Joint Press Release
October 24, 2023
Statement on the Community Reinvestment Act Final Rule By Governor Michelle W. Bowman
I'd like to begin by recognizing the extraordinary efforts of our Division of Consumer and Community Affairs staff, for their work over the past several years to draft the initial Community Reinvestment Act (CRA) proposal and more recently the final rule now before the Board for consideration.
I appreciate that we are continuing to consider important policy issues through open Board meetings. Returning to the pre-Covid practice of deliberating on important regulatory matters transparently within the view of the American public is critical to the legitimacy of the rulemaking process and for all public institutions, including the Federal Reserve. This formerly common practice provides a venue to discuss and debate important issues affecting communities, depository institutions, and other stakeholders in a transparent and accountable way. But, since the matters for consideration during today's and tomorrow's meetings are not urgent, it seems more appropriate to schedule these Board meetings outside of the FOMC blackout period.
An open process for this rule is important because the rule will materially change the way banks think about and choose to make investments in their communities. In an increasingly uncertain economic environment with high inflation and high interest rates, this rule could significantly increase the administrative costs of compliance, diverting funds to the acquisition and ongoing costs of data management compliance systems.
The purpose of the Community Reinvestment Act is to improve access to credit in all communities where banks are located, especially low- and moderate-income communities. The CRA was enacted in 1977 shortly after the civil rights movement and against the backdrop of other significant federal laws designed to address financial inclusion and equal access to credit. At the time Congress passed the CRA, it found that banks had a "continuing and affirmative obligation to help meet the credit needs" of their local communities.1 Congress reinforced this obligation by instructing the federal financial supervisory agencies to encourage banks to help meet the credit needs of those same communities.2 Throughout the years since, Congress has amended the CRA a number of times,3 but at its core, the main objective of the CRA has remained unchanged: banks should be involved in their communities, particularly as it relates to helping meet the credit needs of those communities.
The vast majority of the remaining 4,200 banks in the United States have well under $10 billion in assets.4 These banks have simple, straightforward business models, and they are focused on serving their unique communities – focusing on providing resources, investments, and credit for the betterment of the entire community. Given this landscape, and these banks' important focus on serving their communities, as we approach finalizing the CRA regulation, it is absolutely essential that this final rule is straightforward, clear, and strikes the appropriate balance between encouraging banks to meet the credit needs of their communities and avoiding the creation of a rule that may unintentionally disincentivize or effectively prohibit banks from supporting their communities.
There are several helpful improvements in the final rule, including providing a list of community development activities that qualify for CRA credit. A new pre-approval process also provides banks with a path to ensure that a certain activity will qualify for CRA credit before engaging in a loan or planning an investment activity. Other changes may incentivize banks to consider CRA investments in communities on native lands or that experience persistent poverty. The rule also includes specific provisions related to women- and minority-owned depository institutions and certified Community Development Financial Institutions. In addition, the rule would allow counting community development loans and investments made during a bank's prior evaluation period if they remain on a bank's balance sheet during the current evaluation period. Many of these changes are welcome improvements enhancing the transparency and certainty of CRA requirements and evaluations.
Changes I Cannot Support
Even with these notable improvements, regrettably I cannot support the finalization of this rule. In the final rule, the agencies have arguably exceeded the authority granted by the CRA statute. In addition, the final rule is unnecessarily complex, overly prescriptive, and contains disproportionately greater costs than benefits, adding significantly greater regulatory burden for all banks, but especially for community banks. The premise of the changes being made in this rule is that banks are not doing enough to meet the credit needs of their communities. Yet, there is no evidence provided to support this premise. In light of these issues and the complexity and extraordinary length of this rule, I would have preferred that we more fully address these issues and publish a new proposed rule for comment.
Threshold for Large Banks
The final rule adopts a new $2 billion threshold for a bank to be considered a "large bank" for purposes of CRA. In my view, this threshold is not only far too low, but it also does not sufficiently differentiate between smaller community banks and the largest banks. In no other provision of the regulatory framework is a bank with $2 billion in assets considered a "large" bank. For well over a decade, community banks have been defined to include banks with up to $10 billion in total assets.5 Characterizing these banks as "large banks" simply ignores established definitions and is inconsistent with existing regulatory practice. Further, the final rule essentially applies the same evaluation for a $2 billion bank as it does for a $2 trillion bank. The lack of recognition that these banks are fundamentally different, with different balance sheets and business models, misses an important opportunity to appropriately tailor CRA expectations to a bank's size, risk, service area, and business model.
