Over the past year, the Consumer Financial Protection Bureau (“CFPB”) and other federal financial regulators have issued rules and guidance that will affect a number of aspects of the mortgage industry.
This year’s survey builds on last year’s survey by cataloging some of the recent developments from the interagency Property Appraisal and Valuation Equity (“PAVE”) task force as well as further developments related to Automated Valuation Models (“AVM”).
Over the past year, the Consumer Financial Protection Bureau (“CFPB”) and other federal financial regulators have issued rules and guidance that will affect a number of aspects of the mortgage industry. This year’s survey builds on last year’s survey by cataloging some of the recent developments from the interagency Property Appraisal and Valuation Equity (“PAVE”) task force as well as further developments related to Automated Valuation Models (“AVM”). In the wake of newsworthy developments in artificial intelligence and machine learning, the CFPB and others have focused on the use of artificial intelligence and digital marketing in both housing finance and other consumer financial products. Finally, this survey addresses the CFPB’s recently issued proposed rules regarding Property Assessed Clean Energy (“PACE”) transactions.
Appraisal bias—the artificial inflation or deflation of a home’s appraised value based on a prohibited characteristic, such as race—continues to be a pressing topic in the housing market. Although various federal laws have addressed appraisal bias for decades, this topic has recently received a significantly increased degree of attention.
The Task Force on Property Appraisal and Valuation Equity (“PAVE Task Force”), a “first-of-its-kind interagency effort to address inequity in home appraisals, and [to] conduct[] rulemaking to aggressively combat housing discrimination,” was formed to evaluate the “causes, extent, and consequences of appraisal bias” and provide recommendations “to root out racial and ethnic bias in home valuations.” In March 2022, the PAVE Task Force issued its first report, the PAVE Action Plan, summarizing its findings and containing twenty-one commitments for its member agencies divided across five categories: (1) strengthening guardrails against unlawful discrimination in all stages of residential valuation; (2) enhancing fair housing/fair lending enforcement and driving accountability; (3) building a well-trained, accessible, and diverse appraiser workforce; (4) empowering consumers to take action; and (5) obtaining better data to study and monitor valuation bias.
One year to the date after releasing its first report, the PAVE Task Force described critical progress that had been made toward implementing the PAVE Action Plan in a few key areas. The U.S. Department of Housing and Urban Development (“HUD”) awarded $54 million to 182 fair housing organizations to help fund testing for appraisal bias, enforcement activities, and educating local communities throughout the country on this issue. In January 2023, HUD published draft guidance to make it “easier and quicker” for prospective borrowers applying for Federal Housing Administration (“FHA”) insured loans to request a Reconsideration of Value (“ROV”) if the initial property valuation is “lower because of suspected illegal bias.” These actions are designed to empower consumers to “take action against appraisal bias.” A new online help portal was created to teach consumers who “suspect misvaluations due to racial bias” about their rights and how to file a complaint. Additionally, the Appraisal Subcommittee brought together federal agencies and industry experts and held its first hearing on appraisal bias and potential solutions.
The Federal Housing Finance Agency (“FHFA”) published its first publicly available Uniform Appraisal Dataset (“UAD”) Aggregate Statistics Data File containing statistics on appraisal records to provide public access to a broad set of data points and appraisal report trends. The FHFA also published UAD Aggregate Statistics Dashboards on its website, making this new data easy to visualize and understand utilizing customized maps and charts.
Steps are being taken to address the goal of developing a well-trained pool of professional appraisers that “loo[k] like the communities [they] serve.” For example, the U.S. Department of Veterans Affairs (“VA”) released guidance enhancing oversight procedures to identify potential discriminatory bias in appraisal reports by its fee panel appraisers and highly recommended relevant staff participate in appraisal bias and fair lending training.
Two years after the PAVE Task Force was created, it announced more “meaningful actions” it would take towards achieving its goals. Additional steps to empower consumers to “take action against appraisal bias” included a new working group to “increase coordination and develop more consistent standards for ROV processes” for approximately two-thirds of new mortgage originations in the market. The UAD Aggregate Statistics are being updated quarterly with new data allowing for further research and transparency to strengthen efforts to “eliminate bias and promote equity in the appraisal sector.” Additional efforts are underway that are “breaking down barriers to entry into the appraisal profession.” A new dashboard shows which states impose requirements to become an appraiser that are more rigorous than the “significant” federal minimum requirements in spite of a purported lack of evidence that some of the extra requirements “produce more ethical, accurate, or credible appraisals.” Work by the Appraisal Subcommittee and other related efforts to encourage states to remove barriers is ongoing.
