April 8, 2020
RE: Notice of Proposed Rulemaking, Community Reinvestment Act Regulations, Docket ID OCC-2018-0008 and RIN 3064-AF22
To Whom it May Concern:
The National Community Reinvestment Coalition, an association of 600 community-based organizations that promote access to basic banking services, affordable housing, entrepreneurship, job creation and vibrant communities for America’s working families, believes that the proposed changes to the Community Reinvestment Act (CRA) regulations outlined in Notice of Proposed Rulemaking (NPRM) would weaken CRA and would decrease CRA-related lending, investing, and services to low- and moderate-income (LMI) households and communities.
The Office of the Comptroller of the Currency (OCC) and Federal Deposit Insurance Corporation (FDIC) (the agencies) laud CRA, stating that it has been responsible for trillions of dollars in lending and investing in LMI communities. The agencies assert that their proposal would leverage billions of additional CRA dollars. However, the NPRM would halt if not reverse the progress made under CRA by introducing an overly simplistic and yet convoluted evaluation system that would divert CRA lending and investing away from LMI families and communities.
The agencies emphasize that the current CRA regulations, last updated 25 years ago, have not kept pace with changes in the banking industry. In addition, the agencies assert that CRA exams lack transparency, clarity, and fairness. According to the agencies, the ossification and uncertainty of CRA has discouraged banks from seeking innovative means to meet credit needs. CRA assessment areas, the geographical areas on CRA exams, are out-of-date, because they are based on bank branch locations while banks are increasingly conducting business on-line.
While the agencies identify pressing issues associated with CRA, their prescriptions for reform will not achieve clarity and fairness nor leverage more dollars for those most in need. Instead, in their efforts to inject simplicity in CRA exams, they introduce a dollar-based metric (called the CRA evaluation measure) that would establish benchmarks without empirical support and would be overly complex, too rigid to adjust when economic conditions change, and unable to accurately measure bank responsiveness to local needs. The agencies admit that most of the public comments on the Advance Notice of Proposed Rulemaking (ANPR) opposed the single metric, but the agencies nevertheless proceeded with this unworkable and poorly-documented metric. The inclusion of a few additional metrics since the ANPR does not nullify the criticism of the CRA evaluation measure presented in the ANPR.
As FDIC board member Martin Gruenberg stated, “This is a deeply misconceived proposal that would fundamentally undermine and weaken the Community Reinvestment Act.”
We agree with Mr. Gruenberg’s assessment. Previously, based on Federal Reserve research, NCRC estimated that any proposal that undermines local evaluations of CRA performance could result in a reduction of up to $105 billion in home and small business lending over five years. Since the proposal would allow banks to fail in up to one half of their assessment areas, eliminate evaluations of home lending in LMI tracts, and evaluate banks by a CRA evaluation measure that favors large-scale finance over smaller dollar retail lending, NCRC’s estimate is not likely to overstate the potential harm of this proposal.
In the preamble to the NPRM, the agencies list the goals of the proposal, many of which are laudatory from our perspective. However, the following comments show how the agencies’ proposals would often achieve the opposite result.
- Agency assertion about re-focus on LMI households and communities. The agencies state that they seek to encourage banks to serve areas with the “greatest need for economic development.” However, by broadening what counts on CRA exams beyond activities that primarily benefit LMI households, it is likely that the dollar amount of CRA activities directly benefiting LMI families and communities would decline.
- Agency assertion about reducing displacement by refocusing on LMI individuals and activities – The agencies propose to eliminate retail lending in LMI census tracts as a criterion on a CRA exam. NCRC’s comment below will explain how this is a counterproductive reaction to gentrification and will thwart fair lending policies aimed at achieving integration. Moreover, the agencies would expand the number of activities that would count as community development, such as financing improvements to sports stadiums in Opportunity Zones that would probably increase displacement of LMI families.
- Agency assertion about increasing small business and farm lending – The agencies propose to raise the revenue size of businesses and farms eligible to receive CRA lending and investments. By redefining small businesses and farms to include larger entities, the measures of lending and investing may increase but more of those dollars would not reach the businesses and farms most in need of credit and which empirically have been shown to be the primary engines of economic growth and job creation.
- Agency assertion about reducing CRA deserts and hotspots – The agencies assert that the definition of assessment areas (AA) on CRA exams needs to be amended in order to reduce the number of geographical areas that are “overheated” CRA markets and increase activity in areas of “financial need.” However, the agencies did not present data analysis describing the impacts of their reform proposal on AAs. Moreover, allowing banks to pass CRA exams while failing in up to half of their AAs would exacerbate the problems of CRA deserts and hotspots. Banks would not have strong incentives to serve all of their AAs (“the entire community” in the words of the statute) and could concentrate their efforts on easier places in which to lend and invest.
- Agency assertion about creating incentives to do more – The agencies state that their proposed performance measures would be based on “historical performance” and would be “high enough to increase” the levels of activities. While the agencies proposed empirical benchmarks, they did not present the data analysis that led them to conclude that CRA activities would increase. The fact that many activities proposed to qualify for CRA credit could not be incorporated into historical analysis further calls into question the appropriateness of the proposed benchmarks. Further, since most types of community development financing would be multiplied by a factor of two in CRA exams, actual community development financing would likely fall.
- Agency assertion about preserving the importance of branches – Since areas with bank branches would continue to be designated as AAs, the agencies asserted that they preserved the importance of branches. Yet, the summary of the NPRM does not mention that the agencies propose to delete the retail service test of the large bank exam that explicitly evaluated the number and percent of branches in LMI communities or the provision of basic banking services and deposit accounts to LMI consumers. As a substitute for the service test, the proposal would factor branches in to the new CRA evaluation measure, but the proposal would do so in a way that radically devalues the importance of maintaining branches in LMI census tracts.
- Agency assertion about preserving community voice – The agencies stated that by preserving community input into AA needs and opportunities, they were retaining and promoting community voice in the CRA process. Yet, they did not propose making it easier for communities to identify CRA examiners or agency staff to whom to send comments, nor do they explicitly state that the public could and should comment on the actual CRA performance of banks.
In fact, the NPRM would:
- Reduce accountability – The NPRM would reduce the public accountability of banks to continually serve credit needs in contradiction to the statute. The NPRM proposes to examine banks that achieve the highest rating of Outstanding once every five years instead of the current examination schedule of once every two to three years. This long time period would allow banks to relax their CRA efforts in the early years of the time period. NCRC research found that 30 percent of the largest 50 banks received the highest rating in recent years, meaning that this proposal would significantly relax CRA incentives and accountability for banks that had more than $1.6 trillion in assets at the time of the NCRC study.
- Small banks relieved from community development responsibilities – The agencies proposed to exempt about 85% of banks (those with less than $500 million in assets) from responsibilities to make community development loans and investments, although a significant segment of such banks currently make a considerable level of these loans and investments in rural areas and smaller towns. This would reduce activity in many CRA “deserts,” contrary to the goals the agencies lay out in the NPRM.