Changes for Community Banks
Under the final rule, community banks with more than $600 million in total assets also would see significant changes to their CRA requirements when compared to the current CRA rule. When I agreed to support the initial proposal, I did so with the understanding that all banks under $10 billion in assets would be subject only to existing CRA rules and would have the option to choose to adopt the new CRA framework. Unfortunately, the rule being finalized today disregards this agreement and materially changes requirements for these banks including mandating compliance with a new retail lending test, significantly altering and expanding assessment areas, and increasing data and reporting obligations. Instead of requiring these changes, in my view, community banks should have the option, at their discretion, to opt into the new retail lending test and assessment areas, or to continue with the existing framework.
The changes to CRA requirements for this category of banks will also come with significant increases in burden and cost, despite the lack of evidence that these banks are not currently meeting the credit needs of their communities. Indeed, the opposite seems to be true. Community banks are indispensable in supporting their communities and they are meeting the credit needs of their communities. The new rule will require these banks to make significant changes to policies and systems to comply with the new requirements, all of which will come at significant ongoing cost – solely for the purpose of complying with the rule. I am concerned that the aggregate effect of the increased burden could lead to a reduction in lending to offset the increased costs associated with new data requirements. To better understand this risk, the final rule should include a proportionality measurement to ensure that the cost of compliance neither outweighs the benefit from nor detracts from the investment that would have been made in the absence of the new framework.
Lack of Congressional Authorization
One of the purported goals of the final rule is to modernize the CRA to account for changes in the way banks operate. While I generally support efforts to modernize the CRA to make it align more closely with current practices of extending credit in communities, including acknowledging the increased presence of mobile and online banking, Congress, not the banking agencies, is responsible for modernizing the statute. In doing so, Congress could consider several approaches to modernize the CRA, including reflecting the variety of financial institutions that provide credit and financial services in their communities.6
In my view, some of the changes being made by the agencies in this rule, including those that evaluate banks outside of their deposit-taking footprint, are likely beyond the scope of our authority under the statute. Furthermore, at the same time that we are recognizing the effects of mobile and online banking in this rule, we are not willing to consider similar effects from mobile and online banking in other areas of our authorities, such as the evaluation of competitive effects of merger and acquisition proposals.
I am also concerned that the final rule contains many changes that will have the unintended consequence of reduced community investment. The addition of the retail lending assessment areas and the outside retail lending areas, combined with the new requirement for large banks to include an entire county instead of a partial county as an assessment area, may create disincentives for banks to continue lending in these areas, essentially limiting credit access.
Specifically, if a bank would be required to establish a new assessment area or must satisfy certain activities in an entire county, does this create an incentive for banks to stop or limit lending in these new assessment areas? If banks will be required to satisfy new CRA requirements in new locations, potentially untethered to their deposit-taking footprint, it is entirely possible that banks will reduce their level of lending or operations in those areas. Under the rule, a large bank will need to define a new retail lending assessment area if it originates specific numbers of home mortgage or small business loans in that area. A bank approaching these thresholds may decide that limiting lending would alleviate the need to create a new assessment area.
The final rule does not consider whether this will occur. Furthermore, the rule also does not include a cost analysis to quantify the cost to implement new or update old data management systems for the new data requirements. We simply do not know if the costs of new systems and resulting changes to banks' business plans will similarly result in a reduction in lending and investment to offset the increased costs of compliance.
The final rule would also make it much more difficult for banks to maintain existing CRA ratings without making significant changes to their current activities. As described in the materials before the Board today, based on changes to the retail lending test alone, nearly 10 percent of banks would be rated "Needs to Improve" based on data from 2018 to 2020. Today, the number of banks with a "Needs to Improve" rating stands at roughly one percent. The approach in the final rule assumes that the low number of banks with a "Needs to Improve" rating is a sign of shortcomings in the rule, ignoring that banks have a deep commitment to supporting their communities and are already meeting the letter and spirit of the existing CRA statute.
If a bank receives a "Needs to Improve" rating under the retail lending test, it would not be possible to achieve a "Satisfactory" rating under the final CRA rule. Accordingly, this proposal effectively requires a number of banks to change their business plans to satisfy the new retail lending test. This seems like regulatory overreach, and as I have already noted, there is little evidence that banks are not currently meeting the credit needs of their communities.