As Automated Valuation Models (“AVMs”) continue to gain popularity for assessing property value, there has been a trend towards greater regulation. In 2018, Congress required promulgation of regulations to implement quality control standards for AVMs, including ensuring a high level of confidence, protection against data manipulation, avoiding conflicts of interest, and requiring random sampling and reviews. In June 2023, six federal regulatory agencies announced a long-awaited notice of proposed rulemaking and request for comment regarding the implementation of quality control standards for the use of covered AVMs in determining the worth of a mortgage secured by the borrower’s principal dwelling. Institutions engaging in certain credit decisions or securitization determinations must “adopt policies, practices, procedures, and control systems to ensure that AVMs used in these transactions” follow requirements in the areas required by Congress as well as complying with applicable nondiscrimination laws.
Also in June 2023, five federal agencies proposed interagency guidance with a request for comment on ROVs for residential real estate transactions. The proposal highlights risks with deficient residential real estate valuations, describes how ROV processes and controls may be included in existing risk management functions, and provides examples of policies and procedures that can be used to “help identify, address, and mitigate the risk of discrimination” impacting residential real estate valuations.
In the wake of newsworthy developments in artificial intelligence and machine learning, the CFPB and other federal financial regulators have issued interpretive rules, published guidance, and made several public statements regarding the use of artificial intelligence, machine learning, and other technological tools in mortgage and other lending—and in so doing, have affirmed their intention to apply existing laws to the use of these emerging tools.
In August 2022, the CFPB published an interpretive rule seeking to limit the applicability of the “time or space” exception under section 1002 of the Consumer Financial Protection Act (“CFPA”) for “digital marketing providers that comingle the targeting and delivery of advertisements to consumers” (“Time or Space Interpretive Rule”). The CFPB issued its Time or Space Interpretive Rule in response to its observation of an increase in “digital marketers” that “target and deliver ads to specific consumers using sophisticated analytical techniques, including machine learning and behavioral analytics, to process large amounts of consumer data.” The CFPB noted that digital marketers may “commingle the service of targeting and delivering advertisements with the activities of traditional media sources in providing airtime or physical space.” According to the CFPB, digital marketers obtain data from a variety of sources, including directly from consumers, by monitoring consumers’ web activity, and through “third-party data brokers” or “‘second-party’ partnerships with other companies.” Although the Time or Space Interpretive Rule is not exclusive to the housing finance market, a number of companies have offered—and continue to offer—these related services to mortgage lenders and brokers.
Under the CFPA, a “service provider” is defined as “any person that provides a material service to a covered person in connection with the offering or provision by such covered person of a consumer financial product or service.” In enacting the CFPA, Congress specified that any person who “participates in designing, operating, or maintaining” or “processes transactions relating to” a consumer financial product or service is a service provider under the CFPA. In contrast, any person who solely offers “a support service of a type provided to businesses generally or a similar ministerial service” or “time or space for an advertisement for a consumer financial product or service through print, newspaper, or electronic media” is not a service provider. Service providers to covered persons are subject to the requirements of the CFPA, including its prohibition of unfair, deceptive, or abusive acts and practices.
According to the CFPB’s Time and Space Interpretive Rule, digital marketers that are “materially involved in the development of content strategy” do not provide the same “time or space” advertising services offered by traditional media sources, and thus may qualify as “service providers” under the CFPA. The CFPB noted, however, that whether a particular digital marketing provider is a service provider or qualifies for the time or space exception depends upon the particular circumstances.
To provide further guidance to covered entities and digital marketing providers, the Time and Space Interpretive Rule includes several examples where a digital marketing provider would qualify as a service provider under the CFPA. A digital marketing provider that will “target and deliver . . . advertisements to users with certain characteristics” may not qualify within the “time or space” exception, even if the covered person specifies the particular characteristics to be targeted. In addition, a digital marketing provider that uses “algorithms and data” to identify the target audience may not qualify for the “time or space” exception, since this task traditionally would be performed by the covered entity itself. Even if the covered entity identifies particular users to receive advertisements, if the digital marketing firm is tasked with providing advertisements “at specific times to increase or maximize engagement,” this is akin to a covered entity deciding at what times to show television or radio advertisements and thus the digital marketing firm would not be eligible for the “time or space” exception. Finally, when a digital marketing provider determines or suggests “the appropriate audience for advertisements” for a covered entity, it is not eligible for the “time or space” exception since this function is similar to the actions of the covered entity’s own marketing group—or even a “lead generator” that would be a covered entity in its own right under the CFPA.