NCRC Suggestions for Reform
NCRC supports reform, but NCRC believes that incremental reforms building on the existing regulations would more effectively and transparently clarify what counts, achieve AA reform (where activity counts) and establish how those activities count in determining bank CRA ratings. Below, NCRC provides more detail on our reform proposals in response to the counterproductive proposals of the agencies. In short, NCRC suggests:
- Clarifying what counts – While NCRC supports the concept of a list of qualified activities that the agencies propose, the agencies propose an ad hoc procedure that would not be transparent to the public. The agencies should update the list periodically via public notice and comment. The agencies should not act upon bank requests to add to the list in the time periods in between rounds of public notice and comment. Instead, the agencies should conduct annual requests (twice a year in the early years after a new rule) for comment regarding proposed changes in order to accommodate industry and community organization requests.
In addition, the list should be principles-based rather than simply identifying qualified activities. A list specifying activities, without explaining the principles underlying qualification could lead banks and other stakeholders to believe that it is exclusive, which is not the agencies’ intent. To complement a principles-based list, the agencies could develop an interactive database of qualifying activities on CRA exams.
- Clarifying where activities count – The agencies base their AA reform proposals on data that is not available yet. NCRC, in contrast, has advocated for the addition of AAs beyond areas with bank branches based on publicly available lending data, which readily indicates substantial loan volumes in areas beyond bank branches as well as in areas with bank branches. The agencies also proposed enlarging the current definition of underserved and distressed census tracts as additional places where activities count, but the agency definition would miss the most underserved areas that NCRC’s counter-proposal detailed below would include.
- Clarifying how activities count – NCRC suggests incremental reforms to the existing performance measures, which would include guidelines establishing benchmarks. NCRC suggests that retaining the separate performance measures in the current component tests would be more effective in holding banks accountable for increasing CRA activities instead of the agencies’ proposed CRA evaluation measure which would likely decrease activities most responsive to local need. The agencies currently state they do not have benchmarks for performance under the lending, investment and service tests applicable to large banks. A simpler and operationally more transparent reform would be to establish benchmarks, based on historical data, for the existing tests.
- Mandatory inclusion of affiliates – The agencies did not require mandatory inclusion on exams of bank mortgage company affiliates, some of whom engaged in abusive lending during the financial crisis. This increased oversight would be more effective in enlarging the pool of safe and sound CRA loans and investments in contrast to the agencies’ proposal to allow for more activities that do not significantly benefit LMI communities.
- Consideration of race and fair lending exams – Fair lending exams must be made more rigorous in order to prevent discriminatory and abusive lending. In addition, the agencies must more explicitly consider race on CRA exams. Below, NCRC describes a proposal for underserved census tracts that appropriately focuses on communities of color that receive relatively few loans.
- Improvements to data collection and dissemination – The agencies would require banks to collect more data on consumer lending and community development activities but would not require banks to publicly release this data on a county or census tract level, which NCRC recommends as the most effective way to hold banks accountable for serving local needs and to increase transparency regarding a bank’s activities in between CRA exams.
The agencies assert that the sum of their proposals would improve the quality and consistency of publicly-available CRA performance evaluations. On the contrary, the exams would become opaque and less effective in terms of facilitating public review and comment. In addition, the agencies assert that their proposals would facilitate more timely publication of CRA exams, but the agencies ignore the reason that some exams had not been released in a timely manner, namely that violations of fair lending and consumer protection laws during the financial crisis took a long time to resolve and delayed release of exams. Regular and proactive enforcement of laws and regulations would be the most effective manner to improve the timeliness of CRA exam publication.
Comment Road Map
This comment letter will address what counts, where it counts, and how it counts. Before addressing what counts, a brief legislative and regulatory history of CRA will explain how the proposed rule is contrary to the statute and is arbitrary and capricious. A lack of agency data analysis makes it impossible for the public to ascertain whether the proposed rule would increase or decrease CRA activity. After the legislative and regulatory history section, a section on what counts will emphasize that CRA must remain focused on LMI households and communities instead of the agencies’ proposal to broaden activities that qualify for CRA credit, which have, at best, diffuse benefits for LMI households and communities. This is followed by a section on where it counts that will describe how the agencies’ proposal to reform AAs, which are geographical areas on CRA exams, must be re-worked because it is not based on data that currently exists.
A section then discusses the agencies’ misguided proposal regarding how it counts, which consists of proposed evaluation measures that would over-simplify CRA exams and render bank activities less responsive to community needs. The letter follows with a discussion of additional reforms NCRC believes would increase CRA activities by making banks more accountable to the public. Finally, the recommendations and conclusion section summarizes a series of NCRC recommendations related to what counts, where it counts, and how it counts.
The Proposed Rule, if finalized, is Contrary to Law, Arbitrary and Capricious, and Enacted Without Proper Observance of Required Procedure
As an initial matter, courts are empowered to “hold unlawful and set aside agency action, findings, and conclusions found to be . . . arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law; . . . [and] without observance of procedure required by law.” The Proposed Rule, if finalized, is therefore subject to vacatur as “contrary to law” under the Administrative Procedure Act if it is inconsistent with the requirements of the CRA or any other law.
Moreover, the Proposed Rule must be set aside as arbitrary and capricious if it fails to include a rational connection between the facts found and the choice made.” It must reflect that the OCC and the FDIC examined the relevant data and articulate[d] a satisfactory explanation for its action.” The rule is subject to reversal if the agency has “relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.”
Where an agency deviates from its existing policy, it “must ordinarily display awareness that it is changing position,” and must “provide a more detailed justification” when its new policy rests upon contradictory factual findings or its prior policy has engendered serious reliance interests. And the rule must be set aside where the agency provides “an explanation for [its] action that is incongruent with what the record reveals about the agency’s priorities and decision making process.”
For the reasons described herein and throughout this comment, the Proposed Rule would be contrary to the text, history, structure, and purpose of the CRA and therefore would be subject to vacatur if finalized. Moreover, the Proposed Rule, if finalized, would likely also be arbitrary and capricious as it is not supported by available data and fails entirely to consider critical aspects of the needs of the communities that the CRA was enacted to serve.
The Proposed Rule’s evaluation measure is inconsistent with CRA
The CRA requires Federal financial supervisory agencies to “assess the institution’s record of meeting the credit needs of its entire community” and “to encourage such institutions to help meet the credit needs of the local communities in which they are chartered” in a CRA examination. Further, the CRA provides that the “regulated financial institutions have continuing and affirmative obligation[s] to help meet the credit needs of the local communities in which they are chartered.” Finally, CRA written evaluations are required to “state the appropriate Federal financial supervisory agency’s conclusions for each assessment factor identified in the regulations prescribed by the Federal financial supervisory agencies to implement this chapter [and] discuss the facts and data supporting such conclusions.”
The proposal seeks to establish an “objective” method to measure CRA performance review. At the core of this review is the evaluation measure, which is inconsistent with the CRA’s requirements for evaluation.
The proposal is contrary to the statutory requirement that an evaluation shall assess a bank’s record of meeting the needs of the “entire community.”