Furthermore, we do not know the impact of the community development financing test on banks' CRA ratings. The materials indicate that the agencies do not have the data available to determine how to apply the community development financing test, noting that the agencies will need to issue guidance. However, based on the retail lending test alone, we know that a significant number of banks will see their CRA ratings fall, which will have real and meaningful consequences. Any bank that receives a rating below "Satisfactory" is generally prohibited from merger and acquisition activity. This prohibition could result in significant harm to certain communities, and potentially to the broader economy.
It is not appropriate for the banking agencies to materially increase the requirements on banks resulting in a downgrade of currently satisfactory performance to "Needs to Improve" without a thorough, data-supported analysis that justifies a recalibration evidenced by actual shortcomings in bank activities. The final rule contains no discussion or explanation for why currently satisfactory practices will no longer be satisfactory. As a result, many banks will not be able to receive a "Satisfactory" rating without significant changes that are largely dictated by regulators. Those changes are likely to be directed by several elements of the rule that give banks extra credit only for certain selected activities, loans, or products. By raising the standards for what will qualify as satisfactory performance, regulators are effectively mandating that banks offer preferred products and services to counteract the downward pressure on ratings.
Use of Unknowable Benchmarks
The rule also contains a market benchmark and a community benchmark for the retail lending test evaluation. The market benchmark reflects performance relative to peers, and the community benchmark reflects performance relative to firm and household demographics. Although the market benchmark is designed to account for changes in economic conditions during different business cycles, it is not clear if it will accurately reflect changes to local economic conditions or loan demand as actual conditions change. If local demand for loans falls due to changed economic circumstances, will the market benchmark accurately reflect new conditions in a timely way?
In addition, banks would not be made aware of the market benchmark in advance, and they would not know which benchmark would apply for their evaluation. The lack of transparency regarding these benchmarks raises potential due process concerns. The benchmarks should be published and measurable, not unknown and unknowable. Further, institutions should not be graded on a curve or compared to others with different business models and product offerings.
The final rule also may put banks in a no-win situation with their supervisors. CRA requires banks to help meet the credit needs of their communities, but they must do so consistent with safe and sound operations. Just as a bank may be criticized for not meeting expectations under these new CRA obligations, I worry that the changes imposed by this rule will result in banks being criticized for extending credit to less credit worthy borrowers to account for the increased barriers to achieve a satisfactory rating. Supervisory scrutiny extends not only to CRA considerations but also to safety and soundness considerations. Because the supervisory process for both the CRA and safety and soundness is opaque, it will be important to understand how supervisors will balance the new CRA benchmarks and requirements with safety and soundness.
In general, this rule is unnecessarily long and complex, so much so that most banks and members of the public will find it difficult to understand.7 Taken together, it will be a challenge for banks, particularly smaller banks, to understand what they must do to continue to receive satisfactory ratings under the CRA. This is exacerbated by the fact that over the past few months banks have been overwhelmed with lengthy, complex, and costly regulatory proposals, rules, and guidance that do not address any identified shortcomings in banking regulation, yet add significant burden. At a time when confidence in public institutions is waning, we should focus on rulemakings that aim to solve identified or documented problems.
There are a number of additional provisions in the rule that raise concerns.
The final rule requires the federal banking agencies to publish on their websites data available under the Home Mortgage Disclosure Act ("HMDA"). It is not clear what additional benefit this offers since HMDA data is already published and available for public review. In addition, the public benefit of disclosing this data on the Board's website is outweighed by the likelihood of confusing or misleading the public. The federal banking agencies previously concluded that HMDA data alone provide an incomplete measure of a bank's lending in their communities.8 HMDA data also cannot prove unlawful discrimination because it does not contain critical information including borrower credit history, debt-to-income ratios, or housing prices. The redundant disclosure of this information on the Board's and other agency websites could result in an increase in public comments on Board applications based solely on allegations from incomplete data.
Summary of Deposit Data Reliance
I am also concerned that certain elements of the evaluation framework will result in smaller banks either being penalized or forced to collect and report deposit data that is only "required" for larger banks. Under the final rule, the denominator for each of the Bank Volume Metric and the Bank Assessment Area Community Development Financing Metric rely on deposit data. However, for smaller banks that are allowed to use SOD data, the practical effect of these metrics may be to require these banks to either collect deposit information or risk ratings that are lower because the SOD deposit data may include deposits outside of a banks assessment area, which would artificially inflate the denominator of the metric. As such, these metrics could lower banks' performance, and the only way to remedy the lowered rating would be to incur significant costs to collect and report new deposit data.
