Pricing Factors Applicant/Borrower Credit Score Debt-to-Income Ratio Loan-to-Value Ratio (via Property Type) Combined Loan-to-Value Ratio AUS Recommendation 8

9 Additional HMDA Reportable Data (cont.) Loan Characteristics Detailed Loan Purpose Differentiate b/w Rate/Term & Cash-Out Refinance ARM Features Initial Fixed Rate Period Prepayment Penalty Terms Loan Term Non-Amortizing Features Balloon Loans, Interest Only Loans, etc. Reverse Mortgages HELOCs Business Channel Retail, Broker, Correspondent Rates & Fees Note Rate Total Points & Fees Origination Charges Discount Points Lender Credits 9

10 The Likely Impact of the New Amendments to Regulation C on Litigation The new HMDA data will be mainly used to support disparate impact theories under fair lending laws This gives the government and private parties more data to support fair lending claims The ease of access through the CFPB s website may also advantage complaining parties Conversely, a defendant will also have more data including credit characteristics to respond to a complaint Relying on the Inclusive Communities burden-shifting test this may give defendants a better chance of dismissing the action at an early stage Using history as a guide the revised Regulation C will likely lead to increased discrimination claims Increased fair lending analyses by lenders will be required 10

11 Inclusive Communities Decision Addresses the problem of abusive disparate impact claims announced rules when litigating disparate impact litigation: Courts must promptly assess the viability of a case A mere showing of a statistical disparity is insufficient, as disparate impact litigation is not meant to impose racial quotas This is a significant requirement favoring lenders The plaintiff must show that a statistical disparity was caused by the defendant s policies and practices causality required The defendant may defeat a prima facie disparate impact claim by showing that the policy or practice at issue was necessary to achieve a valid interest, which may include practical business choices and profit-related decisions If the defendant identifies such a valid interest, the burden shifts back to the plaintiff to show that there is an available alternative... practice that has less disparate impact and serves the [entity s] legitimate needs. 11

12 Statistical Analysis of Underwriting and Pricing Based upon a review of underwriting guidelines and rate sheets, develop customized statistical models that may control for factors such as the following: Underwriting Analysis Detailed Loan Purpose Detailed Occupancy Status Detailed Property Type Loan Amount Debt-to-Income Ratio Loan-to-Value Ratio Combined Loan-to-Value Ratio Applicant Credit Score Automated Underwriting Decision Detailed Loan Product Business Channel Pricing Analysis Detailed Loan Purpose Detailed Occupancy Status Detailed Property Type Loan Amount Rate Lock Week Loan-to-Value Ratio Combined Loan-to-Value Ratio Borrower Credit Score Detailed Loan Product Business Channel MSA 12

13 Hypothetical Lending Institution - Fair Lending Analysis of 2014 HMDA Incidence of Denial Protected Class Comparator Group Protected Class Model Total Denials Total Denials Odds Ratio p-value Conventional Mortgage Applications Pseudo R-Squared African American Hispanic Raw 44,126 11,459 3,008 1, Controlled 44,074 11,408 3,001 1, Raw 44,126 11,459 3,999 1, Controlled 44,078 11,412 3,989 1,

14 Hypothetical Lending Institution - Fair Lending Analysis of 2014 HMDA APR Regression Results Protected Class Model Comparator Group Loan Count Protected Class Loan Count Coefficient (bps) p-value Adjusted R-Squared Conventional First Lien Mortgages African American Raw 26,222 1, Controlled 26,222 1, Hispanic Raw 26,222 1, Controlled 26,222 1,

15 Fair Lending Redlining Analyses Redlining investigations focus on geography generally by looking at the percentage minority in particular census tracts or by looking at the market share on one institution compared to others offering the same product in the same geography First look at originations for financial institution as % in majority minority tracts compared to total originations. Next, calculate shares of financial institution in low, or high minority tracts compared to peers shares in those tracts. Measure statistically significant differences. Total Count Majority Minority Tracts Originations Majority African American and/or Hispanic Tracts Low Tract Minority Share High Tract Minority Share Count % p-value Count % p-value % p-value % p-value Low Minority Share / High Minority Share 15