Rather than evaluating a bank based on its record of meeting the needs of the “entire community,” the proposed CRA evaluation measure would establish a “presumptive rating” based on serving only a subset of the community. To receive a bank level rating of Satisfactory or Outstanding, a bank would only need to achieve a Satisfactory or Outstanding rating in a “significant portion” — defined as more than 50 percent — of its assessment areas. Rather than accounting for the bank’s entire performance across its entire community, as the CRA requires, the proposal would award a passing CRA rating even if a bank receives a failing grade in up to half of its assessment areas. Adopting the proposal would contradict the requirement that the agencies consider performance in the “entire community.”
Indeed, during the passage of the CRA, Congress expressly considered and rejected legislative text that would privilege only a portion of a bank’s footprint. The first draft of the CRA legislation contained specific requirements to serve primary service areas. The primary service area was defined as a compact area covering branches from which a bank expected to obtain more than one-half of its deposit customers. Then, when a bank applied to merge or open a branch, it would have been required to indicate what proportion of customer deposits would be reinvested in primary service areas. Some Senators and bank representatives criticized this provision as cumbersome and bureaucratic. They also stated that it would prevent banks from addressing other unmet needs in other areas including rural communities (that presumably would have lower levels of deposits). In response, Senator Proxmire, who spearheaded the legislation, re-worked the legislation and introduced it without this provision. Instead, agencies were required to examine banks and rate them based on how well they met credit needs of their entire communities.
The proposed evaluation measure is not tailored to the “credit needs of the local communities”
The Proposed Rule seeks to establish a primarily dollar-based ratio metric to determine a CRA rating – an approach that Congress rejected when it enacted CRA, and for good reason. The proposed evaluation measure sets specific ratios of CRA activity to deposits that would correspond to CRA ratings at either the bank or assessment area level. The shift to dollar value would not account for local needs, such as small-dollar home and small business lending necessary in many communities, and would instead motivate banks to focus on large-scale financing, which is easier to accomplish with fewer individual investments or loans. Establishing a one-size-fits-all dollar-based metric would override the credit needs of local communities contrary to the CRA and would not account for the individual and unique “credit needs of the local communities in which [banks] are chartered.”
During the hearings preceding passage of the CRA in 1977, bank and community witnesses spoke against a ratio-based framework to evaluate and assess a bank’s overall performance because it would not be reflective of market conditions or representative of needs of local communities. Witnesses stated that a ratio for a bank’s primary service area could not accurately assess whether banks were meeting needs in other areas, including rural areas. Moreover, the ratio could not take into account shifts in demand for loans. Instead of the ratio, witnesses made recommendations to expand the use of multiple measures including reviewing the types of loans made and whether borrowers inside and outside bank service areas were being served so that the governing federal banking regulatory agency could decide if credit needs have been met.
The proposed approach was again rejected during the passage of the CRA when Senator Proxmire and congressional witnesses opposed such a regime for just these reasons. Senator Proxmire expressly stated in his opening statement at the Community Credit Needs Hearing on March 23, 1977, that the CRA “does not provide for credit allocation. To criticize reinvestment incentives as a form of credit allocation is disingenuous. It would not allocate credit, nor would it require any fixed ratio of deposits to loans. But it would provide that a bank charter is indeed a franchise to serve local convenience and needs, including credit needs.”
Until the present proposal, the agencies have acted consistently with Congress’s clear intent in rejecting a ratio-based approach. Accordingly, when the OCC and FDIC revised their CRA regulations in 1995, they rejected a strict loan-to-deposit ratio test for small banks in favor of multiple metrics to rate an institution’s qualifying activity level. The Proposed Rule effectively implements an evaluation method that the CRA itself rejects.
A presumptive, dollar-based evaluation focused on a portion of assessment areas would rewrite the CRA’s statutorily required written evaluation procedure
Reducing a CRA examination to a dollar-based and presumptive evaluation measure that turns on only half of the assessment areas would contradict longstanding agency interpretation of and practice under the CRA’s written evaluations requirement. The CRA requires each written evaluation to “state the appropriate Federal financial supervisory agency’s conclusions for each assessment factor identified in the regulations prescribed by the Federal financial supervisory agencies to implement this chapter [and] discuss the facts and data supporting such conclusions.” The longstanding interpretation of these required procedures, embodied by agency practice, follows the holistic character of evaluation as intended under the CRA.
The proposal undermines the requirement that evaluations “state . . . conclusions for each assessment factor [and] discuss the facts and data supporting such conclusions.” The statute places no restrictions on an examiner’s ability to consider facts and data supporting her conclusions other than that the conclusions be linked to regulatory assessments factors. This approach has provided examiners with the flexibility to capture market trends or shifts within any community. In contrast, the proposed evaluation measure would create a uniform, restricted, and narrow set of criteria across banks, excluding and devaluing facts and data that have appropriately carried greater weight when examiners state their conclusions for each assessment factor and ultimately decide upon final ratings. This new approach would be contrary to the holistic consideration required by statute, which has been endorsed by the agencies for decades, and which is necessary to effectuate the CRA’s objective.
The agencies should make available for public comment the data and analysis that are the basis for the proposal and extend the comment period
The Administrative Procedure Act requires that the agencies “give interested persons an opportunity to participate in the rule making” and justify its decision with reference to the “whole record.” These requirements include providing the public a meaningful opportunity to comment upon the whole set of data and analysis that serve as the basis for the proposed regulations. As detailed below, the agencies have failed on numerous occasions to provide the data and analysis upon which they relied. The agencies should make the data public and extend the comment period to provide a sufficient opportunity for public participation, particularly given the complexity and importance of the proposed changes to fundamental aspects of the CRA regime.
The agencies fail to explain how they arrived at the proposed evaluation measure benchmarks and other important thresholds
Even if the evaluation measure were consistent with the CRA, the agencies have not offered evidence that the proposed CRA evaluation measure would continue to protect the interests of LMI neighborhoods at the current level. In fact, the agencies developed empirical benchmarks for the CRA evaluation measure that would correspond to various ratings without showing via data analysis whether or how these benchmarks would increase CRA-related lending, investing, and services and would thus benefit LMI communities and individuals.
The NPRM also proposes numerical thresholds for the retail lending distribution test and for a minimum community development finance level that is simply described without any justification via data analysis. The public has no way to determine or render judgments regarding whether these proposed thresholds would result in more loans and investments for LMI households and communities.
The agencies used various data sources to derive their empirical benchmarks for their proposed CRA evaluation measure, the thresholds for the proposed retail lending test, and the proposed assessment area procedures. However, the agencies suggested that they did not have full confidence in the current data and that they would issue a request for information (RFI) that would ask banks to voluntarily submit more data that the agencies would use in refining their empirical benchmarks, thresholds, and assessment area procedures before issuing their final rule.