Implementation Phase Length
Finally, the final CRA rule is extremely complicated and will require extensive changes both to banks' CRA practices, and to each of the agencies' CRA examination policies and guidance. In my view, the length, complexity, and number of required changes suggests that we will need an implementation phase that exceeds the two years adopted in the final rule. Additional time to implement this rule would result in better outcomes for banks and their communities.
Although I am not able to support this final rule, I continue to fully support the underlying intent and goals of the CRA. The statute remains as important today as it was in 1977 when Congress first enacted the statute. Even in the absence of CRA, banks should serve the credit needs of their communities, especially minority and low- and moderate-income communities. As banks begin to comply with the new CRA requirements, it would be helpful to understand how the implementation of this rule impacts the provision of credit in communities. Regulators must implement this rule in a way that is transparent and fair to ensure that banks understand how their community investment activities will continue to meet the credit needs of their communities and fulfill their obligations to comply with CRA.9
1. 12 U.S.C. § 2901(a)(3). Return to text
2. 12 U.S.C. § 2901(b). Return to text
3. See Financial Institutions Reform, Recovery, and Enforcement Act of 1989, Pub. L. No. 101-73, 103 Stat. 183 (Aug. 9, 1989); Resolution Trust Corporation Refinancing, Restructuring, and Improvement Act of 1991, Pub. L. No. 102-233, 105 Stat. 1761 (Dec. 12, 1991); Federal Deposit Insurance Corporation Improvement Act of 1991, Pub. L. No. 101-242, 105 Stat. 2236 (Dec. 19, 1991); Housing and Community Development Act of 1992, Pub. L. No. 102-550, 106 Stat. 3874 (Oct. 28, 1992); Riegle-Neal Interstate Banking and Branching Efficiency Act of 1994, Pub. L. No. 103-328, 108 Stat. 2338 (Sept. 29, 1994); and Gramm-Leach Bliley Act of 1999, Pub. L. No. 106-102, 113 Stat. 1338 (Nov. 12, 1999). Return to text
4. See Federal Deposit Insurance Corporation, Quarterly Banking Profile, Second Quarter 2023 (PDF), at 14. Return to text
5. Statement of Thomas M. Hoenig, President, Federal Reserve Bank of Kansas City, before the Subcommittee on Oversight and Investigations of the Committee on Financial Services, Aug. 23, 2010; GAO Report on Community Banks and Credit Unions: Impact of Dodd-Frank Act Depends Largely on Future Rulemaking (PDF), Sept. 2012, n.3 (defining community banks as those with under $10 billion in total assets because the Dodd-Frank Act exempts small institutions from a number of provisions based on that threshold). Return to text
6. As noted above, Congress has amended the CRA several times. See n. 3. Return to text
7. The final rule and explanatory preamble consists of 1500 pages of text, including appendices. Return to text
8. First Michigan Bank Corporation, 80 Federal Reserve Bulletin 632, 633 (July 1994) ("The Board recognizes, however, that HMDA data alone provide an incomplete measure of an institution's lending in its community. The Board also recognizes that HMDA data have limitations that make the data an inadequate basis, absent other information, for concluding that an institution has engaged in illegal discrimination in making lending decisions."); CBTX, Inc., FRB Order No. 2022-19 (PDF) (Sept. 14, 2022) (finding that "other information critical to an institution's lending may not be available solely from public HMDA data); OCC Conditional Approval #454 n.19 (PDF) (April 2001) ("It is important to note that HMDA data alone are inadequate to provide a basis for concluding that a bank is engaged in lending discrimination or in indicating whether its level of lending is sufficient. HMDA data do not take into consideration borrower capacity, housing prices, and other factors relevant in each of the individual markets and do not illustrate the full range of the bank's lending activities or efforts."); Frequently Asked Questions About the New HMDA Data, Question 13 (PDF) (noting that "HMDA data will not alone prove unlawful discrimination" because it "exclude[s] many other potential determinants, such as borrower credit history, borrower debt-to-income ratio, and the ratio of the loan amount to the value of the property securing the loan (loan-to-value ratio)."). Return to text
9. This document was corrected after publication to remove a sentence that was inadvertently added during the publication process and was not part of Governor Bowman's statement. Return to text