In 1977, Congress passed the Community Reinvestment Act (CRA) to encourage banks to meet the credit needs of their local communities.1 The CRA started out as six sections, just over a page of legislative text, and after several amendments over the years, still stands at only eight brief sections.2 The core operative language in the law remains simple and concise: bank regulators are required to assess a bank’s record of meeting the credit needs of its entire community, including low- and moderate-income (LMI) neighborhoods, consistent with the safe and sound operation of the institution,3 and regulators must disclose the assessment to the public.4
The agencies issued an implementing regulation in 1978, and made significant revisions in 1995. Since then, the rule has remained largely unchanged for almost 30 years. The current effort to modernize the CRA rule launched in earnest when the OCC under former Comptroller Joseph Otting issued an Advanced Notice of Proposed Rulemaking in 2018.5 Five years and several twists and turns later, the FDIC is now considering a 1,500 page6 rule to substantially rewrite the CRA regulation. The bulk of the new requirements would apply to large, midsize, and community banks with $2 billion or more in assets, and my remarks will primarily focus on these provisions and these institutions (hereinafter “banks” or “institutions”).7
I agree that the CRA rule needs to be modernized. The rule should reflect the profound changes in banking that have occurred over the past 30 years, including new technologies and new business models, and it should provide more clarity, consistency, and certainty regarding how regulators assess CRA activity.
However, I oppose today’s Final Rule, for the reasons discussed below.
First, in an effort to provide more certainty regarding CRA compliance, the Final Rule establishes a highly complex series of formulas and benchmarks to evaluate a bank’s retail lending. However, because certain benchmarks are based on the activities of other lenders over the same evaluation period, combined with the complexity of the overall evaluation methodology, it will be difficult for an institution to know with confidence if its planned activities will produce the expected assessment results. The result will be more clarity around process, but not necessarily around outcomes.
Second, the new retail lending tests, in conjunction with the new retail lending assessment areas, create several potential disincentives for banks to continue serving certain communities. For example, using loan count to trigger new assessment areas could incentivize banks to retreat from, or reduce lending below the prescribed thresholds in, areas where they lend but lack a physical presence, to avoid bearing the substantial costs of investing in and maintaining the personnel, systems, and relationships necessary for an effective CRA program.8 While there are multiple reasons a bank may not, for CRA reasons, leave an area in which it has physical branches, there is far less to keep a bank in an area where it only makes loans.
Third, the new framework codifies a methodology that effectively grades banks on a curve, which is more stringent and more hardwired than the existing qualitative approach, and which the agencies project will significantly increase the percentage of banks that receive subpar scores.9 I am skeptical that this is a desirable outcome. When we amend our capital rules, for example, we do not make amendments that result in 10 percent of previously well-capitalized banks becoming undercapitalized. Instead, we assume in our analysis of such rules that banks will come into compliance by the compliance date, as they typically have done historically. Relatedly, we should be careful in promoting ever-more CRA lending, particularly in saturated markets.10 I support the goal of expanding access to affordable credit, including within LMI communities, but this always needs to be balanced with other objectives, including, as required by the CRA, safety and soundness.
The Final Rule’s treatment of community development (CD) activities contains several notable improvements over the existing CRA regime, including (1) providing credit for activities outside of banks’ branch areas, which will channel critical funding for projects in LMI areas with comparatively little banking presence, and (2) providing an illustrative list of CD projects that will qualify for CRA credit, along with a process for institutions to seek guidance on whether projects not on the list would receive credit.
However, the Final Rule also includes various restrictions that will discourage many worthwhile projects that we know make a difference for LMI communities. For example, institutions will not receive pro-rata credit for a range of projects that provide critical services to LMI communities, such as hospitals or schools, if they fail to meet the “Majority” or “Bona Fide Intent” standards of the new rule.11 Nor will firms get credit for a variety of projects unless they are done “in conjunction with” a government or non-profit plan, discouraging investment in private sector-initiated projects that would otherwise satisfy all the criteria for credit.
The Final Rule also puts a cap on the overall CRA rating an institution can receive if it fails to achieve high scores on its retail lending tests.12 This means that banks with business models that focus less on retail lending cannot improve their CRA ratings through more CD activity. The result will be to discourage valuable CD projects that would otherwise qualify for credit, without necessarily resulting in any additional retail lending. Decreased flexibility in the use of strategic plans may also have a similar impact.13
The Final Rule also requires banks with $10 billion or more in assets to annually identify every county in which depositors reside and report aggregate deposits for each county. This will be a substantial undertaking for firms. Meanwhile, following the bank failures earlier this year, the banking agencies have learned that having more granular deposit data could be helpful for supervisory and deposit insurance purposes. While I appreciate that more precise geographic information about deposits might be helpful in certain contexts, given that deposits are irrelevant for determining assessment areas under the CRA Final Rule and will only be used as inputs for the performance tests, I question whether we should impose substantial costs on the industry to identify with precision deposits on a county-by-county basis, rather than prioritize deposit data that could be more useful in reducing the risk of bank failures.
Finally, to my understanding, this is by far the longest rulemaking the FDIC has ever issued. By word count, it is more than 75,000 words longer than double the length of the capital proposal approved over the summer.14 I appreciate that rule writers always face the temptation to achieve greater precision and granularity at the cost of greater complexity. But the more complex a rule is, the less likely it is that bankers, the public, and examiners fully understand it; the more time and cost is spent on training, consultants, vendors, lawyers, compliance systems, IT tools, and more training; and the less likely it is that the rule is applied in a consistent and intended manner over time. At some point, the costs of added complexity outweigh the benefits of added precision and granularity, and I think this rule has blown far past that point.
We should also be open to the possibility that, given the rule’s complexity, adjustments to the rule may be warranted as data is collected and the impact on banks and LMI communities becomes clearer.
For the reasons above, I will vote against the Final Rule. I thank staff for all their work going back a number of years.
2 12 U.S.C. §§ 2901-08.
3 12 U.S.C. § 2903(a)(1).
4 12 U.S.C. § 2906.
5 Office of the Comptroller of the Currency, Advanced Notice of Proposed Rulemaking, Reforming the Community Reinvestment Act Regulatory Framework, 83 Fed. Reg. 45053 (Sept. 5, 2018). Certain concepts adopted by the OCC earlier in the process remain in the Final Rule under consideration today, including the publicly available illustrative list of activities eligible for credit under the Community Development Test and the process for seeking guidance on whether projects not on the list would be eligible for credit. See, e.g., Office of the Comptroller of the Currency, Community Reinvestment Act Regulations, 85 Fed. Reg. 34734, § 25.05 (June 5, 2020).
6 The page length would have been substantially longer if the 1,000-page, single-spaced preamble were double-spaced, as is standard practice. See Office of the Federal Register, Document Drafting Handbook, p. 70 (Aug. 2023).
7 Intermediate banks, with between $600 million and $2 billion in assets, will also be evaluated under the new Retail Lending Test, but not the Retail Services and Products Test, the Community Development Financing Test (unless they opt in), or the Community Development Services Test.
8 In addition, the elimination of partial counties from facilities-based assessment areas may incentivize banks to retreat from certain counties in which they have a relatively small presence.
9 See, e.g., Final Rule, Community Reinvestment Act, Supplementary Information, pp. 652-53 (Oct. 24, 2023) (“Final Rule”) (Table 32 to §_.22: Estimated Institution-Level Retail Lending Test Conclusions, 2018-2020).
10 See, e.g., Ben Bernanke, The Community Reinvestment Act: Its Evolution and New Challenges (March 30, 2007) (“[R]ecent problems in mortgage markets illustrate that an underlying assumption of the CRA -- that more lending equals better outcomes for local communities may not always hold.”); Sheila Bair, Did Low-income Homeownership Go Too Far? (Dec. 17, 2008) (“CRA always recognized … that a bank’s capacity and opportunity for safe and sound lending in the LMI community may be limited.”).
11 See Final Rule, supra note 9, §§_.13(a)(1)(i) and (ii).
12 In particular, banks cannot receive an overall State, multistate MSA, or institution rating of “outstanding” or “satisfactory” if the bank fails to achieve (1) at least a “low satisfactory” on the applicable State, multistate MSA, or institution-level retail lending test, or, (2) if the bank has 10 or more facility-based or retail-lending assessment areas, at least “low satisfactory” in 60 percent of assessment areas in the applicable State or multistate MSA or for the institution. See id., §_.28(b)(4); p. 1271 (Appendix D to Part _–Ratings, Section (g)).
13 See id., §_.27(c)(2) (requiring banks to include in strategic plans the same performance tests that would apply in the absence of an approved plan, except as provided in paragraph (g)(1)); §_.27(d) (detailing justifications required for strategic plans); §_.27(g) (establishing plan content requirements).
14 The capital Notice of Proposed Rulemaking was 288,032 words. The CRA Final Rule is 651,627 words.