Data requested by the agencies from banks has not been disclosed to the public and therefore frustrates the public’s ability to evaluate the Proposed Rule
On January 10, the OCC issued an RFI that asked how the proposed rule should be revised “to ensure that the final rule better achieves the statute’s purpose in encouraging banks to help serve their communities.” The RFI asked for a considerable amount of data that would pertain to the formulation of the final rule including deposits and data on CRA qualifying activities such as retail lending and community development financing. The due date for banks to submit this data is March 10, one day after the original deadline for public comments on the NPRM. The data from the RFI has still not been made publicly available.
Thus, the agencies now have the data requested by the RFI to refine their proposal but the public lacks access to the data when responding to the NPRM by the due date of April 8. The public is not able to engage in their own analysis of the data submitted in response to the RFI and does not know how the agencies used the data to alter their proposal.
In addition, it seems clear that the agencies did not understand the impacts of their proposal when they issued the NPRM. Writing about the impact on smaller banks, the FDIC stated:
If the proposed general performance standards are more stringent for some institutions than the current parameters, the proposed rule could pose costs for covered institutions by potentially reducing their CRA examination rating. If the proposed general performance standards are less stringent for some institutions than the current parameters, the proposed rule could benefit covered institutions by potentially increasing their CRA examination rating. It is difficult to accurately quantify these aspects of the proposed rule with the information currently available [bold added] to the FDIC.
This astonishing admission indicates that the agencies do not have a clear sense of whether their proposal would further inflate ratings, thereby leading to stagnant or decreasing levels of loans and investments for LMI communities. Recently, the FDIC inspector general found in a report that the FDIC has not engaged in sufficient cost-benefit analysis of proposed rules. These admonitions would seem to apply to the NPRM that does not sufficiently describe costs, such as possible reductions in bank CRA activity.
The comment period is insufficient to provide the public an opportunity to participate
Because of the lack of public data, the comment period has deprived the public of “an opportunity to participate in the rule making.” This rulemaking therefore violates the objectives of the APA to inform fully federal agencies of the impacts of proposed rules by providing the public with meaningful opportunities to comment. A recent Congressional Research Service report on the APA states that the APA requires a “meaningful opportunity for public comment.” The report states, “although the APA sets the minimum degree of public participation the agency must permit, the legislative history of the APA suggests that matters of great importance, or those where the public submission of facts will be either useful to the agency or a protection to the public, should naturally be accorded more elaborate public procedures.”
The agencies’ changes are complex, transformative, and not supported by a transparent analysis of data for estimating impacts of the changes on LMI families and communities. The comment period should be extended for a period of time after FDIC and OCC release the underlying data in order to permit meaningful public consideration and input, as the APA requires. Alternatively, if the agencies believe that the data gathered by the RFI necessitates changes in their proposal, they need to issue a new NPRM with their changes supported by transparent data analysis.
OCC is interfering in the rulemaking process in a manner designed to influence public comments.
The OCC has undertaken an aggressive campaign during the public comment period to influence public comments. Elements of the campaign include calling community-based organizations and banks, publishing fact sheets, and OCC senior staff improperly appearing in marketing videos produced by stakeholders.
In mid-January, NCRC member organizations informed NCRC that OCC staff had contacted organizations that indicated a support for a NCRC letter asking the FDIC board not to join the OCC in the current rulemaking. Some 500 organizations signed onto this letter. NCRC staff received more than 30 emails from concerned NCRC member organizations. These emails included repeated requests by OCC staff for discussions over the phone and asked the community organizations to respond in a few days. The emails also touted the benefits of the NPRM to the mission of the community organizations such as increasing minority homeownership.
This type of correspondence is unusual and uncomfortable for a number of community-based organizations that are not accustomed to the regulatory comment process. Since our inception in 1990, NCRC has not encountered regulatory staff contacting organizations that had indicated opposition to a proposed rule change in a systematic manner during an open rulemaking in an effort to influence comments. Instead, the customary practice is for regulatory staff to respond to technical questions from the public in an impartial and informational manner.
We also note that the OCC has been contacting banks and seeking to influence their views of the NPRM. This concerted activity has been documented in news reports.
The OCC has also been circulating fact sheets about the proposal that depict it in a favorable light, using descriptions like “clarifying what counts” and “measuring CRA performance more objectively.” Towards the end of the document emailed to stakeholders, a matrix appears with symbols like question marks depicting the current rules and check marks depicting the proposal that conveys a sense of increased clarity and contentedness as a result of the proposal. In addition, the Comptroller and senior OCC staff appear with a stakeholder in a video in which “misperceptions” of the proposed rule are discussed.
What Counts – Agencies are Proposing to Dramatically Shift Focus away from LMI Communities
The CRA permits the agencies to promulgate “[r]egulations to carry out the purposes of” the Community Reinvestment Act, that is, to combat redlining and systematic discrimination against LMI communities and communities of color. Indeed, it is the congressionally stated “purpose of this chapter to require each appropriate Federal financial supervisory agency to use its authority when examining financial institutions, to encourage such institutions to help meet the credit needs of the local communities.” Similarly, the agencies are required to “assess the institution’s record of meeting the credit needs of its entire community, including low- and moderate-income neighborhoods.”
Because of decades of public and private sector discrimination, lending markets in many LMI communities were not functioning. Banks were not familiar with the creditworthiness of borrowers, the quality of the housing stock, or the viability of small businesses in these communities. To overcome the market failures caused by discrimination, CRA imposed an affirmative obligation on all banks to serve LMI communities. By requiring banks to seek out business in LMI communities, CRA sought to break down barriers to information that contributed to a scarcity of lending. All banks were required to learn about and collect information about LMI communities so that they could make safe and sound loans. As a result of this affirmative obligation, Federal Reserve research has concluded that CRA increased home lending by about 20 percent in LMI census tracts and also increased small business lending. As explained in detail below, the proposed regulations deviate from this purpose because they provide CRA credit for activities that do not meet the “needs” of the local communities which are the subject of the CRA: LMI communities and communities of color.
Agency and industry proposals to diminish CRA examination and focus on LMI communities and people would dilute CRA’s effect and undermine its purpose
The proposed rule would contravene the CRA’s text and purpose by crediting non-LMI activities or general community building activities that benefit middle- and even upper-income communities, but may have some partial or theoretical benefit for LMI communities. This change would be inconsistent with the purpose of the Act and the economic rationales underpinning it.
Clearly, redlining motivated CRA’s passage, including the concern about high capital export from local communities where banks were taking deposits. The law provided incentives for banks to seek lending and investing opportunities in their local markets and to overcome the “market failures” that limited lending and investing in LMI areas. Banks were bypassing profitable local lending opportunities in favor of far-off investments, to the detriment of local housing, small business, and small farm credit needs.
In the hearings on CRA before passage of the law in 1977, Senator Proxmire cited an elaborate series of articles in New York-area newspapers on the amount of disinvestment in the City, pointing out that only 11 percent of the money deposited in Brooklyn remained in the borough. In the District of Columbia, researchers found that about 90 percent of the deposits were loaned and invested outside of the community. Chicago, Los Angeles, and St. Louis also experienced a high degree of deposit flight.