16 Monitoring of Redlining Risk There is no standard monitoring approach, but all involve an assessment of the distribution of own institution s lending activity during a given time period within a defined geographic area versus a benchmark. For own institution s lending activity within the defined geographic area determine the proportion that involved properties located in census tracts with relatively high concentrations of minority residents. Compare own institution s proportion with that of lending activity for other lending institutions operating in the same defined geographic area using publicly available HMDA data from the same time period. Prior to public release of HMDA data for given time period, monitor trends by comparing own institution s proportion during given time period with own institution s proportion during prior time period. 16

17 CFPB/DOJ: Redlining Risk Remediation Hudson Savings Bank First joint redlining consent order issued by CFPB/DOJ was for Hudson Savings Bank, September 24, CMP of $5.5 million paid to CFPB $25 million in subsidy fund $200k advertising/marketing $100k Financial Education $750k CD partnerships Required new branches be established Required changes to CRA assessment areas to include full counties in NY and to add Camden NJ and Philadelphia Drew heavily not only on own percentage applications in high minority areas, but explicitly compared to peers for majorityblack-and-hispanic tracts Ignored higher than average approval rates and ignored purchased loans 17

18 HUD: Redlining Risk Remediation Associated Bank Conciliation Agreement dated May 22, 2015 $200 million settlement over focusing on allegedly smaller market share Subsidy assistance ($10 million lower interest rates, down payment assistance, closing costs) $3 million to help home repairs $1.35 million for community reinvestment and fair lending education $1.25 million for marketing/outreach (print media, radio, outreach) 4 new production centers in majority-minority areas (Milwaukee, Chicago, Cook or DuPage counties) 3 new branches in majority-minority areas (Chicago, Milwaukee, Racine) 18

19 Hypothetical Lending Institution Redlining Risk Assessment Given Geographical Area HMDA Lender Name Total Loans Loans in Majority Minority Tracts Share in Majority Minority Tracts Institution A 40,000 2, % Institution B 35,000 2, % Institution C 30,000 1, % Institution D 24,500 1, % Institution E 13, % Institution F 12, % OWN INSTITUTION 7, % Institution G 6, % Institution H 3, % Institution I 3, % Institution J 3, % Institution K 1, % Institution L % All Other Lenders 175,000 10, % Lenders with Similar Volumes 36,500 1, % 19

20 Hypothetical Lending Institution -- Analysis of Differences in Proportions of Lending in Majority Minority Census Tracts Lender Name Total Loans Loans in Majority Minority Tracts Share in Majority Minority Tracts Odds Ratio p-value OWN INSTITUTION 7, % - - All Other Lenders 175,000 10, % Lenders with Similar Volumes 36,500 1, %

21 UDAAP: Unfair, Deceptive and Abusive Unfair causes or is likely to cause substantial injury to consumers; not reasonably avoidable; and injury not outweighed by offsetting benefits (e.g. lower prices, more products). Substantial injury involves monetary harm (e.g. costs or fees even small amount if large number of consumers impacted) Deceptive misleads or is likely to mislead in a material way (central characteristics; expressed claims ); consumer s interpretation is reasonable (e.g. bait & switch). Evaluation with the four P s (prominence, presented in easy to understand format; placement where consumers look; info in close proximity to claim) and may be interpreted relative to a particular target audience 21

22 Abusive Materially interferes with consumer s ability to understand a term or condition Takes unreasonable advantage of Lack of understanding Inability of consumer to protect its interest in choosing or using product Reasonable reliance of consumer on a covered person acting in their interest 22

23 Examination Objectives Assess quality of compliance risk management review of internal controls and policies and procedures Doc review: lists of products, descriptions, fees, disclosures, account statements Procedure manuals and written policies Mgmt and Board meeting minutes Internal control and monitoring information Compensation Scripts, marketing, promotional materials Third party agreements Identify acts or practices that materially increase risk of UDA Review Complaints Gather facts and determine violations 23

24 Transaction Testing High Risk Areas Identified Are products underwritten on basis of ability to repay? Does product profitability depend on penalty fees or back-end rather than upfront fees? Does product have high rates of re-pricing or changes in terms? Does combination of terms increase difficulty in understanding? Are there penalties for terminating relationship? Does consumer bear fees or costs to get information about own accounts? Is product targeting to particular populations without making sure marketing, disclosures suit that population? 24