In February 2023, the CFPB issued an Advisory Opinion addressing how section 8 of the Real Estate Settlement Procedures Act (“RESPA”)—which prohibits referring business under a federally related mortgage loan in exchange for “any fee, kickback, or thing of value”—applies to “digital technology platforms that enable consumers to comparison shop for mortgages and other real estate settlement services” (“Platforms”). In its Digital Mortgage Platform Advisory Opinion, the CFPB focuses on consumer-facing online Platforms that “enable consumers to comparison shop options for mortgages and other settlement services,” particularly those that “generate potential leads for the platform participants through consumers’ interaction with the platform.” Although the CFPB acknowledged that Platforms were already covered by a 1996 HUD policy statement, the CFPB issued the Digital Mortgage Platform Advisory Opinion to discuss and provide examples of when payments to Platforms may violate section 8 of RESPA.
According to the CFPB, a Platform receives a referral fee that is prohibited by section 8 of RESPA when: (i) it “non-neutrally uses or presents information about one or more settlement service providers”; (ii) the non-neutral use or presentation steers “the consumer to use, or otherwise affirmatively influences the selection of, those settlement service providers”; and (iii) the operator of the Platform receives “a payment or other thing of value that is, at least in part, for that referral activity.” While HUD’s 1996 policy statement noted that receiving a higher fee for referring a particular settlement service provider could be evidence of a prohibited referral under section 8 of RESPA, the CFPB’s Opinion expands upon this by noting that “even if the Operator were to receive the same fee from each provider” this could be evidence of a RESPA section 8 violation.
To provide greater context, the Opinion identified five different categories of examples that—absent other evidence—could demonstrate a prohibited referral under section 8 of RESPA. First, the Opinion indicates that if a Platform “non-neutrally uses information to preference the highest bidding lender” this could imply an endorsement and encourage consumers to select a particular lender. Second, the Opinion states that a Platform may violate section 8 of RESPA if it received payments from one or more mortgage lenders that appear in the top spot of a list of lenders on a scheduled or random basis without “tailoring [the results] to a consumer’s needs based on their inputted information.” Third, if a Platform “steers consumers to use settlement service providers that are affiliates of the Operator” and receives payments from the affiliates, this would be a prohibited referral under section 8 of RESPA. Fourth, a Platform may violate section 8 of RESPA if it provides consumers a neutral list of lender options based upon the consumer’s information, but the Platform also contracts with a lender to send messages promoting the lender to the consumer using the information gathered in creating the neutral list. The CFPB notes that such “promotional activity . . . undermines the platform’s neutral presentation of information by steering the consumer to use a particular provider soon after the consumer had searched for comparison information.” Finally, the Opinion states that if a Platform calls the consumer and connects him or her with a live representative of the first lender to respond to the Platform’s information push, there is a referral because (i) the first lender to respond may not be the best fit for the consumer and (ii) this “warm handoff ” suggests an “implied endorsement” of the lender.
In April 2023, the leaders of the CFPB, the U.S. Department of Justice Civil Rights Division, the Equal Employment Opportunity Commission, and the Federal Trade Commission (collectively, “Agency Leaders”) issued a Joint Statement on Enforcement Efforts Against Discrimination and Bias in Automated Systems (“Joint Statement”) that will affect both mortgage and non-mortgage finance alike.
In the Joint Statement, the Agency Leaders acknowledged that innovation can “bring tangible benefits to people in a fair and competitive manner,” but emphasized that automated systems have “the potential to perpetuate unlawful bias, automate unlawful discrimination, and produce other harmful outcomes.” The Agency Leaders—whose agencies are responsible for enforcing a number of anti-discrimination, fair competition, and consumer protection laws—noted that their agencies have previously identified concerns with the use of automated systems.
Focusing specifically on anti-discrimination laws, the authors of the Joint Statement identified three areas where automated systems may create problems: (i) the use of “unrepresentative or imbalanced datasets” influenced by either historical bias or containing errors, as well as automated systems that “correlate data with protected classes” and yield discriminatory outcomes; (ii) so-called “black box” models that lack transparency and make it more difficult to determine whether automated systems are “fair”; and (iii) systems that are designed for one purpose but are used for a different one—and as such can lead to discrimination. The Agency Leaders committed to “vigorously” use their authority to protect individual rights “regardless of whether legal violations occur through traditional means or advanced technologies.”