A 2000 baseline report on CRA by the U.S. Treasury Department captured Congress’s thinking at the time of CRA’s enactment regarding capital export and local revitalization efforts:
Congress intended the CRA to increase credit access and revitalize inner-city communities that were deteriorating with the movement of investment and development elsewhere. In addition, Congress recognized that the success of federal community development, housing assistance and mortgage insurance programs enacted at the same time as the CRA . . . would depend in large part upon the availability of private capital, particularly as made available through local financial institutions.
Congress designed CRA to strengthen these public economic development efforts by facilitating “efforts between private investments and federal grants and insurance in order to increase the viability of our urban communities.”
CRA’s regulatory framework must be focused on LMI borrowers and communities to correct for market failures and externalities
Critically, the law requires regulators to examine whether banks are overcoming market failures and informational externalities associated with the lack of investment in LMI communities. The U.S. Treasury Department explained the positive and negative externalities that existed when banks, in deciding where to lend and invest, did not bear the full costs nor reap the full benefits of their actions. As Treasury stated:
There are significant positive externalities, for example, associated with lending in areas where there are frequent numbers of transactions, such as middle-class or relatively affluent neighborhoods. These transactions produce a steady stream of information about market values for other lenders (and appraisers) to consider when making their credit decisions. A larger number of transactions increases confidence in the appraised value of individual properties, and increases the liquidity of other investments in the neighborhood, thus improving the values of properties that serve as collateral for mortgages. This process lowers lenders’ transactions costs, thereby lowering the cost of credit for all borrowers in the area.
The reverse is true for neighborhoods where there are relatively few transactions. In particular, to the extent that lenders do conduct appraisals in LMI neighborhoods, these appraisals can be more costly and less accurate because “comparable” transactions and appraisers familiar with such neighborhoods are not available. Loans in these areas are therefore riskier, and lenders will compensate by charging higher rates of interest or requiring larger down payments. The stiffer terms on such loans can cause some borrowers either to borrow less or to drop out of the market altogether. For LMI neighborhoods, the end result can be a downward spiral – less lending, fewer appraisals, even less lending, and so forth – producing an effect that resembles redlining.
Being among the first institutions to enter a new or previously underserved market or investing in an innovative but high-impact LMI project, when other lenders are unwilling to lend, requires overcoming negative externalities and lack of information about borrowers and neighborhood characteristics. This could result in delays, caution about perceived risk, and banks charging higher interest rates. Lender expectations of this sort could cause a potentially viable market to suffocate from lack of credit. In the process, borrowers who may otherwise be credit-worthy would be denied credit because of the absence of entry by competitive lenders.
Therefore, the CRA can be understood as a vehicle for facilitating coordination and for assuring banks that they will not be the lone participants in thinly-traded markets . . . the Act can produce positive information externalities and allow all lenders – both those covered by the CRA and those not covered by the CRA – to better assess and price for risk.
Congress recognized these market failures and externalities. Accordingly, CRA requires regulators to examine the data and assess whether banks were overcoming these failures and negative externalities in a “continuing and affirmative” way. No one bank would be the lone pioneer; CRA required all of them to participate in LMI markets, thereby increasing the flow of information and transactions.
Due to CRA, banks have made good strides in LMI markets. They have taken numerous steps, including establishing loan products geared towards LMI borrowers, entering loan pooling arrangements, undertaking lending consortiums, and partnering with local groups, community development corporations, and community development financial institutions (CDFIs) to break down the barriers that impede the efficient flow of capital into LMI communities.
The statutory design of CRA and the existing regulatory framework to date have executed the law in a way to overcome redlining and other market failures. Until this NPRM, the agencies were consistent in keeping a focus on LMI communities and people in order to continue the process of breaking down information barriers and eliminating negative externalities. The agencies emphasized a LMI focus in the 1995 final rulemaking and subsequently in the Interagency Q&A document.
The preamble to the 1995 final rule states:
Under the rule, community development includes activities outside of low- and moderate-income areas if the activities provide affordable housing for, or community services targeted to low- or moderate-income individuals or if they promote economic development by financing small businesses and farms.
The final rule also requires that, in order to be community development loans or services or qualified investments, activities must have community development as their primary purpose. Activities not designed for the express purpose of revitalizing or stabilizing low- or moderate-income areas, providing affordable housing for, or community services targeted to, low- or moderate-income persons, or promoting economic development…are not eligible.
It is clear from the 1995 rule that either LMI communities or LMI people in non-LMI communities were to benefit from CRA activities. The focus was on LMI people and communities.
Developed and refined after the 1995 final rule, the Interagency Q&A provided careful exceptions to the focus on LMI people, such as including mixed-income housing as an example of community development. This was as it should be; exceptions needed to be narrowly targeted to promote positive outcomes like economic mobility that might provide residents with LMI affordable housing in high-opportunity areas. Broad-based exceptions would result in LMI communities being neglected in favor of easier-to-serve affluent communities.
A second example in the Interagency Q&A was a supermarket on the edge of a middle-income census tract but adjacent to an LMI tract. In this case, the Q&A stated that the supermarket development helped stabilize the LMI tract by providing “needed shopping services that are otherwise unavailable in the low-income community.” Again, the point was that it was acceptable for CRA activities to not exclusively benefit LMI areas but activities must have a significant and measurable benefit for LMI areas.
Thus, the overall focus was plainly on LMI people and communities. Additional Q&As in the Interagency Q&A document suggested that generally the substantial majority of recipients of community development activities should be LMI persons (the language varies between “majority” and “primarily” recipients being LMI).
The 1977 CRA hearings and the insertion of LMI neighborhoods in the statute suggested a focus and a priority of LMI people and neighborhoods, which had the greatest unmet needs due to redlining and other factors. If the OCC and FDIC adopt a final rule that fails to recognize that serving LMI people and communities must be the primary objective of CRA activities, it would contravene the stated purpose of the statute and 40 years of agency interpretations. Moreover, it would halt or reverse the progress made in overcoming market failures in LMI communities, which depends on continued bank focus on lending and investing in LMI communities.
Question1. Are the proposed criteria for determining which activities would qualify for credit under the CRA sufficiently clear and consistent with the CRA’s objective of encouraging banks to conduct CRA activities in the communities they serve?
Answer to question 1: proposed list of criteria for counting activities under CRA is not clear and would divert activities away from LMI communities
The FDIC and OCC’s proposal would diminish if not halt the progress of revitalizing LMI communities and their lending markets by allowing banks to turn their attention away from LMI communities. In addition to retail lending (home and small business lending), LMI communities need community development financing which support affordable housing, job creation, small business growth, facilities like child care and health clinics, and larger scale projects like the development of commercial corridors.
Home loans or small business loans would not succeed in creating an economically thriving and vibrant community if the community lacked community development financing necessary for creating and refurbishing community facilities and assets. Neighborhood residents would move out of communities that lack community facilities and access to commercial corridors. If the agencies reduced the focus of community development on LMI communities, the regeneration of lending and housing markets would be halted and market failure would re-emerge.