25 Indirect Auto Lending Indirect auto lending involves a prospective car purchaser s financing of the transaction through a third-party (bank, nonbank affiliate of a bank, captive nonbank) that contracts with the dealer Typical Process Car dealer collects car purchaser s information and forwards information to indirect auto lenders that evaluate purchaser s creditworthiness Indirect lenders decline to provide financing or give the dealer a minimum interest rate at which the lender is willing to purchase the financing agreement originated by the dealer Indirect lenders may use discretion to modify interest rate, allow for underwriting exceptions, or change other terms and conditions Indirect lender may allow dealer to increase interest rate above minimum rate and to receive compensation from the lender based on the difference in the actual rate and minimum rate 25

26 Indirect Auto Lending (cont.) CFPB Bulletin (Mar. 2013) CFPB issued Bulletin applying ECOA to indirect auto lenders even though they do not interact with the car purchaser because they participate in decision to extend credit Lenders may be liable for pricing disparities that exist on the basis of a prohibited class by (1) giving dealers discretion and incentive to increase actual interest rates and/or (2) allowing for discretion in modifying their own criteria and terms and conditions Lenders may be liable under ECOA under disparate treatment and disparate impact theories of discrimination Bulletin recommends steps to mitigate fair lending risk: Impose controls on dealer discretion and compensation policies Eliminate dealer discretion to increase interest rate and provide for alternative compensation structure Robust fair lending compliance management program CFPB Supervisory Highlights (Sept. 2014) Pricing disparities are due to discretionary pricing adjustments 26

27 Indirect Auto Lending (cont.) CFPB Larger Participant Rule In June 2015, CFPB finalized rule to supervise larger nonbank auto finance companies and issued related examination procedures CFPB already had supervised auto lending conducted by larger banks and thrifts Nonbank auto finance companies that originate more than 10,000 loans or leases are subject to CFPB examination for compliance with, among others, ECOA, TILA, UDAAP, and Consumer Leasing Act Examination focal points: Disclosure of auto financing terms/fair marketing Provision of accurate information to credit bureaus Fair debt collection Fair lending Other areas of emphasis: ancillary products, use of service providers, and repossession/bankruptcy 27

28 Indirect Auto Lending (cont.) Indirect auto lending has been major enforcement priority of CFPB: Nonbank lenders Westlake Services, LLC/Wilshire Consumer Credit Security National Automotive Acceptance Company First Investors Financial Services Group Captives Toyota Motor Credit Corporation (CFPB/DOJ) American Honda Finance Corporation (CFPB/DOJ) Dealers Herbies Auto Sales CarHop/Universal Acceptance Corporation DriveTime/DT Acceptance Corporation Banks Fifth Third Bank Ally Financial U.S. Bank/Dealers Financial Services 28

29 Indirect Auto Lending (cont.) CFPB s fair lending enforcement initiative on auto lending includes coordinated action with the DOJ CFPB and DOJ have Memorandum of Understanding to coordinate fair lending enforcement (information sharing, joint investigations, referrals and notifications) CFPB examines institutions for compliance with ECOA and is required to refer matter to the DOJ if there is a pattern or practice of discrimination 29

30 Indirect Auto Lending (cont.) CFPB and DOJ fair lending enforcement actions relating to auto lending typically include following findings and remedial provisions: Minority borrowers typically paid higher dealer markups (i.e., interest rate increases) that were not based on creditworthiness or transaction characteristics Pricing discrepancies affected thousands of borrowers and were attributable to discretionary pricing systems established by indirect lenders with dealers e.g., protected class borrowers charged 30 basis points more than similarly situated non-hispanic whites Remedial provisions Revised dealer compensation policy pricing mark-up limited to % depending on loan term Remediation to customers who were overcharged Retain settlement administrator to distribute funds 30

31 Indirect Auto Lending (cont.) CFPB s efforts to regulate indirect auto lending have been sharply criticized and subject to controversy Auto dealers carved out of CFPB jurisdiction in Dodd-Frank Act Regulation via enforcement Disparate impact Proxy methods for race and gender (see appendix) House passed bill to withdraw Bulletin and subject further efforts to notice and comment rulemaking 31

32 Small Business Lending ECOA applies to commercial credit and consumer credit fair lending protections apply to commercial credit applications and originations CFPB has started fair lending examinations of business lenders while navigating Dodd-Frank statutory framework for CFPB authority Fair lending priorities: mortgages, auto loans, credit cards, small business loans 32

