2020-08409. Home Mortgage Disclosure (Regulation C)

AGENCY:

Bureau of Consumer Financial Protection.

ACTION:

Final rule; official interpretation.

SUMMARY:

The Bureau of Consumer Financial Protection (Bureau) is amending Regulation C to increase the threshold for reporting data about closed-end mortgage loans, so that institutions originating fewer than 100 closed-end mortgage loans in either of the two preceding calendar years will not have to report such data effective July 1, 2020. The Bureau is also setting the threshold for reporting data about open-end lines of credit at 200 open-end lines of credit effective January 1, 2022, upon the expiration of the current temporary threshold of 500 open-end lines of credit.

DATES:

This final rule is effective on July 1, 2020, except for the amendments to § 1003.2 in amendatory instruction 5, the amendments to § 1003.3 in amendatory instruction 6, and the amendments to supplement I to part 1003 in amendatory instruction 7, which are effective on January 1, 2022. See part VI for more information.

FOR FURTHER INFORMATION CONTACT:

Jaydee DiGiovanni, Counsel; or Amanda Quester or Alexandra Reimelt, Senior Counsels, Office of Regulations, at 202-435-7700 or https://reginquiries.consumerfinance.gov. If you require this document in an alternative electronic format, please contact [e-mail].

SUPPLEMENTARY INFORMATION:

I. Summary of the Final Rule

Regulation C, 12 CFR part 1003, implements the Home Mortgage Disclosure Act (HMDA), 12 U.S.C. 2801 through 2810. In an October 2015 final rule (2015 HMDA Rule), the Bureau established institutional and transactional coverage thresholds in Regulation C that determine whether financial institutions are required to collect, record, and report any HMDA data on closed-end mortgage loans or open-end lines of credit (collectively, coverage thresholds).^[1]

The 2015 HMDA Rule set the closed-end threshold at 25 loans in each of the two preceding calendar years, and the open-end threshold at 100 open-end lines of credit in each of the two preceding calendar years. In 2017, before those thresholds took effect, the Bureau temporarily increased the open-end threshold to 500 open-end lines of credit for two years (calendar years 2018 and 2019). In October 2019, the Bureau extended to January 1, 2022, the temporary threshold of 500 open-end lines of credit for open-end coverage.

This final rule adjusts Regulation C's coverage thresholds for closed-end mortgage loans and open-end lines of credit.^[2] Effective July 1, 2020, this final rule permanently raises the closed-end coverage threshold from 25 to 100 closed-end mortgage loans in each of the two preceding calendar years. The final rule also amends § 1003.3(c)(11) and comment 3(c)(11)-2 so that institutions have the option to report closed-end data collected in 2020 if they: (1) Meet the definition of financial institution as of January 1, 2020 but are newly excluded on July 1, 2020 by the increase in the closed-end threshold, and (2) report closed-end data for the full calendar year. The final rule sets the permanent open-end threshold at 200 open-end lines of credit effective January 1, 2022, upon expiration of the temporary threshold of 500 open-end lines of credit.

II. Background

A. HMDA and Regulation C

HMDA requires certain depository institutions and for-profit nondepository institutions to report data about originations and purchases of mortgage loans, as well as mortgage loan applications that do not result in originations (for example, applications that are denied or withdrawn). The purposes of HMDA are to provide the public with loan data that can be used: (i) To help determine whether financial institutions are serving the housing needs of their communities; (ii) to assist public officials in distributing public-sector investment so as to attract private investment to areas where it is needed; and (iii) to assist in identifying possible discriminatory lending patterns and enforcing antidiscrimination statutes.^[3] Prior to the enactment of the Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act), Regulation C required reporting of 22 data points and allowed for optional reporting of the reasons for which an institution denied an application.^[4]

B. Dodd-Frank Act

In 2010, Congress enacted the Dodd-Frank Act, which amended HMDA and transferred HMDA rulemaking authority and other functions from the Board of Governors of the Federal Reserve System (Board) to the Bureau.^[5] Among other changes, the Dodd-Frank Act expanded the scope of information relating to mortgage applications and loans that institutions must compile, maintain, and report under HMDA. Specifically, the Dodd-Frank Act amended HMDA section 304(b)(4) by adding one new data point. The Dodd-Frank Act also added new HMDA section 304(b)(5) and (6), which requires various additional new data points.^[6] New HMDA section 304(b)(6), in addition, authorizes the Bureau to require, “as [it] may determine to be appropriate,” a unique identifier that identifies the loan originator, a universal loan identifier (ULI), and the parcel number that corresponds to the real property pledged as collateral for the mortgage loan.^[7] New HMDA section 304(b)(5)(D) and (6)(J) further provides the Bureau with the authority to mandate reporting of “such other information as the Bureau may require.” ^[8]

C. 2015 HMDA Rule

In October 2015, the Bureau issued the 2015 HMDA Rule implementing the Dodd-Frank Act amendments to HMDA.^[9] Most of the 2015 HMDA Rule took effect on January 1, 2018.^[10] The 2015 HMDA Rule implemented the new data points specified in the Dodd-Frank Act, added a number of additional data points pursuant to the Bureau's discretionary authority under HMDA section 304(b)(5) and (6), and made revisions to certain pre-existing data points to clarify their requirements, provide greater specificity in reporting, and align certain data points more closely with industry data standards, among other changes.

The 2015 HMDA Rule requires some financial institutions to report data on certain dwelling-secured, open-end lines of credit, including home-equity lines of credit. Prior to the 2015 HMDA Rule, Regulation C allowed, but did not require, reporting of home-equity lines of credit.

The 2015 HMDA Rule also established institutional coverage thresholds based on loan volume that limit the definition of “financial institution” to include only those institutions that either originated at least 25 closed-end mortgage loans in each of the two preceding calendar years or originated at least 100 open-end lines of credit in each of the two preceding calendar years.^[11] The 2015 HMDA Rule separately established transactional coverage thresholds that are part of the test for determining which loans are excluded from coverage and were designed to work in tandem with the institutional coverage thresholds.^[12]

D. 2017 HMDA Rule

In April 2017, the Bureau issued a notice of proposed rulemaking to address certain technical errors in the 2015 HMDA Rule, ease the burden of reporting certain data requirements, and clarify key terms to facilitate compliance with Regulation C.^[13] In July 2017, the Bureau issued a notice of proposed rulemaking (July 2017 HMDA Proposal) to increase temporarily the 2015 HMDA Rule's open-end coverage threshold of 100 for both institutional and transactional coverage, so that institutions originating fewer than 500 open-end lines of credit in either of the two preceding calendar years would not have to commence collecting or reporting data on their open-end lines of credit until January 1, 2020.^[14] In August 2017, the Bureau issued the 2017 HMDA Rule, which, inter alia, temporarily increased the open-end threshold to 500 open-end lines of credit for calendar years 2018 and 2019.^[15] In doing so, the Bureau indicated that the two-year period would allow time for the Bureau to decide, through an additional rulemaking, whether any permanent adjustments to the open-end threshold are needed.^[16]

E. Economic Growth, Regulatory Relief, and Consumer Protection Act and 2018 HMDA Rule

On May 24, 2018, the President signed into law the Economic Growth, Regulatory Relief, and Consumer Protection Act (EGRRCPA).^[17] Section 104(a) of the EGRRCPA amends HMDA section 304(i) by adding partial exemptions from HMDA's requirements for certain insured depository institutions and insured credit unions.^[18] New HMDA section 304(i)(1) provides that the requirements of HMDA section 304(b)(5) and (6) shall not apply with respect to closed-end mortgage loans of an insured depository institution or insured credit union if it originated fewer than 500 closed-end mortgage loans in each of the two preceding calendar years. New HMDA section 304(i)(2) provides that the requirements of HMDA section 304(b)(5) and (6) shall not apply with respect to open-end lines of credit of an insured depository institution or insured credit union if it originated fewer than 500 open-end lines of credit in each of the two preceding calendar years. Notwithstanding the new partial exemptions, new HMDA section 304(i)(3) provides that an insured depository institution must comply with HMDA section 304(b)(5) and (6) if it has received a rating of “needs to improve record of meeting community credit needs” during each of its two most recent examinations or a rating of “substantial noncompliance in meeting community credit needs” on its most recent examination under section 807(b)(2) of the CRA.^[19]

On August 31, 2018, the Bureau issued an interpretive and procedural rule (2018 HMDA Rule) to implement and clarify the partial exemptions established by section 104(a) of the EGRRCPA and effectuate the purposes of the EGRRCPA and HMDA.^[20] In the 2018 HMDA Rule, the Bureau stated that it anticipated that, at a later date, it would initiate a notice-and-comment rulemaking to incorporate the interpretations and procedures into Regulation C and further implement the EGRRCPA.

F. May 2019 Proposal and 2019 HMDA Rule

On May 2, 2019, the Bureau issued a notice of proposed rulemaking (May 2019 Proposal) relating to Regulation C's coverage thresholds and the EGRRCPA partial exemptions and requested public comment.^[21] In the May 2019 Proposal, the Bureau proposed two alternatives to amend Regulation C to increase the current threshold of 25 closed-end mortgage loans for reporting data about closed-end mortgage loans so that institutions originating fewer than either 50 closed-end mortgage loans or, alternatively, 100 closed-end mortgage loans in either of the two preceding calendar years would not have to report such data. The May 2019 Proposal proposed an effective date of January 1, 2020, for the amendment to the closed-end coverage threshold. The May 2019 Proposal also proposed to adjust the coverage threshold for reporting data about open-end lines of credit by (a) extending to January 1, 2022 the current temporary coverage threshold of 500 open-end lines of credit, and (b) setting the permanent coverage threshold at 200 open-end lines of credit upon the expiration of the proposed extension of the temporary coverage threshold. In the May 2019 Proposal, the Bureau also proposed to make changes to effectuate section 104(a) of the EGRRCPA, including incorporating into Regulation C the interpretations and procedures from the 2018 HMDA Rule to implement and clarify section 104(a).

The comment period for the May 2019 Proposal closed on June 12, 2019.^[22] The Bureau received over 300 comments during the initial comment period from lenders, industry trade associations, consumer groups, consumers, members of Congress, and others. Among the comments received were a number of letters expressing concern that the national loan level dataset for 2018 and the Bureau's annual overview of residential mortgage lending based on that data (collectively, the 2018 HMDA Data ^[23] ) would not be available until after the close of the comment period for the May 2019 Proposal. Stakeholders asked to submit comments on the May 2019 Proposal that reflect consideration of the 2018 HMDA Data. To allow for the submission of such comments, on July 31, 2019 the Bureau issued a notice to reopen the comment period on certain aspects of the proposal until October 15, 2019 (July 2019 Reopening Notice).^[24] Specifically, the Bureau reopened the comment period with respect to: (1) The Bureau's proposed amendments to the permanent coverage threshold for closed-end mortgage loans, (2) the Bureau's proposed amendments to the permanent coverage threshold for open-end lines of credit, and (3) the appropriate effective date for any amendment to the closed-end coverage threshold.^[25] The Bureau stated that, after reviewing the comments received by the October 15, 2019 deadline, it anticipated that it would issue in 2020 this separate final rule addressing the permanent thresholds for closed-end mortgage loans and open-end lines of credit.

The Bureau concluded that further comment was not necessary with respect to the other aspects of the May 2019 Proposal.^[26] The Bureau therefore did not reopen the comment period with respect to the May 2019 Proposal's proposed two-year extension of the temporary coverage threshold for open-end lines of credit or the provisions in the May 2019 Proposal that would incorporate the EGRRCPA partial exemptions into Regulation C and further effectuate EGRRCPA section 104(a). The Bureau issued a final rule that finalized as proposed these aspects of the May 2019 Proposal on October 10, 2019 (2019 HMDA Rule).^[27] The Bureau explained in the 2019 HMDA Rule that extending the current threshold of 500 open-end lines of credit for an additional two years would allow the Bureau to consider fully the appropriate level for the permanent open-end coverage threshold for data collected beginning January 1, 2022, after reviewing additional comments relating to that aspect of the May 2019 Proposal. The Bureau also stated that such an extension would ensure that any institutions that are covered under the new permanent open-end coverage threshold would have until January 1, 2022, to comply.

G. HMDA Coverage Under Current Regulation C

The Bureau's estimates of HMDA coverage and the sources used in deriving those estimates are explained in detail in the Bureau's analysis under Dodd-Frank Act section 1022(b) in part VII below.^[28] The Bureau estimates that currently there are about 4,860 financial institutions required to report their closed-end mortgage loans and applications under HMDA. The Bureau estimates that approximately 4,120 of these current reporters are depository institutions and approximately 740 are nondepository institutions. The Bureau estimates that together these financial institutions originated about 6.3 million closed-end mortgage loans in calendar year 2018. The Bureau estimates that among the 4,860 financial institutions that are currently required to report closed-end mortgage loans under HMDA, about 3,250 insured depository institutions and insured credit unions are partially exempt for closed-end mortgage loans under the EGRRCPA, and thus are not required to report a subset of the data points currently required by Regulation C for these transactions.

As explained in more detail in part VII.E.3 and table 4 below, under the current temporary threshold of 500 open-end lines of credit, the Bureau estimates that there are about 333 financial institutions required under HMDA to report about 1.23 million open-end lines of credit. Of these institutions, the Bureau estimates that approximately 318 are depository institutions and approximately 15 are nondepository institutions. The Bureau estimates that none of these 333 institutions are partially exempt under the EGRRCPA.

Absent this final rule, if the open-end coverage threshold were to adjust to 100 on January 1, 2022, the Bureau estimates that the number of reporters would be about 1,014, who in total originate about 1.41 million open-end lines of credit. The Bureau estimates that approximately 972 of these open-end reporters would be depository institutions and approximately 42 would be nondepository institutions. The Bureau estimates that, among the 1,014 financial institutions that would be required to report open-end lines of credit under a threshold of 100, about 595 insured depository institutions and insured credit unions are partially exempt for open-end lines of credit under the EGRRCPA, and thus are not required to report a subset of the data points currently required by Regulation C for these transactions. Additional information on the Bureau's estimates for open-end reporting, including the Bureau's estimates at the permanent threshold of 200 lines of credit, is provided in the section-by-section analysis of § 1003.2(g) and part VII below.

III. Summary of the Rulemaking Process

On May 2, 2019, the Bureau issued the May 2019 Proposal, which was published in the Federal Register on May 13, 2019.^[29] As explained in part II.F above, the comment period on the May 2019 Proposal closed on June 12, 2019, and the Bureau subsequently reopened the comment period with respect to certain aspects of the May 2019 Proposal until October 15, 2019.^[30] In total, the Bureau received over 700 comments in response to the May 2019 Proposal and the July 2019 Reopening Notice from lenders, industry trade associations, consumer groups, consumers, and others.^[31] As discussed in more detail below, the Bureau has considered the comments received both during the initial comment period and in response to the July 2019 Reopening Notice in adopting this final rule.

IV. Legal Authority

The Bureau is issuing this final rule pursuant to its authority under the Dodd-Frank Act and HMDA. Section 1061 of the Dodd-Frank Act transferred to the Bureau the “consumer financial protection functions” previously vested in certain other Federal agencies, including the Board.^[32] The term “consumer financial protection function” is defined to include “all authority to prescribe rules or issue orders or guidelines pursuant to any Federal consumer financial law, including performing appropriate functions to promulgate and review such rules, orders, and guidelines.” ^[33] Section 1022(b)(1) of the Dodd-Frank Act authorizes the Bureau's Director to prescribe rules “as may be necessary or appropriate to enable the Bureau to administer and carry out the purposes and objectives of the Federal consumer financial laws, and to prevent evasions thereof.” ^[34] Both HMDA and title X of the Dodd-Frank Act are Federal consumer financial laws.^[35] Accordingly, the Bureau has authority to issue regulations to implement HMDA.

HMDA section 305(a) broadly authorizes the Bureau to prescribe such regulations as may be necessary to carry out HMDA's purposes.^[36] These regulations may include classifications, differentiations, or other provisions, and may provide for such adjustments and exceptions for any class of transactions, as in the judgment of the Bureau are necessary and proper to effectuate the purposes of HMDA, and prevent circumvention or evasion thereof, or to facilitate compliance therewith.^[37]

V. Section-by-Section Analysis

Section 1003.2 Definitions

2(g) Financial Institution

Regulation C requires financial institutions to report HMDA data. Section 1003.2(g) defines financial institution for purposes of Regulation C and sets forth Regulation C's institutional coverage criteria for depository financial institutions and nondepository financial institutions.^[38] In the 2015 HMDA Rule, the Bureau adjusted the institutional coverage criteria under Regulation C so that depository institutions and nondepository institutions are required to report HMDA data if they: (1) Originated at least 25 closed-end mortgage loans or 100 open-end lines of credit in each of the two preceding calendar years, and (2) meet all of the other applicable criteria for reporting. In the 2017 HMDA Rule, the Bureau amended § 1003.2(g) and related commentary to increase temporarily from 100 to 500 the number of open-end originations required to trigger reporting responsibilities.^[39] In the May 2019 Proposal, the Bureau proposed (1) to amend §§ 1003.2(g)(1)(v)(A) and (g)(2)(ii)(A) and 1003.3(c)(11) and related commentary to raise the closed-end coverage threshold to either 50 or 100 closed-end mortgage loans, and (2) to amend §§ 1003.2(g)(1)(v)(B) and (g)(2)(ii)(B) and 1003.3(c)(12) and related commentary to extend to January 1, 2022, the current temporary open-end coverage threshold of 500 open-end lines of credit and then to set the threshold permanently at 200 open-end lines of credit beginning in calendar year 2022. In the 2019 HMDA Rule, as discussed in part II.F above, the Bureau finalized the proposed amendments relating to the two-year extension of the temporary open-end coverage threshold. The Bureau stated at that time that it anticipated that it would issue a separate final rule in 2020 addressing the permanent thresholds for closed-end mortgage loans and open-end lines of credit.^[40] For the reasons discussed below, the Bureau is raising the closed-end coverage threshold to 100, effective July 1, 2020, and is finalizing the proposed permanent open-end coverage threshold of 200, effective January 1, 2022, upon expiration of the current temporary open-end coverage threshold of 500.^[41]

Legal Authority for Changes to § 1003.2(g)

In the 2015 HMDA Rule, the Bureau adopted the thresholds for certain depository institutions in § 1003.2(g)(1) pursuant to its authority under section 305(a) of HMDA to provide for such adjustments and exceptions for any class of transactions that in the judgment of the Bureau are necessary and proper to effectuate the purposes of HMDA. Pursuant to section 305(a) of HMDA, for the reasons given in the 2015 HMDA Rule, the Bureau found that the exception in § 1003.2(g)(1) is necessary and proper to effectuate the purposes of and facilitate compliance with HMDA. The Bureau found that the provision, by reducing burden on financial institutions and establishing a consistent loan-volume test applicable to all financial institutions, would facilitate compliance with HMDA's requirements.^[42] Additionally, as discussed in the 2015 HMDA Rule, the Bureau adopted the thresholds for certain nondepository institutions in § 1003.2(g)(2) pursuant to its interpretation of HMDA sections 303(3)(B) and 303(5), which require persons other than banks, savings associations, and credit unions that are “engaged for profit in the business of mortgage lending” to report HMDA data. The Bureau stated that it interprets these provisions, as the Board also did, to evince the intent to exclude from coverage institutions that make a relatively small number of mortgage loans.^[43] Pursuant to its authority under HMDA section 305(a), and for the reasons discussed below, the Bureau believes that the final rule's amendments to the thresholds in § 1003.2(g)(1) and (2) are necessary and proper to effectuate the purposes of HMDA and facilitate compliance with HMDA by reducing burden and establishing a consistent loan-volume test, while still providing significant market coverage.^[44]

2(g)(1) Depository Financial Institution

2(g)(1)(v)

2(g)(1)(v)(A)

Section 1003.2(g) defines financial institution for purposes of Regulation C and conditions Regulation C's institutional coverage, in part, on the institution's closed-end mortgage loan origination volume. In the 2015 HMDA Rule, the Bureau added the threshold of 25 closed-end mortgage loans to the pre-existing regulatory coverage scheme for depository institutions.^[45] In the May 2019 Proposal, the Bureau proposed to amend § 1003.2(g)(1)(v)(A) and related commentary to increase the closed-end threshold for depository institutions from 25 to 50 or, alternatively, 100 closed-end mortgage loans. For the reasons discussed below, the Bureau is now amending § 1003.2(g)(1)(v)(A) and related commentary to raise the threshold to 100 closed-end mortgage loans.^[46]

Background on Closed-End Mortgage Loan Threshold for Institutional Coverage of Depository Institutions

HMDA and its implementing regulation, Regulation C, require certain depository institutions (banks, savings associations, and credit unions) to report data about originations and purchases of mortgage loans, as well as mortgage loan applications that do not result in originations (for example, applications that are denied or withdrawn). In adopting the threshold of 25 closed-end mortgage loans in the 2015 HMDA Rule, the Bureau stated that it believed that the institutional coverage criteria should balance the burden on financial institutions of reporting HMDA data against the value of the data reported and that a threshold should be set that did not impair HMDA's ability to achieve its purposes but also did not impose burden on institutions if their data are of limited value.^[47] The Bureau also stated that the closed-end threshold of 25 would meaningfully reduce burden by relieving an estimated 1,400 depository institutions, or 22 percent of depository institutions that previously reported HMDA data, of their obligations to report HMDA data on closed-end mortgage loans.^[48] The Bureau acknowledged that it would be possible to maintain reporting of a significant percentage of the national mortgage market with a closed-end threshold set higher than 25 loans annually and that data reported by some institutions that would satisfy the threshold of 25 closed-end mortgage loans may not be as useful for statistical analysis as data reported by institutions with much higher loan volumes.^[49] However, the Bureau determined that a higher closed-end threshold would have a material negative impact on the availability of data about patterns and trends at the local level and the data about local communities are essential to achieve HMDA's purposes.^[50] The Bureau concluded that, if it were to set the closed-end threshold higher than 25, the resulting loss of data at the local level would substantially impede the public's and public officials' ability to understand access to credit in their communities.^[51]

However, after issuing the 2015 HMDA Rule and the 2017 HMDA Rule, the Bureau heard concerns that lower-volume institutions continue to experience significant burden with the threshold set at 25 closed-end mortgage loans.^[52] For example, several depository institutions recommended that the Bureau use its exemption authority to increase the closed-end loan threshold and stated that the costs of HMDA reporting and its impact on the operations of lower-volume financial institutions do not justify the small amount of data such institutions would report.^[53] In light of the concerns expressed by industry stakeholders regarding the considerable burden associated with reporting the new data points on closed-end mortgage loans required by the 2015 HMDA Rule, in the May 2019 Proposal the Bureau proposed to increase the closed-end threshold for institutions to ensure that it appropriately balances the benefits of the HMDA data reported by lower-volume institutions in furthering HMDA's purposes with the burden on such institutions associated with reporting closed-end data. The Bureau stated in the proposal that increasing the closed-end threshold may provide meaningful burden relief for lower-volume depository institutions without reducing substantially the data reported under HMDA. The Bureau sought comments on how the proposed increase to the closed-end threshold would affect the number of depository institutions required to report data on closed-end mortgage loans, the significance of the data that would not be available for achieving HMDA's purposes as a result of the proposed increase, and the reduction in burden that would result from the proposed increase for depository institutions that would not be required to report.

Comments Received on Closed-End Threshold for Institutional Coverage of Depository Institutions

The Bureau received many comments regarding the proposed alternatives for increasing the closed-end threshold from 25 to 50 or, alternatively, 100 in proposed § 1003.2(g)(1)(v)(A). Except for comments related to the EGRRCPA, commenters typically did not distinguish between their recommended closed-end threshold for depository institutions under § 1003.2(g)(1)(v)(A) and their recommended closed-end threshold for nondepository institutions under § 1003.2(g)(2)(ii)(A).

Many commenters, including most financial institutions and national and State trade associations that commented, supported increasing the closed-end threshold. Most of these commenters discussed the burden of collecting and reporting HMDA data, and despite acknowledging the importance of HMDA data, stated that the cost of complying with regulations has affected their ability to serve their communities. Many industry commenters stated that the burden of complying with HMDA requirements is exacerbated in smaller financial institutions due to fewer staff and a lack of automated processes. A number of small financial institutions stated that they have only a few employees who work in mortgage lending and that these employees spend a considerable amount of time on HMDA compliance, including collecting and entering HMDA data into the appropriate software system and reviewing the data for accuracy. One national trade association added that many small financial institutions operate in geographic areas with shortages of compliance professionals. Several industry commenters also noted that the economies of scale that larger financial institutions can leverage are generally not available to small financial institutions. Many small financial institutions stated that, if the Bureau increased the closed-end threshold and thus excluded them from HMDA's coverage, the significant burden relief would allow their staff to focus on serving customers.

A national trade association stated that a significant number of small financial institutions limit or no longer offer specific mortgage products due to the increased regulatory burden and legal risks associated with such loans. This commenter stated that certain institutions manage their mortgage lending to stay below the threshold for HMDA reporting, which ultimately leaves customers with fewer lending options, and suggested that an increase in the closed-end threshold could increase the flow of credit by small banks into their communities. For example, one small financial institution suggested that, if it did not have to collect HMDA data, the resulting decrease in compliance costs would allow it to maintain a program that provides mortgages to low-to-moderate income families.

Several industry commenters stated that the loss of HMDA data as a result of an increase in the closed-end threshold would not impact the ability to identify potentially discriminatory lending or areas in need of public sector investment. One national trade association stated that institutions that would qualify for the exclusion under the thresholds proposed by the Bureau were extremely small market participants with limited loan volumes and that data on their lending patterns could be obtained through the examination process. This commenter also suggested that any type of fair lending peer comparisons using the HMDA data could still be accurate because, under the proposed threshold of 100, at least 96 percent of total originations would be retained. Many small financial institutions stated that, in their fair lending exams, their regulators have relied on HMDA data, but also on transaction testing data and staff interviews, partly because of the limited number of mortgage applications these institutions receive. These commenters stated further that their regulators also retain full access to their lending files and data needed for their fair lending assessment. A number of commenters, including many small financial institutions and a State trade association, stated that small financial institutions are eligible for partial exemptions under the EGRRCPA and thus are already exempt from reporting much of the data on their credit decisions that would signal lending disparities, such as pricing information and credit scores. The State trade association further stated that examiners already need to review such institutions' files, rather than relying on their HMDA data, to identify potentially discriminatory lending patterns. A State trade association expressed the belief that, due to budget constraints for some local governments and housing authorities, HMDA data are not considered in the distribution of public sector investments in certain areas. Moreover, this commenter stated that, in areas where government authorities do consider HMDA data in making public investment decisions, HMDA data from lower-volume institutions make up a small percentage of the overall lending data within the area and thus do not impact such investment decisions.

Relying on the reasons described above, most of the commenters supporting the proposed increase to the closed-end threshold stated that they preferred the proposed threshold of 100 closed-end mortgage loans over the proposed threshold of 50 closed-end mortgage loans. In some cases, commenters urged the Bureau to consider increasing the closed-end threshold even higher, such as to 250, 500, or 1,000. A national trade association recommended increasing the closed-end threshold to 500 to harmonize HMDA's coverage requirements with the threshold for the EGRRCPA partial exemptions. A trade association and some small financial institutions suggested that the Bureau increase the threshold to 1,000 closed-end mortgage loans, expressing the belief that the Bureau's proposal did not go far enough to distinguish small lending institutions from larger institutions in the mortgage market. The trade association reasoned that increasing the threshold to at least 1,000 closed-end mortgage loans would meaningfully reduce regulatory burden associated with HMDA compliance and allow institutions to direct their cost savings towards improving customer service, increasing consumer-friendly products, and continuing to invest in their communities.

Other commenters, including many community organizations, consumer advocates, research organizations, and individuals, opposed the Bureau's proposal to increase the closed-end threshold. These commenters stated that an increase to the closed-end threshold, which would result in a decrease in HMDA data, would imperil HMDA's purpose of assessing whether financial institutions are meeting the housing needs of their communities. For example, one research organization stated that robust, quality data are critical to the work of regulators and community reinvestment advocates in assessing how well institutions are serving their communities. One commenter stated that lenders rely on HMDA data for internal fair lending and community reinvestment compliance efforts and to assess their performance relative to that of their peers and competitors. A comment letter from 19 U.S. Senators stated that the Bureau's proposal ignores existing exemptions that already reduce the usefulness of HMDA data in many communities. These Senators pointed out that the 2015 HMDA Rule exempted 22 percent of depository institutions that had previously been required to report HMDA data, which resulted in a significant loss of data in certain census tracts. They also noted that an underlying purpose of HMDA is to show how institutions are serving local communities and stated that, even if the loss of data from smaller-volume institutions would be limited when compared to the overall market, the loss of the data would have a real and meaningful impact for residents of affected communities. In a joint comment letter, a large number of consumer groups, civil rights groups, and other organizations expressed a similar concern that increasing the threshold would result in a large loss of HMDA reporting that would otherwise provide a view of lending trends in underserved areas.

Many consumer groups, civil rights groups, and other organizations also discussed the importance of HMDA data for transparency and accountability, noting that the public visibility of HMDA data has motivated financial institutions to increase lending to traditionally underserved borrowers and communities. They expressed concern that the smaller institutions that would no longer be required to report closed-end data at the proposed higher thresholds disproportionately lend in underserved neighborhoods and that the proposed threshold increase would result in a more notable decrease in closed-end data for distressed urban areas, rural areas, tribal areas, communities of color, and neighborhoods that have a high number of immigrants. These commenters asserted that for decades the public has used HMDA data to uncover and address redlining and other fair lending and fair housing violations and stated that an increase in the threshold would make identifying such practices more difficult. A consumer advocacy organization stated that lenders offering unfavorable and unsustainable loan terms and refinance loans in the period just before the financial crisis disproportionally targeted certain groups. Many commenters also stated that an increase in the closed-end threshold would impact the public's ability to evaluate whether public investments are successful in revitalizing struggling areas. For example, one community group noted that the loss of HMDA data from banks and credit unions operating in rural towns and communities would result in less information about where capital is being deployed in those areas.

Many commenters who were opposed to increasing the closed-end threshold pointed out the impact that an increase to the closed-end threshold would have on the work of Federal and State agencies. They stated that raising the thresholds would compromise enforcement work against unfair and deceptive lending because there would be less data available to monitor such activity. Similarly, commenters stated that examinations pursuant to the Community Reinvestment Act (CRA) would become more burdensome because examiners would likely need to be onsite to review data rather than using publicly available HMDA data. A State attorney general commenter suggested that under the Bureau's proposal, major lenders would be exempt from HMDA reporting and that the lack of data from such lenders would affect its ability to ensure that all of its residents are able to access affordable credit free of discrimination. In addition, this commenter stated that the proposed closed-end threshold increase would all but eliminate its ability to enforce fair lending laws in “hyper-localized” markets in rural areas because small, local lenders are disproportionately represented in rural areas (the commenter did not define the term “hyper-localized,” but the Bureau understands this term as referring to markets that are limited to a small geographic area).

Several commenters also expressed concerns about the impact that an increase in the closed-end threshold would have on visibility into specific loan products, such as loans for multifamily housing and manufactured housing. For example, a State attorney general expressed concerns that a threshold higher than the current threshold of 25 would limit data reported on multifamily dwellings that provide a significant source of affordable housing in urban areas across the State. Another commenter opposed an increase to the closed-end threshold because of concerns that it would decrease visibility into manufactured housing loans, reasoning that there are a small number of lenders that make such loans.

Many commenters who opposed an increase in the closed-end threshold also stated that the cost savings that would result from excluding lenders from HMDA reporting would be modest. These commenters asserted that the burden of HMDA reporting is not so significant as to make up for the loss of data that would otherwise be available at the current threshold of 25 closed-end mortgage loans for the public and regulators to monitor fair lending compliance. These commenters stated further that the estimates of potential cost savings provided by the Bureau in the proposal were too high and that the cost of HMDA reporting is low. First, they stated that most of the lenders that would be excluded under the Bureau's proposed rule are already exempt from reporting many of the new HMDA data points because they qualify for partial exemptions under the EGRRCPA and would therefore be reporting data that lenders have been reporting for decades. Second, these commenters noted that much of the data reportable under HMDA must be collected for other rules, including the TILA-RESPA Integrated Disclosure and Ability-to-Repay/Qualified Mortgage rules, and ordinary underwriting standards. Finally, these commenters stated that HMDA data should be collected as a matter of sound banking practices and asserted that reporting the data is unlikely to require substantial resources given modern technological advancements.

Final Rule

Pursuant to its authority under HMDA section 305(a) as discussed above, the Bureau is finalizing the closed-end threshold for depository institutions at 100 in § 1003.2(g)(1)(v)(A). As discussed below, the Bureau believes that increasing the closed-end threshold to 100 will provide meaningful burden relief for lower-volume depository institutions while maintaining reporting sufficient to achieve HMDA's purposes.

Since the 2015 HMDA Rule was issued, a few developments have affected the Bureau's analyses of the costs and benefits associated with the closed-end threshold. The Bureau has gathered extensive information regarding stakeholders' experience with the 2015 HMDA Rule, through comments received in this rulemaking and other feedback. As stated above, the Bureau has heard that financial institutions have encountered significant burdens in complying with the rule, and the Bureau is particularly concerned about the increased burdens faced by smaller institutions. Additionally, the Bureau now has access to HMDA data from 2018, which was the first year that financial institutions collected data under the 2015 HMDA Rule, and has used these data in updating and generating the estimates provided in this final rule. With the benefit of this additional information about the 2015 HMDA Rule, and the new data to supplement the Bureau's analyses, the Bureau is now in a better position to assess both the benefits and burdens of the reporting required under the 2015 HMDA Rule.

Another development since the 2015 HMDA Rule is the enactment of the EGRRCPA, which created partial exemptions from HMDA's requirements that certain insured depository institutions and insured credit unions may now use.^[54] The partial exemption for closed-end mortgage loans under the EGRRCPA relieves certain insured depository institutions and insured credit unions that originated fewer than 500 closed-end mortgage loans in each of the two preceding calendar years of the obligation to report many of the data points generally required by Regulation C.^[55] While the EGRRCPA relieves burden for some depository institutions, it does not relieve smaller depository institutions from the burdens of reporting entirely.

The Bureau has considered the appropriate closed-end threshold in light of these developments and the comments received. On balance, the Bureau determines that the threshold of 100 closed-end mortgage loans provides sufficient information on closed-end mortgage lending to serve HMDA's purposes, while appropriately reducing ongoing costs that smaller institutions are incurring under the current threshold. These considerations are discussed in turn below, and additional explanation of the Bureau's cost estimates is provided in the Bureau's analysis under Dodd-Frank Act section 1022(b) in part VII.E.2 below.

Effect on Market Coverage

For this final rule, the Bureau reviewed multiple data sources, including recent HMDA data ^[56] and Reports of Condition and Income (Call Reports), and developed estimates for the two thresholds the Bureau proposed in the alternative, 50 and 100, as well as thresholds of 250 and 500, which many commenters suggested the Bureau consider. The Bureau notes that many of the estimates provided in this final rule differ slightly from the initial estimates provided in the May 2019 Proposal. As discussed below in part VII.E.2, the estimates in this final rule update the initial estimates provided in the May 2019 Proposal with the 2018 HMDA data, which were not available at the time the Bureau developed the May 2019 Proposal. For the May 2019 Proposal, the Bureau used data from 2016 and 2017 with a two-year look-back period covering calendar years 2016 and 2017 to estimate potential reporters and projected the lending activities of financial institutions using their 2017 data as proxies. In generating the updated estimates provided in this final rule, the Bureau has used data from 2017 and 2018 with a two-year look-back period covering calendar years 2017 and 2018 to estimate potential reporters and has projected the lending activities of financial institutions using their 2018 data as proxies. In addition, for the estimates provided in the May 2019 Proposal and in this final rule, the Bureau restricted the projected reporters to only those that actually reported data in the most recent year of HMDA data considered (2017 for the May 2019 Proposal and 2018 for this final rule).^[57]

The estimates below compare coverage under these thresholds to coverage under the current threshold of 25 closed-end mortgage loans. The estimated effect that increasing the threshold from 25 closed-end mortgage loans to various higher thresholds would have on the overall HMDA data, local-level HMDA data, and specific loan products reported are discussed in turn below.

Effect on covered depository institutions and reportable originations. For this final rule, the Bureau has considered the impact that the two alternative proposed thresholds and other possible thresholds would have on the number of depository institutions that would report HMDA data and how many originations they would report.^[58] The Bureau estimates that if the closed-end threshold were increased from 25 to 50, approximately 3,400 out of approximately 4,120 depository institutions covered under the current threshold of 25 (or approximately 85 percent) would continue to be required to report HMDA data on closed-end mortgage loans. Further, the Bureau estimates that if the threshold were increased from 25 to 50, approximately 98.9 percent or approximately 2.89 million total originations of closed-end mortgage loans in current market conditions reported by depository institutions under the current Regulation C coverage criteria would continue to be reported.^[59]

The Bureau estimates that with the closed-end threshold set at 100 under the final rule, approximately 2,480 out of approximately 4,120 depository institutions covered under the current threshold of 25 (or approximately 60 percent) will continue to be required to report HMDA data on closed-end mortgage loans. Further, the Bureau estimates that when the final rule increases the closed-end threshold from 25 to 100 loans, approximately 96 percent or approximately 2.79 million total originations of closed-end mortgage loans in current market conditions reported by depository institutions under the current Regulation C coverage criteria will continue to be reported.^[60]

The Bureau also generated estimates for closed-end thresholds higher than those that the Bureau proposed. These estimates indicate that the decrease in the number of depository institutions that would be required to report HMDA data and the resulting decrease in the HMDA data that would be reported becomes more pronounced at thresholds higher than 100. For example, if the closed-end threshold were set at 250, the Bureau estimates that approximately 1,340 out of approximately 4,120 depository institutions covered under the current threshold of 25 (or approximately 32 percent) would continue to be required to report HMDA data on closed-end mortgage loans. Further, the Bureau estimates that, if the threshold were set at 250 closed-end mortgage loans, approximately 89 percent or approximately 2.57 million total originations of closed-end mortgage loans in current market conditions reported by depository institutions under the current Regulation C coverage criteria would continue to be reported.^[61]

The Bureau estimates that if the closed-end threshold were set at 500, approximately 720 out of approximately 4,120 depository institutions covered under the current threshold of 25 (or approximately 18 percent) would continue to be required to report HMDA data on closed-end mortgage loans. Further, the Bureau estimates that, if the threshold were set at 500, approximately 81 percent or approximately 2.34 million total originations of closed-end mortgage loans in current market conditions reported by depository institutions under the current Regulation C coverage criteria would continue to be reported.^[62]

As described above, many commenters opposed increasing the closed-end threshold because of concerns that there would be less data with which to further HMDA's statutory purposes. While the Bureau recognizes that the increase in the threshold to 100 closed-end mortgage loans will reduce market coverage compared to the current threshold of 25, the Bureau estimates that information covering approximately 96 percent of loans currently reported will still be available to further HMDA's statutory purposes. The Bureau believes that the small amount of HMDA data obtained from lower-volume depository institutions does not justify the costs imposed on those institutions to comply with HMDA data reporting requirements.

Although a commenter suggested the Bureau increase the closed-end threshold to 500 to harmonize the thresholds with the EGRRCPA provisions, the Bureau determines that it is not appropriate to set the closed-end threshold at 500. Doing so would provide a complete exclusion from reporting all closed-end data for institutions below the threshold of 500, even though Congress opted to provide only a partial exemption at the threshold of 500, and would extend that complete exclusion to institutions that Congress did not include in even the partial exemption. The EGRRCPA partial exemption already relieves most lenders originating fewer than 500 closed-end loans in each of the two preceding calendar years from the requirement to report many data points associated with their closed-end transactions.^[63] Providing a complete exclusion at 500 closed-end mortgage loans would exclude visibility into approximately 82 percent of institutions covered under the current threshold of 25 closed-end mortgage loans, which would result in a significant loss of coverage in closed-end lending and negatively impact the utility of HMDA data.

Effect on HMDA data at the local level. For the proposal and this final rule, the Bureau reviewed estimates at varying closed-end thresholds to examine the potential effect on available data at the census tract level. The Bureau's estimates of the effect on reportable HMDA data at the census tract level comprise both depository institutions and nondepository institutions. The Bureau estimates that, if the closed-end threshold were raised from 25 to 50, approximately 74,300 out of the approximately 74,600 total census tracts in which HMDA data are currently reported, or over 99 percent, would retain more than 80 percent of reportable HMDA data, relative to the current threshold. The Bureau estimates there would be a decrease of at least 20 percent of reportable HMDA data on closed-end mortgage loans relative to the current threshold in approximately 300 out of approximately 74,600 total census tracts in which HMDA data are currently reported, or less than one-half of 1 percent. With respect to low-to-moderate income census tracts, if the closed-end threshold were raised from 25 to 50, the Bureau estimates that, relative to the current threshold of 25, over 99 percent of low-moderate income census tracts would retain more than 80 percent of reportable HMDA data, and there would be at least a 20 percent decrease in reportable HMDA data on closed-end mortgage loans in less than 1 percent of such tracts. In addition, the Bureau examined the effects on rural census tracts and estimates that, relative to the current threshold of 25, more than 98 percent of rural tracts would retain more than 80 percent of reportable HMDA data, and there would be at least a 20 percent decrease in reportable HMDA data in just over 1 percent of rural tracts.^[64]

With the threshold of 100 closed-end mortgage loans established by this final rule, the Bureau estimates that, relative to the current threshold of 25, approximately 73,400 census tracts out of approximately 74,600 total census tracts in which HMDA data are currently reported, or over 98 percent, would retain more than 80 percent of reportable HMDA data. The Bureau estimates that there will be a decrease of at least 20 percent of reportable HMDA data on closed-end mortgage loans relative to the current threshold in about 1,200 out of approximately 74,600 total census tracts in which HMDA data are currently reported, or under 2 percent. For low-to-moderate income census tracts, with the threshold of 100 closed-end mortgage loans, the Bureau estimates that, relative to the current threshold of 25, approximately 97 percent of such tracts will retain more than 80 percent of reportable HMDA data, and there will be a decrease of at least 20 percent of reportable HMDA data in approximately 3 percent of such tracts. The Bureau also estimates that, relative to the current threshold of 25, approximately 95 percent of rural tracts will retain more than 80 percent of reportable HMDA data, and there will be a decrease of at least 20 percent of reportable HMDA data in approximately 5 percent of such tracts.^[65]

The Bureau's estimates also reflect that the effect on data available at the census tract level would become more pronounced at closed-end mortgage loan thresholds above 100. For example, the Bureau estimates that, if the threshold were increased from 25 to 250 loans, approximately 68,800 out of the approximately 74,600 total census tracts in which HMDA data are currently reported, or over 92 percent, would retain more than 80 percent of reportable HMDA data, relative to the current threshold. The Bureau estimates there would be a decrease of at least 20 percent of reportable HMDA data on closed-end mortgage loans in about 5,800 out of approximately 74,600 total census tracts in which HMDA data are currently reported, or about 8 percent of those census tracts, relative to the current threshold. For low-to-moderate income census tracts, if the threshold were increased from 25 to 250, the Bureau estimates that approximately 90 percent of tracts would retain more than 80 percent of reportable HMDA data, and there would be a decrease of at least 20 percent of reportable HMDA data in approximately 10 percent of such tracts, relative to the current threshold. For rural tracts, the Bureau estimates that approximately 81 percent of tracts would retain more than 80 percent of reportable HMDA data, and there would be a decrease of at least 20 percent of reportable HMDA data in approximately 19 percent of such tracts, relative to the current threshold.^[66]

Further, the Bureau estimates that, if the closed-end threshold were increased from 25 to 500 loans, approximately 60,500 out of approximately 74,600 total census tracts in which HMDA data are currently reported, or approximately 81 percent, would retain more than 80 percent of reportable HMDA data. The Bureau estimates there would be a decrease of at least 20 percent of reportable HMDA data on closed-end mortgage loans in approximately 14,100, or 19 percent of the total number of census tracts in which HMDA data are currently reported, relative to the current threshold of 25. For low-to-moderate income census tracts, the Bureau estimates that, if the threshold were increased from 25 to 500, over 78 percent of such tracts would retain more than 80 percent of reportable HMDA data, and there would be a decrease of at least 20 percent of reportable HMDA data in over 21 percent of such tracts. For rural census tracts, the Bureau estimates that approximately 62 percent of such tracts would retain more than 80 percent of reportable HMDA data, and there would be a decrease of at least 20 percent of reportable HMDA data in approximately 37 percent of such tracts, relative to the current threshold.^[67]

The Bureau recognizes that any loan-volume threshold will affect individual markets differently, depending on the extent to which smaller creditors service individual markets and the market share of those creditors. The Bureau concludes, however, based on the estimates provided above, that the threshold of 100 closed-end loans adopted in this final rule will provide substantial visibility into rural and low-to-moderate income tracts and permit the public and public officials to identify patterns and trends at the local level. At the same time, the Bureau is concerned that the higher closed-end mortgage loan-volume thresholds above 100 suggested by industry commenters could have a material negative impact on the availability of data about patterns and trends at the local level and could affect the availability of data necessary to achieve HMDA's purposes.

Specific types of data. The Bureau has also considered the impact that increasing the threshold could have on data related to specific types of closed-end lending mentioned by commenters, such as applications and originations related to multifamily housing and manufactured housing lending. The Bureau estimates that with the closed-end threshold increased from 25 to 100 under the final rule, approximately 87 percent of multifamily loan applications and originations will continue to be reported by depository and nondepository institutions combined, when compared to the current threshold of 25 closed-end mortgage loans in today's market conditions. Regarding the effect on manufactured housing data, the Bureau estimates that at a threshold of 100 closed-end mortgage loans, approximately 96 percent of loans and applications related to manufactured housing will continue to be reported by depository and nondepository institutions combined, when compared to the current threshold of 25 closed-end mortgage loans in today's market conditions. Increasing the threshold above 100 would have a more pronounced impact on data regarding both multifamily housing and manufactured housing lending. Although less data will be available regarding multifamily housing and manufactured housing lending at the threshold of 100 than at the current threshold, the Bureau believes that the limited decreases in the amount of data are justified by the benefits of relieving smaller-volume institutions of the burdens of HMDA reporting.

Ongoing Cost Reduction From Threshold of 100

As noted above, small financial institutions and trade associations commented on the cost of HMDA reporting, suggesting that compliance costs have had an impact on the ability of small financial institutions to serve their customers and communities. For the proposal and this final rule, the Bureau developed estimates for depository and nondepository institutions combined to determine the savings in annual ongoing costs at various thresholds.^[68] These estimates illustrate the cost savings under the various thresholds when compared to the current threshold of 25.

The Bureau estimates that if the closed-end threshold were set at 50, institutions that originate between 25 and 49 closed-end mortgage loans would save approximately $3.7 million per year in total annual ongoing costs, relative to the current threshold of 25.^[69] The Bureau estimates that with a threshold of 100 closed-end mortgage loans established by the final rule, institutions that originate between 25 and 99 closed-end mortgage loans will save approximately $11.2 million per year, relative to the current threshold of 25.^[70] With a threshold of 250 or 500 closed-end mortgage loans, the Bureau estimates that institutions would save approximately $27.2 million and $45.4 million, respectively, relative to the current threshold of 25. Based on the Bureau's estimates, the Bureau believes that the cost reduction from increasing the threshold from 25 to 100 closed-end mortgage loans is significant and more than double the cost savings that a threshold of 50 closed-end mortgage loans would have provided, providing meaningful cost savings to institutions.

The Bureau recognizes that the estimated ongoing costs savings associated with increasing the threshold from 25 to 100 closed-end loans are less than they would have been absent the relief provided by the EGRRCPA. Nonetheless, the Bureau determines that these ongoing cost savings will provide meaningful burden reduction to smaller institutions that are currently covered at the threshold of 25 closed-end loans but will be excluded from closed-end reporting under the increased threshold in this final rule. Avoiding the imposition of such costs for these affected institutions may also enable smaller institutions to focus on lending activities and serving their communities, as suggested by some commenters.

The Bureau concludes that increasing the closed-end threshold to 100 will provide meaningful burden relief for lower-volume depository institutions while maintaining reporting sufficient to achieve HMDA's purposes. As discussed above, the Bureau has heard of significant burdens in complying with the 2015 HMDA Rule, especially from smaller institutions, and the Bureau has been able to confirm the impact of the rule and any potential changes to the closed-end threshold, based on the new 2018 HMDA data. The Bureau recognizes that there is some loss of data at this threshold but believes that it strikes the right balance between the burden of collecting and reporting and the benefit of HMDA data. The Bureau's estimates reflect an estimated decrease of about 4 percent of total originations by depository institutions reportable under the current closed-end threshold of 25 in today's market conditions. According to the Bureau's estimates, about 60 percent of current HMDA reporters that are depository institutions will continue to report HMDA data, and only approximately 1,200 out of 74,600 census tracts will reflect a decrease of at least 20 percent in HMDA data from depository and nondepository institutions. Therefore, the Bureau believes that the decrease in data from institutions that will be newly excluded with the closed-end threshold set at 100 is justified by the significant reduction in burden for the approximately 1,640 lower-volume depository institutions that will no longer be required to report HMDA data when compared to the current threshold of 25. The threshold of 100 closed-end mortgage loans balances the benefits and burdens of covering institutions engaged in closed-end mortgage lending by retaining significant coverage of the closed-end market while excluding from coverage smaller institutions whose limited closed-end data would be of lesser utility in furthering HMDA's purposes. For the reasons stated above, the Bureau is amending § 1003.2(g)(1)(v)(A) and comments 2(g)-1 and 2(g)-5 to adjust the threshold to 100 closed-end mortgage loans. As discussed in part VI.A below, the change to the closed-end threshold will take effect on July 1, 2020, to provide relief quickly.^[71]

2(g)(1)(v)(B)

Section 1003.2(g) defines financial institution for purposes of Regulation C and conditions Regulation C's institutional coverage, in part, on the institution's open-end line of credit origination volume. In the 2015 HMDA Rule, the Bureau established the threshold at 100 open-end lines of credit and required financial institutions that originate at least 100 open-end lines of credit in each of the two preceding calendar years to report data on open-end lines of credit.^[72] In the 2017 HMDA Rule, the Bureau amended § 1003.2(g) to increase for two years (calendar years 2018 and 2019) the open-end threshold from 100 to 500 open-end lines of credit. In the May 2019 Proposal, the Bureau proposed to amend § 1003.2(g)(1)(v)(B) and comments 2(g)-3 and -5, effective January 1, 2020, to extend until January 1, 2022, the temporary open-end institutional coverage threshold for depository institutions of 500 open-end lines of credit. Upon expiration of this temporary threshold, the Bureau proposed to increase the permanent threshold from 100 to 200 open-end lines of credit.^[73] The Bureau sought comments on how the proposed temporary and permanent increases to the open-end threshold would affect the number of financial institutions required to report data on open-end lines of credit, the significance of the data that would not be available for achieving HMDA's purposes as a result of the proposed increases, and the reduction in burden that would result from the proposed increases for institutions that would not be required to report. In the 2019 HMDA Rule, the Bureau finalized the proposed extension of the temporary open-end institutional coverage threshold for depository institutions of 500 open-end lines of credit in § 1003.2(g)(1)(v)(B) until January 1, 2022.^[74] For the reasons discussed below, the Bureau is now finalizing the proposed amendments to § 1003.2(g)(1)(v)(B) and comments 2(g)-3 and -5 to increase the permanent threshold from 100 to 200 open-end lines of credit, effective January 1, 2022.

Background on Reporting Data Concerning Open-End Lines of Credit Under the 2015 HMDA Rule and the 2017 HMDA Rule

By its terms, the definition of “mortgage loan” in HMDA covers all loans secured by residential real property and home improvement loans, whether open- or closed-end.^[75] However, home-equity lines of credit were uncommon in the 1970s and early 1980s when Regulation C was first issued, and the Board's definition of mortgage loan covered only closed-end loans. In 2000, in response to the increasing importance of open-end lending in the housing market, the Board proposed to revise Regulation C to require mandatory reporting of all home-equity lines of credit, which lenders had the option to report.^[76] However, the Board's 2002 final rule left open-end reporting voluntary, as the Board determined that the benefits of mandatory reporting relative to other then-proposed amendments (such as collecting information about higher-priced loans) did not justify the increased burden.^[77]

As discussed in the 2015 HMDA Rule, open-end mortgage lending continued to increase in the years following the Board's 2002 final rule, particularly in areas with high home-price appreciation.^[78] In light of that development and the role that open-end lines of credit may have played in contributing to the financial crisis,^[79] the Bureau decided in the 2015 HMDA Rule to require reporting of dwelling-secured, consumer purpose open-end lines of credit,^[80] concluding that doing so was a reasonable interpretation of “mortgage loan” in HMDA and necessary and proper to effectuate the purposes of HMDA and prevent evasions thereof.^[81]

As noted in the 2015 HMDA Rule, in expanding coverage to include mandatory reporting of open-end lines of credit, the Bureau recognized that doing so would impose one-time and ongoing operational costs on reporting institutions, that the one-time costs of modifying processes and systems and training staff to begin open-end line of credit reporting likely would impose significant costs on some institutions, and that institutions' ongoing reporting costs would increase as a function of their open-end lending volume.^[82] The Bureau sought to avoid imposing these costs on small institutions with limited open-end lending, where the benefits of reporting the data did not justify the costs of reporting.^[83] In seeking to draw such a line, the Bureau acknowledged that it was handicapped by the lack of available data concerning open-end lending.^[84] This created challenges both in estimating the distribution of open-end origination volume across financial institutions and in estimating the one-time and ongoing costs that institutions of various sizes would be likely to incur in reporting data on open-end lending.

To estimate the one-time and ongoing costs of reporting data under HMDA in the 2015 HMDA Rule, the Bureau identified seven “dimensions” of compliance operations and used those to define three broadly representative financial institutions according to the overall level of complexity of their compliance operations: “tier 1” (high-complexity), “tier 2” (moderate-complexity), and “tier 3” (low-complexity).^[85] The Bureau then sought to estimate one-time and ongoing costs for a representative institution in each tier.^[86]

The Bureau recognized in the 2015 HMDA Rule that the one-time cost of reporting open-end lines of credit could be substantial because most financial institutions had not reported open-end lines of credit and thus would have to develop completely new systems to begin reporting these data. As a result, there would be one-time costs to create processes and systems for open-end lines of credit.^[87] However, for low-complexity tier 3 institutions, the Bureau believed that the additional one-time costs of open-end reporting would be relatively low. Because these institutions are less reliant on information technology systems for HMDA reporting and they may process open-end lines of credit on the same system and in the same business unit as closed-end mortgage loans, their one-time costs would be derived mostly from new training and procedures adopted for the overall changes in the final rule, not distinct from costs related to changes in reporting of closed-end mortgage loans.^[88]

The Bureau acknowledged in the 2015 HMDA Rule that ongoing costs for open-end reporting vary by institutions due to many factors, such as size, operational structure, and product complexity, and that this variance makes it impossible to provide complete and definitive cost estimates.^[89] At the same time, the Bureau stated that it believed that the HMDA reporting process and ongoing operational cost structure for open-end reporting would be fundamentally similar to closed-end reporting.^[90] Thus, using the ongoing cost estimates developed for closed-end reporting, the Bureau estimated that for a representative high-complexity tier 1 institution the ongoing operational costs would be $273,000 per year; for a representative moderate-complexity tier 2 institution $43,400 per year; and for a representative low-complexity tier 3 institution $8,600 per year.^[91] These translated into costs per HMDA record of approximately $9, $43, and $57 respectively.^[92] The Bureau acknowledged that, precisely because no good source of publicly available data existed concerning open-end lines of credit, it was difficult to predict the accuracy of the Bureau's cost estimates but also stated its belief that these estimates were reasonably reliable.^[93]

Drawing on all of these estimates, the Bureau decided in the 2015 HMDA Rule to establish an open-end threshold that would require institutions that originate 100 or more open-end lines of credit in each of the two preceding calendar years to report data on such lines of credit. The Bureau estimated that this threshold would avoid imposing the burden of establishing mandatory open-end reporting on approximately 3,000 predominantly smaller-sized institutions with low-volume open-end lending ^[94] and would require reporting by 749 financial institutions, all but 24 of which would also report data on their closed-end mortgage lending.^[95] The Bureau explained in the 2015 HMDA Rule that it believed this threshold appropriately balanced the benefits and burdens of covering institutions based on their open-end mortgage lending.^[96] However, as discussed in the 2017 HMDA Rule, the Bureau lacked robust data for the estimates that it used to establish the open-end threshold in the 2015 HMDA Rule.^[97]

The 2017 HMDA Rule explained that, between 2013 and 2017, the number of dwelling-secured open-end lines of credit financial institutions originated had increased by 36 percent.^[98] The Bureau noted that, to the extent institutions that had been originating fewer than 100 open-end lines of credit shared in that growth, the number of institutions at the margin that would be required to report under an open-end threshold of 100 lines of credit would also increase.^[99] Additionally, in the 2017 HMDA Rule, the Bureau explained that information received by the Bureau since issuing the 2015 HMDA Rule had caused the Bureau to question its assumption that certain low-complexity institutions ^[100] process home-equity lines of credit on the same data platforms as closed-end mortgages, on which the Bureau based its assumption that the one-time costs for these institutions would be minimal.^[101] After issuing the 2015 HMDA Rule, the Bureau heard reports suggesting that one-time costs to begin reporting open-end lines of credit could be as high as $100,000 for such institutions.^[102] The Bureau likewise heard reports suggesting that the ongoing costs for these institutions to report open-end lines of credit, which the Bureau estimated would be under $10,000 per year and add under $60 per line of credit, could be at least three times higher than the Bureau had estimated.^[103]

Based on this information regarding one-time and ongoing costs and new data indicating that more institutions would have reporting responsibilities under the 100-loan open-end threshold than estimated in the 2015 HMDA Rule, the Bureau increased for two years (i.e., until January 1, 2020) the open-end threshold to 500 in the 2017 HMDA Rule.^[104] Specifically, the Bureau amended § 1003.2(g)(1)(v)(B) and comments 2(g)-3 and -5, effective January 1, 2018, to increase temporarily the open-end threshold from 100 to 500 and, effective January 1, 2020, to revert to a permanent threshold of 100. This temporary increase was intended to allow the Bureau to collect additional data and assess what open-end threshold would best balance the benefits and burdens of covering institutions.

In the May 2019 Proposal, the Bureau proposed to extend until January 1, 2022, the temporary open-end institutional coverage threshold for depository institutions of 500 open-end lines of credit. Upon expiration of this temporary threshold, the Bureau proposed to set the permanent threshold at 200 open-end lines of credit.^[105] In the 2019 HMDA Rule, the Bureau finalized the proposed two-year extension of the temporary threshold of 500 open-end lines of credit.^[106] The Bureau explained that the extension of the temporary threshold would provide additional time for the Bureau to issue this final rule in 2020 on the permanent open-end threshold and for affected institutions to prepare for compliance with the final rule.^[107]

Comments Received on Permanent Open-End Line of Credit Threshold for Institutional Coverage of Depository Institutions

The Bureau received a number of comments relating to the proposed permanent increase in the open-end threshold from 100 to 200 open-end lines of credit in §§ 1003.2(g) and 1003.3(c)(12). Commenters typically discussed the open-end threshold without distinguishing between the threshold applicable to depository institutions under § 1003.2(g)(1)(v)(B) and the threshold applicable to nondepository institutions under § 1003.2(g)(2)(ii)(B).

Industry commenters generally expressed support for an increase in the permanent open-end threshold, indicating that a threshold of 200 open-end lines of credit would be preferable to the threshold of 100 open-end lines of credit that would otherwise take effect beginning in 2022. Many industry commenters described the significant costs that HMDA data collection and reporting impose on small institutions, and some expressed concern that they might not be able to offer open-end lines of credit at all if the threshold of 100 open-end lines of credit were to take effect. One national trade association and many small financial institutions stated that open-end lines of credit are crucial products for borrowers and expressed concern that the costs associated with reporting such lines of credit would make them unprofitable, leading banks to either discontinue offering such loans or to pass on cost increases to consumers. Many industry commenters also suggested that higher thresholds would allow institutions to focus on making loans in the communities they serve rather than diverting resources to HMDA compliance. Another national trade association stated that lenders may seek to manage their origination volumes to stay below the applicable open-end threshold, which could limit consumers' access to credit. This commenter stated that compliance with HMDA requires specialized staffing and training as well as dedicated software, policies, and procedures, and that an institution's decision to exceed the origination volume that triggers open-end reporting would involve a careful assessment of the time, cost, and risk associated with implementing and supporting ongoing open-end reporting. Several commenters stated that there would be significant costs to implementing open-end reporting for institutions that have not previously reported such transactions, with one small financial institution stating that it originated between 100 and 200 open-end lines of credit annually and that reporting such loans would entail considerable effort because these transactions are processed by a different department and system than its reportable closed-end mortgage loans.

Some industry commenters advocating for an increase in the open-end threshold asserted that data on open-end lines of credit are of limited value in serving HMDA's statutory purposes. A few national trade associations stated that open-end lines of credit provide little information about whether lenders are serving the housing needs of their communities because such loans are generally used for non-housing related purposes, such as paying for educational expenses or consolidating outstanding debt. One national trade association stated that open-end lending data are of limited value for fair lending purposes because of the unique features typically present in such transactions and because only certain borrowers—existing homeowners with equity in their homes—can obtain them.

A large number of industry commenters recommended that the Bureau make the temporary threshold of 500 open-end lines of credit permanent or raise the threshold even higher, such as to 1,000. These commenters noted that based on the Bureau's estimates, maintaining the current threshold of 500 open-end lines of credit would relieve approximately 280 institutions from reporting open-end data with only a 6 percent decrease in the overall number of open-end lines of credit reported relative to the proposed permanent threshold of 200. One State trade association expressed concern that the limited increase in open-end data reported at a permanent threshold of 200 as compared to 500 open-end lines of credit would not justify the costs for the institutions that would be newly required to report open-end data. One national trade association stated that many smaller institutions originate close to 500 open-end lines of credit annually and that if the permanent threshold were set at 200 these lenders might curtail their open-end lending to avoid incurring the additional compliance costs associated with open-end reporting. A few industry commenters stated that the continuity that would be provided by a permanent 500 open-end threshold would be valuable and questioned why the Bureau would set the open-end threshold at 500 for several years but not retain this threshold. Although not part of the May 2019 Proposal, many industry commenters recommended that the Bureau return to optional rather than mandatory reporting of open-end lines of credit.

Other commenters, including many consumer and civil rights groups, a bank, a State attorney general, and some members of Congress, expressed opposition to the proposed increase from 100 to 200 in the permanent open-end threshold based on their concerns about the consequences of excluding more institutions and open-end lines of credit from HMDA reporting. Many of these commenters stated that, in the years before the 2008 financial crisis, abuses pervaded in open-end lending that resulted in distress or foreclosure for large numbers of homeowners. A State attorney general noted that open-end lines of credit were often extended simultaneously with closed-end home purchase loans in place of down payments, thus bypassing the need for borrowers to obtain private mortgage insurance and creating higher debt obligations that increased the risk to both closed-end mortgage lenders and borrowers. A large number of consumer groups, civil rights groups, and other organizations noted in a joint comment letter the Bureau's estimates in the May 2019 Proposal that increasing the permanent threshold from 100 to 200 open-end lines of credit would exempt 401 lenders originating 69,000 open-end lines of credit from reporting such data under HMDA. These commenters expressed concern that too many lenders and open-end lines of credit might escape public scrutiny at such a higher permanent threshold and thus make it more likely that events similar to those that led to the 2008 financial crisis would occur again.

These consumer groups, civil rights groups, and other organizations also stated that the 2018 HMDA Data indicated that open-end lines of credit have a high incidence of features that can be risky for borrowers, particularly when layered on top of one another. They explained that the 2018 HMDA Data show that 77 percent of open-end lines of credit have adjustable rates, 50 percent feature interest-only payments, and 28 percent include prepayment penalties. These commenters also stated that the 2018 HMDA Data show that the median interest rate, as well as the interest rate at the 95th percentile, was significantly higher for open-end lines of credit than for closed-end mortgage loans and suggested that the most vulnerable borrowers were obtaining the open-end lines of credit with the highest interest rates.

These commenters stated further that the increase in open-end lending between 2013 and 2017 discussed in the May 2019 Proposal supports maintaining the permanent threshold of 100 open-end lines of credit to increase visibility into open-end lending. A State attorney general expressed concern that the May 2019 Proposal did not provide a rationale as to how decreasing open-end reporting by increasing the permanent open-end threshold serves the purposes of HMDA. This commenter stated that the Bureau's analysis instead focused almost entirely on the cost to lenders associated with open-end reporting. Some members of Congress stated that data on open-end lines of credit remain limited and noted that the Bureau had to consult multiple sources to estimate the impact of the proposed changes to the open-end threshold. These commenters expressed concern that the Bureau would reduce future open-end reporting based on limited data, particularly in light of the local and national concerns related to open-end lending prior to the financial crisis in 2008 cited by the Bureau in the 2015 HMDA Rule.

Final Rule

The Bureau has considered the comments received and, pursuant to its authority under HMDA section 305(a) as discussed above, has decided to increase the permanent open-end threshold to 200 open-end lines of credit, as proposed. As discussed below, the increase in the permanent threshold from 100 to 200 open-end lines of credit will provide meaningful burden relief for smaller institutions while still providing significant market coverage of open-end lending.

As discussed in the May 2019 Proposal, several developments since the Bureau issued the 2015 HMDA Rule have affected the Bureau's analyses of the costs and benefits associated with the open-end threshold. As explained in more detail in part VII below, the estimates the Bureau used in the 2015 HMDA Rule may understate the burden that open-end reporting would impose on smaller institutions if they were required to begin reporting on January 1, 2022. For example, in developing the one-time cost estimates for open-end lines of credit in the 2015 HMDA Rule, the Bureau had envisioned that there would be cost sharing between the line of business that conducts open-end lending and the line of business that conducts closed-end lending at the corporate level, as the implementation of open-end reporting that became mandatory under the 2015 HMDA Rule would coincide with the implementation of the changes to closed-end reporting under the 2015 HMDA Rule. However, this type of cost sharing is less likely now since financial institutions have already implemented almost all of the closed-end reporting changes required under the 2015 HMDA Rule. As explained in more detail in part VII.E.3, the Bureau's coverage estimates also indicate that the total number of institutions exceeding the threshold of 100 open-end lines of credit in 2018 would be approximately 1,014, which is significantly higher than the estimate of 749 in the 2015 HMDA Rule that was based on 2013 data.^[108]

Another development since the Bureau finalized the 2015 HMDA Rule is the enactment of the EGRRCPA, which created partial exemptions from HMDA's requirements that certain insured depository institutions and insured credit unions may now use.^[109] The partial exemption for open-end lines of credit under the EGRRCPA relieves certain insured depository institutions and insured credit unions that originated fewer than 500 open-end mortgage loans in each of the two preceding calendar years of the obligation to report many of the data points generally required by Regulation C.^[110] The EGRRCPA has thus changed the costs and benefits associated with different coverage thresholds, as the partial exemptions are available to the vast majority of the depository financial institutions that originate fewer than 500 open-end lines of credit annually.^[111]

The Bureau has considered the appropriate permanent open-end threshold in light of these developments and the comments received in response to the May 2019 Proposal and the July 2019 Reopening Notice. On balance, the Bureau determines that the permanent threshold of 200 open-end lines of credit provides sufficient information on open-end lending to serve HMDA's purposes while appropriately reducing one-time and ongoing costs for smaller institutions that would be incurred if the threshold of 100 open-end lines of credit were to take effect.^[112] These considerations are discussed in turn below, and additional explanation of the Bureau's cost estimates is provided in the Bureau's analysis under Dodd-Frank Act section 1022(b) in part VII.E.3 below.^[113]

Effect on market coverage. While the increase in the permanent threshold to 200 open-end lines of credit will reduce market coverage compared to the threshold of 100 that would otherwise take effect, information about a sizeable portion of the open-end lending market will still be available. The Bureau has used multiple data sources, including credit union Call Reports, Call Reports for banks and thrifts, HMDA data, and Consumer Credit Panel data, to develop estimates about open-end originations for institutions that offer open-end lines of credit and to assess the impact of various thresholds on the numbers of institutions that report and the number of lines of credit about which they report under various scenarios.^[114] Based on this information, the Bureau estimates that, as of 2018, approximately 333 financial institutions originated at least 500 open-end lines of credit in each of the two preceding years, approximately 613 financial institutions originated at least 200 open-end lines of credit in each of the two preceding years, and approximately 1,014 financial institutions originated at least 100 open-end lines of credit in each of the two preceding years.^[115] Under the permanent threshold of 200 open-end lines of credit, the Bureau estimates about 1.34 million lines of credit or approximately 84 percent of origination volume will be reported by about 9 percent of all institutions providing open-end lines of credit.^[116] By comparison, the Bureau estimates that about 1.41 million lines of credit or approximately 89 percent of origination volume would be reported by about 15 percent of all institutions providing open-end lines of credit if the permanent threshold were to adjust to 100 open-end lines of credit. The Bureau determines that the benefits of the one-time and ongoing cost savings for the estimated 401 affected institutions originating between 100 and 199 open-end lines of credit, all but 17 of which are depository financial institutions, justify the limited decrease in the data reported about open-end lending that will result from this threshold increase. The permanent threshold of 200 open-end lines of credit balances the benefits and burdens of covering institutions engaged in open-end mortgage lending by retaining significant coverage of the open-end market while excluding from coverage smaller institutions whose limited open-end data would be of lesser utility in furthering HMDA's purposes.

Additionally, the effect of a threshold of 200 open-end lines of credit will be limited because the EGRRCPA now provides a partial exemption that exempts approximately 378 of the estimated 401 institutions that the permanent threshold increase will affect from any obligation to report many of the data points generally required by Regulation C for open-end lines of credit. In light of the EGRRCPA's partial exemption from reporting certain data for open-end lines of credit for certain insured depository institutions and insured credit unions, setting the permanent threshold at 200 open-end lines of credit will result in a much smaller decrease in data than the Bureau anticipated when it adopted a threshold of 100 open-end lines of credit in the 2015 HMDA Rule or when it revisited the open-end line of credit threshold in the 2017 HMDA Rule.

The Bureau declines to increase the permanent threshold further, as suggested by several commenters. Under a threshold of 500 open-end lines of credit, the Bureau estimates that about 1.23 million lines of credit or approximately 78 percent of origination volume would be reported by about 5 percent of all institutions providing open-end lines of credit. The Bureau determines that the more significant reduction in open-end reporting that would result if the current threshold of 500 open-end lines of credit were permanent, or if the Bureau increased the threshold to a level above 500, is not warranted. The temporary threshold of 500 open-end lines of credit was intended to allow the Bureau time to collect additional data and assess the appropriate level of the permanent threshold.^[117] Although the Bureau appreciates some commenters' suggestions regarding the benefits of continuity that would result from a permanent threshold of 500 open-end lines of credit, it determines that the permanent threshold of 200 open-end lines of credit adopted in this rule best balances the benefits and burdens of covering institutions based on their open-end lending volume. The data about open-end lines of credit that will be reported at this threshold will assist HMDA data users in understanding how financial institutions are serving the housing needs of their communities and assist in the distribution of public sector investments. The Bureau recognizes, as noted by several commenters, that open-end lines of credit may be used for non-housing related purposes, but the Bureau believes the data on these dwelling-secured loans will further HMDA purposes. The visibility into this segment of the mortgage market that will result from the permanent threshold of 200 open-end lines of credit, as opposed to a higher threshold, will also allow for a better understanding of these products and monitoring of the potential risks, as noted by many commenters, that could be associated with such loans. Such data could also help to assist in identifying possible discriminatory lending patterns if, for example, risky lending practices were concentrated among certain borrowers or communities.^[118]

Additionally, and as discussed above, the EGRRCPA partial exemption already relieves most lenders originating fewer than 500 open-end lines of credit in each of the two preceding years from the requirement to report many data points associated with their open-end transactions. In light of the concerns discussed above and the existing relief provided by the EGRRCPA at a threshold of 500, the Bureau determines that it is not appropriate to set the permanent threshold for open-end lines of credit at 500 or higher. Doing so would provide a complete exclusion from reporting all open-end data for institutions below the threshold of 500, even though Congress opted to provide only a partial exemption at the threshold of 500, and would extend that complete exclusion to institutions that Congress did not include in even the partial exemption. For the reasons stated above, the Bureau also declines to adopt the recommendation of several commenters to return to voluntary open-end reporting, which it did not propose.^[119]

Reduction in one-time costs from permanent threshold of 200. The Bureau's increase in the permanent open-end threshold to 200 open-end lines of credit after the temporary extension expires in 2022 will avoid imposing one-time costs of reporting open-end lines of credit on institutions originating between 100 and 199 open-end lines of credit. The Bureau estimates that setting the permanent threshold at 200 rather than 100 open-end lines of credit will exclude 401 institutions from reporting open-end lines of credit starting in 2022. According to the Bureau's estimates, about 309 of those 401 financial institutions are low-complexity tier 3 open-end reporters, about 92 are moderate-complexity tier 2 open-end reporters, and none are high-complexity tier 1 reporters.^[120]

The Bureau recognizes that, as a small financial institution commenter discussed, financial institutions may process applications for open-end lines of credit in different departments and on different systems than those used for closed-end loans. Many institutions that would have had to report with a threshold of 100 after the extension of the temporary threshold of 500 expires in 2022 do not currently report open-end lines of credit. These institutions might have to develop completely new reporting infrastructures to comply with mandatory reporting if the threshold of 100 lines of credit were to take effect, including new training, software, and policies and procedures. As a result, these institutions would incur one-time costs to create processes and systems for reporting open-end lines of credit in addition to the one-time costs to modify processes and systems used for reporting other mortgage products.

As explained in part VII below, the Bureau estimates that increasing the threshold from 100 to 200 open-end lines of credit starting in 2022 will result in a one-time cost savings of approximately $3,000 for low-complexity tier 3 reporters and $250,000 for moderate-complexity tier 2 reporters, for an aggregate savings of about $23.9 million in avoided one-time costs associated with reporting open-end lines of credit.^[121] The Bureau determines that avoiding the burden on smaller institutions of implementing open-end reporting, which as commenters noted could involve setting up entirely new reporting infrastructures distinct from those used for closed-end mortgage loans, is justified by the limited decrease in open-end data that will be reported under this final rule, as discussed in more detail above.

Ongoing cost reduction from permanent threshold of 200. The increase in the open-end threshold from 100 to 200 open-end lines of credit starting in 2022 will permanently relieve institutions that originate between 100 and 199 open-end lines of credit of the ongoing costs associated with reporting open-end lines of credit that they might otherwise incur if the threshold of 100 open-end lines of credit established in the 2015 HMDA Rule were to take effect. As noted above, many industry commenters expressed how costly and resource-intensive HMDA compliance can be on an ongoing basis for smaller institutions.

As discussed in more detail in part VII below, the Bureau estimates that increasing the permanent threshold from 100 to 200 open-end lines of credit will result in annual ongoing cost savings of approximately $4,300 for low-complexity tier 3 institutions eligible for the EGRRCPA partial exemption and $21,900 for moderate-complexity tier 2 institutions eligible for the EGRRCPA partial exemption. For the low-complexity tier 3 and moderate-complexity tier 2 institutions that are not eligible for the EGRRCPA partial exemption, the Bureau estimates that the increase in the permanent threshold from 100 to 200 open-end lines of credit will result in annual ongoing cost savings of approximately $8,800 and $44,700, respectively. The Bureau estimates that the increase in the permanent threshold will result in aggregate savings on the ongoing operational costs associated with open-end lines of credit of about $3.7 million per year starting in 2022.^[122] The Bureau recognizes that the estimated ongoing costs savings associated with increasing the permanent threshold from 100 to 200 open-end lines of credit are less than they would have been absent the relief provided by the EGRRCPA. Nonetheless, the Bureau determines that these ongoing cost savings, coupled with the one-time cost savings discussed above, will provide meaningful burden reduction to smaller institutions that would have been covered at the threshold of 100 open-end lines of credit but will be excluded from open-end reporting under this final rule. Avoiding the imposition of such costs for these affected institutions will also limit any potential for cost increases to borrowers or other disruptions in open-end lending that could result from HMDA coverage, as discussed by some commenters.

For the reasons discussed above, the Bureau amends § 1003.2(g)(1)(v)(B) and comments 2(g)-3 and -5, to set the open-end institutional coverage threshold for depository institutions at 200, effective January 1, 2022.

2(g)(2) Nondepository Financial Institution

HMDA extends reporting responsibilities to certain nondepository institutions, defined as any person engaged for profit in the business of mortgage lending other than a bank, savings association, or credit union.^[123] HMDA section 309(a) authorizes the Bureau to adopt an exemption for covered nondepository institutions that are comparable within their respective industries to banks, savings associations, and credit unions with $10 million or less in assets in the previous fiscal year.^[124] HMDA sections 303(3)(B) and 303(5) require persons other than banks, savings associations, and credit unions that are “engaged for profit in the business of mortgage lending” to report HMDA data. As the Bureau stated in the 2015 HMDA Rule, the Bureau interpreted these provisions, as the Board did, to evince the intent to exclude from coverage institutions that make a relatively small volume of mortgage loans.^[125] In the 2015 HMDA Rule, the Bureau interpreted “engaged for profit in the business of mortgage lending” to include nondepository institutions that originated at least 25 closed-end mortgage loans or 100 open-end lines of credit in each of the two preceding calendar years. Due to the questions raised about potential risks posed to applicants and borrowers by nondepository institutions and the lack of other publicly available data sources about nondepository institutions, the Bureau believed that requiring additional nondepository institutions to report HMDA data would better effectuate HMDA's purposes. The Bureau estimated in 2015 that these changes to institutional coverage could result in HMDA coverage for up to an additional 450 nondepository institutions. The Bureau stated in the 2015 HMDA Rule its belief that it was important to increase visibility into the lending practices of nondepository institutions because of their history of making riskier loans than depository institutions, including their role in the financial crisis and lack of available data about the mortgage lending practices of lower-volume nondepository institutions. The Bureau also stated that expanded coverage of nondepository institutions would ensure more equal visibility into the practices of nondepository institutions and depository institutions.

For the reasons discussed below, the Bureau has now determined that higher thresholds for closed-end mortgage loans and open-end lines of credit will more appropriately cover nondepository institutions that “are engaged for profit in the business of mortgage lending” and maintain visibility into the lending practices of such institutions.

2(g)(2)(ii)(A)

Regulation C implements HMDA's coverage criteria for nondepository institutions in § 1003.2(g)(2). The Bureau revised the coverage criteria for nondepository institutions in the 2015 HMDA Rule by requiring such institutions to report HMDA data if they met the statutory location test and exceeded either the closed-end or open-end thresholds.^[126] In the May 2019 Proposal, the Bureau proposed to amend § 1003.2(g)(2)(ii)(A) and related commentary to raise the closed-end threshold for nondepository institutions to either 50 or, alternatively, 100 closed-end mortgage loans. For the reasons discussed below, the Bureau is amending § 1003.2(g)(1)(ii)(A) and related commentary to raise the threshold to 100 closed-end mortgage loans.

Background on Closed-End Mortgage Loan Threshold for Institutional Coverage of Nondepository Institutions

After issuing the 2015 HMDA Rule and the 2017 HMDA Rule, the Bureau heard concerns that lower-volume institutions experience significant burden with the current threshold of 25 closed-end mortgage loans.^[127] Various industry stakeholders advocated for an increase to the closed-end threshold in order to reduce burden on additional lower-volume financial institutions. In light of the concerns raised by industry stakeholders, the Bureau proposed to raise the closed-end threshold for nondepository institutions and indicated that it was considering whether a higher threshold would more appropriately cover nondepository institutions that “are engaged for profit in the business of mortgage lending” and maintain visibility into the lending practices of such institutions. The Bureau sought comment on whether an increase to the threshold would more appropriately balance the benefits and burdens of covering lower-volume nondepository institutions based on their closed-end lending.

The Bureau proposed two alternatives to the closed-end mortgage loan threshold, and both proposed alternatives would have maintained a uniform closed-end threshold for depository and nondepository institutions.^[128] The Bureau sought specific comment on how the proposed increase to the closed-end threshold would affect the number of nondepository institutions required to report data on closed-end mortgage loans, the significance of the data that would not be available as a result of the proposed increase to the closed-end threshold, and the reduction in burden that would result from the proposed increase to the closed-end threshold for nondepository institutions that would not be required to report.

Comments Received on Closed-End Threshold for Institutional Coverage for Nondepository Institutions

As mentioned above, commenters typically discussed the closed-end threshold without distinguishing between the threshold applicable to depository institutions under § 1003.2(g)(1)(v)(A) and the threshold applicable to nondepository institutions under § 1003.2(g)(2)(ii)(A). Comments received regarding the Bureau's proposal to increase the closed-end threshold are discussed in more detail in the section-by-section analysis of § 1003.2(g)(1)(v)(A) above.

Final Rule

Pursuant to its authority under HMDA section 305(a), the Bureau is finalizing the closed-end threshold for nondepository institutions at 100 in § 1003.2(g)(1)(ii)(A). The Bureau has considered the comments received in response to the May 2019 Proposal and updated estimates in determining the appropriate threshold. As discussed below, the Bureau believes that it is reasonable to interpret “engaged for profit in the business of mortgage lending” to include nondepository institutions that originated at least 100 closed-end mortgage loans in each of the two preceding calendar years and that doing so will effectuate the purposes of HMDA and facilitate compliance. The Bureau believes that increasing the closed-end threshold to 100 will provide meaningful burden relief for lower-volume nondepository institutions while maintaining sufficient reporting to achieve HMDA's purposes. The final rule's uniform loan-volume threshold applicable to depository and nondepository institutions also maintains the simplicity of this aspect of the reporting regime, thereby facilitating compliance.^[129]

As explained in the section-by-section analysis of § 1003.2(g)(1)(v)(A) above, a few developments have affected the Bureau's analyses of the costs and benefits associated with the closed-end threshold for depository and nondepository institutions since the 2015 HMDA Rule was issued. The Bureau has gathered extensive information regarding stakeholders' experience with the 2015 HMDA Rule through comments received in this rulemaking and other feedback. As described above, the Bureau has heard that financial institutions have encountered significant burdens in complying with the 2015 HMDA Rule, and the Bureau is particularly concerned about the increased burdens faced by smaller institutions. Additionally, the Bureau now has access to HMDA data from 2018, which was the first year that financial institutions collected data under the 2015 HMDA Rule, and has used these data in updating and confirming the estimates included in this final rule. With the benefit of this additional information about the 2015 HMDA Rule and the new data to supplement the Bureau's analyses, the Bureau is now in a better position to assess both the benefits and burdens of the reporting required under the 2015 HMDA Rule.

The Bureau has considered the appropriate closed-end threshold for nondepository institutions in light of these developments and the comments received. The Bureau determines that the threshold of 100 closed-end mortgage loans for nondepository institutions provides sufficient information on closed-end mortgage lending to serve HMDA's purposes and maintains uniformity with the closed-end threshold for depository institutions established in this rule, while appropriately reducing ongoing costs that smaller nondepository institutions are incurring under the current threshold. These considerations are discussed in turn below, and additional explanation of the Bureau's cost estimates is provided in the Bureau's analysis under Dodd-Frank Act section 1022(b) in part VII.E.2 below.

Effect on Market Coverage

Similar to the estimates in the section-by-section analysis of § 1003.2(g)(1)(v)(A), the Bureau developed estimates for nondepository institution coverage at varying thresholds. Using multiple data sources, including recent HMDA data ^[130] and Call Reports, the Bureau developed estimates for the two thresholds the Bureau proposed in the alternative, 50 and 100, as well as the thresholds of 250 and 500, which many commenters suggested the Bureau consider.^[131] These estimates compare coverage under these thresholds to coverage under the current threshold of 25.

Similar to the estimates described above for the closed-end threshold for depository institutions, many of the estimates provided for the closed-end threshold for nondepository institutions differ slightly from the initial estimates provided in the May 2019 Proposal. The estimates in this final rule update the initial estimates provided in the May 2019 Proposal using the 2018 HMDA data, which were not available at the time the Bureau developed the May 2019 Proposal. For the May 2019 Proposal, the Bureau used HMDA data from 2016 and 2017 with a two-year look-back period covering calendar years 2016 and 2017 to estimate potential reporters and projected the lending activities of financial institutions using their 2017 HMDA data as proxies. In generating the updated estimates provided in this final rule, the Bureau used data from 2017 and 2018 with a two-year look-back period covering calendar years 2017 and 2018 to estimate potential reporters and has projected the lending activities of financial institutions using their 2018 data as proxies. In addition, for the estimates provided in the May 2019 Proposal and in this final rule, the Bureau restricted the projected reporters to only those that actually reported data in the most recent year of HMDA data considered (2017 for the May 2019 Proposal and 2018 for this final rule).^[132]

Effect on covered nondepository institutions and reportable originations. As discussed above, many commenters opposed increasing the closed-end threshold because of concerns that there would be less data with which to evaluate whether HMDA's statutory purposes are being met. However, the Bureau's analysis indicates that the proposed thresholds of either 50 or 100 closed-end mortgage loans will maintain sufficient reporting to achieve HMDA's purposes.

If the threshold were increased to 50 closed-end mortgage loans, the Bureau estimates that approximately 720 out of approximately 740 nondepository institutions covered under the current threshold of 25 (or approximately 97 percent) would continue to be required to report HMDA data on closed-end mortgage loans. Further, the Bureau estimates that if the threshold were increased from 25 to 50, this would result in about 99.97 percent of closed-end mortgage loan originations currently reported or approximately 3.428 million total closed-end mortgage loan originations, under current market conditions, that would continue to be reported by nondepository institutions.^[133]

The Bureau estimates that with the closed-end threshold set at 100 under the final rule, approximately 680 out of approximately 740 nondepository institutions covered under the current threshold of 25 (or approximately 92 percent) will continue to be required to report HMDA data on closed-end mortgage loans. Further, the Bureau estimates that when the final rule increases the threshold to 100, about 99.9 percent of originations of closed-end mortgage loans currently reported or approximately 3.425 million total originations of closed-end mortgage loans reported by nondepository institutions, under current market conditions, will continue to be reported.^[134]

The Bureau also generated estimates for closed-end thresholds higher than the ones the Bureau proposed, as many commenters suggested that the Bureau consider thresholds higher than proposed. These estimates reflect the decrease in the number of nondepository institutions that would be required to report HMDA data and the resulting decrease in the HMDA data that would be reported becomes more pronounced at thresholds higher than 100. For example, if the closed-end threshold were set at 250, the Bureau estimates that approximately 590 out of approximately 740 nondepository institutions covered under the current threshold of 25 (or approximately 80 percent) would continue to be required to report HMDA data on closed-end mortgage loans. Further, the Bureau estimates that if the threshold were increased from 25 to 250 loans, this would result in about 99.4 percent of closed-end mortgage loan originations currently reported or approximately 3.408 million total closed-end mortgage loan originations, under current market conditions, that would continue to be reported by nondepository institutions.^[135] If the closed-end threshold were set at 500, the Bureau estimates that approximately 480 out of approximately 740 nondepository institutions covered under the current threshold of 25 (or approximately 65 percent) would continue to be required to report HMDA data on closed-end mortgage loans. Further, the Bureau estimates that if the threshold were increased from 25 to 500 loans, this would result in about 98.2 percent of closed-end mortgage loan originations currently reported or approximately 3.367 million total closed-end mortgage loan originations, under current market conditions, that would continue to be reported by nondepository institutions.^[136]

The Bureau recognizes the importance of maintaining data about the lending practices of nondepository institutions. The Bureau also acknowledges the concerns raised by commenters that setting the threshold higher than 25 will result in less data, but the Bureau believes that the modest decrease in data at the threshold of 100 (estimated at less than one percent of data currently reported) will not undermine enforcement of fair lending laws or regulators' ability to identify potentially discriminatory lending through HMDA data. The Bureau believes that retaining approximately 99.9 percent of the data that would be reported under the current threshold of 25 will result in nondepository institutions reporting sufficient HMDA data to enable the public and regulators to monitor risks posed by nondepository institutions.

Effect on HMDA data at the local level. The Bureau recognizes that any loan-volume threshold will affect individual markets differently, depending on the extent to which smaller creditors service individual markets and the market share of those creditors. For the proposal and this final rule, the Bureau reviewed estimates at varying closed-end thresholds to examine the potential effect on available data at the census tract level. The estimates of the effect on reportable HMDA data at the census tract level are discussed in the section-by-section analysis of § 1003.2(g)(1)(v)(A) above. Based on the Bureau's review of the estimates, the Bureau believes that an increase to the closed-end threshold from 25 to 100 will result in sufficient data at the local level, including with respect to rural and low-to-moderate income census tracts, to further HMDA's purposes.

Specific types of data. As mentioned in the section-by-section analysis in § 1003.2(g)(1)(v)(A), a number of commenters expressed concerns about the impact that an increase to the closed-end threshold would have on HMDA data regarding specific types of loan products, such as loans for multifamily housing and manufactured housing. The Bureau believes that data on these types of loan products will still be available at the threshold of 100 set by this final rule. For example, the Bureau estimates that with the closed-end threshold increased from 25 to 100 under the final rule, approximately 87 percent of multifamily loan applications and originations will continue to be reported by depository and nondepository institutions combined, when compared to the current threshold of 25 closed-end mortgage loans in today's market conditions. Regarding the effect on manufactured housing data, the Bureau estimates that at a threshold of 100 closed-end mortgage loans, approximately 96 percent of loans and applications related to manufactured housing will continue to be reported by depository and nondepository institutions combined, when compared to the current threshold of 25 closed-end mortgage loans in today's market conditions. The Bureau's estimates indicate that a significant number of multifamily housing and manufactured housing loans and applications under today's market conditions will continue to be reported under the final rule's threshold of 100.

Ongoing Cost Reduction From Threshold of 100

As noted above, small financial institutions and trade associations commented on the cost of HMDA reporting, suggesting that compliance costs have had an impact on the ability of small financial institutions to serve their communities. For the May 2019 Proposal and this final rule, the Bureau developed estimates for depository and nondepository institutions combined to determine the savings in annual ongoing costs at various thresholds.^[137] The Bureau estimates that if the closed-end threshold were set at 50, institutions that originate between 25 and 49 closed-end mortgage loans would save approximately $3.7 million per year in total annual ongoing costs relative to the current threshold of 25.^[138] The Bureau estimates that with the threshold of 100 closed-end mortgage loans established by the final rule, institutions that originate between 25 and 99 closed-end mortgage loans will save approximately $11.2 million per year, relative to the current threshold of 25.^[139] With a threshold of 250 or 500 closed-end mortgage loans, the Bureau estimates that institutions would save approximately $27.2 million and $45.4 million, respectively, relative to the current threshold of 25.

The Bureau concludes that increasing the closed-end threshold to 100 will provide meaningful burden reduction for lower-volume nondepository institutions, while maintaining sufficient reporting to achieve HMDA's purposes. In the 2015 HMDA Rule, the Bureau expressed concern that if it were to set the threshold higher than 100, the resulting decrease into the visibility of the lending practices of nondepository institutions might hamper the ability of the public and regulators to monitor risks posed to consumers by those nondepository institutions.^[140] The Bureau maintains the same concern under this final rule based on its analysis of various closed-end thresholds using more recent estimates. For example, if the Bureau were to set the closed-end threshold for nondepository institutions at 500 as suggested by a number of commenters, while an estimated 98.2 percent of closed-end mortgage originations currently reported under today's market conditions would continue to be reported, there would be data reported about the lending patterns of only 65 percent of nondepository institutions that are reporters under the current threshold of 25. The Bureau is concerned that an increase to the closed-end threshold higher than 100 could hamper the ability of the public and regulators to monitor risks posed to consumers by nondepository institutions. In comparison, with the Bureau finalizing the closed-end threshold at 100, an estimated 99.9 percent of nondepository closed-end mortgage originations currently reported under today's market conditions will continue to be reported, and there will be data reported about 92 percent of nondepository institutions relative to the current threshold of 25. The Bureau's estimates suggest that, at the threshold of 100 closed-end mortgage loans established by the final rule, the cost savings for financial institutions will be meaningful while maintaining substantial HMDA data for analysis at the national and local levels.^[141]

Based on its analysis, the Bureau believes it is reasonable to interpret “engaged for profit in the business of mortgage lending” to include nondepository institutions that originated at least 100 closed-end mortgage loans in each of the two preceding calendar years. The Bureau determines that this final rule's amendments to § 1003.2(g)(2)(ii)(A) will also effectuate the purposes of HMDA by ensuring significant coverage of nondepository mortgage lending. A threshold of 100 closed-end mortgage loans also facilitates compliance with HMDA by reducing burden on smaller institutions and excluding nondepository institutions that are not engaged for profit in the business of mortgage lending. In addition, the final rule's uniform loan-volume threshold applicable to depository and nondepository institutions maintains the simplicity of this aspect of the reporting regime, thereby facilitating compliance.^[142] For the reasons stated above, the Bureau is amending § 1003.2(g)(2)(ii)(A) to adjust the closed-end threshold to 100. As discussed in part VI.A below, the change to the closed-end threshold will take effect on July 1, 2020, to provide more immediate relief to affected institutions.^[143]

2(g)(2)(ii)(B)

The 2015 HMDA Rule established a coverage threshold of 100 open-end lines of credit in § 1003.2(g)(2)(ii)(B) as part of the definition of nondepository financial institution. As discussed in more detail in the section-by-section analysis of § 1003.2(g)(1)(v)(B) above, the 2017 HMDA Rule amended §§ 1003.2(g)(1)(v)(B) and (g)(2)(ii)(B) and 1003.3(c)(12) and related commentary to raise temporarily the threshold to 500 open-end lines of credit for calendar years 2018 and 2019.^[144] In the May 2019 Proposal, the Bureau proposed to extend to January 1, 2022, Regulation C's temporary threshold of 500 open-end lines of credit for institutional and transactional coverage of both depository and nondepository institutions. The Bureau also proposed to increase the permanent threshold from 100 to 200 open-end lines of credit at the end of the extension. In the 2019 HMDA Rule, the Bureau extended for two years the temporary open-end institutional coverage threshold for nondepository institutions in § 1003.2(g)(2)(ii)(B). The Bureau is now finalizing as proposed the increase in the permanent threshold to 200 open-end lines of credit effective January 1, 2022.

As noted above, commenters typically discussed the open-end threshold without distinguishing between the threshold applicable to depository institutions under § 1003.2(g)(1)(v)(B) and the threshold applicable to nondepository institutions under § 1003.2(g)(2)(ii)(B). Comments received regarding the proposed increase in the permanent open-end threshold are discussed in the section-by-section analysis of § 1003.2(g)(1)(v)(B).

According to the Bureau's estimates, nondepository institutions account for only a small percentage of the institutions and loans in the open-end line of credit market.^[145] Table 4 in the Bureau's analysis under Dodd-Frank Act section 1022(b) in part VII.E.3 below provides coverage estimates for nondepository institutions at the permanent threshold of 200 open-end lines of credit that the Bureau is finalizing. Under the permanent threshold of 200 open-end lines of credit, the Bureau estimates that about 48,000 open-end lines of credit or approximately 84 percent of nondepository open-end origination volume will be reported by approximately 25 nondepository institutions or about 11 percent of all nondepository institutions providing open-end lines of credit. By comparison, the Bureau estimates that if the permanent threshold were set at 100 open-end lines of credit, about 51,000 lines of credit or approximately 89 percent of nondepository open-end origination volume would be reported by approximately 42 nondepository institutions or about 19 percent of all nondepository institutions providing open-end lines of credit.

For the reasons discussed in the section-by-section analysis of § 1003.2(g)(1)(v)(B), and to ensure the thresholds are consistent for depository and nondepository institutions, the Bureau is finalizing as proposed an increase to Regulation C's permanent open-end threshold upon expiration of the current temporary open-end threshold. This final rule increases to 200 the permanent open-end line of credit threshold for institutional coverage of nondepository institutions in § 1003.2(g)(2)(ii)(B) effective January 1, 2022. This amendment to the open-end threshold for institutional coverage of nondepository institutions in § 1003.2(g)(2)(ii)(B) conforms to the amendments that the Bureau is finalizing with respect to the permanent open-end threshold for institutional coverage of depository institutions in § 1003.2(g)(1)(v)(B) and the permanent open-end threshold for transactional coverage in § 1003.3(c)(12).

Pursuant to its authority under HMDA section 305(a) as discussed above, the Bureau is finalizing an increase to the permanent threshold for open-end lines of credit in § 1003.2(g)(2)(ii)(B). Based on its analysis, the Bureau believes it is reasonable to interpret “engaged for profit in the business of mortgage lending” to include nondepository institutions that originated at least 200 open-end lines of credit in each of the two preceding calendar years. The Bureau determines that this final rule's amendments to § 1003.2(g)(2)(ii)(B) will also effectuate the purposes of HMDA by ensuring significant coverage of nondepository mortgage lending. This increase to the permanent threshold also facilitates compliance with HMDA by reducing burden on smaller institutions and excluding nondepository institutions that are not engaged for profit in the business of mortgage lending. The Bureau determines that the reasons provided for changing the permanent threshold for depository institutions in the section-by-section analysis of § 1003.2(g)(1)(v)(B) above apply to the permanent threshold for nondepository institutions as well. Additionally, the increase in the permanent threshold in § 1003.2(g)(2)(ii)(B) to 200 open-end lines of credit will promote consistency, and thereby facilitate compliance, by subjecting nondepository institutions to the same threshold that applies to the depository institutions that make up the majority of the open-end line of credit market.

Section 1003.3 Exempt Institutions and Excluded and Partially Exempt Transactions

3(c) Excluded Transactions

3(c)(11)

Section 1003.3(c)(11) provides an exclusion from the requirement to report closed-end mortgage loans for institutions that originated fewer than 25 closed-end mortgage loans in either of the two preceding calendar years. This transactional coverage threshold complements the closed-end mortgage loan reporting threshold included in the definition of financial institution in § 1003.2(g). In the May 2019 Proposal, the Bureau proposed to increase Regulation C's closed-end threshold for institutional and transactional coverage from 25 to either 50 or 100. Comments regarding the proposed increase to the closed-end coverage threshold are discussed in the section-by-section analysis of § 1003.2(g)(1)(v)(A). For the reasons discussed in the section-by-section analysis of § 1003.2(g)(1)(v)(A), the Bureau is now increasing Regulation C's closed-end threshold for institutional and transactional coverage from 25 to 100. Therefore, the Bureau is finalizing the amendments it proposed to § 1003.3(c)(11) and comments 3(c)(11)-1 and -2 to increase the closed-end threshold for transactional coverage from 25 to 100, with minor clarifying changes.^[146] These amendments conform to the related changes the Bureau is finalizing with respect to the closed-end threshold for institutional coverage in § 1003.2(g).

Although not part of the May 2019 Proposal, the Bureau is also finalizing additional related amendments to § 1003.3(c)(11) and comment 3(c)(11)-2 to ensure that institutions affected by the threshold increase have the option to report data collected in 2020 should they choose to do so. As discussed in part VI.A below, the change to the closed-end threshold will take effect on July 1, 2020. The Bureau selected this effective date to ensure that the relief provided in this final rule is available quickly to affected institutions, after a number of industry commenters expressed support for a mid-year effective date. However, a number of commenters also noted that institutions may have difficulty making changes to their HMDA operations and might need time to implement changes in response to the final rule. For example, a State trade association that expressed support for a mid-year effective date noted that there may be operational issues that make it difficult for institutions to avail themselves of the threshold increase in mid-year and requested that the Bureau allow a transition period for any institution that may need additional time. The Bureau also recognizes that, since 2020 data collection is already underway, some affected institutions may wish to report the HMDA data that they have collected. The Bureau believes that voluntary reporting of 2020 closed-end data from such institutions would be beneficial to the HMDA dataset, as long as the data are submitted for the entire calendar year.

The Bureau is amending § 1003.3(c)(11) and comment 3(c)(11)-2 to allow institutions newly excluded by the final rule the option to report their 2020 closed-end data. Specifically, the Bureau is amending § 1003.3(c)(11) to clarify that, for purposes of information collection in 2020, “financial institution” as used in the discussion of optional reporting in § 1003.3(c)(11) includes an institution that was a financial institution as of January 1, 2020. Thus, for purposes of information collection in 2020, an institution that was a financial institution as of January 1, 2020, may collect, record, report, and disclose information, as described in §§ 1003.4 and .5, for closed-end mortgage loans excluded under § 1003.3(c)(11), as though they were covered loans, provided that the institution complies with such requirements for all applications for closed-end mortgage loans that it receives, closed-end mortgage loans that it originates, and closed-end mortgage loans that it purchases that otherwise would have been covered loans during calendar year 2020. The amendment to comment 3(c)(11)-2 clarifies that an institution that was a financial institution as of January 1, 2020, but is not a financial institution on July 1, 2020, because it originated fewer than 100 closed-end mortgage loans in either 2018 or 2019 is not required in 2021 to report, but may report, applications for, originations of, or purchases of closed-end mortgage loans for calendar year 2020 that are excluded transactions because the institution originated fewer than 100 closed-end mortgage loans in either 2018 or 2019. The amendment to comment 3(c)(11)-2 further clarifies that an institution that was a financial institution as of January 1, 2020, and chooses to report such excluded applications for, originations of, or purchases of closed-end mortgage loans in 2021 must report all such applications for closed-end mortgage loans that it receives, closed-end mortgage loans that it originates, and closed-end mortgage loans that it purchases that otherwise would be covered loans for all of calendar year 2020. These amendments thus permit an institution that was a financial institution as of January 1, 2020, but is not a financial institution on July 1, 2020, because it originated fewer than 100 closed-end mortgage loans in either 2018 or 2019 to report voluntarily in 2021 its closed-end HMDA data collected in 2020, as long as the institution reports such closed-end data for the full calendar year 2020.^[147] The Bureau believes that these amendments are appropriate to ensure that institutions newly excluded by the mid-year increase in the closed-end threshold in this final rule have the option to report their 2020 closed-end data should they choose to do so.

As explained in part VI below, the amendments to increase the closed-end threshold, including the amendments to § 1003.3(c)(11) and comment 3(c)(11)-2 permitting optional reporting of data collected in 2020, take effect on July 1, 2020. Because the deadline for reporting of data collected in 2020 (voluntary or otherwise) is March 1, 2021, the amendments to § 1003.3(c)(11) and comment 3(c)(11)-2 relating to optional reporting of data collected in 2020 will no longer be necessary after 2021. To streamline Regulation C, the final rule therefore removes these amendments effective January 1, 2022.

3(c)(12)

As adopted in the 2015 HMDA Rule, § 1003.3(c)(12) provides an exclusion from the requirement to report open-end lines of credit for institutions that did not originate at least 100 such loans in each of the two preceding calendar years. This transactional coverage threshold complements the open-end threshold included in the definition of financial institution in § 1003.2(g), which sets forth Regulation C's institutional coverage. The 2017 HMDA Rule replaced “each” with “either” in § 1003.3(c)(11) and (12) to correct a drafting error and to ensure that the exclusions provided in § 1003.3(c)(11) and (12) mirror the loan-volume thresholds for financial institutions in § 1003.2(g).^[148] As discussed in more detail in the section-by-section analysis of § 1003.2(g), in the 2017 HMDA Rule the Bureau also amended §§ 1003.2(g) and 1003.3(c)(12) and related commentary to raise temporarily the open-end threshold in those provisions to 500 lines of credit for calendar years 2018 and 2019.^[149] In the May 2019 Proposal, the Bureau proposed to extend to January 1, 2022, Regulation C's current temporary open-end threshold for institutional and transactional coverage of 500 open-end lines of credit and then to increase the permanent threshold from 100 to 200 open-end lines of credit upon the expiration of the proposed extension of the temporary threshold. The Bureau stated in the 2019 HMDA Rule that it intended to address in a separate final rule in 2020 the May 2019 Proposal's proposed amendment to the permanent threshold for open-end lines of credit. Comments regarding the proposed permanent increase in the open-end threshold are discussed in the section-by-section analysis of § 1003.2(g)(1)(v)(B) above.

For the reasons discussed in the section-by-section analysis of § 1003.2(g)(1)(v)(B), the Bureau is now finalizing as proposed the increase in the permanent threshold to 200 open-end lines of credit, effective January 1, 2022. To align the permanent increase in the open-end threshold for institutional coverage in § 1003.2(g) with the transactional coverage threshold, the Bureau is also finalizing the permanent increase in the open-end threshold for transactional coverage in § 1003.3(c)(12) and comments 3(c)(12)-1 and -2, as proposed, with minor changes for clarity.^[150]

VI. Effective Dates

In consideration of comments received and for the reasons discussed below, the Bureau is finalizing the increase in the closed-end threshold to 100 effective July 1, 2020,^[151] and the adjustment to the permanent open-end threshold to 200 effective January 1, 2022, when the current temporary open-end threshold expires.^[152]

A. Closed-End Threshold

In the May 2019 Proposal, the Bureau proposed to amend §§ 1003.2(g)(1)(v)(A) and (g)(2)(ii)(A) and 1003.3(c)(11) and related commentary to raise the closed-end threshold for institutional and transactional coverage effective January 1, 2020. In the July 2019 Reopening Notice the agency issued in 2019, the Bureau explained that, due to the reopening of the comment period on the closed-end threshold, it would not be possible to finalize any change to the closed-end threshold in time to take effect on the Bureau's originally proposed effective date of January 1, 2020. The Bureau therefore requested additional comment on the appropriate effective date for any change to the closed-end threshold, should the Bureau decide to finalize a change. The Bureau specifically requested comment on the costs and benefits of a mid-year effective date during 2020 versus a January 1, 2021 effective date. With respect to the alternative of a mid-year effective date during 2020, the Bureau also requested comment on the costs and benefits of specific days of the week or times of the month, quarter, or year for a new closed-end threshold to take effect and whether there are any other considerations that the Bureau should address in a final rule if it were to adopt a mid-year effective date.

Most commenters did not address the effective date question, and those that did expressed differing views. A number of banks requested an immediate effective date upon publication of the final rule and requested elimination of any reporting requirement for 2020 data collected prior to that date to provide more immediate regulatory relief. These commenters also asked that the Bureau clarify in the final rule that applications taken prior to a mid-year effective date, which would no longer be HMDA reportable after the threshold increase but for which demographic information was collected, do not violate the demographic collection rules in Regulations B and C. In a joint comment letter, a group of trade associations urged the Bureau to finalize the rule before March 1, 2020 (the deadline for reporting data that financial institutions collected in 2019) and have it take immediate or even retroactive effect. A State trade association that supports a mid-year effective date noted that there may be operational issues that make it difficult for institutions to avail themselves of the threshold increase in mid-year and requested that the Bureau allow a transition for any institution that may need additional time. Other commenters, including at least one bank, a State credit union league, and a joint comment submitted on behalf of consumer groups, civil rights groups, and other organizations, favored a January 1, 2021 effective date. These commenters noted that institutions may have difficulty making changes quickly and that implementing changes at the beginning of the year makes data more consistent and minimizes confusion.

The Bureau has considered the comments received and concludes that a mid-year effective date for the closed-end threshold is appropriate to provide burden relief quickly to institutions that would have to report under the threshold of 25 closed-end mortgage loans but will not have to report under the threshold of 100 closed-end mortgage loans.^[153] The amendments relating to the closed-end threshold in §§ 1003.2(g)(1)(v)(A) and (g)(2)(ii)(A) and 1003.3(c)(11) and related commentary are effective on July 1, 2020.^[154] Thus, in calendar year 2020, an institution could have been subject to HMDA's closed-end requirements as of January 1, 2020 because it originated at least 25 closed-end mortgage loans in 2018 and 2019 and meets all of the other requirements under § 1003.2(g), but no longer subject to HMDA's closed-end requirements as of July 1, 2020 (a newly excluded institution) because it originated fewer than 100 closed-end mortgage loans during 2018 or 2019. The final rule relieves newly excluded institutions of the obligation to collect, record, and report data for their 2020 closed-end mortgage loans effective July 1, 2020. Newly excluded institutions may cease collecting 2020 data for closed-end mortgage loans as of July 1, 2020. Pursuant to § 1003.4(f), newly excluded institutions must still record data on a loan/application register for the first quarter of 2020 by 30 calendar days after the end of the first quarter. They will not, however, be required to record closed-end data for the second quarter of 2020 because the deadline under § 1003.4(f) for recording such data falls after July 1, 2020. Because newly excluded institutions collecting HMDA data in 2020 would not otherwise report those data until early 2021, the final rule also relieves newly excluded institutions of the obligation to report by March 1, 2021 data collected in 2020 on closed-end mortgage loans (including closed-end data collected in 2020 before July 1, 2020).

The Bureau appreciates that some newly excluded institutions will need to continue to collect certain data due to other regulatory requirements or may wish to continue collecting data for other reasons. For example, Regulation B includes an independent requirement to collect information regarding the applicant's ethnicity, race, sex, marital status, and age where the credit sought is primarily for the purchase or refinancing of a dwelling that is or will be the applicant's principal residence and will secure the credit.^[155] Institutions may also decide to continue collecting after the effective date for other reasons—for example, in order to assess whether they will be subject to HMDA data collection requirements in 2021 or to continue monitoring their own operations. As noted above, some commenters indicated that many smaller institutions might not be able to change their collection practices quickly.

To accommodate such institutions, the amendments to § 1003.3(c)(11) and comment 3(c)(11)-2 permit an institution that was a financial institution as of January 1, 2020 but is not a financial institution on July 1, 2020 because it originated fewer than 100 closed-end mortgage loans in 2018 or 2019 to report voluntarily in 2021 its closed-end HMDA data collected in 2020, as long as the institution reports such closed-end data for the full calendar year 2020.^[156] These changes take effect with the other closed-end changes on July 1, 2020 and are discussed in more detail in the section-by-section analysis of § 1003.3(c)(11). Because the deadline for reporting of 2020 data (voluntary or otherwise) is in 2021, the final rule also removes the amendments to § 1003.3(c)(11) and comment 3(c)(11)-2 relating to optional reporting of 2020 data effective January 1, 2022.

As noted above, a number of industry commenters asked that the Bureau clarify in its final rule that applications taken prior to a mid-year effective date, which would no longer be HMDA reportable after the threshold increase but for which demographic information was collected, do not violate the rules governing demographic information collection in Regulations B and C. Newly excluded institutions do not violate Regulation B or C by collecting demographic information about applicants before the effective date in accordance with their legal obligations under Regulation C as that regulation is in effect before the effective date. As noted above, even after the effective date, creditors will still be required under Regulation B to collect information regarding ethnicity, race, sex, marital status, and age where the credit sought is primarily for the purchase or refinancing of a dwelling that is or will be the applicant's principal residence and will secure the credit.^[157] The Bureau recognizes that some newly excluded institutions may also continue collecting demographic information for other loans in 2020 after the effective date—for example, if they need additional time to update their systems and forms and to retrain employees or if they decide to continue collecting for the full year and report 2020 data voluntarily in accordance with § 1003.3(c)(11) and comment 3(c)(11)-2. Although Regulation B, 12 CFR 1002.5(b), prohibits creditors from inquiring about the race, color, religion, national origin, or sex of a credit applicant except under certain circumstances, the Bureau notes that even after the effective date, applicable exceptions in Regulation B will permit newly excluded institutions ^[158] to collect information in 2020 about the ethnicity, race, and sex of applicants for loans that would have been covered loans absent this final rule.^[159]

B. Open-End Threshold

In the May 2019 Proposal, the Bureau proposed to amend §§ 1003.2(g)(1)(v)(B) and (g)(2)(ii)(B) and 1003.3(c)(12) and related commentary to extend to January 1, 2022, the current temporary open-end threshold of 500 open-end lines of credit and then to set the threshold permanently at 200 open-end lines of credit beginning in calendar year 2022. In the 2019 HMDA Rule, the Bureau finalized as proposed the two-year extension of the temporary open-end threshold, effective January 1, 2020. In this final rule the Bureau is adjusting the permanent open-end threshold to 200 open-end lines of credit. For these amendments, the Bureau is finalizing the effective date of January 1, 2022, as proposed, to coincide with the expiration of the current temporary open-end threshold of 500 open-end lines of credit. As explained below, the amendments to the open-end threshold apply to covered loans and applications with respect to which final action is taken beginning January 1, 2022.

Consistent with feedback provided by industry stakeholders in connection with the 2015 HMDA Rule and the 2017 HMDA Rule, a number of commenters indicated in response to the May 2019 Proposal that a long implementation period is necessary when coverage changes result in new institutions having reporting obligations under HMDA. A few trade associations and industry commenters suggested the Bureau adopt a transition period or good-faith efforts standard for compliance with HMDA requirements in consultation with other regulators.

The Bureau determines that the period of more than 20 months between this final rule's issuance and the January 1, 2022 effective date for the adjustment to the open-end threshold will provide newly covered financial institutions with sufficient time to revise and update policies and procedures, implement any necessary systems changes, and train staff before beginning to collect open-end data in 2022. As the Bureau explained in the 2019 HMDA Rule, the two-year extension of the temporary threshold of 500 open-end lines of credit ensures that institutions that will be required to report under the new permanent threshold that takes effect in 2022 will have time to adapt their systems and prepare for compliance.

Under the permanent open-end threshold of 200 open-end lines of credit, beginning in calendar year 2022, financial institutions that originated at least 200 open-end lines of credit in each of the two preceding calendar years must collect and record data on their open-end lines of credit pursuant to § 1003.4 and report such data by March 1 of the following calendar year pursuant to § 1003.5(a)(i). As noted above, this requirement applies to covered loans and applications with respect to which final action is taken on or after January 1, 2022. For example, if a financial institution described in § 1003.2(g) originated at least 200 open-end lines of credit each year in 2020 and 2021 and takes final action on an application for an open-end line of credit on February 15, 2022, the financial institution must collect and record data on that application and report such data by March 1, 2023. This is true regardless of when the financial institution received the application.^[160] However, if an institution originated fewer than 200 open-end lines of credit in either of the two preceding calendar years, that institution is not required to collect, record, or report data on its open-end lines of credit for that calendar year. For example, if an institution originated at least 200 open-end lines of credit in 2020 but fewer than 200 open-end lines of credit in 2021, that institution does not need to collect, record, or report any data on open-end lines of credit for which it takes final action in 2022.

VII. Dodd-Frank Act Section 1022(b) Analysis

The Bureau has considered the potential benefits, costs, and impacts of the final rule.^[161] In developing the final rule, the Bureau has consulted with or offered to consult with the prudential regulators (the Board, the Federal Deposit Insurance Corporation (FDIC), the National Credit Union Administration, and the Office of the Comptroller of the Currency), the Department of Agriculture, the Department of Housing and Urban Development (HUD), the Department of Justice, the Department of the Treasury, the Department of Veterans Affairs, the Federal Housing Finance Agency, the Federal Trade Commission, and the Securities and Exchange Commission regarding, among other things, consistency with any prudential, market, or systemic objectives administered by such agencies.

As discussed in greater detail elsewhere throughout this supplementary information, in this rulemaking the Bureau is amending Regulation C, effective July 1, 2020, to increase the threshold for reporting data about closed-end mortgage loans to 100 originated closed-end mortgage loans in each of the two preceding years. In addition, the Bureau is amending Regulation C to set the open-end threshold at 200 originated open-end lines of credit in each of the two preceding years beginning in calendar year 2022.

A. Provisions To Be Analyzed

The final rule contains regulatory or commentary language (provisions). The discussion below considers the benefits, costs, and impacts of the following sets of major provisions of the final rule to:

1. Increase the threshold for reporting data about closed-end mortgage loans from 25 to 100 originations in each of the two preceding calendar years, effective July 1, 2020; and

2. Set the threshold for reporting data about open-end lines of credit at 200 originations in each of the two preceding calendar years, effective January 1, 2022.

With respect to each major provision, the discussion below considers the benefits, costs, and impacts to consumers and covered persons. The discussion also addresses comments the Bureau received on the proposed Dodd-Frank Act section 1022(b) analysis, as well as certain other comments on the benefits or costs of the relevant provisions of the May 2019 Proposal that the Bureau is finalizing in this rule, when doing so is helpful to understanding the Dodd-Frank Act section 1022(b) analysis. Some commenters that mentioned the benefits or costs of a provision of the May 2019 Proposal in the context of commenting on the merits of that provision are addressed in the relevant section-by-section analysis, above. In this respect, the Bureau's discussion under Dodd-Frank Act section 1022(b) is not limited to this discussion in part VII of the final rule.

B. Baselines for Consideration of Costs and Benefits

The Bureau has discretion in any rulemaking to choose an appropriate scope of analysis with respect to potential benefits, costs, and impacts and an appropriate baseline.

For the purposes of this analysis, references to the “first set of provisions” in this final rule are to those that increase the threshold for reporting data about closed-end mortgage loans from 25 to 100 originations in each of the two preceding calendar years, effective July 1, 2020. Under current Regulation C, absent this final rule, financial institutions that originated no fewer than 25 closed-end mortgage loans in each of the two preceding calendar years and meet other reporting criteria are required to report their closed-end activity under HMDA; furthermore, depository institutions and credit unions that originated fewer than 500 closed-end mortgage loans in each of the two preceding calendar years are generally exempt under the EGRRCPA from reporting certain data points under HMDA. That is the baseline adopted for this set of provisions throughout the analyses presented below.

For the purposes of this analysis, references to the “second set of provisions” in this final rule are to those that set the threshold for reporting data about open-end lines of credit at 200 originations in each of the two preceding calendar years, effective January 1, 2022. In the 2019 HMDA Rule, the Bureau granted two-year temporary relief (specifically, for 2020 and 2021) for financial institutions that did not originate at least 500 open-end lines of credit in each of the two preceding calendar years. The 2019 HMDA Rule provides that, absent any future rulemaking, the open-end threshold will revert to 100 open-end lines of credit starting in 2022, as established in the 2015 HMDA Rule. This final rule sets the threshold for reporting data about open-end lines of credit at 200 originations in each of the two preceding calendar years, effective January 1, 2022. Meanwhile, the EGRRCPA's partial exemption for open-end lines of credit of eligible insured depository institutions and insured credit unions took effect on May 24, 2018. Therefore, for the consideration of benefits and costs of this second set of provisions the Bureau is adopting a baseline in which the open-end threshold starting in year 2022 is reset at 100 open-end lines of credit in each of the two preceding calendar years, with some depository institutions and credit unions partially exempt under the EGRRCPA.

The Bureau notes that the May 2019 proposal's analysis relied on three separate baselines for each of the three sets of provisions in the proposal. With regard to the provision to increase the closed-end threshold from 25 to 100, the Bureau had explained that the appropriate baseline for this provision is a post-EGRRCPA world in which eligible financial institutions under the EGRRCPA are already partially exempt from the reporting of certain data points for closed-end mortgage loans. And with regard to the provision to set the permanent open-end threshold at 200, the Bureau had adopted a baseline in which the open-end coverage threshold starting in year 2020 is reset at 100 open-end lines of credit in each of the two preceding calendar years with some depository institutions and credit unions partially exempt under the EGRRCPA. But for the purpose of this final rule, with the 2019 HMDA Rule already having incorporated the EGRRCPA changes and finalized the two-year extension of the temporary open-end threshold, the Bureau can simplify by using a baseline that includes the 2019 HMDA Rule for the two provisions being finalized now, with the closed-end analysis using July 1, 2020 as the baseline, and the open-end analysis using January 1, 2022 as the baseline.

C. Coverage of the Final Rule

Both sets of provisions relieve certain financial institutions from HMDA's requirements for data points regarding closed-end mortgage loans or open-end lines of credit that they originate or purchase, or for which they receive applications, as described further in each section below. The final rule affects all financial institutions below certain thresholds as discussed in detail below.

D. Basic Approach of the Bureau's Consideration of Benefits and Costs and Data Limitations

This discussion relies on data that the Bureau has obtained from industry, other regulatory agencies, and publicly available sources. However, as discussed further below, the Bureau's ability to fully quantify the potential costs, benefits, and impacts of this final rule is limited in some instances by a scarcity of necessary data.

1. Benefits to Covered Persons

This final rule relates to the institutions and transactions that are excluded from HMDA's reporting requirements. Both sets of provisions in this final rule will reduce the regulatory burdens on covered persons while also decreasing the data reported to serve the statute's purposes. Therefore, the benefits of these provisions to covered persons are mainly the reduction of the costs to covered persons relative to the compliance costs the covered persons would have to incur under each baseline scenario.

The Bureau's 2015 HMDA Rule, as well as the 2014 proposed rule for the 2015 HMDA Rule and the material provided to the Small Business Review Panel leading to the 2015 HMDA Rule, presented a basic framework of analyzing compliance costs for HMDA reporting, including ongoing costs and one-time costs for financial institutions. Based on the Bureau's then study of the HMDA compliance process and costs, with the help of additional information gathered and verified through the Small Business Review Panel process, the Bureau classified the operational activities that financial institutions use for HMDA data collection and reporting into 18 discrete compliance “tasks” which can be grouped into four “primary tasks.” ^[162] Recognizing that the cost per loan of complying with HMDA's requirements differs by financial institution, the Bureau further identified seven key dimensions of compliance operations that were significant drivers of compliance costs, including the reporting system used, the degree of system integration, the degree of system automation, the compliance program, and the tools for geocoding, performing completeness checks, and editing. The Bureau found that the compliance operations of financial institutions tended to have similar levels of complexity across all seven dimensions. For example, if a given financial institution had less system integration, then it tended to use less automation and less complex tools for geocoding. Financial institutions generally did not use less complex approaches on one dimension and more complex approaches on another. The small entity representatives validated this perspective during the Small Business Review Panel meeting convened under the Small Business Regulatory Enforcement Fairness Act.^[163]

The Bureau realizes that costs vary by institution due to many factors, such as size, operational structure, and product complexity, and that this variance exists on a continuum that is impossible to fully represent. To consider costs in a practical and meaningful way, in the 2015 HMDA Rule the Bureau adopted an approach that focused on three representative tiers of financial institutions. In particular, to capture the relationships between operational complexity and compliance cost, the Bureau used the seven key dimensions noted above to define three broadly representative financial institutions according to the overall level of complexity of their compliance operations. Tier 1 denotes a representative financial institution with the highest level of complexity, tier 2 denotes a representative financial institution with a moderate level of complexity, and tier 3 denotes a representative financial institution with the lowest level of complexity. For each tier, the Bureau developed a separate set of assumptions and cost estimates.

Table 1 below provides an overview of all three representative tiers across the seven dimensions of compliance operations: ^[164]

Table 1—Types of HMDA Reporters ¹

	Tier 3 FIs tend to . . .	Tier 2 FIs tend to . . .	Tier 1 FIs tend to . . .
Systems	Enter data in Excel loan/application register Formatting Tool	Use LOS and HMS; Submit data via the HMDA Platform	Use multiple LOS, central SoR, HMS; Submit data via the HMDA Platform.
Integration	(None)	Have forward integration (LOS to HMS)	Have backward and forward integration; Integration with public HMDA APIs.
Automation	Manually enter data into loan/application register Formatting Tool; review and verify edits in the HMDA Platform	Loan/application register file produced by HMS; review edits in HMS and HMDA platform; verify edits via HMDA Platform	Loan/application register file produced by HMS; high automation compiling file and reviewing edits; verify edits via the HMDA platform.
Geocoding	Use FFIEC tool (manual)	Use batch processing	Use batch processing with multiple sources.
Completeness Checks	Check in HMDA Platform only	Use LOS, which includes completeness checks	Use multiple stages of checks.
Edits	Use FFIEC Edits only	Use FFIEC and customized edits	Use FFIEC and customized edits run multiple times.
Compliance Program	Have a joint compliance and audit office	Have basic internal and external accuracy audit	Have in-depth accuracy and fair lending audit.
1 FI is “financial institution”; LOS is “Loan Origination System”; HMS is “HMDA Data Management Software”; SoR is “System of Record.”

For a representative institution in each tier, in the 2015 HMDA Rule the Bureau produced a series of estimates of the costs of compliance, including the ongoing costs that financial institutions incurred prior to the implementation of the 2015 HMDA Rule, and the changes to the ongoing costs due to the 2015 HMDA Rule. The Bureau further provided the breakdown of the changes to the ongoing costs due to each major provision in the 2015 HMDA Rule, which includes the changes to the scope of the institutional coverage, the change to the scope of the transactional coverage, the revisions to the existing data points (as before the 2015 HMDA Rule) and the addition of new data points by the 2015 HMDA Rule.

For the impact analysis in this final rule, the Bureau is utilizing the cost estimates provided in the 2015 HMDA Rule for the representative financial institution in each of the three tiers, with some updates, mainly to reflect the inflation rate. In addition, for the financial institutions eligible for partial exemptions under the EGRRCPA, the Bureau is making updates to align the partially exempt data points (and data fields used to report these data points) with the cost impact analyses discussed in the impact analyses for the 2015 HMDA Rule. The Bureau's analyses below also take into account the operational improvements that have been implemented by the Bureau regarding HMDA reporting since the issuance of the 2015 HMDA Rule. The details of such analyses are contained in the following sections addressing the two sets of provisions of this final rule.

The Bureau received a number of comments relating to the benefits to covered persons of the May 2019 Proposal, both in response to the original proposal and in response to the July 2019 Reopening Notice, and it has considered these comments in finalizing this rule. Many industry commenters reported that they expend substantial resources on HMDA compliance that they could instead use for other purposes or that they have structured their lines of business to ensure they are not required to report under HMDA. Some cited, for example, the burden of establishing procedures, purchasing reporting software, and training staff to comply with HMDA, and noted that compliance can be particularly difficult for smaller institutions with limited staff. A trade association commented that the Bureau's estimates do not account for the reduction in examination burdens and the resources diverted to HMDA compliance from other more productive activities. It also asserted that the Bureau's burden analysis did not properly address data security costs associated with HMDA collection and reporting. Another trade association suggested that the three-tiered approach to estimating costs does not seem to account for the unique challenges of adapting business and multifamily lending to HMDA regulations and HMDA reporting infrastructure designed with single-family consumer mortgage lending in mind.

In their comments, consumer groups, civil rights groups, and other nonprofit organizations stated that Federal agency fair lending and CRA exams will become more burdensome for Federal agencies and the HMDA-exempt lenders since the agencies will now have to ask for internal data from the lenders instead of being able to use the HMDA data. They also noted that smaller-volume lenders already benefit from the EGRRCPA's partial exemptions and stated that almost all of the data that such institutions must report under HMDA would already need to be collected to comply with other statutes like the Truth in Lending Act, to sell loans to Fannie Mae or Freddie Mac, or to acquire Federal Housing Administration insurance for loans. A nonprofit organization that does HMDA-related research commented that it is hard to imagine that a bank would not keep an electronic record of its lending, even if it were not subject to HMDA reporting.

The Bureau has considered these comments and concludes, as it did in the 2019 HMDA Rule, that they do not undermine the Bureau's approach or cost parameters used in part VI of the May 2019 Proposal. For example, the activities that many industry commenters described as burdensome—including scrubbing data, training personnel, and preparing for HMDA-related examinations—are consistent with and captured by the 18 discrete compliance “tasks” that the Bureau identified through its study of the HMDA compliance process and costs in the 2015 HMDA rulemaking. As part of its analysis, the Bureau also recognized that costs vary by institution due to many factors, such as size, operational structure, and product complexity, and adopted a tiered framework to capture the relationships between operational complexity and compliance cost. While some products are more costly than others to report, the three-tiered framework uses representative institutions to capture this type of variability and estimate overall costs of HMDA reporting. In estimating compliance costs associated with HMDA reporting through this framework, the Bureau also recognized that much of the information required for HMDA reporting is information that financial institutions would need to collect, retain, and secure as part of their lending process, even if they were not subject to HMDA reporting. The Bureau therefore does not believe that the comments received provide a basis for departing from the approach for analyzing costs and benefits for covered persons used in part VI of the May 2019 Proposal.

The next step of the Bureau's consideration of the reduction of costs for covered persons involved aggregating the institution-level estimates of the cost reduction under each set of provisions up to the market-level. This aggregation required estimates of the total number of potentially impacted financial institutions and their total number of loan/application register records. The Bureau used a wide range of data in conducting this task, including recent HMDA data,^[165] Call Reports, and Consumer Credit Panel data. These analyses were challenging, because no single data source provided complete coverage of all the financial institutions that could be impacted and because there is varying data quality among the different sources.

To perform the aggregation, the Bureau mapped the potentially impacted financial institutions to the three tiers described above. For each of the provisions analyzed, the Bureau assumed none of the changes would affect the high-complexity tier 1 reporters. The Bureau then assigned the potentially impacted financial institutions to either tier 2 or tier 3. In doing so, the Bureau relied on two constraints: (1) The estimated number of impacted institutions in tiers 2 and 3, combined, must equal the estimated number of impacted institutions for the applicable provision, and (2) the number of loan/application register records submitted annually by the impacted financial institutions in tiers 2 and 3, combined, must equal the estimated number of loan/application register records for the applicable provision. As in the 2015 HMDA Rule, the Bureau assumed for closed-end reporting that a representative low-complexity tier 3 financial institution has 50 closed-end mortgage loan HMDA loan/application register records per year and a representative moderate-complexity tier 2 financial institution has 1,000 closed-end mortgage loan HMDA loan/application register records per year. Similarly, the Bureau assumed for open-end reporting that a representative low-complexity tier 3 financial institution has 150 open-end HMDA loan/application register records per year and a representative moderate-complexity tier 2 financial institution has 1,000 open-end HMDA loan/application register records per year. Constraining the total number of impacted institutions and the number of impacted loan/application register records across tier 2 and tier 3 to the aggregate estimates thus enables the Bureau to calculate the approximate numbers of impacted institutions in tiers 2 and 3 for each set of provisions.^[166]

Multiplying the impact estimates for representative financial institutions in each tier by the estimated number of impacted institutions, the Bureau arrived at the market-level estimates.

2. Costs to Covered Persons

In general, and as discussed in part VII.D.1 above, both sets of provisions in this final rule will reduce the ongoing operational costs associated with HMDA reporting for the affected covered persons. In the interim, it is possible that to adapt to the rule, covered persons may incur certain one-time costs. Such one-time costs are mostly related to training and system changes in covered persons' HMDA reporting/loan origination systems. Based on the Bureau's outreach to industry, however, the Bureau believes that such one-time costs are fairly small. Commenters did not indicate that covered persons who would be excluded completely from reporting HMDA data would incur significant costs, either for closed-end mortgage loans or for open-end lines of credit, or both.

3. Benefits to Consumers

Having generated estimates of the changes in ongoing costs and one-time costs to covered financial institutions, the Bureau then can attempt to estimate the potential pass-through of such cost reduction from these institutions to consumers, which could benefit consumers and affect credit access. According to economic theory, in a perfectly competitive market where financial institutions are profit maximizers, the affected financial institutions would pass on to consumers the marginal, i.e., variable, cost savings per application or origination, and absorb the one-time and increased fixed costs of complying with the rule. The Bureau estimated in the 2015 HMDA Rule the impacts on the variable costs of the representative financial institutions in each tier due to various provisions of that rule. Similarly, the estimates of the pass-through effect from covered persons to consumers due to the provisions under this rule are based on the relevant estimates of the changes to the variable costs in the 2015 HMDA Rule with some updates. The Bureau notes that the market structure in the consumer mortgage lending markets may differ from that of a perfectly competitive market (for instance due to information asymmetry between lenders and borrowers) in which case the pass-through to the consumers would most likely be smaller than the pass-through under the perfect competition assumption.^[167]

In the May 2019 Proposal, the Bureau requested additional comments on the potential pass-through from financial institutions to consumers due to the reduction in reporting costs. A trade association commented that it believed that the proposed higher thresholds will move mortgage markets to more perfect competition. It suggested that institutions that currently manage their origination volumes to stay below HMDA reporting thresholds will be incentivized to increase operations and that, by being able to offer savings on fees and pricing, and by being more competitive due to lower productions costs, smaller banks will be able to enter the mortgage market at more profitable levels. However, as the Bureau noted in the 2019 HMDA Rule, this comment did not provide specific estimates that the Bureau can utilize in refining the analyses.

4. Cost to Consumers

HMDA is a sunshine statute. The purposes of HMDA are to provide the public with loan data that can be used: (i) To help determine whether financial institutions are serving the housing needs of their communities; (ii) to assist public officials in distributing public-sector investment so as to attract private investment to areas where it is needed; and (iii) to assist in identifying possible discriminatory lending patterns and enforcing antidiscrimination statutes.^[168] The provisions in this final rule, as adopted, lessen the reporting requirements for excluded financial institutions by relieving them of the obligation to report all data points related to either closed-end mortgage loans or open-end lines of credit. As a sunshine statute regarding data reporting and disclosure, most of the benefits of HMDA are realized indirectly. With less data required to be collected and reported under HMDA, the HMDA data available to serve HMDA's statutory purposes will decline.^[169] However, to quantify the reduction of such benefits to consumers presents substantial challenges. The Bureau sought comment on the magnitude of the loss of HMDA benefits from these changes to the available data and/or methodologies for measuring these effects in the May 2019 Proposal.

The Bureau received a number of comments emphasizing the loss of HMDA benefits from decreased information lenders would report under HMDA were the May 2019 Proposal to be finalized. For example, a group of 148 local and national organizations stated that raising reporting thresholds will lead to another round of abusive and discriminatory lending similar to abuses that occurred in the years before the financial crisis. These commenters also stated that the general public, researchers, and Federal agencies will have an incomplete picture of lending trends in thousands of census tracts and neighborhoods if affected institutions no longer report HMDA data. Additionally, a State attorney general stated that the May 2019 Proposal failed to fully account for the harms that would be imposed by the proposal, including the costs to States in losing access to helpful data. However, as the Bureau noted in the 2019 HMDA Rule, none of these commenters provided specific quantifiable estimates of the loss of benefits from decreased information lenders would report under HMDA.

The Bureau acknowledges that quantifying and monetizing benefits of HMDA to consumers would require identifying all possible uses of HMDA data, establishing causal links to the resulting public benefits, and then quantifying the magnitude of these benefits. For instance, quantification would require measuring the impact of increased transparency on financial institution behavior, the need for public and private investment, the housing needs of communities, the number of financial institutions potentially engaging in discriminatory or predatory behavior, and the number of consumers currently being unfairly disadvantaged and the level of quantifiable damage from such disadvantage. Similarly, for the impact analyses of this final rule, the Bureau is unable to readily quantify the loss of some of the HMDA benefits to consumers with precision, both because the Bureau does not have the data to quantify all HMDA benefits and because the Bureau is not able to assess completely how this final rule will reduce those benefits.

In light of these data limitations, the discussion below generally provides a qualitative (not quantitative) consideration of the costs, i.e., the potential loss of HMDA benefits to consumers from the rule.

E. Potential Benefits and Costs to Consumers and Covered Persons

1. Overall Summary

In this section, the Bureau presents a concise, high-level table summarizing the benefits and costs considered in the remainder of the discussion. This table is not intended to capture all details and nuances that are provided both in the rest of the analysis and in the section-by-section discussion above. Instead, it provides an overview of the major benefits and costs of the final rule, including the provisions to be analyzed, the baseline chosen for each set of provisions, the sub-provisions to be analyzed, the implementation dates of the sub-provisions, the annual savings on the operational costs of covered persons due to the sub-provisions, the impact on the one-time costs of covered persons due to the sub-provision, and generally how the provisions in the final rule affect HMDA's benefits.

Table 2—Overview of Major Potential Benefits and Costs of the Final Rule

Provisions to be analyzed	Baseline	Baseline threshold	New threshold	Implementation date	Savings on annual operational costs	Impact on one time costs	Loss of data coverage
Increasing Closed-end Mortgage Loan Coverage Threshold	2015 and 2017 HMDA Rules, EGRRCPA, 2019 HMDA Rule	25 originations in each of two preceding calendar years	100 originations in each of two preceding calendar years	Effective July 1, 2020	$6.4 M	Negligible	Complete exclusion of reporting of approximately 1,700 reporters with about 112,000 closed-end mortgage loans.
Increasing Open-end Line of Credit Coverage Threshold	2015 and 2017 HMDA Rules, EGRRCPA, 2019 HMDA Rule	100 originations in each of two preceding calendar years	200 originations in each of two preceding calendar years	Effective January 1, 2022	$3.7 M	Savings of $23.9 M	Complete exclusion of reporting of approximately 400 reporters with about 69,000 open-end line of credit originations.

2. Provisions to Increase the Closed-End Threshold

Scope of the Provisions

The final rule increases the thresholds for reporting data about closed-end mortgage loans so that financial institutions originating fewer than 100 closed-end mortgage loans in either of the two preceding calendar years are excluded from HMDA's requirements for closed-end mortgage loans effective July 1, 2020.

The 2015 HMDA Rule requires institutions that originated at least 25 closed-end mortgage loans in each of the two preceding calendar years and meet all other reporting criteria to report their closed-end mortgage applications and loans. The EGRRCPA provides a partial exemption for insured depository institutions and insured credit unions that originated fewer than 500 closed-end mortgage loans in each of the two preceding years and do not have certain less than satisfactory CRA examination ratings. This section considers the provisions in the final rule that increase the closed-end loan threshold to 100 so that only financial institutions that originated at least 100 closed-end mortgage loans in each of the two preceding years must report data on their closed-end mortgage applications and loans under HMDA.

Using data from various sources, including past HMDA submissions, Call Reports, Credit Union Call Reports, Summary of Deposits, and the National Information Center, the Bureau applied all current HMDA reporting requirements, including Regulation C's existing complete regulatory exclusion for institutions that originated fewer than 25 closed-end mortgage loans in either of the two preceding calendar years, and estimates that currently there are about 4,860 financial institutions required to report their closed-end mortgage loans and applications under HMDA. Together, these financial institutions originated about 6.3 million closed-end mortgage loans in calendar year 2018. The Bureau observes that the total number of institutions that were engaged in closed-end mortgage lending in 2018, regardless of whether they met all HMDA reporting criteria, was about 11,600, and the total number of closed-end mortgage originations in 2018 was about 7.2 million. In other words, under the current 25 closed-end loan threshold, about 41.9 percent of all mortgage lenders are required to report HMDA data, and they account for about 87.8 percent of all closed-end mortgage originations in the country. The Bureau estimates that among those 4,860 financial institutions that are currently required to report closed-end mortgage loans under HMDA, about 3,250 insured depository institutions and insured credit unions are partially exempt for closed-end mortgage loans under the EGRRCPA, and thus are not required to report a subset of the data points currently required by Regulation C for these transactions.

The Bureau stated in the May 2019 Proposal that it intended to review the 2018 HMDA data more closely in connection with this rulemaking once the 2018 submissions were more complete. The Bureau released the national loan level dataset for 2018 and the Bureau's annual overview of residential mortgage lending based on HMDA data ^[170] (collectively the 2018 HMDA Data) in August 2019, and reopened the comment period on aspects of the May 2019 Proposal until October 15, 2019, to give commenters an opportunity to comment on the 2018 HMDA Data. The estimates reflected in this final rule are based on the HMDA data collected in 2017 and 2018 as well as other sources. The Bureau notes that the estimates provided above update the initial estimates provided in the May 2019 Proposal with the 2018 HMDA data.^[171] In particular, as the 2018 HMDA data analyses were not available at the time when the Bureau developed the May 2019 Proposal, the Bureau used HMDA data from 2016 and 2017 with a two-year look-back period in calendar years 2016 and 2017 for its estimates of potential reporters and projected the lending activities of financial institutions using their 2017 activities as proxies. In generating the updated estimates for this final rule, the Bureau has used HMDA data from 2017 and 2018 with a two-year look-back period in calendar years 2017 and 2018 for its estimates of potential reporters and projected the lending activities of financial institutions using their 2018 activities as proxies. In addition, for the estimates provided in the May 2019 Proposal and in this final rule, the Bureau restricted the projected reporters to only those that actually reported data in the most recently available year of HMDA data (2017 for the May 2019 proposal and 2018 for this final rule).^[172]

The Bureau estimates that when the closed-end threshold increases to 100 under this final rule, the total number of financial institutions required to report closed-end mortgage loans will drop to about 3,160, a decrease of about 1,700 financial institutions compared to the current level. These 1,700 newly excluded institutions originated about 112,000 closed-end mortgage loans in 2018. The Bureau estimates that there will be about 6.2 million closed-end mortgage loan originations reported under the threshold of 100 closed-end mortgage loans, which will account for about 86.3 percent of all closed-end mortgage loan originations in the entire mortgage market. The Bureau further estimates that about 1,630 of the 1,700 newly excluded closed-end reporters that will be excluded under the threshold of 100 closed-end mortgage loans are eligible for a partial exemption for closed-end mortgage loans under the EGRRCPA.

The Bureau notes that the estimates presented above update the corresponding estimates from the May 2019 Proposal ^[173] for the reasons explained above, reflecting more recent data. The updated estimates overall are consistent with the Bureau's analysis in the May 2019 Proposal and continue to support the Bureau's view regarding the impacts of a threshold of 100 closed-end mortgage loans.

Table 3 below shows the Bureau's estimates of the number of closed-end reporters that would be required to report under various potential thresholds, and the number of closed-end originations reported by these financial institutions, both in total and broken down by whether they are depository institutions or non-depository institutions, and among depository institutions whether they are partially exempt under the EGRRCPA.^[174]

Table 3—Estimated Number of Closed-End Reporters and Closed-End Mortgage Loans Reported Under Various Thresholds

Threshold	Non-depository institution	Depository institution		Total
		Not partially exempt	Partially exempt
25:
# of Reporters	740	870	3,250	4,860
# of Reported Loans (in thousands)	3,429	2,419	475	6,323
50
# of Reporters	720	870	2,530	4,120
# of Reported Loans (in thousands)	3,428	2,419	443	6,290
100
# of Reporters	680	860	1,620	3,160
# of Reported Loans (in thousands)	3,425	2,417	369	6,211

Benefits to Covered Persons

The final rule's complete exclusion from closed-end mortgage reporting for institutions that originated fewer than 100 closed-end mortgage loans in either of the two preceding calendar years conveys a direct benefit to the excluded covered persons by reducing the ongoing costs of having to report closed-end mortgage loans and applications that were previously required.

In the impact analysis of the 2015 HMDA Rule, prior to the adoption of the changes in the 2015 HMDA Rule and implementation of the Bureau's operational improvements, the Bureau estimated that the annual operational costs for financial institutions of reporting under HMDA were approximately $2,500 for a representative low-complexity tier 3 financial institution with a loan/application register size of 50 records; $35,600 for a representative moderate-complexity tier 2 financial institution with a loan/application register size of 1,000 records; and $313,000 for a representative high-complexity tier 1 financial institution with a loan/application register size of 50,000 records. The Bureau estimated that accounting for the operational improvements, the net impact of the 2015 HMDA Rule on ongoing operational costs for closed-end reporters would be approximately $1,900, $7,800, and $20,000 ^[175] per year, for representative low-, moderate-, and high-complexity financial institutions, respectively. This means that with all components of the 2015 HMDA Rule implemented and accounting for the Bureau's operational improvements, the estimated annual operational costs for closed-end mortgage reporting would be approximately $4,400 for a representative low-complexity tier 3 reporter, $43,400 for a representative moderate-complexity tier 2 reporter, and $333,000 for a representative high-complexity tier 1 reporter.

For purposes of this final rule, updating the above numbers to account for inflation, the Bureau estimates that if a financial institution is required to report under the 2015 HMDA Rule and is not partially exempt under the EGRRCPA, the savings on the annual operational costs from not reporting any closed-end mortgage data under the final rule is as follows: approximately $4,500 for a representative low-complexity tier 3 institution, $44,700 for a representative moderate-complexity tier 2 institution, and $343,000 for a representative high-complexity tier 1 institution. On the other hand, the Bureau estimates that if a financial institution is eligible for a partial exemption on its closed-end mortgage loans under the EGRRCPA, the annual savings in the ongoing costs from the partial exemption alone would be approximately $2,300 for a representative low-complexity tier 3 institution, $11,900 for a representative moderate-complexity tier 2 institution and $33,900 for a representative high-complexity tier 1 institution. Therefore, the Bureau estimates that if a financial institution is required to report under the 2015 HMDA Rule, but is partially exempt under the EGRRCPA, the savings in the annual operational costs from not reporting any closed-end mortgage data would be as follows: approximately $2,200 for a representative low-complexity tier 3 institution, $32,800 for a representative moderate-complexity tier 2 institution, and $309,000 for a representative high-complexity tier 1 institution. These estimates have already been adjusted for inflation.

In part VI of the May 2019 Proposal, the Bureau specifically requested information relating to the costs financial institutions incurred in collecting and reporting 2018 data in compliance with the 2015 HMDA Rule. The Bureau stated this information might be valuable in estimating costs in the Dodd-Frank Act section 1022(b) analysis issued with the final rule. The Bureau received a number of comments regarding the costs of collecting and reporting data in compliance with the 2015 HMDA Rule. Among the comments that provided specific cost estimates of compliance, most are related to closed-end reporting. The degree of details of such comments vary. Some provided the estimates of HMDA operational costs in dollar terms, some provided estimates of hours employees spent on each loan/application register record on average, some provided the cost of purchasing software, some provided consulting and auditing costs, and some provided the number of loan/application register records they processed while others did not.

The Bureau has reviewed these cost estimates provided in comments and compared them with the Bureau's estimates of HMDA operational costs using the three representative tier approach. Most of these commenters had low loan/application register volume similar to the representative tier 3 financial institutions. For example, one small financial institution commented that it was spending approximately $12,000 in employee expenses alone to generate its loan/application register or approximately $68 to $100 per loan/application register record. Based on the information provided by this commenter, the Bureau estimates the annual loan/application register size for this commenter is between 175 and 200 records, which is close to the Bureau's assumption for a representative low-complexity tier 3 financial institution in the estimates provided in the 2015 HMDA Rule. Specifically, the Bureau estimated that for a representative low-complexity tier 3 financial institution with 50 HMDA loan/application register records, the total ongoing costs with operational improvements the Bureau has implemented since issuing the 2015 HMDA Rule would be about $4,400, or about $88 per loan/application register record. Overall, the cost and hourly estimates provided by the commenters vary. Figure 1 plots the average costs per loan/application register record (on the vertical axis) against the number of loan/application register records (on vertical axis) for the low-complexity tier 3 financial institutions that provided cost estimates in their comments and that the Bureau was able to match to their 2018 HMDA loan/application register records. Other than a few outliers, they are all within the reasonable range that the Bureau anticipated and close to (though not exactly equal to) the Bureau's cost estimates for representative low-complexity tier 3 institutions. The Bureau notes that variation of operational costs among different financial institutions is not surprising. As the Bureau recognized in the 2015 HMDA Rule and the May 2019 Proposal, costs vary by institution due to many factors, such as size, operational structure, and product complexity, and that is the reason the Bureau adopted a tiered framework to capture the relationships between operational complexity and compliance cost. The three-tiered framework uses representative institutions to capture this type of variability and estimate overall costs of HMDA reporting.

The Bureau has considered the comments it received on compliance costs and concludes that they do not undermine the Bureau's approach or cost parameters used in part VI of the May 2019 Proposal. The Bureau therefore does not believe that the comments received provide a basis for departing from the approach for analyzing costs for covered persons used in part VI of the May 2019 Proposal.

Using the methodology discussed above in part VI.D.1, the Bureau estimates that with the threshold of 100 closed-end mortgage loans under the final rule, about 1,700 institutions will be completely excluded from reporting closed-end mortgage data compared to the current level. About 1,630 of the 1,700 are eligible for the partial exemption for closed-end mortgage loans under the EGRRCPA. Approximately 1,640 of these newly-excluded institutions are depository institutions, and approximately 60 are nondepository institutions.

The Bureau estimates that, of the approximately 1,630 institutions that are (1) required to report closed-end mortgage loans under the 2015 HMDA Rule, (2) partially exempt under the EGRRCPA, and (3) completely excluded under the threshold of 100 closed-end mortgage loans, about 1,560 are similar to the representative low-complexity tier 3 institution and about 70 are similar to the representative moderate-complexity tier 2 institution. Of the approximately 70 remaining institutions that are required to report closed-end mortgage data under the 2015 HMDA Rule and are not partially exempt under the EGRRCPA but will be completely excluded under the threshold of 100 closed-end mortgage loans, about 60 are similar to the representative low-complexity tier 3 institution and about 10 are similar to the representative moderate-complexity tier 2 institution.^[176]

Based on the estimates of the savings of annual ongoing costs for closed-end reporting per representative institution, grouped by whether or not it is partially exempt under the EGRRCPA, and the estimated tier distribution of these financial institutions that will be excluded under the 100 closed-end loan threshold, the Bureau estimates that the total savings in the annual ongoing costs from HMDA reporting by excluded firms that are already partially exempt for closed-end mortgage loans under the EGRRCPA will be about $5.9 million. The Bureau also estimates that the total savings in the annual ongoing costs from HMDA reporting by fully excluded firms that are not eligible for a partial exemption under the EGRRCPA will be about $0.5 million. Together the annual savings in the operational costs of firms newly excluded under the threshold of 100 closed-end loans will be about $6.4 million.^[177]

Alternative Considered: 50 Closed-End Threshold

The threshold of 100 closed-end mortgage loans adopted in this final rule is one of the two alternative closed-end thresholds that the Bureau proposed in the May 2019 Proposal. The other alternative threshold proposed in the May 2019 Proposal was 50.

The Bureau estimates that if the closed-end threshold were increased to 50, the total number of financial institutions that would be required to report closed-end mortgage loans would drop to about 4,120, a decrease of about 740 financial institutions compared to the current level at 25. The 740 institutions that would be excluded originated about 33,000 closed-end mortgage loans in 2018. There would be about 6.29 million closed-end mortgage originations reported under the alternative threshold of 50 closed-end mortgage loans that the Bureau considered, which would account for about 87.4 percent of all closed-end mortgage loan originations in the entire mortgage market.

The Bureau further estimates that about 720 of the 740 closed-end mortgage reporters that would be excluded under the alternative threshold of 50 closed-end mortgage loans would be eligible for a partial exemption for closed-end mortgage loans under the EGRRCPA.

The Bureau estimates that, of the approximately 720 financial institutions that are (1) required to report closed-end mortgage loans under the 2015 HMDA Rule, (2) partially exempt under the EGRRCPA, and (3) completely excluded under the alternative threshold of 50 closed-end mortgage loans, about 710 are similar to the representative low-complexity tier 3 institution and about 10 are similar to the representative moderate-complexity tier 2 institution. Of the approximately 20 remaining financial institutions that are required to report closed-end mortgage loans under the 2015 HMDA Rule and are not partially exempt under the EGRRCPA but would be completely excluded under the alternative threshold of 50 closed-end mortgage loans, all are similar to the representative low-complexity tier 3 institution.

As described above, the Bureau first estimates the savings of annual ongoing costs for closed-end reporting per representative institution, grouped by whether or not it is partially exempt for closed-end reporting under the EGRRCPA, and the tier distribution of these institutions that would be excluded under the alternative threshold of 50 closed-end mortgage loans. Using that information, the Bureau then estimates that, under the alternative threshold of 50 closed-end mortgage loans, the total savings in annual ongoing costs from HMDA reporting by fully excluded institutions that are already partially exempt under the EGRRCPA would be about $1.9 million, and the total savings in the annual ongoing costs from HMDA reporting by fully excluded firms that are not eligible for a partial exemption under the EGRRCPA would be about $0.1 million. Together the annual savings in the operational costs of firms excluded under the alternative threshold of 50 closed-end mortgage loans would be about $2.0 million.

The Bureau notes the estimates provided above for the alternative threshold of 50 update the estimates for the proposed threshold of 50 in the May 2019 Proposal for the reasons explained above.^[178]

The Bureau further notes that, because the Bureau is finalizing the closed-end threshold at 100 instead of 50, the covered persons will realize additional annual savings in their operational costs of about $4.4 million.

Costs to Covered Persons

It is possible that, like any new regulation or revision to an existing regulation, financial institutions will incur certain one-time costs adapting to the changes to the regulation. Based on the Bureau's outreach to stakeholders, the Bureau understands that most of these one-time costs consists of interpreting and implementing the regulatory changes and not from purchasing software upgrades or turning off the existing reporting functionality that the newly excluded institutions already built or purchased prior to the new changes taking effect.

The Bureau sought comments on any costs to institutions that would be newly excluded under either of the alternative proposed increases to the closed-end threshold. No commenter expressed concern that the costs for newly excluded reporters would be substantial.

Benefits to Consumers

Having generated estimates of the reduction in ongoing costs on covered financial institutions due to the increase in the closed-end loan threshold, the Bureau then attempts to estimate the potential pass-through of such cost reduction from these institutions to consumers, which could benefit consumers and affect credit access. According to economic theory, in a perfectly competitive market where financial institutions are profit maximizers, the affected financial institutions would pass on to consumers the marginal, i.e., variable, cost savings per application or origination, and absorb the one-time and increased fixed costs of complying with the rule.

The Bureau estimated in the 2015 HMDA Rule that the final rule would increase variable costs by $23 per closed-end mortgage application for representative low-complexity tier 3 institutions and $0.20 per closed-end mortgage application for representative moderate-complexity tier 2 institutions. The Bureau estimated that prior to the 2015 HMDA Rule, the variable costs of HMDA reporting were about $18 per closed-end mortgage application for representative low-complexity tier 3 institutions and $6 per closed-end mortgage application for representative moderate-complexity tier 2 institutions. For purposes of this final rule, adjusting the above numbers for inflation, the Bureau estimates the savings on the variable cost per closed-end application for a representative low-complexity tier 3 financial institution that is not partially exempt under the EGRRCPA but excluded from closed-end reporting under this final rule will be about $42 per application; the savings on the variable cost per application for a representative moderate-complexity tier 2 financial institution that is not partially exempt under the EGRRCPA but excluded from closed-end reporting under the final rule will be about $6.40 per application.

The Bureau estimates that the partial exemption for closed-end mortgage loans under the EGRRCPA for eligible insured depository institutions and insured credit unions reduces the variable costs of HMDA reporting by approximately $24 per closed-end mortgage application for representative low-complexity tier 3 institutions, $0.68 per closed-end mortgage application for representative moderate-complexity tier 2 institutions, and $0.05 per closed-end mortgage application for representative high-complexity tier 1 institutions. The savings on the variable cost per application for a representative low-complexity tier 3 financial institution that is partially exempt under the EGRRCPA and also fully excluded from closed-end reporting under the final rule will be about $18.30 per application. The savings on the variable cost per application for a representative moderate-complexity tier 2 financial institution that is partially exempt under the EGRRCPA and fully excluded from closed-end reporting under the final rule will be about $5.70 per application. These are the cost reductions that excluded institutions under the final rule might pass through to consumers, assuming the market is perfectly competitive. This potential reduction in the expense consumers face when applying for a mortgage will be amortized over the life of the loan and may represent a very small amount relative to the cost of a mortgage loan. As a point of reference, the median total loan costs for closed-end mortgages was $6,056 according to the 2018 HMDA Data.^[179] The Bureau notes that the market structure in the consumer mortgage lending market may differ from that of a perfectly competitive market (for instance due to information asymmetry between lenders and borrowers) in which case the pass-through to the consumers would most likely be smaller than the pass-through under the perfect competition assumption.^[180]

Costs to Consumers

The increase in the closed-end threshold to 100 loans will relieve excluded financial institutions from the reporting requirements for all closed-end mortgage loans and applications. As a result, HMDA data on these institutions' closed-end mortgage loans and applications will no longer be available to regulators, public officials, and members of the public. The decreased data about excluded institutions may lead to adverse outcomes for some consumers. For instance, HMDA data, if reported, could help regulators and public officials better understand the type of funds that are flowing from lenders to consumers and consumers' needs for mortgage credit. The data may also help improve the processes used to identify possible discriminatory lending patterns and enforce antidiscrimination statutes. A State attorney general commenter expressed concern that the May 2019 Proposal did not fully account for these costs, including the costs to States in losing access to helpful data, while a consumer organization commenter stated that the public would face an increased burden in understanding and accurately mapping the flow of credit. The Bureau did not, however, receive any comments that quantify the losses.

The Bureau recognizes that the costs to consumers from increasing the threshold to 100 loans will be higher than it would be if the Bureau were to increase the threshold to 50 loans. The Bureau currently lacks sufficient data to quantify these costs other than the estimated numbers of covered loans and covered institutions under the two alternative proposed thresholds, as discussed above and reported in Table 3.

3. Provisions to Increase the Open-End Threshold

Scope of the Provisions

The final rule will permanently set the threshold for reporting data about open-end lines of credit at 200 open-end lines of credit in each of the two preceding calendar years starting in 2022.

The 2015 HMDA Rule generally requires financial institutions that originated at least 100 open-end lines of credit in each of the two preceding years to report data about their open-end lines of credit and applications. The 2017 HMDA Rule temporarily increased the open-end threshold to 500 open-end lines of credit for two years, and the 2019 HMDA Rule extended the temporary threshold for two additional years. Thus, only financial institutions that originated at least 500 open-end lines of credit in each of the two preceding years are subject to HMDA's requirements for their open-end lines of credit for 2018 through 2021. The EGRRCPA generally provides a partial exemption for insured depository institutions and insured credit unions that originated fewer than 500 open-end lines of credit in each of the two preceding years and do not have certain less than satisfactory CRA examination ratings. However, for 2018 through 2021, all insured depository institutions and insured credit unions that are eligible for a partial exemption for open-end lines of credit by the EGRRCPA are also fully excluded from HMDA's requirements for their open-end lines of credit. Absent this final rule, starting in 2022 the open-end threshold would have reverted to 100, and eligible institutions that exceeded the threshold of 100 open-end lines of credit would have been able to use the EGRRCPA's open-end partial exemption if they originated fewer than 500 open-end lines of credit in each of the two preceding years. Thus, the appropriate baseline for the consideration of benefits and costs of the change to the open-end threshold is a situation in which the open-end threshold is set at 100 for each of two preceding years for data collection starting in 2022, with a partial exemption threshold of 500 open-end lines of credit.

The Bureau has used multiple data sources, including credit union Call Reports, Call Reports for banks and thrifts, HMDA data, and Consumer Credit Panel data, to develop estimates about open-end originations for lenders that offer open-end lines of credit and to assess the impact of various thresholds on the number of reporters and on market coverage under various scenarios.^[181]

In part VI of the May 2019 Proposal, the Bureau estimated that if the threshold were set at 100 open-end lines of credit, the number of reporters would be about 1,014, who in total originated about 1.41 million open-end lines of credit, representing about 88.7 percent of all originations and 15.3 percent of all lenders in the market. In comparison, if the threshold were set at 200 open-end lines of credit, the Bureau estimated that the number of reporters would be about 613, who in total originated about 1.34 million open-end lines of credit. In terms of market coverage, this would represent about 84.2 percent of all originations and 9.2 percent of all lenders in the open-end line of credit market. In other words, if the threshold were increased to 200, in comparison to the default baseline where the threshold was set at 100 in 2022, the Bureau estimated that the number of institutions affected would be about 401, who in total originated about 69,000 open-end lines of credit. Among those 401 institutions, the Bureau estimated that about 378 already qualify for a partial exemption for their open-end lines of credit under the EGRRCPA and in total they originate about 61,000 open-end lines of credit.

As the 2018 HMDA data analyses were not available at the time of the May 2019 Proposal, 2017 was the most recent year of HMDA data the Bureau used for the analyses in the May 2019 Proposal. For this part of the final rule, the Bureau has supplemented the analyses with the 2018 HMDA data now available. In the 2018 HMDA data, which used an open-end reporting threshold of 500, about 957 reporters actually reported any open-end line of credit transactions. In total, these institutions reported about 1.15 million open-end originations, which is close to what the Bureau projected in its estimate of 1.23 million originations to be reported in the May 2019 Proposal. Even though the number of open-end reporters in the 2018 HMDA data (957) is greater than the number the Bureau forecasted would be required to report (333) in the May 2019 Proposal, only 307 of the institutions that reported open-end transactions in the 2018 HMDA data actually reported greater than 500 open-end originations, which is close to the Bureau's projection that there would be 333 required open-end reporters with a reporting threshold of 500. The Bureau's projection in the May 2019 Proposal for the temporary threshold of 500 open-end originations was based on the projected number of open-end reporters whose open-end origination volumes were greater than 500 in each of the preceding two years (which is how the HMDA reporting requirements are structured), and not on the volume from the current HMDA activity year. In addition, that projection cannot account for the number of reporters who would report voluntarily even though they are not required to do so. Given these factors, it is possible that some lenders with open-end line of credit origination volumes exceeding 500 in both 2016 and 2017 originated fewer than 500 open-end lines of credit in 2018, but were nevertheless required to report their 2018 data under the HMDA reporting requirements. On the other hand, it is also possible that some reporters opted to report their open-end lending activities in the 2018 HMDA data even though they were not required to report. Regardless, these 2018 open-end reporters with a reported origination volume of fewer than 500 open-end lines of credit in 2018 are not required to collect data on their open-end activity in 2020 after the two-year temporary extension of the 500 open-end threshold of the 2019 HMDA Rule took effect, based on the two-year look-back period for the reporting requirements. Therefore, the Bureau believes that its estimates of the number of impacted institutions provided in the May 2019 Proposal were and are reasonable and consistent with the actual number of open-end reporters in the 2018 HMDA data.

Moreover, there are two additional considerations that support the Bureau's continued reliance in this final rule on its estimates on open-end coverage under various thresholds developed in the May 2019 Proposal. First, the permanent threshold of 200 open-end lines of credit starting in 2022 adopted in this final rule is lower than the temporary threshold of 500 open-end lines of credit that was in effect for the 2018 HMDA data. Hence, most institutions that originated fewer than 500 open-end lines of credit but at least 200 open-end lines of credit likely were not captured by the 2018 HMDA data, but they would have been required to report if the threshold had been set at 200. Second, the HMDA reporting requirements consider a two-year look-back period, and only 2018 HMDA data analyses were available as of the time of this final rule's development. For these reasons, the Bureau believes that its estimates of open-end coverage under various thresholds developed for the May 2019 Proposal continue to provide the most reliable estimates for this final rule.

On the other hand, because the number of open-end applications was not available in any data sources prior to the 2018 HMDA data, in past HMDA rulemakings related to open-end reporting, the Bureau relied on the projected number of originations as a proxy for the number of loan/application register records for the analyses. With the 2018 HMDA data reported, the Bureau now can evaluate the impact of the final rule using the projected loan/application register records instead of projected originations for the first time. Because most of the data points under HMDA are required for all loan/application register records, not just originated loans, the Bureau has updated the estimates of cost and cost savings for open-end lines of credit based on the number of loan/application register records instead of originations. The Bureau's coverage estimates, however, continue to be based on originations because the thresholds are based on origination volume, and thus, as noted immediately above, the estimates previously provided continue to be reasonable. The analyses below have been supplemented to reflect the new 2018 HMDA data that includes applications, originations, and purchased loans. Table 4 below shows the estimated number of reporters of open-end lines of credit, their estimated origination volume, and the market share under thresholds of 100, 200 and 500 open-end lines of credit.^[182] The Bureau notes that the threshold of 100 open-end lines of credit is the baseline of the analyses adopted for purposes of this final rule, the threshold of 200 open-end lines of credit is the threshold adopted under the final rule, and the threshold of 500 open-end lines of credit is the temporary threshold in place for 2020 and 2021 under the 2019 HMDA Rule.

Table 4—Estimated Number of Open-End Reporters and Open-End Lines of Credit Reported under Various Thresholds

Open-end Lines of Credit	Universe	Reporting Threshold
		100	200	500
# of Loans (in 1000's):
All	1,590	1,410	1,341	1,233
Market Coverage		88.7%	84.4%	77.6%
Type:
Banks Thrifts	880	814	787	753
Credit Unions	653	545	506	437
Non-DIs	57	51	48	44
# of Institutions:
All	6,615	1,014	613	333
Type:
Banks Thrifts	3,819	391	212	113
Credit Unions	2,578	581	376	205
Non-DIs	218	42	25	15

Benefits to Covered Persons

The increase in the permanent threshold from 100 to 200 open-end lines of credit in each of the two preceding calendar years starting in 2022, conveys a direct benefit to covered persons that originated fewer than 200 open-end lines of credit in either of the two preceding years but originated at least 100 open-end lines of credit in each of the two preceding years in reducing the ongoing costs of having to report their open-end lines of credit. The Bureau estimates that increasing the permanent threshold to 200 open-end lines of credit will relieve approximately 384 depository institutions and approximately 17 non-depository institutions from reporting open-end lines of credit as compared to having the threshold decrease to 100.

The Bureau estimates that, with the threshold increased to 200 as compared to decreasing to 100 starting in 2022, about 401 financial institutions will be excluded from reporting open-end lines of credit starting in 2022. About 378 of those 401 financial institutions are eligible for the partial exemption for open-end lines of credit under the EGRRCPA, and about 23 of them are not eligible for the partial exemption for open-end lines of credit because in one of the preceding two years their open-end origination volume exceeded 500. Of the 378 institutions that are already partially exempt under the EGRRCPA but will be fully excluded from open-end reporting starting in 2022 under this final rule, the Bureau estimates that about 301 are low-complexity tier 3 open-end reporters, about 77 are moderate-complexity tier 2 open-end reporters, and none are high-complexity tier 1 reporters. In addition, of the 23 institutions that are not eligible for the partial exemption under the EGRRCPA but will be fully excluded from open-end reporting starting in 2022 under this final rule, the Bureau estimates that about 8 are low-complexity tier 3 open-end reporters, about 15 are moderate-complexity tier 2 open-end reporters, and none are high-complexity tier 1 reporters.^[183] Using the estimates of savings on ongoing costs for open-end lines of credit for representative financial institutions, grouped by whether the lender is already eligible for the partial exemption under the EGRRCPA, as described above, the Bureau estimates that by increasing the threshold to 200 open-end lines of credit starting in 2022, the excluded financial institutions that are already partially exempt under the EGRRCPA will receive an aggregate reduction in operational cost associated with open-end lines of credit of about $3.0 million per year starting in 2022, while the excluded financial institutions that are not already partially exempt under the EGRRCPA will receive an aggregate reduction in operational cost associated with open-end lines of credit of about $0.7 million per year starting in 2022. In total, increasing the threshold from 100 to 200 open-end lines of credit will result in savings in the operational costs associated with open-end lines of credit of about $3.7 million per year starting in 2022.^[184] The increase in the threshold to 200 open-end lines of credit starting in calendar year 2022, as compared to having the threshold revert to 100, also conveys a direct benefit to covered persons that originated fewer than 200 open-end lines of credit in either of the two preceding years but originated at least 100 open-end lines of credit in each of the two preceding years in removing the one-time costs of having to report their open-end lines of credit, had the reporting threshold decreased to 100 according to the 2017 HMDA Rule.

It is the Bureau's understanding that most of the financial institutions that were temporarily excluded for 2018 through 2021 under the temporary threshold of 500 open-end lines of credit established in the 2017 HMDA Rule and 2019 HMDA Rule have not fully prepared for open-end reporting because they have been waiting for the Bureau to decide on the permanent open-end reporting threshold that will apply after the temporary threshold expires in 2022. Under the baseline in this impact analysis, absent this final rule, some of those financial institutions would have to start reporting their open-end lines of credit starting in 2022, and hence incur one-time costs to create processes and systems for open-end lines of credit. If the proposal to increase the open-end threshold to 200 starting in 2022 were not finalized, financial institutions that originated fewer than 200 open-end lines of credit in either of the two preceding years but originated at least 100 open-end lines of credit in each of the two preceding years would eventually have incurred one-time costs of having to report their open-end lines of credit, once the reporting threshold reverted to the permanent threshold of 100.

As noted in the 2015 HMDA Rule, the Bureau recognizes that many financial institutions, especially larger and more complex institutions, process applications for open-end lines of credit in their consumer lending departments using procedures, policies, and data systems separate from those used for closed-end loans. In the 2015 HMDA Rule, the Bureau assumed that the one-time costs for reporting information on open-end lines of credit required under the 2015 HMDA Rule would be roughly equal to 50 percent of the one-time costs of reporting information on closed-end mortgages. This translates to one-time costs of about $400,000 and $125,000 for open-end reporting for representative high- and moderate-complexity financial institutions, respectively, that will be required to report open-end lines of credit while also reporting closed-end mortgage loans. This assumption accounted for the fact that reporting open-end lines of credit will require some new systems, extra start-up training, and new compliance procedures and manuals, while recognizing that some fixed, one-time costs would need to be incurred anyway in making systemic changes to bring institutions into compliance with Regulation C and could be shared with closed-end lines of business. The assumption was consistent with the Bureau's estimate that an overwhelming majority of open-end reporters would also be reporting simultaneously closed-end mortgage loans and applications. In the 2015 HMDA Rule, the Bureau also assumed that the additional one-time costs of open-end reporting would be relatively low for low-complexity tier 3 financial institutions because they are less reliant on information technology systems for HMDA reporting and may process open-end lines of credit on the same system and in the same business unit as closed-end mortgage loans. Therefore, for low-complexity tier 3 financial institutions, the Bureau had assumed that the additional one-time cost created by open-end reporting is minimal and is derived mostly from new training and procedures adopted for the overall changes in the 2015 HMDA Rule.

In the proposal leading to the 2015 HMDA Rule, the Bureau asked for public comments and specific data regarding the one-time cost of reporting open-end lines of credit. Although some commenters on that proposal provided generic feedback on the additional burden of reporting data on these products, very few provided specific estimates of the potential one-time costs of reporting open-end lines of credit. After issuing the 2015 HMDA Rule, the Bureau heard anecdotal reports that one-time costs to begin reporting information on open-end lines of credit could be higher than the Bureau's estimates in the 2015 HMDA Rule. In the May 2019 Proposal, the Bureau indicated that it had reviewed the 2015 estimates and believed that the one-time cost estimates for open-end lines of credit provided in 2015, if applied to the proposed rule, would most likely be underestimates, for two reasons.

First, in developing the one-time cost estimates for open-end lines of credit in the 2015 HMDA Rule, the Bureau had envisioned that there would be cost sharing between the line of business that conducts open-end lending and the line of business that conducts closed-end lending at the corporate level, as the implementation of open-end reporting that became mandatory under the 2015 HMDA Rule would coincide with the implementation of the changes to closed-end reporting under the 2015 HMDA Rule. For instance, the resources of the corporate compliance department and information technology department could be shared and utilized simultaneously across different lines of business within the same lender in its efforts to set up processes and systems adapting to the 2015 HMDA Rule. Therefore, the Bureau assumed the one-time cost due to open-end reporting would be about one-half of the one-time costs due to closed-end reporting, in order to both reasonably count for the costs for reporting open-end lines of credit and avoid double counting. However, as the Bureau noted in the May 2019 Proposal, circumstances have somewhat changed since the 2015 HMDA Rule. The 2017 HMDA Rule temporarily increased the open-end lines of credit threshold from 100 to 500 for two years (2018 and 2019). The 2019 HMDA Rule further extended the temporary threshold of 500 open-end lines of credit for two additional years (2020 and 2021). Thus, there will be a considerable lag between the implementation of closed-end reporting changes under the 2015 HMDA Rule and the implementation of mandatory open-end reporting for those open-end lenders that have been temporarily excluded under the 2017 HMDA Rule and the 2019 HMDA Rule, but will be required to comply with HMDA's requirements for their open-end lines of credit starting in 2022 with the 200 origination threshold taking effect. As a result, the efficiency gain from one-time cost sharing between the closed-end and open-end reporting that was envisioned in the cost-benefit analysis of the 2015 HMDA Rule likely will not be applicable, if some of the temporarily excluded open-end reporters under the 2017 HMDA Rule and the 2019 HMDA Rule were to start preparing for open-end reporting several years after the implementation of closed-end changes.

Therefore, the Bureau now believes the one-time costs of starting to report information on open-end lines of credit, if the financial institution is to start reporting open-end lines of credit in 2022 and beyond, will be higher than the Bureau's initial estimates of one-time costs of open-end reporting provided in the 2015 HMDA Rule. Thus, for this impact analysis, the Bureau assumes for a representative moderate-complexity tier 2 open-end reporter that the one-time costs of starting open-end reporting in 2022 will be approximately equal to the one-time cost estimate for closed-end reporting that the Bureau estimated in the 2015 HMDA Rule, instead of being about one half of the one-time cost estimate for closed-end reporting. This translates to about $250,000 per representative moderate-complexity tier 2 open-end reporter, instead of $125,000 as the Bureau estimated in the 2015 HMDA Rule regarding the one-time costs of open-end reporting. This is the case regardless of whether the open-end reporters also report closed-end mortgage loans under HMDA. The Bureau notes that the moderate-complexity tier 2 financial institutions that will be permanently excluded from open-end reporting under this final rule will no longer have to incur such one-time costs.

Second, the temporary threshold that the 2017 HMDA Rule and 2019 HMDA Rule established delayed open-end reporting for those low-complexity tier 3 financial institutions that originated between 100 and 499 open-end lines of credit in either of the two preceding years. This delay means that those institutions would have had to incur the one-time costs to restart the training process for staff directly responsible for open-end data collection and reporting and update compliance procedures and manuals if the open-end threshold had reverted to 100 starting in 2022. In the 2015 HMDA Rule, the Bureau estimated the total one-time cost estimate for low-complexity tier 3 financial institutions would be approximately $3,000 regardless of whether the financial institution reports open-end lines of credit. Under this final rule, the Bureau thus assumes that the low-complexity tier 3 financial institutions that will be completely excluded from open-end reporting will be able to avoid incurring a one-time cost of about $3,000.

The Bureau estimates that, with the permanent threshold increased to 200 starting in 2022 as compared to reverting to 100, about 401 more institutions will be excluded from reporting open-end lines of credit starting in 2022. About 309 of those 401 institutions are low-complexity tier 3 open-end reporters, about 92 are moderate-complexity tier 2 open-end reporters, and none are high-complexity tier 1 reporters. Using the estimates of savings on one-time costs for open-end lines of credit for representative financial institutions discussed above, the Bureau estimates that with the increase in the threshold to 200 open-end lines of credit starting in 2022, the excluded institutions will receive an aggregate savings in avoided one-time cost associated with open-end lines of credit of about $23.9 million. This is an upward revision from the estimated savings of about $3.7 million in avoided one-time costs in the May 2019 Proposal, mainly because the Bureau has supplemented its analysis with new information from the 2018 HMDA data. As discussed above, these data allow the Bureau to develop estimates based on the total number of open-end loan/application register records rather than the number of open-end originations, and as a result the Bureau has shifted more affected institutions from tier 3 to tier 2. The overall analysis, however, is consistent with the Bureau's methodology and conclusions from the May 2019 Proposal.

Costs to Covered Persons

Like any new regulation or revision to the existing regulations, financial institutions may incur certain one-time costs adapting to the changes to the regulation. Based on the Bureau's outreach to stakeholders, the Bureau understands that most of such one-time costs would result from interpreting and implementing the regulatory changes, not from purchasing software upgrades or turning off the existing reporting functionality that the excluded institutions already built or purchased prior to the new changes taking its effect.

The Bureau sought comment on the costs and benefits to institutions that the rule would exclude pursuant to the proposed increases to the open-end threshold. No commenter expressed concern that the costs for newly excluded reporters would be substantial.

Benefits to Consumers

Having generated estimates of the reduction in ongoing costs on covered financial institutions due to the increase in the open-end threshold, the Bureau then attempts to estimate the potential pass-through of such cost reduction from the lenders to consumers, which could benefit consumers and affect credit access. According to economic theory, in a perfectly competitive market where financial institutions are profit maximizers, the affected financial institutions would pass on to consumers the marginal, i.e., variable, cost savings per application or origination, and absorb the one-time and increased fixed costs of complying with the rule.

The Bureau estimated in the 2015 HMDA Rule that the rule would increase variable costs by $41.50 per open-end line of credit application for representative low-complexity tier 3 institutions and $6.20 per open-end line of credit application for representative moderate-complexity tier 2 institutions. If the market is perfectly competitive, all of these savings on variable costs by the excluded open-end reporters could potentially be passed through to the consumers. These expenses will be amortized over the life of a loan and may represent a negligible reduction in the cost of a mortgage loan. As a point of reference, the median loan amount of open-end lines of credit (excluding reverse mortgages) in the 2018 HMDA data was $75,000.^[185] The Bureau notes that the market structure in the consumer mortgage lending market may differ from that of a perfectly competitive market (for instance due to information asymmetry between lenders and borrowers) in which case the pass-through to the consumers would most likely be smaller than the pass-through under the perfect competition assumption.^[186]

Costs to Consumers

Setting the permanent open-end threshold at 200 starting in 2022 will reduce the open-end data submitted under HMDA. As a result, HMDA data on these institutions' open-end lines of credit and applications will no longer be available to regulators, public officials, and members of the public. The decreased data concerning affected financial institutions may lead to adverse outcomes for some consumers. For instance, reporting data on open-end line of credit applications and originations and on certain demographic characteristics of applicants and borrowers could help the regulators and public officials better understand the type of funds that are flowing from lenders to consumers and consumers' need for mortgage credit. Open-end line of credit data that may be relevant to underwriting decisions may also help improve the processes used to identify possible discriminatory lending patterns and enforce antidiscrimination statutes. The Bureau has no quantitative data that can sufficiently measure the magnitude of any such impact of setting the permanent open-end threshold at 200. Additionally, the Bureau sought comment on the costs to consumers associated with the proposed increase to the open-end threshold but did not receive any comments that quantify the losses.

F. Potential Specific Impacts of the Final Rule

1. Depository Institutions and Credit Unions With $10 Billion or Less in Total Assets, as Described in Section 1026

As discussed above, the final rule will increase the threshold for reporting data about closed-end mortgage loans from 25 to 100 originations in both of the preceding two calendar years and increase the permanent threshold for reporting data about open-end lines of credit from 100 to 200 open-end lines of credit in both of the preceding two calendar years starting in 2022.

Both sets of provisions focus on burden reduction for smaller institutions. Therefore, the Bureau believes that the benefits of this final rule to depository institutions and credit unions with $10 billion or less in total assets will be similar to the benefit to creditors as a whole, as discussed above.

For the closed-end threshold provision, the Bureau estimates that for depository institutions and credit unions with $10 billion in assets or less that would have been required to report under the 2015 HMDA Rule, and are not partially exempt under the EGRRCPA, the savings on the annual operational costs from being excluded from closed-end reporting under the proposal will be approximately $4,500 for a representative low-complexity tier 3 institution, $44,700 for a representative moderate-complexity tier 2 institution, and $343,000 for a representative high-complexity tier 1 institution that fall below the threshold of 100. For depository institutions and credit unions with $10 billion in assets or less that would have been required to report under the 2015 HMDA Rule, but are partially exempt under the EGRRCPA, the Bureau estimates the savings on the annual operational costs from not reporting any closed-end mortgage data under the final rule will be approximately $2,200 for a representative low-complexity tier 3 institution, $32,800 for a representative moderate-complexity tier 2 institution, and $309,000 for a representative high-complexity tier 1 institution. For purposes of this final rule, the Bureau estimates that about 1,640 of the approximately 1,700 institutions that will be excluded by the reporting threshold of 100 closed-end mortgage loans are small depository institutions or credit unions with assets at or below $10 billion, and all but three of them are already partially exempt under the EGRRCPA. About 1,560 of them are similar to representative low-complexity tier 3 institution, with the rest being moderate-complexity tier 2 institutions. Combined, the annual savings on operational costs for depository institutions and credit unions with $10 billion or less in assets newly excluded under the threshold of 100 closed-end mortgage loans will be about $6.0 million.^[187]

For the open-end threshold provisions, the Bureau estimates that for depository institutions and credit unions with $10 billion in assets or less that will not have to report open-end lines of credit under the final rule, the reduction in annual ongoing operational costs for the excluded institutions not eligible for the partial exemption for open-end lines of credit under the EGRRCPA will be approximately $8,800, $44,700, and $281,000 per year, for representative low-, moderate-, and high-complexity financial institutions, respectively. The Bureau estimates that the reduction in annual ongoing operational costs for excluded institutions already partially exempt for open-end lines of credit under the EGRRCPA will be approximately $4,300, $21,900, and $138,000 annually, for representative low-, moderate-, and high-complexity financial institutions, respectively. The Bureau estimates that about 378 of the approximately 401 institutions that will be excluded from open-end reporting starting in 2022 under the final rule are small depository institutions or credit unions with assets at or below $10 billion, and about 372 of them are already partially exempt under the EGRRCPA. Combined, the Bureau estimates that the annual saving on operational costs for depository institutions and credit unions with $10 billion or less in assets newly excluded from open-end reporting under the threshold of 200 open-end lines of credit in this final rule would be about $3.5. million per year starting in 2022. Using the estimates of savings on one-time costs for open-end lines of credit for representative financial institutions discussed above, the Bureau estimates that by increasing the open-end threshold to 200 starting in 2022, the excluded depository institutions and credit unions with $10 billion or less in assets will receive an aggregate savings in avoided one-time costs associated with open-end lines of credit of about $20.9 million.^[188]

2. Impact of the Provisions on Consumers in Rural Areas

The final rule will not directly impact consumers in rural areas. However, as with all consumers, consumers in rural areas may be impacted indirectly. This would occur if financial institutions serving rural areas are HMDA reporters (in which case the final rule will lead to decreased information in rural areas) and if these institutions pass on some or all of the cost reduction to consumers (in which case, some consumers could benefit).

Recent research suggests that financial institutions that primarily serve rural areas are generally not HMDA reporters.^[189] The Housing Assistance Council (HAC) suggests that the current asset and geographic coverage criteria already in place disproportionately exempt small lenders operating in rural communities. For example, HAC uses 2009 Call Report data to show that approximately 700 FDIC-insured lending institutions had assets totaling less than the HMDA institutional coverage threshold and were headquartered in rural communities. These institutions, which would not be HMDA reporters, may represent one of the few sources of credit for many rural areas. Some research also suggests that limited HMDA data are currently reported for rural areas, especially areas further from Metropolitan Statistical Areas (MSAs).^[190] If a large portion of the rural housing market is serviced by financial institutions that are already not HMDA reporters, any indirect impact of the changes on consumers in rural areas would be limited, as the changes directly involve none of those financial institutions.

However, although some research suggests that HMDA currently does not cover a significant number of financial institutions serving the rural housing market, HMDA data do contain information for some covered loans involving properties in rural areas. These data can be used to estimate the number of HMDA reporters servicing rural areas, and the number of consumers in rural areas that might potentially be affected by the changes to Regulation C. For this analysis, the Bureau uses non-MSA areas as a proxy for rural areas, with the understanding that portions of MSAs and non-MSAs may contain urban and rural territory and populations. In 2018, 4,773 HMDA reporters reported applications or purchased loans for property located in geographic areas outside of an MSA. In total, these 5,207 financial institutions reported 1,562,399 applications or purchased loans for properties in non-MSA areas. This number provides an upper-bound estimate of the number of consumers in rural areas that could be impacted indirectly by the changes. In general, individual financial institutions report small numbers of covered loans from non-MSAs, as approximately 76 percent reported fewer than 100 covered loans from non-MSAs.

Following microeconomic principles, the Bureau believes that financial institutions will pass on reduced variable costs to future mortgage applicants, but absorb one-time costs and increased fixed costs if financial institutions are profit maximizers and the market is perfectly competitive.^[191] The Bureau defines variable costs as costs that depend on the number of applications received. Based on initial outreach efforts, the following five operational steps affect variable costs: Transcribing data, resolving reportability questions, transferring data to an HMS, geocoding, and researching questions. The primary impact of the final rule on these operational steps is a reduction in time spent per task. Overall, the Bureau estimates that the impact of the final rule on variable costs per application is to reduce variable costs by no more than $42 for a representative low-complexity tier 3 financial institution, $6 for a representative moderate-complexity tier 2 financial institution, and $3 for a representative high-complexity tier 1 financial institution.^[192] The 4,773 financial institutions that serviced rural areas could attempt to pass these reduced variable costs on to all future mortgage customers, including the estimated 1.6 million consumers from rural areas. Amortized over the life of the loan, this expense likely represents a negligible reduction in the cost of a mortgage loan. The Bureau notes that the market structure in the consumer mortgage lending market may differ from that of a perfectly competitive market (for instance due to information asymmetry between lenders and borrowers) in which case the pass-through to the consumers would most likely be smaller than the pass-through under the perfect competition assumption.^[193]

The rural market may differ from non-rural markets in terms of market structure, demand, supply, and competition level. For instance, local or community banks may be more likely to serve some rural markets than national lenders. Therefore, consumers in rural areas may experience benefits and costs from the final rule that are different than those experienced by consumers in general. To the extent that the impacts of the final rule on creditors differ by type of creditor, this may affect the costs and benefits of the final rule on consumers in rural areas.

The Bureau also recognizes, as discussed in the section-by-section analysis of § 1003.2(g) above, that rural and low-to-moderate income census tracts will lose proportionately more data as the threshold increases than other areas. However, the Bureau currently lacks sufficient data to quantify the impact of this decrease in data.

VIII. Final Regulatory Flexibility Act Analysis

The Regulatory Flexibility Act ^[194] as amended by the Small Business Regulatory Enforcement Fairness Act of 1996 ^[195] (RFA) requires each agency to consider the potential impact of its regulations on small entities, including small businesses, small governmental units, and small not-for-profit organizations.^[196] The RFA defines a “small business” as a business that meets the size standard developed by the Small Business Administration pursuant to the Small Business Act.^[197]

The RFA generally requires an agency to conduct an initial regulatory flexibility analysis (IRFA) and a final regulatory flexibility analysis (FRFA) of any rule subject to notice-and-comment rulemaking requirements, unless the agency certifies that the rule will not have a significant economic impact on a substantial number of small entities.^[198] The Bureau also is subject to certain additional procedures under the RFA involving the convening of a panel to consult with small business representatives prior to proposing a rule for which an IRFA is required.^[199]

As discussed above, this final rule increases the threshold for reporting data about closed-end mortgage loans from 25 to 100 originations in each of the two preceding calendar years and sets the permanent open-end threshold at 200 originations when the temporary threshold of 500 originations expires in 2022. The section 1022(b)(2) analysis above describes how this final rule reduces the costs and burdens on covered persons, including small entities. Additionally, as described in the analysis above, a small entity that is in compliance with the law at such time when this final rule takes effect does not need to take any additional action to remain in compliance other than choosing to switch off all or parts of reporting systems and functions. Based on these considerations, the final rule does not have a significant economic impact on any small entities.

Accordingly, the Director hereby certifies that this final rule will not have a significant economic impact on a substantial number of small entities. Thus, neither an FRFA nor a small business review panel is required for this final rule.

IX. Paperwork Reduction Act

Under the Paperwork Reduction Act of 1995 (PRA) (44 U.S.C. 3501 et seq.), Federal agencies are generally required to seek the Office of Management and Budget's (OMB's) approval for information collection requirements prior to implementation. The collections of information related to Regulation C have been previously reviewed and approved by OMB and assigned OMB Control number 3170-0008. Under the PRA, the Bureau may not conduct or sponsor and, notwithstanding any other provision of law, a person is not required to respond to an information collection unless the information collection displays a valid control number assigned by OMB. The Bureau has determined that this final rule would not impose any new or revised information collection requirements (recordkeeping, reporting or disclosure requirements) on covered entities or members of the public that would constitute collections of information requiring OMB approval under the PRA.

X. Congressional Review Act

Pursuant to the Congressional Review Act,^[200] the Bureau will submit a report containing this rule and other required information to the U.S. Senate, the U.S. House of Representatives, and the Comptroller General of the United States prior to the rule's published effective date. The Office of Information and Regulatory Affairs has designated this rule as not a “major rule” as defined by 5 U.S.C. 804(2).

XI. Signing Authority

The Director of the Bureau, having reviewed and approved this document is delegating the authority to electronically sign this document to Laura Galban, a Bureau Federal Register Liaison, for purposes of publication in the Federal Register.

List of Subjects in 12 CFR Part 1003

• Banks
• Banking
• Credit unions
• Mortgages
• National banks
• Reporting and recordkeeping requirements
• Savings associations

Authority and Issuance

For the reasons set forth above, the Bureau amends Regulation C, 12 CFR part 1003, as follows:

PART 1003—HOME MORTGAGE DISCLOSURE (REGULATION C)

1. The authority citation for part 1003 continues to read as follows:

Authority: 12 U.S.C. 2803, 2804, 2805, 5512, 5581.

2. Effective July 1, 2020, § 1003.2 is amended by revising paragraphs (g)(1)(v)(A) and (g)(2)(ii)(A) to read as follows:

§ 1003.2

Definitions.

* * * * *

(g) * * *

(1) * * *

(v) * * *

(A) In each of the two preceding calendar years, originated at least 100 closed-end mortgage loans that are not excluded from this part pursuant to § 1003.3(c)(1) through (10) or (c)(13); or

* * * * *

(2) * * *

(ii) * * *

(A) In each of the two preceding calendar years, originated at least 100 closed-end mortgage loans that are not excluded from this part pursuant to § 1003.3(c)(1) through (10) or (c)(13); or

* * * * *

3. Effective July 1, 2020, § 1003.3 is amended by revising paragraph (c)(11) to read as follows:

§ 1003.3

Exempt institutions and excluded and partially exempt transactions.

* * * * *

(c) * * *

(11) A closed-end mortgage loan, if the financial institution originated fewer than 100 closed-end mortgage loans in either of the two preceding calendar years; a financial institution (including, for purposes of information collected in 2020, an institution that was a financial institution as of January 1, 2020) may collect, record, report, and disclose information, as described in §§ 1003.4 and 1003.5, for such an excluded closed-end mortgage loan as though it were a covered loan, provided that the financial institution complies with such requirements for all applications for closed-end mortgage loans that it receives, closed-end mortgage loans that it originates, and closed-end mortgage loans that it purchases that otherwise would have been covered loans during the calendar year during which final action is taken on the excluded closed-end mortgage loan;

* * * * *

4. Effective July 1, 2020, supplement I to part 1003 is amended as follows:

a. Under Section 1003.2—Definitions, revise 2(g) Financial Institution.

b. Under Section 1003.3—Exempt Institutions and Excluded and Partially Exempt Transactions, under 3(c) Excluded Transactions, revise Paragraph 3(c)(11).

The revisions read as follows:

Supplement I to Part 1003—Official Interpretations

* * * * *

Section 1003.2—Definitions

* * * * *

2(g) Financial Institution

1. Preceding calendar year and preceding December 31. The definition of financial institution refers both to the preceding calendar year and the preceding December 31. These terms refer to the calendar year and the December 31 preceding the current calendar year. For example, in 2021, the preceding calendar year is 2020, and the preceding December 31 is December 31, 2020. Accordingly, in 2021, Financial Institution A satisfies the asset-size threshold described in § 1003.2(g)(1)(i) if its assets exceeded the threshold specified in comment 2(g)-2 on December 31, 2020. Likewise, in 2021, Financial Institution A does not meet the loan-volume test described in § 1003.2(g)(1)(v)(A) if it originated fewer than 100 closed-end mortgage loans during either 2019 or 2020.

2. Adjustment of exemption threshold for banks, savings associations, and credit unions. For data collection in 2020, the asset-size exemption threshold is $47 million. Banks, savings associations, and credit unions with assets at or below $47 million as of December 31, 2019, are exempt from collecting data for 2020.

3. Merger or acquisition—coverage of surviving or newly formed institution. After a merger or acquisition, the surviving or newly formed institution is a financial institution under § 1003.2(g) if it, considering the combined assets, location, and lending activity of the surviving or newly formed institution and the merged or acquired institutions or acquired branches, satisfies the criteria included in § 1003.2(g). For example, A and B merge. The surviving or newly formed institution meets the loan threshold described in § 1003.2(g)(1)(v)(B) if the surviving or newly formed institution, A, and B originated a combined total of at least 500 open-end lines of credit in each of the two preceding calendar years. Likewise, the surviving or newly formed institution meets the asset-size threshold in § 1003.2(g)(1)(i) if its assets and the combined assets of A and B on December 31 of the preceding calendar year exceeded the threshold described in § 1003.2(g)(1)(i). Comment 2(g)-4 discusses a financial institution's responsibilities during the calendar year of a merger.

4. Merger or acquisition—coverage for calendar year of merger or acquisition. The scenarios described below illustrate a financial institution's responsibilities for the calendar year of a merger or acquisition. For purposes of these illustrations, a “covered institution” means a financial institution, as defined in § 1003.2(g), that is not exempt from reporting under § 1003.3(a), and “an institution that is not covered” means either an institution that is not a financial institution, as defined in § 1003.2(g), or an institution that is exempt from reporting under § 1003.3(a).

i. Two institutions that are not covered merge. The surviving or newly formed institution meets all of the requirements necessary to be a covered institution. No data collection is required for the calendar year of the merger (even though the merger creates an institution that meets all of the requirements necessary to be a covered institution). When a branch office of an institution that is not covered is acquired by another institution that is not covered, and the acquisition results in a covered institution, no data collection is required for the calendar year of the acquisition.

ii. A covered institution and an institution that is not covered merge. The covered institution is the surviving institution, or a new covered institution is formed. For the calendar year of the merger, data collection is required for covered loans and applications handled in the offices of the merged institution that was previously covered and is optional for covered loans and applications handled in offices of the merged institution that was previously not covered. When a covered institution acquires a branch office of an institution that is not covered, data collection is optional for covered loans and applications handled by the acquired branch office for the calendar year of the acquisition.

iii. A covered institution and an institution that is not covered merge. The institution that is not covered is the surviving institution, or a new institution that is not covered is formed. For the calendar year of the merger, data collection is required for covered loans and applications handled in offices of the previously covered institution that took place prior to the merger. After the merger date, data collection is optional for covered loans and applications handled in the offices of the institution that was previously covered. When an institution remains not covered after acquiring a branch office of a covered institution, data collection is required for transactions of the acquired branch office that take place prior to the acquisition. Data collection by the acquired branch office is optional for transactions taking place in the remainder of the calendar year after the acquisition.

iv. Two covered institutions merge. The surviving or newly formed institution is a covered institution. Data collection is required for the entire calendar year of the merger. The surviving or newly formed institution files either a consolidated submission or separate submissions for that calendar year. When a covered institution acquires a branch office of a covered institution, data collection is required for the entire calendar year of the merger. Data for the acquired branch office may be submitted by either institution.

5. Originations. Whether an institution is a financial institution depends in part on whether the institution originated at least 100 closed-end mortgage loans in each of the two preceding calendar years or at least 500 open-end lines of credit in each of the two preceding calendar years. Comments 4(a)-2 through -4 discuss whether activities with respect to a particular closed-end mortgage loan or open-end line of credit constitute an origination for purposes of § 1003.2(g).

6. Branches of foreign banks—treated as banks. A Federal branch or a State-licensed or insured branch of a foreign bank that meets the definition of a “bank” under section 3(a)(1) of the Federal Deposit Insurance Act (12 U.S.C. 1813(a)) is a bank for the purposes of § 1003.2(g).

7. Branches and offices of foreign banks and other entities—treated as nondepository financial institutions. A Federal agency, State-licensed agency, State-licensed uninsured branch of a foreign bank, commercial lending company owned or controlled by a foreign bank, or entity operating under section 25 or 25A of the Federal Reserve Act, 12 U.S.C. 601 and 611 (Edge Act and agreement corporations) may not meet the definition of “bank” under the Federal Deposit Insurance Act and may thereby fail to satisfy the definition of a depository financial institution under § 1003.2(g)(1). An entity is nonetheless a financial institution if it meets the definition of nondepository financial institution under § 1003.2(g)(2).

* * * * *

Section 1003.3—Exempt Institutions and Excluded and Partially Exempt Transactions

* * * * *

3(c) Excluded Transactions

* * * * *

Paragraph 3(c)(11)

1. General. Section 1003.3(c)(11) provides that a closed-end mortgage loan is an excluded transaction if a financial institution originated fewer than 100 closed-end mortgage loans in either of the two preceding calendar years. For example, assume that a bank is a financial institution in 2021 under § 1003.2(g) because it originated 600 open-end lines of credit in 2019, 650 open-end lines of credit in 2020, and met all of the other requirements under § 1003.2(g)(1). Also assume that the bank originated 75 and 90 closed-end mortgage loans in 2019 and 2020, respectively. The open-end lines of credit that the bank originated or purchased, or for which it received applications, during 2021 are covered loans and must be reported, unless they otherwise are excluded transactions under § 1003.3(c). However, the closed-end mortgage loans that the bank originated or purchased, or for which it received applications, during 2021 are excluded transactions under § 1003.3(c)(11) and need not be reported. See comments 4(a)-2 through -4 for guidance about the activities that constitute an origination.

2. Optional reporting. A financial institution may report applications for, originations of, or purchases of closed-end mortgage loans that are excluded transactions because the financial institution originated fewer than 100 closed-end mortgage loans in either of the two preceding calendar years. However, a financial institution that chooses to report such excluded applications for, originations of, or purchases of closed-end mortgage loans must report all such applications for closed-end mortgage loans that it receives, closed-end mortgage loans that it originates, and closed-end mortgage loans that it purchases that otherwise would be covered loans for a given calendar year. Note that applications which remain pending at the end of a calendar year are not reported, as described in comment 4(a)(8)(i)-14. An institution that was a financial institution as of January 1, 2020 but is not a financial institution on July 1, 2020 because it originated fewer than 100 closed-end mortgage loans in 2018 or 2019 is not required in 2021 to report, but may report, applications for, originations of, or purchases of closed-end mortgage loans for calendar year 2020 that are excluded transactions because the institution originated fewer than 100 closed-end mortgage loans in 2018 or 2019. However, an institution that was a financial institution as of January 1, 2020 and chooses to report such excluded applications for, originations of, or purchases of closed-end mortgage loans in 2021 must report all such applications for closed-end mortgage loans that it receives, closed-end mortgage loans that it originates, and closed-end mortgage loans that it purchases that otherwise would be covered loans for all of calendar year 2020.

* * * * *

5. Effective January 1, 2022, § 1003.2, as amended at 84 FR 57946, October 29, 2019, is further amended by revising paragraphs (g)(1)(v)(B) and (g)(2)(ii)(B) to read as follows:

§ 1003.2

Definitions.

* * * * *

(g) * * *

(1) * * *

(v) * * *

(B) In each of the two preceding calendar years, originated at least 200 open-end lines of credit that are not excluded from this part pursuant to § 1003.3(c)(1) through (10); and

* * * * *

(2) * * *

(ii) * * *

(B) In each of the two preceding calendar years, originated at least 200 open-end lines of credit that are not excluded from this part pursuant to § 1003.3(c)(1) through (10).

* * * * *

6. Effective January 1, 2022, § 1003.3, is amended by revising paragraph (c)(11) and as amended at 84 FR 57946, October 29, 2019, is further amended by revising paragraph (c)(12) to read as follows:

§ 1003.3

Exempt institutions and excluded and partially exempt transactions.

* * * * *

(c) * * *

(11) A closed-end mortgage loan, if the financial institution originated fewer than 100 closed-end mortgage loans in either of the two preceding calendar years; a financial institution may collect, record, report, and disclose information, as described in §§ 1003.4 and 1003.5, for such an excluded closed-end mortgage loan as though it were a covered loan, provided that the financial institution complies with such requirements for all applications for closed-end mortgage loans that it receives, closed-end mortgage loans that it originates, and closed-end mortgage loans that it purchases that otherwise would have been covered loans during the calendar year during which final action is taken on the excluded closed-end mortgage loan;

(12) An open-end line of credit, if the financial institution originated fewer than 200 open-end lines of credit in either of the two preceding calendar years; a financial institution may collect, record, report, and disclose information, as described in §§ 1003.4 and 1003.5, for such an excluded open-end line of credit as though it were a covered loan, provided that the financial institution complies with such requirements for all applications for open-end lines of credit that it receives, open-end lines of credit that it originates, and open-end lines of credit that it purchases that otherwise would have been covered loans during the calendar year during which final action is taken on the excluded open-end line of credit; or

* * * * *

7. Effective January 1, 2022, supplement I to part 1003, as amended at 84 FR 57946, October 29, 2019, is further amended as follows:

a. Under Section 1003.2—Definitions, revise 2(g) Financial Institution; and

b. Under Section 1003.3—Exempt Institutions and Excluded and Partially Exempt Transactions, under 3(c) Excluded Transactions, revise Paragraphs 3(c)(11) and 3(c)(12).

The revisions read as follows:

Supplement I to Part 1003—Official Interpretations

* * * * *

Section 1003.2—Definitions

* * * * *

2(g) Financial Institution

1. Preceding calendar year and preceding December 31. The definition of financial institution refers both to the preceding calendar year and the preceding December 31. These terms refer to the calendar year and the December 31 preceding the current calendar year. For example, in 2021, the preceding calendar year is 2020, and the preceding December 31 is December 31, 2020. Accordingly, in 2021, Financial Institution A satisfies the asset-size threshold described in § 1003.2(g)(1)(i) if its assets exceeded the threshold specified in comment 2(g)-2 on December 31, 2020. Likewise, in 2021, Financial Institution A does not meet the loan-volume test described in § 1003.2(g)(1)(v)(A) if it originated fewer than 100 closed-end mortgage loans during either 2019 or 2020.

2. [Reserved]

3. Merger or acquisition—coverage of surviving or newly formed institution. After a merger or acquisition, the surviving or newly formed institution is a financial institution under § 1003.2(g) if it, considering the combined assets, location, and lending activity of the surviving or newly formed institution and the merged or acquired institutions or acquired branches, satisfies the criteria included in § 1003.2(g). For example, A and B merge. The surviving or newly formed institution meets the loan threshold described in § 1003.2(g)(1)(v)(B) if the surviving or newly formed institution, A, and B originated a combined total of at least 200 open-end lines of credit in each of the two preceding calendar years. Likewise, the surviving or newly formed institution meets the asset-size threshold in § 1003.2(g)(1)(i) if its assets and the combined assets of A and B on December 31 of the preceding calendar year exceeded the threshold described in § 1003.2(g)(1)(i). Comment 2(g)-4 discusses a financial institution's responsibilities during the calendar year of a merger.

4. Merger or acquisition—coverage for calendar year of merger or acquisition. The scenarios described below illustrate a financial institution's responsibilities for the calendar year of a merger or acquisition. For purposes of these illustrations, a “covered institution” means a financial institution, as defined in § 1003.2(g), that is not exempt from reporting under § 1003.3(a), and “an institution that is not covered” means either an institution that is not a financial institution, as defined in § 1003.2(g), or an institution that is exempt from reporting under § 1003.3(a).

i. Two institutions that are not covered merge. The surviving or newly formed institution meets all of the requirements necessary to be a covered institution. No data collection is required for the calendar year of the merger (even though the merger creates an institution that meets all of the requirements necessary to be a covered institution). When a branch office of an institution that is not covered is acquired by another institution that is not covered, and the acquisition results in a covered institution, no data collection is required for the calendar year of the acquisition.

ii. A covered institution and an institution that is not covered merge. The covered institution is the surviving institution, or a new covered institution is formed. For the calendar year of the merger, data collection is required for covered loans and applications handled in the offices of the merged institution that was previously covered and is optional for covered loans and applications handled in offices of the merged institution that was previously not covered. When a covered institution acquires a branch office of an institution that is not covered, data collection is optional for covered loans and applications handled by the acquired branch office for the calendar year of the acquisition.

iii. A covered institution and an institution that is not covered merge. The institution that is not covered is the surviving institution, or a new institution that is not covered is formed. For the calendar year of the merger, data collection is required for covered loans and applications handled in offices of the previously covered institution that took place prior to the merger. After the merger date, data collection is optional for covered loans and applications handled in the offices of the institution that was previously covered. When an institution remains not covered after acquiring a branch office of a covered institution, data collection is required for transactions of the acquired branch office that take place prior to the acquisition. Data collection by the acquired branch office is optional for transactions taking place in the remainder of the calendar year after the acquisition.

iv. Two covered institutions merge. The surviving or newly formed institution is a covered institution. Data collection is required for the entire calendar year of the merger. The surviving or newly formed institution files either a consolidated submission or separate submissions for that calendar year. When a covered institution acquires a branch office of a covered institution, data collection is required for the entire calendar year of the merger. Data for the acquired branch office may be submitted by either institution.

5. Originations. Whether an institution is a financial institution depends in part on whether the institution originated at least 100 closed-end mortgage loans in each of the two preceding calendar years or at least 200 open-end lines of credit in each of the two preceding calendar years. Comments 4(a)-2 through -4 discuss whether activities with respect to a particular closed-end mortgage loan or open-end line of credit constitute an origination for purposes of § 1003.2(g).

6. Branches of foreign banks—treated as banks. A Federal branch or a State-licensed or insured branch of a foreign bank that meets the definition of a “bank” under section 3(a)(1) of the Federal Deposit Insurance Act (12 U.S.C. 1813(a)) is a bank for the purposes of § 1003.2(g).

7. Branches and offices of foreign banks and other entities—treated as nondepository financial institutions. A Federal agency, State-licensed agency, State-licensed uninsured branch of a foreign bank, commercial lending company owned or controlled by a foreign bank, or entity operating under section 25 or 25A of the Federal Reserve Act, 12 U.S.C. 601 and 611 (Edge Act and agreement corporations) may not meet the definition of “bank” under the Federal Deposit Insurance Act and may thereby fail to satisfy the definition of a depository financial institution under § 1003.2(g)(1). An entity is nonetheless a financial institution if it meets the definition of nondepository financial institution under § 1003.2(g)(2).

* * * * *

Section 1003.3—Exempt Institutions and Excluded and Partially Exempt Transactions

* * * * *

3(c) Excluded Transactions

* * * * *

Paragraph 3(c)(11)

1. General. Section 1003.3(c)(11) provides that a closed-end mortgage loan is an excluded transaction if a financial institution originated fewer than 100 closed-end mortgage loans in either of the two preceding calendar years. For example, assume that a bank is a financial institution in 2022 under § 1003.2(g) because it originated 300 open-end lines of credit in 2020, 350 open-end lines of credit in 2021, and met all of the other requirements under § 1003.2(g)(1). Also assume that the bank originated 75 and 90 closed-end mortgage loans in 2020 and 2021, respectively. The open-end lines of credit that the bank originated or purchased, or for which it received applications, during 2022 are covered loans and must be reported, unless they otherwise are excluded transactions under § 1003.3(c). However, the closed-end mortgage loans that the bank originated or purchased, or for which it received applications, during 2022 are excluded transactions under § 1003.3(c)(11) and need not be reported. See comments 4(a)-2 through-4 for guidance about the activities that constitute an origination.

2. Optional reporting. A financial institution may report applications for, originations of, or purchases of closed-end mortgage loans that are excluded transactions because the financial institution originated fewer than 100 closed-end mortgage loans in either of the two preceding calendar years. However, a financial institution that chooses to report such excluded applications for, originations of, or purchases of closed-end mortgage loans must report all such applications for closed-end mortgage loans that it receives, closed-end mortgage loans that it originates, and closed-end mortgage loans that it purchases that otherwise would be covered loans for a given calendar year. Note that applications which remain pending at the end of a calendar year are not reported, as described in comment 4(a)(8)(i)-14.

Paragraph 3(c)(12)

1. General. Section 1003.3(c)(12) provides that an open-end line of credit is an excluded transaction if a financial institution originated fewer than 200 open-end lines of credit in either of the two preceding calendar years. For example, assume that a bank is a financial institution in 2022 under § 1003.2(g) because it originated 100 closed-end mortgage loans in 2020, 175 closed-end mortgage loans in 2021, and met all of the other requirements under § 1003.2(g)(1). Also assume that the bank originated 175 and 185 open-end lines of credit in 2020 and 2021, respectively. The closed-end mortgage loans that the bank originated or purchased, or for which it received applications, during 2022 are covered loans and must be reported, unless they otherwise are excluded transactions under § 1003.3(c). However, the open-end lines of credit that the bank originated or purchased, or for which it received applications, during 2022 are excluded transactions under § 1003.3(c)(12) and need not be reported. See comments 4(a)-2 through -4 for guidance about the activities that constitute an origination.

2. Optional reporting. A financial institution may report applications for, originations of, or purchases of open-end lines of credit that are excluded transactions because the financial institution originated fewer than 200 open-end lines of credit in either of the two preceding calendar years. However, a financial institution that chooses to report such excluded applications for, originations of, or purchases of open-end lines of credit must report all such applications for open-end lines of credit which it receives, open-end lines of credit that it originates, and open-end lines of credit that it purchases that otherwise would be covered loans for a given calendar year. Note that applications which remain pending at the end of a calendar year are not reported, as described in comment 4(a)(8)(i)-14.

* * * * *

Laura Galban,

Federal Register Liaison, Bureau of Consumer Financial Protection.

Footnotes

1.  Home Mortgage Disclosure (Regulation C), 80 FR 66128 (Oct. 28, 2015). HMDA requires financial institutions to collect, record, and report data. To simplify review of this document, the Bureau generally refers herein to the obligation to report data instead of listing all of these obligations in each instance.

2.  When amending the Bureau's commentary, the Office of the Federal Register requires reprinting of certain subsections being amended in their entirety rather than providing more targeted amendatory instructions and commentary. The subsections of regulatory text and commentary included in this document show the complete language of those subsections. In addition, the Bureau is releasing an unofficial, informal redline to assist industry and other stakeholders in reviewing the changes that it is finalizing to the regulatory text and commentary of Regulation C. This redline can be found on the Bureau's regulatory implementation page for the HMDA Rule at https://www.consumerfinance.gov/​policy-compliance/​guidance/​hmda-implementation/​. If any conflicts exist between the redline and this final rule, this final rule is the controlling document.

3.  12 CFR 1003.1.

4.  As used in this final rule, the term “data point” refers to items of information that entities are required to compile and report, generally listed in separate paragraphs in Regulation C. Some data points are reported using multiple data fields.

5.  Public Law 111-203, 124 Stat. 1376, 1980, 2035-38, 2097-101 (2010).

6.  Dodd-Frank Act section 1094(3), amending HMDA section 304(b), 12 U.S.C. 2803(b).

7.  Id.

8.  Id.

9.  80 FR 66128 (Oct. 28, 2015).

10.  Id. at 66128, 66256-58.

11.  Id. at 66148-50, 66309 (codified at 12 CFR 1003.2(g)(1)(v)). The 2015 HMDA Rule excludes certain transactions from the definition of covered loans, and those excluded transactions do not count towards the threshold. Id.

12.  Id. at 66173, 66310, 66322 (codified at 12 CFR 1003.3(c)(11) and (12)).

13.  Technical Corrections and Clarifying Amendments to the Home Mortgage Disclosure (Regulation C) October 2015 Final Rule, 82 FR 19142 (Apr. 25, 2017).

14.  Home Mortgage Disclosure (Regulation C) Temporary Increase in Institutional and Transactional Coverage Thresholds for Open-End Lines of Credit, 82 FR 33455 (July 20, 2017).

15.  Home Mortgage Disclosure (Regulation C), 82 FR 43088 (Sept. 13, 2017).

16.  Id. at 43095. The 2017 HMDA Rule also, among other things, replaced “each” with “either” in § 1003.3(c)(11) and (12) to correct a drafting error and to ensure that the exclusion provided in that section mirrors the loan-volume threshold for financial institutions in § 1003.2(g). Id. at 43100, 43102. Recognizing the significant systems and operations challenges needed to adjust to the revised regulation, the Bureau also issued a statement in December 2017 indicating that, for HMDA data collected in 2018 and reported in 2019, the Bureau did not intend to require data resubmission unless data errors were material. Among other things, the Bureau also indicated that it intended to engage in a rulemaking to reconsider various aspects of the 2015 HMDA Rule, such as the institutional and transactional coverage tests and the rule's discretionary data points. Bureau of Consumer Fin. Prot., “Statement with Respect to HMDA Implementation” (Dec. 21, 2017), available at https://files.consumerfinance.gov/​f/​documents/​cfpb_​statement-with-respect-to-hmda-implementation_​122017.pdf. The Board, the Federal Deposit Insurance Corporation, the National Credit Union Administration, and the Office of the Comptroller of the Currency released similar statements relating to their supervisory examinations.

17.  Public Law 115-174, 132 Stat. 1296 (2018).

18.  For purposes of HMDA section 104, the EGRRCPA provides that the term “insured credit union” has the meaning given the term in section 101 of the Federal Credit Union Act, 12 U.S.C. 1752, and the term “insured depository institution” has the meaning given the term in section 3 of the Federal Deposit Insurance Act, 12 U.S.C. 1813.

19.  12 U.S.C. 2906(b)(2).

20.  Partial Exemptions from the Requirements of the Home Mortgage Disclosure Act Under the Economic Growth, Regulatory Relief, and Consumer Protection Act (Regulation C), 83 FR 45325 (Sept. 7, 2018).

21.  Home Mortgage Disclosure (Regulation C), 84 FR 20972 (May 13, 2019). The Bureau also issued concurrently with the May 2019 Proposal an Advance Notice of Proposed Rulemaking (ANPR) to solicit comment, data, and information from the public about the data points that the 2015 HMDA Rule added to Regulation C or revised to require additional information and Regulation C's coverage of certain business- or commercial-purpose transactions. Home Mortgage Disclosure (Regulation C) Data Points and Coverage, 84 FR 20049 (May 8, 2019). The Bureau anticipates that it will issue a notice of proposed rulemaking later this year to follow up on the ANPR.

22.  A separate comment period related to the Paperwork Reduction Act closed on July 12, 2019. 84 FR 20972 (May 13, 2019).

23.  This document uses “2018 HMDA Data” to refer to the publicly available national loan level dataset for 2018 and the Bureau's annual overview of residential mortgage lending, and “2018 HMDA data” to refer to the HMDA data submitted for collection year 2018.

24.  See Home Mortgage Disclosure (Regulation C); Reopening of Comment Period, 84 FR 37804 (Aug. 2, 2019).

25.  Id. at 37806.

26.  Id.

27.  Home Mortgage Disclosure (Regulation C), 84 FR 57946 (Oct. 29, 2019).

28.  See infra part VII.D.1. All coverage numbers provided in this document are estimates based on available data, as explained in part VII below. Due to rounding there may be some minor discrepancies within the estimates provided. As discussed further in part VII below, the Bureau's analyses in the May 2019 Proposal were based on HMDA data collected in 2016 and 2017 and other sources. In part VII of this final rule, the Bureau has supplemented the analyses from the May 2019 Proposal with the 2018 HMDA data. See infra part VII.E.2 & VII.E.3. In the May 2019 Proposal, the Bureau estimated that there were about 4,960 financial institutions required to report their closed-end mortgage loans and applications under HMDA, with about 4,263 of those reporters being depository institutions and about 697 being nondepository institutions. The Bureau estimated that together, these financial institutions originated about 7 million closed-end mortgage loans in calendar year 2017. The Bureau also estimated that among the estimated 4,960 closed-end reporters, about 3,300 insured depository institutions and insured credit unions were eligible for a partial exemption under the EGRRCPA. The estimates in this final rule differ slightly from those in the May 2019 Proposal due to the supplementation of the analyses with 2018 HMDA data.

29.  Home Mortgage Disclosure (Regulation C), 84 FR 20972 (May 13, 2019).

30.  Home Mortgage Disclosure (Regulation C); Reopening of Comment Period, 84 FR 37804 (Aug. 2, 2019).

31.  A large number of consumer groups, civil rights groups, and other organizations stated in a joint comment letter in response to the July 2019 Reopening Notice that, because the 2018 HMDA Data were released in late August 2019, the reopened comment period did not provide sufficient time for the public to analyze the proposed changes prior to October 15, 2019. These commenters stated further that the public was not provided a meaningful opportunity to comment on the May 2019 Proposal and that the Bureau would not have the benefit of fully informed comments that took into consideration the 2018 HMDA Data. The Bureau determines that additional time to comment on the aspects of the May 2019 Proposal addressed in this final rule is not necessary. The Bureau believes the more than 45-day period between the release of the 2018 HMDA Data and the close of the reopened comment period on October 15, 2019, provided interested persons sufficient time to meaningfully review the proposed changes relating to the permanent open-end and closed-end thresholds and provide comment informed by the 2018 HMDA Data. Regarding commenters' separate concerns over the Bureau's dissemination of the data, the Bureau made available with the 2018 HMDA Data a HMDA data browser that facilitates analysis in spreadsheet software, such as Excel, and allows users to filter the national loan-level data to create summary tables and custom datasets. The HMDA data browser allows users to, for example, create summary tables that can be used to show which institutions are active in a particular Metropolitan Statistical Area (MSA), state, or county. Users can develop some understanding of the effect of various loan-volume thresholds, for individual institutions, by analyzing the publicly available modified loan/application register data or, for all institutions, by analyzing the publicly available national loan-level dataset.

32.  12 U.S.C. 5581. Section 1094 of the Dodd-Frank Act also replaced the term “Board” with “Bureau” in most places in HMDA. 12 U.S.C. 2803 et seq.

33.  12 U.S.C. 5581(a)(1)(A).

34.  12 U.S.C. 5512(b)(1).

35.  Dodd-Frank Act section 1002(14), 12 U.S.C. 5481(14) (defining “Federal consumer financial law” to include the “enumerated consumer laws” and the provisions of title X of the Dodd-Frank Act); Dodd-Frank Act section 1002(12), 12 U.S.C. 5481(12) (defining “enumerated consumer laws” to include HMDA).

36.  12 U.S.C. 2804(a).

37.  Id.

38.  12 CFR 1003.2(g)(1) (definition of depository financial institution); 1003.2(g)(2) (definition of nondepository financial institution).

39.  82 FR 43088, 43095 (Sept. 13, 2017).

40.  84 FR 57946, 57949 (Oct. 29, 2019).

41.  For discussion of the effective dates, see part VI.

42.  80 FR 66128, 66150 (Oct. 28, 2015).

43.  Id. at 66153.

44.  A State attorney general suggested in its comments that increasing the thresholds exceeds the Bureau's legal authority, but as discussed above, the Bureau is adopting the increased thresholds based on its authority under section 305(a) of HMDA.

45.  80 FR 66128, 66129 (Oct. 28, 2015). Prior to the 2015 HMDA Rule, a bank, savings association, or credit union was covered under Regulation C if: (1) On the preceding December 31, it satisfied an asset-size threshold; (2) on the preceding December 31, it had a home or branch office in an MSA; (3) during the previous calendar year, it originated at least one home purchase loan or refinancing of a home purchase loan secured by a first lien on a one- to four-unit dwelling; and (4) the institution is federally insured or regulated, or the mortgage loan referred to in item (3) was insured, guaranteed, or supplemented by a Federal agency or intended for sale to the Federal National Mortgage Association or the Federal Home Loan Mortgage Corporation. 12 CFR 1003.2 (2016).

46.  In addition to finalizing changes that the Bureau proposed to comment 2(g)-1 and changes related to optional reporting that are discussed below, the final rule makes minor changes to comment 2(g)-1 to update the years and loan-volumes in an example that illustrates how the closed-end mortgage loan threshold works.

47.  80 FR 66128, 66147 (Oct. 28, 2015).

48.  Id. at 66148, 66277.

49.  Id. at 66147.

50.  Id.

51.  Id. at 66148.

52.  The Bureau temporarily raised the threshold for open-end lines of credit in the 2017 HMDA Rule because of concerns based on new information that the estimates the Bureau used in the 2015 HMDA Rule may have understated the burden that open-end reporting would impose on smaller institutions if they were required to begin reporting on January 1, 2018. However, the Bureau declined to raise the threshold for closed-end mortgage loans at that time and stated that, in developing the 2015 HMDA Rule, it had robust data to make a determination about the number of transactions that would be reported at the threshold of 25 closed-end mortgage loans as well as the one-time and ongoing costs to industry. 82 FR 43088, 43095-96 (Sept. 13, 2017).

53.  In the May 2019 Proposal, the Bureau stated that it received this recommendation in response to the Bureau's 2018 Request for Information Regarding the Bureau's Adopted Regulations and New Rulemaking Authorities (RFI) although the 2015 HMDA Rule was outside the scope of the RFI. See 84 FR 20972, 20976 (May 13, 2019).

54.  Public Law 115-174, 132 Stat. 1296 (2018).

55.  See 84 FR 57946 (Oct. 29, 2019).

56.  The Bureau stated in the May 2019 Proposal that it intended to review the 2018 HMDA data more closely in connection with this rulemaking once the 2018 submissions were more complete. The Bureau released the 2018 HMDA Data including the two data point articles on August 30, 2019, and reopened the comment period until October 15, 2019, to give commenters an opportunity to comment on the 2018 HMDA Data. The estimates reflected in this final rule are based on the HMDA data collected in 2017 and 2018 as well as other sources.

57.  The Bureau recognizes that the coverage estimates generated using this restriction may omit certain financial institutions that should have reported but did not report in the most recent HMDA reporting year. However, the Bureau applied this restriction to ensure that institutions included in its coverage estimates are in fact financial institutions for purposes of Regulation C because it recognizes that institutions might not meet the Regulation C definition of financial institution for reasons that are not evident in the data sources that it utilized.

58.  The estimates for coverage and reportable originations described in this section cover only depository institutions. Estimates for coverage of nondepository institutions and reportable originations of nondepository institutions are described in the section-by-section analysis of § 1003.2(g)(2)(ii)(A) below. For estimates that are comprehensive of depository and nondepository institutions, see part VII.E.2 below.

59.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 50, about 3,518 out of about 4,263 depository institutions covered under the current threshold of 25 (or approximately 83 percent) would continue to report HMDA data on closed-end mortgage loans, and approximately 99 percent or approximately 3.54 million total originations of closed-end mortgage loans in current market conditions reported by depository institutions under the current Regulation C coverage criteria would continue to be reported. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

60.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 100, about 2,581 out of about 4,263 depository institutions covered under the current threshold of 25 (or approximately 61 percent) would continue to report HMDA data on closed-end mortgage loans, and approximately 90 percent or approximately 3.43 million total originations of closed-end mortgage loans in current market conditions reported by depository institutions under the current Regulation C coverage criteria would continue to be reported. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

61.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 250, about 1,413 out of about 4,263 depository institutions covered under the current threshold of 25 (or approximately 33 percent) would continue to report HMDA data on closed-end mortgage loans, and approximately 90 percent or approximately 3.21 million total originations of closed-end mortgage loans in current market conditions reported by depository institutions under the current Regulation C coverage criteria would continue to be reported. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

62.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 500, about 798 out of about 4,263 depository institutions covered under the current threshold of 25 (or approximately 19 percent) would continue to report HMDA data on closed-end mortgage loans, and approximately 83 percent or approximately 2.97 million total originations of closed-end mortgage loans in current market conditions reported by depository institutions under the current Regulation C coverage criteria would continue to be reported. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

63.  As discussed in more detail in part VII.E.2 below, about 1,620 of the estimated 2,480 financial institutions that the Bureau estimates will report closed-end loans at the threshold of 100 are eligible for a partial exemption under the EGRRCPA. The Bureau estimates that these partially exempt institutions report only 369,000 out of the estimated 2.7 million closed-end mortgage loans that will be reported at the threshold of 100.

64.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 50, there would be a loss of at least 20 percent of reportable HMDA data in just under 300 out of approximately 74,000 total census tracts, or less than one-half of 1 percent of the total number of census tracts, relative to the current threshold. For low-to-moderate income census tracts, the Bureau estimated that there would be a loss of at least 20 percent of reportable HMDA data in less than 1 percent of such tracts relative to the current threshold, and for rural census tracts, the Bureau estimated there would be at least a 20 percent loss of reportable HMDA data in less than one-half of 1 percent of such tracts relative to the current threshold. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

65.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 100, there would be a loss of at least 20 percent of reportable HMDA data in about 1,100 out of approximately 74,000 total census tracts, or 1.5 percent of the total number of census tracts, relative to the current threshold. For low-to-moderate income census tracts, the Bureau estimated that there would be a loss of at least 20 percent of reportable HMDA data in 3 percent of such tracts relative to the current threshold, and for rural census tracts, the Bureau estimated there would be at least a 20 percent loss of reportable HMDA data in less than 3 percent of such tracts relative to the current threshold. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

66.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 250, there would be a loss of at least 20 percent of reportable HMDA data in over 4,000 out of approximately 74,000 total census tracts, or 5.4 percent of the total number of census tracts, relative to the current threshold. For low-to-moderate income census tracts, the Bureau estimated that there would be a loss of at least 20 percent of reportable HMDA data in just over 8 percent of such tracts relative to the current threshold, and for rural census tracts, the Bureau estimated there would be at least a 20 percent loss of reportable HMDA data in about 14 percent of such tracts relative to the current threshold. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

67.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 500, there would be a loss of at least 20 percent of reportable HMDA data in approximately 11,000 out of approximately 74,000 total census tracts, or 14.9 percent of the total number of census tracts, relative to the current threshold. For low-to-moderate income census tracts, the Bureau estimated that there would be a loss of at least 20 percent of reportable HMDA data in 17 percent of such tracts relative to the current threshold, and for rural census tracts, the Bureau estimated there would be at least a 20 percent loss of reportable HMDA data in 32 percent of such tracts relative to the current threshold. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

68.  These cost estimates reflect the combined ongoing reduction in costs for depository and nondepository institutions. These estimates also take into account the enactment of the EGRRCPA, which created partial exemptions from HMDA's requirements that certain insured depository institutions and insured credit unions may use, and reflect updates made to the cost estimates since the May 2019 Proposal. See part VII.E.2 below for a more comprehensive discussion of the cost estimates.

69.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 50, the aggregate savings on the operational costs associated with reporting closed-end mortgage loans would be approximately $2.2 million per year. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

70.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 100, the aggregate savings on the operational costs associated with reporting closed-end mortgage loans would be approximately $8.1 million per year. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

71.  Thus, as comment 2(g)-1 explains, in 2021, a financial institution does not meet the loan-volume test described in § 1003.2(g)(1)(v)(A) if it originated fewer than 100 closed-end mortgage loans during either 2019 or 2020. See part VI.A below for a discussion of the HMDA obligations for the 2020 data collection year of institutions affected by the closed-end threshold change, and see the section-by-section analysis of § 1003.3(c)(11) in this part for a discussion of optional reporting of 2020 closed-end data permitted for such institutions.

72.  Section 1003.3(c)(12) includes a complementary transactional coverage threshold set at the same level that determines whether a financial institution is required to collect and report data on open-end lines of credit.

73.  The Bureau also proposed conforming changes to the institutional coverage threshold for nondepository institutions in § 1003.2(g)(2)(ii)(B) and to the transactional coverage threshold in § 1003.3(c)(12), as discussed below.

74.  The Bureau also finalized conforming amendments to extend for two years the temporary open-end institutional coverage threshold for nondepository institutions in § 1003.2(g)(2)(ii)(B) and to align the timeframe of the temporary open-end transactional coverage threshold in § 1003.3(c)(12). Because the extension of the temporary threshold lasts two years, and the Bureau had not yet made a determination about its proposed permanent threshold when it issued the 2019 HMDA Rule, that rule would have restored effective January 1, 2022 the threshold set in the 2015 HMDA Rule of 100 open-end lines of credit in §§ 1003.2(g) and 1003.3(c)(12) absent this final rule.

75.  HMDA section 303(2), 12 U.S.C. 2802(2).

76.  65 FR 78656, 78659-60 (Dec. 15, 2000). In 1988, the Board had amended Regulation C to permit, but not require, financial institutions to report certain home-equity lines of credit. 53 FR 31683, 31685 (Aug. 19, 1988).

77.  67 FR 7222, 7225 (Feb. 15, 2002).

78.  80 FR 66128, 66160 (Oct. 28, 2015).

79.  Id. The Bureau stated in the 2015 HMDA Rule that research indicated that some real estate investors used open-end, home-secured lines of credit to purchase non-owner-occupied properties, which correlated with higher first-mortgage defaults and home-price depreciation during the financial crisis. Id. In the years leading up to the crisis, such home-equity lines of credit often were made and fully drawn more or less simultaneously with first-lien home purchase loans, essentially creating high loan-to-value home purchase transactions that were not visible in the HMDA dataset. Id.

80.  The Bureau also required reporting of applications for, and originations of, dwelling-secured commercial-purpose lines of credit for home purchase, home improvement, or refinancing purposes. Id. at 66171.

81.  Id. at 66157-62. HMDA and Regulation C are designed to provide citizens and public officials sufficient information about mortgage lending to ensure that financial institutions are serving the housing needs of their communities, to assist public officials in distributing public-sector investment so as to attract private investment to areas where it is needed, and to assist in identifying possible discriminatory lending patterns and enforcing antidiscrimination statutes. The Bureau believes that collecting information about all dwelling-secured, consumer-purpose open-end lines of credit serves these purposes.

82.  Id. at 66128, 66161.

83.  Id. at 66149.

84.  Id.

85.  Id. at 66261, 66269-70. In the 2015 HMDA Rule and the 2017 HMDA Rule, the Bureau assigned financial institutions to tiers by adopting cutoffs based on the estimated open-end line of credit volume. Id. at 66285; 82 FR 43088, 43128 (Sept. 13, 2017). Specifically, the Bureau assumed the lenders that originated fewer than 200 but more than 100 open-end lines of credit were tier 3 (low-complexity) open-end reporters; lenders that originate between 200 and 7,000 open-lines of credit were tier 2 (moderate-complexity) open-end reporters; and lenders that originated more than 7,000 open-end lines of credit were tier 1 (high-complexity) open-end reporters. 80 FR 66128, 66285 (Oct. 28, 2015); 82 FR 43088, 43128 (Sept. 13, 2017). As explained below in part VII.D.1, for purposes of this final rule, the Bureau has used a more precise methodology to assign excluded financial institutions to tiers 2 and 3 for their open-end reporting, which relies on constraints relating to the estimated numbers of impacted institutions and loan/application register records for the applicable provision.

86.  80 FR 66128, 66264-65 (Oct. 28, 2015); see also id. at 66284.

87.  Id. at 66264; see also id. at 66284-85.

88.  Id. at 66265; see also id. at 66284.

89.  Id. at 66285.

90.  Id.

91.  Id. at 66264, 66286.

92.  Id.

93.  Id. at 66162.

94.  Id. The estimate of the number of institutions that would be excluded from reporting open-end lines of credit by the transactional coverage threshold was relative to the number that would have been covered under the Bureau's proposal that led to the 2015 HMDA Rule. Under that proposal, a financial institution would have been required to report its open-end lines of credit if it had originated at least 25 closed-end mortgage loans in each of the two preceding years without regard to how many open-end lines of credit the institution originated. See Home Mortgage Disclosure (Regulation C), 79 FR 51732 (Aug. 29, 2014).

95.  80 FR 66128, 66281 (Oct. 28, 2015).

96.  Id. at 66162.

97.  82 FR 43088, 43094 (Sept. 13, 2017).

98.  Id.

99.  Id.

100.  See supra notes 85-93 and accompanying text.

101.  82 FR 43088, 43094 (Sept. 13, 2017).

102.  Id.

103.  Id.

104.  Id. at 43088. Comments received on the July 2017 HMDA Proposal to change temporarily the open-end threshold are discussed in the 2017 HMDA Rule. Id. at 43094-95. In the 2015 HMDA Rule and the 2017 HMDA Rule, the Bureau declined to retain optional reporting of open-end lines of credit, after concluding that improved visibility into this segment of the mortgage market is critical because of the risks posed by these products to consumers and local markets and the lack of other publicly available data about these products. Id. at 43095; 80 FR 66128, 66160-61 (Oct. 28, 2015). However, Regulation C as amended by the 2017 HMDA Rule permits voluntary reporting by financial institutions that do not meet the open-end threshold. 12 CFR 1003.3(c)(12).

105.  The Bureau proposed conforming amendments to § 1003.3(c)(12).

106.  84 FR 57946 (Oct. 29, 2019).

107.  Id. at 57953.

108.  82 FR 43088, 43094 (Sept. 13, 2017).

109.  Public Law 115-174, 132 Stat. 1296 (2018).

110.  See 84 FR 57946 (Oct. 29, 2019).

111.  See infra part VII.E.3.

112.  One commenter expressed concern as to how the increase in the open-end threshold would serve HMDA's purposes. As discussed above, this increase in the permanent open-end threshold effectuates the purposes of HMDA and facilitates compliance with HMDA by reducing burden, while still providing significant market coverage.

113.  As explained in part VII below, the Bureau derived these estimates using estimates of savings for open-end lines of credit for representative financial institutions.

114.  As noted by several members of Congress, the Bureau consulted multiple sources to develop its open-end estimates for the May 2019 Proposal. Because collection of data on open-end lines of credit only became mandatory starting in 2018 under the 2015 HMDA Rule and the 2017 HMDA Rule, no single data source existed as of the time of the May 2019 Proposal that could accurately capture the number of originations of open-end lines of credit in the entire market and by lenders. In part VII of this final rule, the Bureau has supplemented the analyses from the May 2019 Proposal with the 2018 HMDA data. For information about the HMDA data used in developing and supplementing the Bureau estimates, see infra part VII.E.3.

115.  See infra part VII.E.3 at table 4 for estimates of coverage among all lenders that are active in the open-end line of credit market at open-end coverage thresholds of 100, 200, and 500.

116.  Id.

117.  See 84 FR 57946, 57953 (Oct. 29, 2019); 82 FR 43088, 43095-96 (Sept. 13, 2017).

118.  One commenter suggested that data on open-end lines of credit are less valuable for fair lending analyses because these products are limited to borrowers with existing equity in their homes. In the 2015 HMDA Rule, the Bureau recognized that borrowers may not be evaluated for open-end credit in the same manner as for traditional mortgage loans, with adequate home equity being a factor. It stated further, however, that lending practices during the financial crisis demonstrated that during prolonged periods of home-price appreciation lenders became increasingly comfortable originating home-equity products to borrowers with less and less equity to spare. 80 FR 66128, 66161 (Oct. 28, 2015). The Bureau continues to believe that the more leveraged the borrower, the more at risk the borrower is of losing his or her home. Id.

119.  See supra note 104. In establishing partial exemptions for reporting data on open-end lines of credit in the EGRRCPA, Congress appears to have assumed that open-end lines of credit should be reported, building upon the Bureau's decision in the 2015 HMDA Rule to require reporting of open-end lines of credit.

120.  For an explanation of the Bureau's assumptions in assigning institutions to tiers 1, 2, and 3, see supra note 85 and infra part VII.D.1.

121.  As discussed in more detail in part VII below, the Bureau has supplemented its analyses from the May 2019 Proposal with 2018 HMDA data. These data allow the Bureau to develop estimates based on the total number of open-end loan/application register records, rather than the number of open-end originations. As a result, the Bureau has assigned more of the estimated 401 institutions affected by the increase in the threshold from 100 to 200 open-end lines of credit to the tier 2 category and fewer to the tier 3 category as compared to the May 2019 Proposal. This increase in the estimated number of affected tier 2 institutions results in a higher estimated aggregate one-time cost savings associated with the threshold increase than the Bureau's estimate of $3.8 million in the May 2019 Proposal. For information about the HMDA data used in developing and supplementing the Bureau estimates, see infra part VII.E.3.

122.  As noted above, as compared to the May 2019 Proposal the Bureau now assigns more of the estimated 401 institutions affected by the increase in the threshold from 100 to 200 to the tier 2 category and fewer to the tier 3 category. As a result, the Bureau's estimated aggregate savings on ongoing operational costs associated with the threshold increase is higher than the Bureau's estimate of $2.1 million in the May 2019 Proposal. For information about the HMDA data used in developing and supplementing the Bureau estimates, see infra part VII.E.3.

123.  HMDA section 303(5) (defining “other lending institutions”).

124.  HMDA section 309(a), 12 U.S.C. 2808(a).

125.  80 FR 66128, 66153 (Oct. 28, 2015) (citing 54 FR 51356, 51358-59 (Dec. 15, 1989)).

126.  Prior to the 2015 HMDA Rule, for-profit nondepository institutions that met the location test only had to report if: (1) In the preceding calendar year, the institution originated home purchase loans, including refinancings of home purchase loans, that equaled either at least 10 percent of its loan-origination volume, measured in dollars, or at least $25 million; and (2) On the preceding December 31, the institution had total assets of more than $10 million, counting the assets of any parent corporation; or in the preceding calendar year, the institution originated at least 100 home purchase loans, including refinancings of home purchase loans. 12 CFR 1003.2 (2017).

127.  The Bureau temporarily raised the threshold for open-end lines of credit in the 2017 HMDA Rule because of concerns based on new information that the estimates the Bureau used in the 2015 HMDA Rule may have understated the burden that open-end reporting would impose on smaller institutions if they were required to begin reporting on January 1, 2018. However, the Bureau declined to raise the threshold for closed-end mortgage loans and stated that in developing the 2015 HMDA Rule, it had robust data to make a determination about the number of transactions that would be reported with a threshold of 25 closed-end mortgage loans as well as the one-time and ongoing costs to industry. 82 FR 43088, 43095-96 (Sept. 13, 2017).

128.  For a discussion on the proposed closed-end coverage threshold for depository institutions, see the section-by-section analysis of § 1003.2(g)(1)(v)(A) above.

129.  The 2015 HMDA Rule simplified the reporting regime by establishing a uniform loan-volume threshold applicable to depository and nondepository institutions.

130.  The Bureau stated in the May 2019 Proposal that it intended to review the 2018 HMDA data more closely in connection with this rulemaking once the 2018 submissions were more complete. The estimates reflected in this final rule are based on the HMDA data collected in 2017 and 2018 as well as other sources. These estimates are discussed further in the analysis under Dodd-Frank Act section 1022(b) in part VII below.

131.  Except for the estimates provided at the census tract level, the estimates provided for potential thresholds in this section cover only nondepository institutions. Estimates for depository institutions are described in the section-by-section analysis of § 1003.2(g)(1)(v)(A). For estimates that are comprehensive of depository and nondepository institutions, see part VII.E.2 below.

132.  The Bureau recognizes that the coverage estimates generated using this restriction may omit certain financial institutions that should have reported but did not report in the most recent HMDA reporting year. However, the Bureau applied this restriction to ensure that institutions included in its coverage estimates are in fact financial institutions for purposes of Regulation C because it recognizes that institutions might not meet the Regulation C definition of financial institution for reasons that are not evident in the data sources that it utilized.

133.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 50, about 683 out of about 697 nondepository institutions covered under the current threshold of 25 (or approximately 98 percent) would continue to report HMDA data on closed-end mortgage loans, and over 99 percent or approximately 3.44 million total originations of closed-end mortgage loans in current market conditions reported by depository institutions under the current Regulation C coverage criteria would continue to be reported. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

134.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 100, about 661 out of about 697 nondepository institutions covered under the current threshold of 25 (or approximately 95 percent) would continue to report HMDA data on closed-end mortgage loans, and over 99 percent or approximately 3.44 million total originations of closed-end mortgage loans in current market conditions reported by depository institutions under the current Regulation C coverage criteria would continue to be reported. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

135.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 250, about 573 out of about 697 nondepository institutions covered under the current threshold of 25 (or approximately 82 percent) would continue to report HMDA data on closed-end mortgage loans, and about 99 percent or approximately 3.42 million total originations of closed-end mortgage loans in current market conditions reported by depository institutions under the current Regulation C coverage criteria would continue to be reported. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

136.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 500, about 477 out of about 697 nondepository institutions covered under the current threshold of 25 (or approximately 68 percent) would continue to report HMDA data on closed-end mortgage loans, and about 98 percent or approximately 3.38 million total originations of closed-end mortgage loans in current market conditions reported by depository institutions under the current Regulation C coverage criteria would continue to be reported. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

137.  These cost estimates reflect the combined ongoing reduction in costs for depository and nondepository institutions. These estimates also take into account the enactment of the EGRRCPA, which created partial exemptions from HMDA's requirements that certain insured depository institutions and insured credit unions may use, and reflect updates made to the cost estimates since the May 2019 Proposal. See part VII.E.2 below for a more comprehensive discussion of the cost estimates.

138.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 50, the aggregate savings on the operational costs associated with reporting closed-end mortgage loans would be approximately $2.2 million per year. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

139.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased from 25 to 100, the aggregate savings on the operational costs associated with reporting closed-end mortgage loans would be approximately $8.1 million per year. As explained above and in greater detail in part VII.E.2 below, the differences in the estimates between the May 2019 Proposal and this final rule are mostly due to updates made to incorporate the newly available 2018 HMDA data.

140.  80 FR 66128, 66281 (Oct. 28, 2015).

141.  These cost estimates reflect the combined ongoing reduction in costs for depository and nondepository institutions. These estimates also take into account the enactment of the EGRRCPA, which created partial exemptions from HMDA's requirements that certain insured depository institutions and insured credit unions may now use. See part VII.E.2 below for a more comprehensive analysis on cost estimates.

142.  The 2015 HMDA Rule established the uniform loan-volume threshold applicable to depository and nondepository institutions to simplify the reporting regime.

143.  See section VI for a discussion of the effective dates.

144.  82 FR 43088, 43095 (Sept. 13, 2017).

145.  See infra part VII.E.3 at table 4.

146.  In light of the new closed-end and open-end thresholds adopted in this final rule, the final rule makes minor changes in comment 3(c)(11)-1 to adjust the years and loan volumes in examples that illustrate how the thresholds work.

147.  Section 1003.3(c)(11) (both currently and as revised) and comment 3(c)(11)-2 permit a financial institution whose closed-end mortgage loans are excluded by § 1003.3(c)(11) to report voluntarily its closed-end data, as long as the financial institution reports such closed-end data for the full calendar year.

148.  82 FR 43088, 43102 (Sept. 13, 2017).

149.  Id. at 43095.

150.  In addition to finalizing the proposed changes to comment 3(c)(12)-1, the final rule makes minor changes in comment 3(c)(12)-1 to adjust the years and loan volumes in an example that illustrates how the open-end line of credit threshold works.

151.  As explained below, institutions that are subject to HMDA's closed-end requirements prior to the effective date but that do not meet the new closed-end threshold for calendar year 2020 as of July 1, 2020 are relieved of the obligation to collect, record, and report data for their 2020 closed-end mortgage loans effective July 1, 2020.

152.  As explained below, the amendments to the open-end threshold apply to covered loans and applications with respect to which final action is taken beginning on January 1, 2022.

153.  The Bureau declines, however, the suggestion of some commenters that the rule should take immediate or even retroactive effect. The Bureau believes that the roughly 75-day period between the final rule's issuance and effective date will provide time for affected institutions to review the final rule and adjust their operations in accordance with it.

154.  As explained below, the final rule also reverses the amendments relating to optional reporting of 2020 data in § 1003.3(c)(11) and comment 3(c)(11)-2 effective January 1, 2022 because those amendments will have become obsolete by that time.

155.  12 CFR 1002.13.

156.  Section 1003.3(c)(11) and comment 3(c)(11)-2 (both currently and as revised) permit a financial institution whose closed-end mortgage loans are excluded by § 1003.3(c)(11) to report voluntarily its closed-end data, as long as the financial institution reports such closed-end data for the full calendar year.

157.  12 CFR 1002.13.

158.  As used here, a newly excluded institution means an institution that was subject to HMDA's closed-end requirements as of January 1, 2020 because it originated at least 25 closed-end mortgage loans in 2018 and 2019 and met all of the other requirements under § 1003.2(g) but is no longer subject to HMDA's closed-end requirements as of July 1, 2020 due to the final rule's change to the closed-end threshold.

159.  For example, § 1002.5(a)(4)(iii) permits creditors that submitted HMDA data for any of the preceding five calendar years but that are not currently a financial institution to collect information regarding the ethnicity, race, and sex of applicants for loans that would otherwise be covered loans if not excluded by § 1003.3(c)(11) or (12). Section 1002.5(a)(4)(i) permits creditors that are currently financial institutions due to their open-end originations to collect information regarding the ethnicity, race, and sex of an applicant for a closed-end mortgage loan that is an excluded transaction under § 1003.3(c)(11) if they either: (1) Report data concerning such closed-end mortgage loans and applications, or (2) reported closed-end HMDA data for any of the preceding five calendar years. Additionally, § 1002.5(a)(4)(iv) permits a “creditor that exceeded an applicable loan volume threshold in the first year of the two-year threshold period provided in 12 CFR 1003.2(g), 1003.3(c)(11), or 1003.3(c)(12)” to collect in the second year information regarding the ethnicity, race, and sex of an applicant for a loan that would otherwise be a covered loan if not excluded by § 1003.3(c)(11) or (12). In the unusual circumstances present here, where Regulation C's closed-end threshold is changing in the middle of 2020, the Bureau interprets “an applicable loan volume threshold” as used in § 1002.5(a)(4)(iv) to include, even after the July 1, 2020 effective date, 25 closed-end mortgage loans for the year 2019 during the 2019-2020 period.

160.  The Bureau understands that final action taken on an application may occur in a different year than the application date.

161.  Specifically, section 1022(b)(2)(A) of the Dodd-Frank Act calls for the Bureau to consider the potential benefits and costs of a regulation to consumers and covered persons, including the potential reduction of access by consumers to consumer financial products or services; the impact on depository institutions and credit unions with $10 billion or less in total assets as described in section 1026 of the Dodd-Frank Act; and the impact on consumers in rural areas.

162.  These tasks include: (1) Data collection: Transcribing data, resolving reportability questions, and transferring data to HMDA Management System (HMS); (2) Reporting and resubmission: Geocoding, standard annual edit and internal checks, researching questions, resolving question responses, checking post-submission edits, filing post-submission documents, creating modified loan/application register, distributing modified loan/application register, distributing disclosure statement, and using vendor HMS software; (3) Compliance and internal audits: Training, internal audits, and external audits; and (4) HMDA-related exams: Examination preparation and examination assistance.

163.  See Bureau of Consumer Fin. Prot., “Final Report of the Small Business Review Panel on the CFPB's Proposals Under Consideration for the Home Mortgage Disclosure Act (HMDA) Rulemaking” 22, 37 (Apr. 24, 2014), http://files.consumerfinance.gov/​f/​201407_​cfpb_​report_​hmda_​sbrefa.pdf.

164.  The Bureau notes this description has taken into account the operational improvements the Bureau has implemented regarding HMDA reporting since issuing the 2015 HMDA Rule and differs slightly from the original taxonomy in the 2015 HMDA Rule that reflected the technology at the time of the study.

165.  The majority of the analyses in part VI of the May 2019 Proposal were conducted prior to the official submission deadline of the 2018 HMDA data on March 1, 2019, and 2017 was the most recent year of HMDA data the Bureau used for the analyses presented in the May 2019 Proposal. For this part of the final rule, the Bureau has supplemented the analyses with the 2018 HMDA data. The majority of the analyses for this final rule were conducted prior to the official submission deadline of the 2019 HMDA data on March 2, 2020. As of the date of issuance of this final rule, the Bureau is processing the 2019 HMDA loan/application register submissions and checking data quality, and some financial institutions are continuing to revise and resubmit their 2019 HMDA data. Accordingly, the Bureau has not considered the 2019 HMDA data in the analyses for this final rule. The Bureau notes the market may fluctuate from year to year, and the Bureau's rulemaking is not geared towards such transitory changes on an annual basis but is instead based on larger trends.

166.  See supra note 85.

167.  The further the market moves away from a perfectly competitive market, the smaller the pass-through would be.

168.  12 CFR 1003.1(b).

169.  The changes in this final rule will reduce public information regarding whether financial institutions are serving the needs of their communities. To the extent that these data are used for other purposes, the loss of data could result in other costs.

170.  The Bureau's overview is available in two articles. Bureau of Consumer Fin. Prot., “Data Point: 2018 Mortgage Market Activity and Trends: A First Look at the 2018 HMDA Data” (Aug. 2019), https://www.consumerfinance.gov/​data-research/​research-reports/​data-point-2018-mortgage-market-activity-and-trends/;;; Bureau of Consumer Fin. Prot., “Introducing New and Revised Data Points in HMDA: Initial Observations from New and Revised Data Points in 2018 HMDA” (Aug. 2019), https://www.consumerfinance.gov/​data-research/​research-reports/​introducing-new-revised-data-points-hmda/​.

171.  In the May 2019 Proposal, the Bureau estimated that there were about 4,960 financial institutions required to report their closed-end mortgage loans and applications under HMDA. Together, these financial institutions originated about 7.0 million closed-end mortgage loans in calendar year 2017. The Bureau observed that the total number of financial institutions that were engaged in closed-mortgage lending in 2017, regardless of whether they met all HMDA reporting criteria, was about 12,700, and the total number of closed-end mortgage originations in 2017 was about 8.2 million. The Bureau estimated then that under the current threshold of 25 closed-end mortgage loans, about 39 percent of all mortgage lenders were required to report HMDA data, and they accounted for about 85.6 percent of all closed-end mortgage originations in the country; among those 4,960 financial institutions that were required to report closed-end mortgage loans under HMDA, about 3,300 insured depository institutions and insured credit unions were partially exempt for closed-end mortgage loans under the EGRRCPA and the 2018 HMDA Rule.

172.  The Bureau recognizes that the estimates generated using this restriction may omit certain financial institutions that should have reported but did not report in the most recent HMDA reporting year. However, the Bureau applied this restriction to ensure that institutions included in its estimates are in fact financial institutions for purposes of Regulation C because it recognizes that institutions might not meet the Regulation C definition of financial institution for reasons that are not evident in the data sources that it reviewed.

173.  In the May 2019 Proposal, the Bureau estimated that if the closed-end threshold were increased to 100, the total number of financial institutions that would be required to report closed-end mortgage loans would drop to about 3,240, a decrease of about 1,720 financial institutions compared to the current level. These 1,720 newly excluded institutions originated about 147,000 closed-end mortgage loans in 2017. There would be about 6.87 million closed-end mortgage loan originations reported under the threshold of 100 closed-end mortgage loans, which would account for about 83.7 percent of all closed-end mortgage originations in the entire mortgage market.

The Bureau further estimated that all but about 50 of the 1,720 newly excluded closed-end mortgage loan reporters that would be excluded under the proposed threshold of 100 closed-end mortgage loans would be eligible for a partial exemption for closed-end mortgage loans as provided by the EGRRCPA and the 2018 HMDA Rule.

174.  In the May 2019 Proposal, the Bureau provided a similar table that included a breakdown of reporters by agency. For the final rule, as more relevant here, the Bureau has instead used this table to summarize the Bureau's estimates broken down by whether the reporters are depository institutions or non-depository institutions and, among depository institutions, whether they are partially exempt under the EGRRCPA.

175.  This does not include the costs of quarterly reporting for financial institutions that have annual origination volume greater than 60,000. Those quarterly reporters are all high-complexity tier 1 institutions, and the Bureau estimates none of the quarterly reporters will be excluded under this final rule.

176.  The Bureau estimated in the May 2019 Proposal that approximately 1,720 institutions would be newly excluded from the closed-end reporting under the proposed 100 loan threshold, of which about 1,670 are already partially exempt under EGRRCPA, and among those 1,670 financial institutions, about 1,540 are low-complexity tier 3 institutions and 130 are moderate-complexity tier 2 institutions. The Bureau also estimated in the May 2019 Proposal that of the approximately 50 remaining institutions that are required to report closed-end mortgage data under the 2015 HMDA Rule and are not partially exempt under the EGRRCPA but would be completely excluded under the threshold of 100 closed-end mortgage loans, about 45 are similar to the representative low-complexity tier 3 institution and about 5 are similar to the representative moderate-complexity tier 2 institution. These estimates are updated in the final rule and presented here.

177.  The Bureau estimated in the May 2019 Proposal that the total savings in the annual ongoing costs from HMDA reporting by excluded firms that are already partially exempt for closed-end mortgage loans under the EGRRCPA would be about $7.7 million. The Bureau also estimated that the total savings in the annual ongoing costs from HMDA reporting by fully excluded firms that are not eligible for a partial exemption under the EGRRCPA would be about $0.4 million. The Bureau estimated that together the annual savings in the operational costs of firms newly excluded under the threshold of 100 closed-end loans would be about $6.4 million. These estimates are updated in the final rule and presented here.

178.  In the May 2019 Proposal, the Bureau estimated that with the proposed threshold of 50 closed-end mortgage loans, about 760 institutions would be completely excluded from reporting closed-end mortgage data compared to the current level. All but about 20 of these 760 institutions would be eligible for a partial exemption under the EGRRCPA and the 2018 HMDA Rule. The Bureau estimated then that, of the approximately 740 financial institutions that are (1) required to report closed-end mortgages under the 2015 HMDA Rule, (2) partially exempt under the EGRRCPA, and (3) completely excluded under the proposed 50 loan threshold, about 727 were similar to the representative low-complexity tier 3 institution and about 13 were similar to the representative moderate-complexity tier 2 institution. Of the approximately 20 remaining financial institutions that are required to report closed-end mortgages under the 2015 HMDA Rule and are not partially exempt under the EGRRCPA but would be completely excluded under the proposed threshold of 50 closed-end mortgage loans, about 19 were similar to the representative low-complexity tier 3 institution and only one was similar to the representative moderate-complexity tier 2 institution. The Bureau estimated that the total savings in annual ongoing costs from HMDA reporting by fully excluded institutions that are already partially exempt under the EGRRCPA would be about $2 million, and the total savings in the annual ongoing costs from HMDA reporting by fully excluded firms that previously were not eligible for a partial exemption under the EGRRCPA would be about $140,000. Together the annual savings in the operational costs of firms excluded under the proposed threshold of 50 closed-end mortgage loans would be about $2.2 million.

179.  See Bureau of Consumer Fin. Prot., “Introducing New and Revised Data Points in HMDA: Initial Observations from New and Revised Data Points in 2018 HMDA” (Aug. 2019), https://www.consumerfinance.gov/​data-research/​research-reports/​introducing-new-revised-data-points-hmda/​.

180.  The further the market moves away from a perfectly competitive market, the smaller the pass-through would be.

181.  In general, credit union Call Reports provide the number of originations of open-end lines of credit secured by real estate but exclude lines of credit in the first-lien status. Call Reports for banks and thrifts report only the balance of the home-equity lines of credit at the end of the reporting period but not the number of originations in the period.

182.  In the May 2019 Proposal, the Bureau provided a similar table that included a breakdown of open-end reporters by agency. For the final rule, as more relevant here, the Bureau has instead included the breakdown by depository institution versus non-depository institution.

183.  The Bureau notes that more reporters are estimated to be in tier 2 in this updated analysis in the final rule than the number of reporters that the Bureau estimated to be in tier 2 in the May 2019 Proposal. This is mainly due to the fact that the Bureau now is able to supplement new information from the 2018 HMDA data, which allows the Bureau to conduct the estimates based on the number of open-end loan/application register records rather than the number of originations. Each institution is estimated to have more loan/application register records than in the May 2019 Proposal, because the Bureau is considering applications as well as originations, thus more institutions that were previously assigned to the tier 3 category are shifted into the tier 2 category.

184.  In the May 2019 Proposal, the Bureau estimated that the annual savings on operational costs would be about $1.8 million if the open-end threshold were increased from 100 to 200 in 2022. The higher estimate presented above for the final rule is mainly due to the fact that the Bureau now is able to supplement new information from the 2018 HMDA data, which allows the Bureau to conduct the estimates based on the number of open-end loan/application register records rather than the number of originations, resulting in more affected moderate-complexity tier 2 institutions and higher operational cost savings. Although the estimated total cost reduction is higher than it was in the proposal based on the additional 2018 HMDA data, the overall analysis is consistent with the Bureau's methodology and conclusions from the May 2019 Proposal.

185.  See Bureau of Consumer Fin. Prot., “Introducing New and Revised Data Points in HMDA: Initial Observations from New and Revised Data Points in 2018 HMDA” (Aug. 2019), https://www.consumerfinance.gov/​data-research/​research-reports/​introducing-new-revised-data-points-hmda/​.

186.  The further the market moves away from a perfectly competitive market, the smaller the pass-through would be.

187.  In comparison, in the May 2019 Proposal, the Bureau estimated that about 1,666 of the approximately 1,720 institutions that would be excluded from the proposed alternative 100 loan closed-end reporting threshold were small depository institutions or credit unions with assets at or below $10 billion, and all but two of them were already partially exempt under the EGRRCPA. About 1,573 of them are similar to representative low-complexity tier 3 institution, with the rest being moderate-complexity tier 2 institutions. Due to a transcription error, the Bureau indicated in the May 2019 Proposal that, combined, the annual saving on operational costs for depository institutions and credit unions with $10 billion or less in assets newly excluded under the proposed threshold of 100 closed-end mortgage loans would be about $4.8 million; however, upon review, the Bureau has determined that that estimate should instead have been $6.7 million based on the analysis in the May 2019 Proposal. As noted above, using the 2018 HMDA data, the Bureau now estimates that there will be approximately $6.0 million estimated savings in annual operational costs under threshold of 100 closed-end mortgage loans.

188.  In comparison, in the May 2019 Proposal, the Bureau estimated that the annual saving on operational costs for depository institutions and credit unions with $10 billion or less in assets newly excluded from open-end reporting under the threshold of 200 open-end lines of credit would be about $19. million. Also, in the May 2019 Proposal, the Bureau estimated the aggregate savings in avoided one-time cost associated with the threshold of 200 open-end lines of credit would be $3.8 million. The increases in the estimated cost savings in this final rule for both annual ongoing costs and one-time costs are due to the fact that the Bureau's updated estimates are able to incorporate the number of applications instead of originations based on information supplemented by the 2018 HMDA data.

189.  See, e.g., Keith Wiley, “What Are We Missing? HMDA Asset-Excluded Filers,” Hous. Assistance Council (2011), http://ruralhome.org/​storage/​documents/​smallbanklending.pdf;​; Lance George Keith Wiley, “Improving HMDA: A Need to Better Understand Rural Mortgage Markets,” Hous. Assistance Council (2010), http://www.ruralhome.org/​storage/​documents/​notehmdasm.pdf.

190.  See Robert B. Avery et al., “Opportunities and Issues in Using HMDA Data,” 29 J. of Real Est. Res. 352 (2007).

191.  If markets are not perfectly competitive or financial institutions are not profit maximizers, then what financial institutions pass on may differ. For example, they may attempt to pass on one-time costs and increases in fixed costs, or they may not be able to pass on variable costs.

192.  These cost estimates represent the highest estimates among the estimates presented in previous sections and form the upper bound of possible savings.

193.  The further the market moves away from a perfectly competitive market, the smaller the pass-through would be.

194.  Public Law 96-354, 94 Stat. 1164 (1980).

195.  Public Law 104-21, section 241, 110 Stat. 847, 864-65 (1996).

196.  5 U.S.C. 601-612. The term “ `small organization' means any not-for-profit enterprise which is independently owned and operated and is not dominant in its field, unless an agency establishes [an alternative definition under notice and comment].” 5 U.S.C. 601(4). The term “ `small governmental jurisdiction' means governments of cities, counties, towns, townships, villages, school districts, or special districts, with a population of less than fifty thousand, unless an agency establishes [an alternative definition after notice and comment].” 5 U.S.C. 601(5).

197.  5 U.S.C. 601(3). The Bureau may establish an alternative definition after consulting with the Small Business Administration and providing an opportunity for public comment. Id.

198.  5 U.S.C. 601-612.

199.  5 U.S.C. 609.

200.  5 U.S.C. 801 et seq.