The current regulatory definition of community development (CD) includes financing affordable housing for LMI households, economic development focused on small businesses under $1 million in revenue, community facilities, and the revitalization and stabilization of LMI communities. The OCC and FDIC would delete the criteria of economic development and revitalization and stabilization from the definition of community development in proposed § 25.04(c). In its place, the agencies would provide more detail on community development criteria, including examples of economic development and revitalization in the list of eligible activities.
This proposed movement of CD activities from the regulation to a list of activities would diminish the importance of the activities. An activity that is in a regulatory definition has more weight on CRA exams than an activity described in a list that is not part of the regulation itself. Bank compliance with CRA is focused on responding to criteria enumerated in the regulation as opposed to an eligible activity included as an example in a list. For example, the regulation currently makes it a high priority to finance activities that revitalize or stabilize LMI neighborhoods. In order to ensure that banks respond to the CD criterion of revitalize and stabilize, CRA exam tables include data with the dollar amount and number of loans and investments that revitalize and stabilize LMI neighborhoods. In contrast, a list suggests a diminished focus on these activities since banks and other stakeholders would find only a few scattered examples in the proposed list of revitalization activities, such as donating bank-owned real estate to a local government-owned land bank.
In practice, this would mean that banks would reduce the financing of certain infrastructure associated with economic development, such as business incubators that target start-up businesses. Also, banks would likely decrease funding for workforce development programs that qualify under the criterion of economic development. Before this proposal, the agencies had distributed guidebooks and webinars promoting the importance of workforce development as bolstering skills for vulnerable populations such as young adults without college educations or people with disabilities. The proposed list of qualified activities would not be a suitable substitute for workforce development because it has just one mention of workforce development that focusses more on financial counseling than workforce development. This sparse attention exacerbates the delegation of workforce development from a regulatory definition to the list of qualified activities.
The change in the regulatory definition of CD would also result in banks reducing the financing of revitalization and stabilization activities, such as reclaiming abandoned housing or instituting foreclosure prevention programs. These activities, which target LMI communities, are vital to the continued development of vibrant housing and lending markets in LMI communities.
In addition to deleting economic development and revitalization/stabilization from the regulation, the agencies would introduce CD criteria in the regulation that would explicitly dilute the focus on LMI communities. For example, a new criterion would be essential infrastructure in proposed § 25.04(c)(6). This refers to major projects like roads and bridges. The difficulty is that these projects would not necessarily be located in LMI communities. Moreover, by adding a criterion of essential infrastructure to the regulation and removing criteria that were focused on supporting smaller businesses and revitalizing LMI communities, the agencies would encourage banks to gravitate to financing large-scale infrastructure rather than targeted revitalization initiatives most needed by LMI communities.
The NPRM discusses a public hospital as a project eligible for CRA consideration under the community development definition. The preamble to the NPRM, however, did not discuss whether such a hospital needs to be in an LMI community or within easy access of it. If the definition of community development encourages banks to favor financing of infrastructure that is needed by a city as a whole (reflecting an overbroad definition of community incompatible with the statutory objective) instead of infrastructure targeted to LMI communities, then the revitalization of LMI communities financed by CRA would slow down. Indeed, the NPRM explicitly states that, “The addition (of infrastructure) also would recognize that essential infrastructure projects are often community-wide projects for which it is not feasible to allocate the benefit to specific populations or areas.”
In addition, the FDIC and OCC would include community development financing that “partially” instead of “primarily” benefits LMI households as eligible for CRA consideration in proposed § 25.04(c). A pro rata procedure would apply. The language “primary purpose” referring to a majority of activity dollars benefiting LMI people or communities has been part of the Interagency Q&A document and other regulatory documents for several years. If the agencies remove the “primary purpose or benefits” standard and replace it with a “partially benefits” standard with an inadequately-defined pro rata procedure that could generously provide credit for projects that may have marginal or theoretical benefits for LMI people, then the number and dollar amount of CRA loans for LMI families and communities would likely decline.
In cases in which a pro rata procedure could be verified with data and clearly described benefits for LMI households, the pro rata procedure would be workable. For example, if a mixed income housing development had 40% of the units reserved for LMI households, the total dollar amount would be multiplied by 40% to determine the dollar amount considered on CRA exams. CRA examiners currently apply pro rata procedures in circumstances like this. Nevertheless, the agencies need to clarify that a pro rata procedure would continue to be data-driven and applied in these circumstances in this manner. The list of qualifying activities introduces confusion since it lists affordable housing with various percentages of the units being affordable (by formula) for LMI households, and does not indicate that a pro rata procedure based on occupants or beneficiaries would be applied.
The introduction of large-scale infrastructure would also make the pro rata procedure less reliable. When a pro rata percentage is applied, how would the agencies calculate a pro rata share for LMI populations of a major road or bridge used by millions of people? The result of back-of-the envelope guesses of pro rata share in these cases would likely over-estimate LMI benefit.
For example, in the Washington DC area, newspapers recently covered proposals to widen major highways and bridges, including the American Legion Bridge. This infrastructure would provide a diffuse benefit to LMI people since it would not directly serve LMI neighborhoods or promote the economic development of LMI neighborhoods. Without any reliable method for estimating the benefit to LMI neighborhoods and people, the agencies must not allow such general infrastructure projects to be considered “essential” nor attempt to apply a pro rata procedure to this type of large-scale infrastructure.
Any allowance for infrastructure investments should be limited to projects that have a discrete and quantifiable benefit to LMI tracts and/or people. For example, a transit line such as light rail that travels through ten census tracts and serves four LMI tracts with multiple stations, would clearly have a benefit for LMI tracts. In such a case, a pro rata procedure could reasonably apply 40% to the dollar amount of a bank construction loan for the project.
Credit for large scale infrastructure conjures up memories of “urban renewal” from the 1960s that focused on bridges and highways that destroyed neighborhoods. In Washington, DC, a proposal to build the Three Sisters Bridge would have demolished homes, displaced thousands of residents and decimated communities of color. Community protests accompanied by a lawsuit halted this project in the early 1970s. The agencies did not address how their “essential” infrastructure proposal would not repeat this history. In this context, replacing the CD criterion of revitalization and stabilization of neighborhoods with essential infrastructure is ominous and sadly ironic.
Proposed credit for sports stadiums in Opportunity Zones must not be allowed
In addition, the agencies must not relax the regulatory definition of community development in the case of Opportunity Zones, even if the census tracts are low-income. The Opportunity Zone program lacks documentation or data to specify who benefits from the financing. Thus, if CRA financing was not constrained to meet the current definition of community development, such as affordable housing, the financing could be for luxury condominiums.
The NPRM does not appear to include these safeguards. In fact, the NPRM would give credit for financing improvements to “athletic stadiums” in LMI tracts. Recent articles document that more than twelve football stadiums are located in Opportunity Zones. Would CRA examiners provide CRA credit for financing improvements to “jumbotrons” or the huge electronic scoreboards in the center of the football field? Moreover, scholarly research has demonstrated that athletic stadiums generally have little impact in terms of economic development of neighborhoods unless they are used for at least 250 sporting contests and other events annually, which is a level much higher than for most stadiums. Stadiums have also led to gentrification and displacement. Given the expressed statutory purpose of the CRA, it is difficult to understand the basis on which the agencies would propose to eliminate consideration of mortgage loans in LMI census tracts but credit stadium improvements in LMI census tracts.