33 Small Business Lending (cont.) Section 1071 of Dodd-Frank Act Amends ECOA to establish a robust HMDA-like data collection requirement for loan applications from women-owned, minorityowned, and small businesses Requires itemization of data fields such as application date, loan purpose, action taken, census tract, gross annual revenue of business applicant, race/sex/ethnicity of principal owners Purpose is to facilitate enforcement of fair lending laws and identify business and community development needs of certain businesses 33

34 Small Business Lending (cont.) Section 1071 of Dodd-Frank Act CFPB has delayed action on section 1071 rules several times CFPB outreach and research Interdisciplinary panel for rulemaking Assistant Director, Small Business Lending position Recent CFPB rulemaking agenda slated pre-rule activities for September 2016 SBREFA panel ANPR or NPR? Public interest groups have pushed for expansive regulations implementing section

35 Questions D. Jean Veta Michael Nonaka Marsha J. Courchane Charles River Associates Jean Veta is described by Chambers USA as one of the premier banking and financial regulatory enforcement litigators in the country. She defends financial institutions and their officers and directors in civil and regulatory enforcement matters, government investigations, internal corporate investigations, and congressional investigations. One client said, [s]he brought a discipline and toughness that was necessary in dealing with private litigants and was very experienced at dealing with government litigations. Based on her success in defending IndyMac Bank s former CEO, Jean was named Litigator of the Week by The American Lawyer, and was featured when Covington was named an Am Law 2014 Litigation Department of the Year finalist. Michael Nonaka advises banks, financial services providers, and non-bank companies on a broad range of compliance, enforcement, transactional, and legislative matters. He has worked extensively with federal and state banking agencies and with other federal agencies authorized to regulate financial services. Mr. Nonaka has significant experience advising clients on issues arising under financial services legislation such as the Dodd-Frank Wall Street Reform and Consumer Protection Act. He has advised clients on, among other areas in Dodd-Frank, regulation as a systemically important financial institution, resolution planning, the Federal Deposit Insurance Corporation s orderly liquidation authority under Title II, and the scope of the Consumer Financial Protection Bureau s authority. Dr. Marsha J. Courchane, Practice Leader of Financial Economics, specializes in financial institution analyses for regulatory reviews and in support of litigation. Dr. Courchane is a leading expert in the areas of mortgage and consumer lending and has worked with many of the largest lenders in the US. Her client work and research focus on issues including fair lending, affordable lending, credit scoring and the origination, pricing, securitization, and servicing of mortgages. Dr. Courchane held a number of academic and professional positions prior to her consulting experience. She served as Director of Financial Strategy and Research in the housing economics and financial research department at Freddie Mac, and she was Senior Financial Economist at the Office of the Comptroller of the Currency, focusing on fair lending and credit scoring matters for large national banks. 35

36 Appendix BISG Proxy Calculations Step 1: Surname Race/Ethnicity Probabilities for surname "Johnson" Race/Ethnicity Share Hispanic 1.5% African American 33.8% Asian/PI 0.4% American Indian 0.9% White 61.6% 2+ Races 1.8% Total 100.0% Source: Census Bureau 36

37 Appendix BISG Proxy Calculations Step 2: Geography 18+ Population of Tract Washington, DC Intra-Tract U.S. 18+ Race/Ethnicity Tract Counts Shares* Population Count Share of U.S. Hispanic 1, % 36,138, % African American 1, % 27,327, % Asian/PI % 11,637, % American Indian % 1,600, % White 2, % 157,123, % 2+ Races % 3,177, % Total 5, % 237,004, % Source: Census Bureau * Important BISG does not use the Intra-tract shares commonly used in other geography-based proxies. 37

38 Appendix BISG Proxy Calculations Step 3: BISG 1 Probabilities BISG Calculation Example BISG Surname Tract Probability Race/Ethnicity "Johnson" Wash, DC Vector Hispanic 1.5% % 2.3% African American 33.8% % 51.1% Asian/PI 0.4% % 0.5% American Indian 0.9% % 0.3% White 61.6% % 43.2% 2+ Races 1.8% % 2.6% Total 100.0% % 100.0% Source: Census Bureau 1 Elliott, Marc N. et al, Using the Census Bureau s Surname List to Improve Estimates of Race Ethnicity and Associated Disparities, Health Serv Outcomes Res Method (2009) 9: