Speeches & Testimony
Statement by Martin J. Gruenberg, Member, FDIC Board of Directors; Notice of Proposed Rulemaking: Community Reinvestment Act Regulations
December 12, 2019
The Notice of Proposed Rulemaking (NPR) before the FDIC Board today is a deeply misconceived proposal that would fundamentally undermine and weaken the Community Reinvestment Act.1 For that reason I will vote against this NPR.
The Community Reinvestment Act was signed into law over forty years ago. Over the course of its history it has become the foundation of finance for low- and moderate-income communities in the United States. As the preamble to this NPR states, "Since becoming law in 1977, the Community Reinvestment Act (CRA) has encouraged insured depository institutions (banks) to invest trillions of dollars into the communities they serve, including low- and moderate-income (LMI) neighborhoods."2
The regulation implementing CRA has not been significantly revised since 1995. Given the evolution of the banking industry over that period, there is certainly reason to consider revisions to the CRA regulation to enhance and strengthen its effectiveness.
For example, there is general agreement that, as technology has expanded opportunities for access to financial services, consideration should be given to how CRA could be adapted to include communities in which banks do substantial business but do not fall within existing branch-based assessment areas.
In addition, under the current CRA regulation, community development is considered in the lending, services, and investment tests. It may make sense to create a single community development test that would include lending, services, and investment, complemented by a retail test, which would include lending and services.
Further, finding ways to ensure greater consistency and predictability in CRA evaluations would also have value. Providing greater certainty to bankers and community organizations as to whether proposed investments or loans would receive CRA credit is a commonly shared goal, although it poses challenges in practice.
The NPR before the FDIC Board today would seek to address these issues. However it would do so in a misconceived, unworkable, and damaging way to CRA.
The Proposed CRA Evaluation Framework
At the core of this proposal is the establishment of a single metric CRA ratio or "CRA evaluation measure"3, at the bank and assessment area level, that would be the ultimate determinant of CRA performance.
Under the proposal, there would be a "bank-level CRA evaluation measure" and an "assessment area CRA evaluation measure."4
The bank-level and assessment area measures would be the sum of the value of all of the CRA qualifying activities divided by the value of retail domestic deposits at the bank and assessment area levels respectively.5
The proposed rule prescribes presumptive measures to determine a bank's CRA rating. For example, at the bank and assessment area level, a CRA evaluation measure of 11 percent would be required for an outstanding rating, 6 percent for a satisfactory rating, 3 percent for a needs to improve rating, and less than 3 percent for a substantial noncompliance rating.6
In addition, the bank would need to achieve a rating of outstanding or satisfactory in a "significant portion" of its assessment areas in order to receive an overall outstanding or satisfactory rating. The proposed regulation does not define the term "significant portion", but the preamble suggests "such as more than 50 percent."7
The proposal provides for pass-fail retail lending and community development tests, but these are really supplements to the CRA evaluation measure.8
It is the presumptive CRA evaluation measure or single metric that is the dominant determinant of the CRA rating.
Problems with the Proposed CRA Evaluation Framework
As the preamble to the proposed rule points out in regard to the commenters on the Advance Notice of Proposed Rulemaking on CRA that the OCC issued, "The majority support objective measurement of CRA performance, although they oppose a single metric."9
Yet a single metric at the bank and assessment area level is the foundation of the evaluation framework for this proposed rule.
It is problematic in several ways.
First, under this single metric approach, a bank must calculate the value of its CRA qualifying activities, at the bank and assessment area level, based on the dollar value of qualifying activities originated, made, and purchased by the bank.10
The problem with this approach is that adding up the dollar value of qualifying activities -- lending, community development investments, and community development services -- into a single metric undermines the evaluation of the bank's performance in each of these areas. It is a "count the widgets" approach that does not take into account the quality and character of the bank's activities and its responsiveness to local needs.
Second, taking this dollar value and dividing it by the value of the bank's retail domestic deposits, whether at the bank or assessment area level, is also problematic.
As the preamble to the NPR points out, "Deposit data …have limitations because the current reporting framework records deposits by attributing them to a branch location, rather than the account holder's address and uses a different definition of deposits than the proposed rule. The proposed rule would remedy these deficiencies by leveraging data that are readily available but not currently reported in an integrated and accessible manner. Over time, the data collection, recordkeeping, and reporting requirements in this proposal would remedy the current limitations." 11
The assumption that the data may improve in the future is not an appropriate basis for proposing changes to the CRA regulations now based on currently available data that is known to have deficiencies.
Third, the presumptive numerical standards for the CRA evaluation measure are not adequately supported in the proposal.
As the preamble to the NPR states,
"…Although these data sources have some limitations, by using all the sources together, … and making a limited number of assumptions, … the agencies were able to estimate what each bank's average CRA evaluation measure would have been from 2011-2018 under the framework in the proposal for all banks that filed a Call Report."12
The preamble goes on to state,
"Based on the agencies' review of these factors thus far, the agencies believe that the average CRA evaluation measure benchmarks associated with each rating category should be set at between 10 and 15 percent for outstanding, 5 and 10 percent for satisfactory, and 2 and 5 percent for needs to improve. … the proposal would set 11 percent as the initial benchmark for outstanding, 6 percent as the initial benchmark for satisfactory, and 3 percent as the initial benchmark for needs to improve. An average CRA evaluation measure of less than 3 percent would be associated with the substantial noncompliance category."13
No explanation is given as to how these specific benchmarks were determined, and none of the analysis referenced is provided. They appear to be arbitrary. Yet they are the basis for establishing new presumptive standards for CRA performance.
These presumptive standards undermine one of the most important benefits of CRA - the incentive for banks to develop partnerships with local community organizations and other stakeholders to address community needs - because the banks can satisfy their CRA obligations by simply hitting the metric. Further, the single, dollar value-based metrics favor large, easy-to-accomplish investments and loans over more complex and innovative activities that may take longer to develop but have a higher impact on the community.
Finally, this proposal would allow a bank to achieve a less than satisfactory rating in nearly half of its assessment areas and still receive a satisfactory or even outstanding rating. Banks would have the flexibility to focus their stronger community reinvestment-qualifying efforts on as few as half of their assessment areas while minimizing their efforts elsewhere.
The challenges presented by the single metric CRA evaluation framework are compounded by other aspects of the proposal. I will highlight three.
First, the NPR expands eligible and qualifying CRA activities to include some of what banks already do in the ordinary course of business, thereby diluting the effectiveness of CRA. For example, community development activities – loans, investments, and services – would no longer have to have a primary purpose of community development targeted on low- and moderate- income individuals and areas, small business or small farms, or underserved or distressed rural areas.14 This broadening of what counts in the proposal comes at the cost of CRA's historic focus on serving low- and moderate-income communities and individuals, while giving the appearance of expanding the overall level of CRA activity.
A second issue is how the proposal would address assessment areas. Under the proposal, a bank that receives 50 percent or more of its domestic deposits from outside of its current branch-based assessment areas would be required to delineate deposit-based assessment areas where it receives five percent or more of its total retail based deposits.
However, as previously noted, the deposit data needed to implement this are deficient. We do not know how many or where these deposit-based assessment areas might be, or how they would benefit low-and moderate-income communities. It is not clear that communities that are so-called "credit deserts" would necessarily benefit from the five percent threshold. In fact, the effect of this proposal may be to add assessment areas in places that are already "CRA hotspots", such as New York City or Silicon Valley
Finally, the NPR would virtually eliminate the retail services test, retaining minimal recognition for branches in low- and moderate-income areas.15 There would be no consideration of a bank's efforts to provide affordable products and services intended to expand access to the banking system to low- and moderate-income individuals who are currently unbanked. This would undermine the FDIC's long-term effort to address this issue. Low-cost transaction and savings accounts, which the FDIC has helped to promote, will no longer be considered for CRA credit simply because these accounts cannot be quantified under the single metric system that would be set up under the NPR.
In conclusion, this is a deeply misconceived proposal.
It would establish a CRA evaluation framework relying on a single metric approach that would allow a bank to concentrate its CRA activity in as little as 50 percent of its assessment areas, disinvest in the other 50 percent, and still receive a satisfactory or even outstanding CRA rating.
The metric itself would establish presumptive percentage standards for CRA ratings for which no basis is given. The metric, as well as new CRA assessment areas, is reliant on retail domestic deposits for which reliable data is deficient.
The proposal would expand current CRA eligible activities and thereby reduce the focus of CRA on low- and moderate-income individuals and communities. It would virtually eliminate the retail services test that currently encourages low cost bank accounts to expand access to the banking system to those who are currently unbanked.
Finally, the proposal undermines what has been a core strength of CRA for forty years – the encouragement of bank engagement and dialogue with stakeholders in local communities to understand and better serve historically underserved areas.
Taken together, this proposal would fundamentally undermine and weaken the Community Reinvestment Act. For that reason, I will vote against this Notice of Proposed Rulemaking.
1 12 U.S.C. 2901 et seq.
2 Preamble to the Notice of Proposed Rulemaking: Community Reinvestment Act Regulations at 1.
3 Proposed Section 345.10 in the NPR.
6 Proposed Section 345.12 in the NPR.
7 Preamble to the NPR at 50.
8 Proposed Sections 345.11 and 345.12, discussed in the Preamble to the NPR at 54-59 and 63-64. The proposal also allows for consideration of performance context factors, but only to adjust the bank-level and assessment area presumptive ratings. The proposed approach bears no relationship to the use of performance context in the current regulations. Proposed Section 345.14. In addition, the NPR allows for consideration of discriminatory conduct in the CRA evaluation. Proposed Section 345.15.
9 Preamble to the NPR at 13.
10 Proposed Section 345.10.
11 Preamble to the NPR at 62-63.
12 Preamble to the NPR at 59-60 (footnotes omitted).
13 Preamble to the NPR at 62.
14 Proposed Section 345.04, discussed in the Preamble to the NPR at 24-33.
15 Proposed Section 345.11, discussed in the Preamble to the NPR at 49-50.