The stadium proposal could become a major CRA resource drain and would divert resources from projects responding to more pressing needs in LMI communities. NCRC’s research found 37 stadiums that were either proposed or under construction. Of these, 15 or 40% were in Opportunity Zones and LMI tracts and would qualify under the proposal. An additional 10 or 27% were in LMI tracts and two or 5.4% were in Opportunity Zones. It is likely that these 12 or an additional 32% would qualify for CRA credit. All told, more than 70% of the new stadiums would appear to be creditable. This would gobble up CRA resources.
This assessment underestimates the magnitude of the resource hog, because the stadium proposal is not confined to professional teams but also includes collegiate teams. Also, the definition of stadiums could expand in the future to include a variety of athletic facilities. If so, on what basis would LMI benefit be determined – the income of the students using the facilities or the income of their parents?
The City of Jacksonville presents a useful illustration of the potential misuse of CRA resources. The City borrowed $45 million to pay for upgrades (a new outdoor amphitheater and indoor practice facility) next to the stadium in which the Jacksonville Jaguars play. One of the two largest banks in terms of deposit market share in Jacksonville had community development lending ($15 million) on a recent CRA exam that was of a lower dollar amount than the loan for the improvements next to the stadium. The community development lending of this bank supported 72 units of affordable housing and economic development. Another bank, which was also one of the two largest banks in terms of deposit market share, made $21 million in community development lending that supported over 200 units of affordable housing.
Thus, if the NPRM allows financing for stadiums, banks would have an incentive to replace important community development lending directly benefiting LMI neighborhoods with stadium financing conferring uncertain benefits to LMI communities. Put another way, the proposal would allow banks to get just as much CRA credit by financing a single stadium project as by providing millions of dollars of housing and economic development loans, despite the obvious mismatch between the former and the goals of the CRA. The agency has not estimated the effects of such an option, and it is difficult to see how any analysis could support it.
Recently, the Comptroller of the Currency testified before Congress and asserted that banks have received CRA credit for financing stadiums for decades. However, the testimony provided examples involving community park facilities or facilities that also benefited high schools with predominantly LMI students or students of color. This is a far cry from the broad-based proposal to include financing large-scale stadiums in Opportunity Zones as CRA eligible activities.
Answer to question 1 continued: the agencies’ proposed criteria will divert attention away from LMI people
In addition to shifting focus away from LMI communities, the agencies proposed criteria for determining CRA activity would shift the focus away from LMI people. In proposed §25.04(c)(1)(D) or (E), the agencies would qualify housing as affordable if it provides housing for middle-income people in high-cost areas.
The NPRM could facilitate mixed income housing occupied by middle- and upper-income housing. Current CRA guidelines in the Interagency Question and Answer (Q&A) provide partial CRA credit to mixed income housing that is LMI and middle- or upper-income. This Q&A stated that the pro rata dollar amount of the total activity would be based on the percentage of units set-aside for affordable housing for LMI individuals. 
In contrast, the proposed list of CRA qualified activities described affordable housing to include housing “that partially or primarily benefits middle-income individuals or families in high-cost areas as demonstrated by an affordable housing set-aside required by a federal, state, local, or tribal government.” Also, affordable housing would include “an investment in a project in a high cost area where 30 percent of the rental units are set aside as affordable to middle-income individuals through local inclusionary zoning.” These examples mentioned a portion of housing reserved for middle-income households but none for LMI households.
Further, the NPRM mentions that a CRA community development activity is “affordable housing that partially or primarily benefits LMI individuals or families or middle-income individuals or families in high-cost areas.” The use of “or” suggests that affordable housing in high cost areas could conceivably be entirely middle-income.
The NPRM sprinkles examples about mixed income and middle-income housing throughout the text. This scattered discussion does not clarify important issues such as whether mixed-income housing that is entirely middle- and upper-income could qualify as affordable housing for CRA purposes. Moreover, the NPRM does not discuss whether the agencies would implement an overall limit on the amount of middle-income housing financed by a bank or banks that could count as affordable housing nor does it allow CRA exams to provide more points to a bank that focusses largely on LMI affordable housing.
The NPRM discusses the needs of teachers and police officers, who certainly have difficulties affording housing near where they work in high-cost jurisdictions. However, the NPRM does not appreciate the need for an approach that would recognize that LMI households would still have the highest housing cost burdens and that CRA financing must primarily serve them. Since LMI housing is more difficult to finance and often requires deeper subsidies, the middle-income housing provision, if implemented, would likely result in less LMI affordable housing.
The NPRM offered a definition of a high-cost county but did not describe the impact of the definition. The NPRM defines a high-cost county where 40% of households pay 30% or more of their monthly income for housing. NCRC decided to conduct research that the agencies did not. According to NCRC findings, the proposed change would affect 32 counties across the U.S., including some of the most populous ones, encompassing 43 million people or 13% of the United States population. The impacted places are cities like Los Angeles, San Diego, Miami and most of metropolitan New York. What would be the incentive for banks to provide CD loans or investments to build rental units for LMI families in these areas, when they could finance more profitable rental housing for middle-income families and get the same credit?
NCRC also found that some counties like Humboldt, Madera and Tulare in California had more than 40% of households experiencing cost burden not because of high housing costs but because of high poverty rates. The agencies would be diverting affordable housing towards middle-income households in these counties not because the middle-income households confront disproportionate cost burdens, but because LMI households confront disproportionate levels of high housing cost burdens. This would be an inappropriate outcome.
The result of this proposal could be hundreds of thousands of units financed for middle-income households and nominal amounts financed for LMI households. The agencies would be allowing banks to not respond to the most pressing housing needs in wide swaths of the country in violation of the statute’s intent.
Also, the agencies would allow naturally occurring affordable housing (NOAH) to be considered as affordable housing. NOAH is described in the proposal as rental housing in which households would have to pay no more than 30% of 80% of area median income on monthly rent. The agencies need to consider how to make sure LMI households actually occupy the units. The NPRM does not discuss any procedures for verifying that LMI households would be the occupants (For example, some stakeholders have proposed that borrowers of multifamily loans commit to reserving some or all units for LMI occupants). Thus, the public cannot comment on the adequacy of any proposed procedures. By casually proposing that NOAH be allowed, the agencies seem to be focused on making it easier for banks to qualify activities for CRA credit rather than ensuring that housing financed by CRA actually benefits LMI people.
In addition to shifting affordable housing away from LMI households, the agencies proposed to eliminate income restrictions for financial education programs. The agencies would be once again ignoring the purpose and intent of CRA to focus on historically redlined communities and the impact of this discrimination on generations of LMI people. As documented by the FDIC, LMI households were the most likely to be under- or un-banked. They had the lowest homeownership rates and the lowest level of assets. Therefore, financial education must be targeted towards LMI households in order to correct for the impacts of discrimination and less access to banking for this population. Financial education must be targeted for LMI households in order to most effectively promote healthy lending and housing markets in LMI communities with educated consumers. Diverting limited resources away from financial education of LMI people is not justified and counterproductive.