In May 2023, the CFPB issued a long-awaited proposed rule outlining regulatory requirements under the Truth in Lending Act (“TILA”) for Property Assessed Clean Energy (“PACE”) transactions (“PACE Proposed Rule”). PACE transactions allow “property owners to finance certain upgrades (typically energy efficiency or natural disaster preparation upgrades) to real property through an assessment on their real property” that is paid through their tax bill. Section 307 of the Economic Growth, Regulatory Relief, and Consumer Protection Act (“EGRRCPA”) directed the CFPB to issue regulations to implement an ability- to-repay analysis for PACE assessments and apply section 130 of TILA—which provides civil liability for violations of TILA—to PACE assessments. The CFPB looked to the EGRRCPA, as well as the CFPB’s general authority under TILA and RESPA, as its basis for issuing the PACE Proposed Rule.
Under the CFPB’s PACE Proposed Rule, Regulation Z, which implements TILA, would apply to PACE financing. Currently, the Official Interpretations to Regulation Z exclude “tax liens” and “tax assessments” from the definition of credit. If enacted, the PACE Proposed Rule would revise this interpretation to read that “involuntary tax liens [and] involuntary tax assessments” are excluded from the definition of credit under Regulation Z—and by implication, that the CFPB takes the position that a “voluntary” tax lien or assessment as part of a PACE assessment would be “credit” subject to the substantive and disclosure requirements of TILA. In the preamble to the PACE Proposed Rule, the CFPB asserts that Congress did not intend to exclude voluntary tax liens from TILA, and that the original text of the Official Staff Commentary (issued by the Board of Governors of the Federal Reserve in 1981) was not “intended to exclude voluntary transactions such as PACE.” Although industry stakeholders noted several reasons why PACE assessments should not be subject to Regulation Z—including that it serves important public policy purposes and that these transactions are secured via the tax system rather than through a traditional mortgage security instrument—the CFPB nevertheless proposed to modify the Official Interpretations to include PACE assessments as credit.
As a result of this definitional change, the CFPB’s PACE Proposed Rule also makes several other changes to Regulation Z. If enacted, the PACE Proposed Rule would provide that a PACE assessment is not a Qualified Mortgage under Regulation Z and thus must go through an Ability to Repay (“ATR”) analysis, and would prohibit PACE assessments from qualifying for a presumption of compliance with ATR requirements. The PACE Proposed Rule also would revise the Official Interpretations for the ATR provisions to provide that the entity facilitating a PACE transaction “knows or has reason to know of any simultaneous . . . PACE transactions” that are included “in any existing database or registry of PACE transactions that includes the geographic area in which the property is located.” If enacted, the PACE Proposed Rule would clarify that escrow accounts would not be necessary for PACE transactions that otherwise meet the requirements of a higher-priced mortgage loan, and would exempt PACE transactions from the Regulation Z’s periodic statement requirements. In addition, the PACE Proposed Rule would modify the initial mortgage loan disclosures and closing disclosures under sections 1026.37 and 1026.38 of Regulation Z to account for the unique nature of PACE transactions, specifically including statements related to how a PACE transaction may affect a loan assumption.
Furthermore, the CFPB noted that the PACE Proposed Rule would not alter the high-cost loan requirements in sections 1026.32 and 1026.34 of Regulation Z, and as such the CFPB believes that these requirements would apply to PACE assessments as well.
In addition to applying Regulation Z to PACE assessments, the CFPB’s PACE Proposed Rule would apply Regulation Z’s civil liability provisions to PACE administrators—entities that assist state and local governments with implementing PACE programs—to the extent that they are “substantially involved” in the decision to extend financing. Notably, in enacting the EGRRCPA, Congress did not alter the definition of “creditor” under TILA to include PACE administrators, and historically Congress has limited liability under TILA to creditors, with a few narrow and express statutory exceptions. Nevertheless, the CFPB’s PACE Proposed Rule purports to extend TILA’s civil liability provisions to this broader class of non-creditor PACE administrators without Congress revising TILA to specify that PACE administrators are in fact broadly subject to the law’s civil liability provisions.
If finalized, the CFPB’s PACE Proposed Rule would be effective at least one year after publication in the Federal Register, but in no event earlier than the October 1 that is at least six months after the date of the final rule’s promulgation.
The views expressed in this survey are those of the authors alone.