Question 2. Are there other criteria for determining which activities would qualify for CRA credit that the agencies should consider?
Answer to Question 2: agencies must reinstate LMI tracts as a criterion on the lending test
Currently, CRA exams assess home lending to LMI borrowers and LMI tracts separately. The proposal would retain an examination of home lending to LMI borrowers but would delete an examination of lending in LMI tracts. The agencies are concerned that examining home lending in LMI tracts would encourage banks to make loans disproportionately to non-LMI borrowers, which would result in gentrification and displacement of LMI residents.
Proposed § 25.04(b)(1) would remove home mortgage lending in LMI tracts as an activity eligible for receiving credit on CRA exams. This change is a counterproductive reaction to possible displacement associated with gentrification in LMI communities, and must not be enacted. Overheated markets in gentrifying communities are a significant barrier to affordable housing; however, NCRC has documented that gentrification is not a nationwide phenomenon. Rather, it is mostly confined to large coastal metropolitan areas. The predominant problem LMI communities faced is economic distress and poverty, not gentrification.
By excluding home lending in LMI communities on CRA exams, the agencies would prolong economic distress in LMI communities. Banks would have less incentive to make home loans in LMI communities. Public sector and nonprofit initiatives to revitalize struggling communities by producing affordable housing would experience more difficulties in succeeding because banks could be more reluctant to make home loans to LMI borrowers in LMI communities. Reduced bank lending in LMI communities would reintroduce negative externalities associated with lack of information about borrower and neighborhood characteristics discussed previously.
Although the criterion of retail lending to LMI households would remain, banks could respond by focusing their lending to LMI borrowers in non-LMI neighborhoods with more developed lending markets, in contradiction to the emphasis discussed above of CRA reviving markets in underserved LMI communities. As stated above, Federal Reserve research revealed that CRA had increased home lending in LMI communities by up to 20%. Other Federal Reserve sponsored research has found a similar impact in small business lending in LMI communities. By abruptly removing home lending in LMI communities as a criterion on the retail lending test, the FDIC’s and OCC’s proposal could significantly decrease, if not wipe out, this CRA impact on retail lending in LMI tracts.
Although small business lending in LMI tracts would still remain in the retail test, it might become harder for banks to make small business loans in LMI tracts because the housing market could become depressed due to fewer home loans to LMI borrowers in LMI tracts. A less viable housing market would likely depress overall economic development, including small business development and expansion.
If the agencies’ goal is to improve CRA regulations to ensure that CRA lending is not diverted to gentrifying borrowers, the agencies should adopt a different approach. In LMI communities that are gentrifying, the agencies could limit how they consider retail lending. They could put a cap on the number and percentages of home loans made to non-LMI borrowers. For example, the agencies could allow no more than half the industry average percentage of loans to non-LMI borrowers to count. For example, if the industry average in LMI gentrifying tracts was 70% of their home loans to non-LMI borrowers in an assessment area, the agencies could allow banks to claim CRA credit for no more than 35% of their loans in these tracts to non-LMI borrowers. Alternatively, the agencies could not allow any loans to non-LMI borrowers in gentrifying tracts to count. Still another approach would be not to count a loan in a LMI tract if it finances a home that has a value greater than the AA or metropolitan average. NCRC, as well as other researchers, has developed a methodology for identifying LMI tracts that are gentrifying, which the agencies could use or adapt when deciding how to consider non-LMI lending in gentrifying tracts.
Disallowing home loans in gentrifying LMI census tracts to count at all thwarts the CRA’s key objectives of encouraging economic integration. Income segregation and concentrated poverty was increasing in the United States, which had perverse outcomes in terms of access to jobs, transportation, and quality housing for LMI households. Raj Chetty had also documented that children from LMI families succeed to a greater extent in school when they live in integrated communities. If the agencies disallow home lending in LMI communities from counting on CRA exams, they would reduce the incentives for banks to make loans in LMI communities and would likely make it harder for LMI borrowers to obtain home loans in gentrifying communities, including in those communities in which the public or nonprofit sectors are seeking to preserve affordable housing for LMI households. In other words, the agencies might hasten the displacement of LMI households from gentrifying communities by making it harder for them to obtain home loans so they can remain in these communities.
In addition, the proposal would discontinue CRA exam analysis of multifamily lending in LMI tracts. This is counterproductive because CRA exams must ensure that banks are financing affordable and decent multifamily housing in LMI tracts so that LMI households have a variety of housing choices to fit their needs.
Question 3. Under the proposal, CD activities conducted in targeted areas, such as Indian country or distressed areas, would qualify for CRA credit. Should there be any additional criteria applicable to the types of CD activities that qualify for CRA credit in these areas? If so, what should those criteria be?
Answer to Question 3: NCRC developed a definition of underserved tracts based on lending levels
NCRC believes that the agencies need to revise their definition of underserved and distressed tracts to better target CD activities (see proposed § 25.03). The agencies created a new category of tracts called underserved areas that were middle-income tracts with low population levels and without easy access to branches. The agencies also revised their definition of distressed tracts as middle-income tracts with high levels of unemployment and poverty rates.
The agencies conducted no research to justify the creation of these new tracts eligible for CRA activities. They did not describe how many middle-income census tracts would be eligible, where these census tracts would be, and whether these tracts would be underserved based on low loan levels. Thus, the public has no way to assess whether these new designations would target areas of great need per the intent of the CRA statute. The agencies have therefore denied the public a meaningful opportunity to evaluate and comment upon this potentially substantial change in the CRA evaluation method.
In contrast to the agency proposal, NCRC has developed a proposal based on data analysis and GIS mapping that focuses on low levels of lending. In a recent report, NCRC described a methodology for identifying the quintile of tracts in metropolitan areas that had the lowest levels of home and small business loans per housing unit and businesses. If community development data on lending and investing were available on a census tract level, the analysis could also include a measure of low levels of community development financing on a per capita basis.
When NCRC grouped tracts into quintiles based on lending levels, the tracts in the lowest quintiles were disproportionately LMI, though not entirely so. Thus, this approach would include middle-income communities in need of additional lending and investment. A significant subset of these tracts exhibited indicators of economic distress as shown by high poverty and unemployment levels.
This proposal is more effective in identifying not only lower income and middle-income communities but also communities of color that are underserved. A CRA reform effort must include a concerted effort to include communities of color more explicitly on CRA exams since a large body of research, including NCRC analyses, has found continuing and stark racial disparities in lending.
The tracts identified via NCRC’s analysis had been disproportionately redlined by the Federal Home Owners’ Loan Corporation (HOLC) classifying them as definitely declining or hazardous. The HOLC classifications from the 1930s continue to disadvantage these tracts to the present day. The NCRC approach, therefore, successfully targeted redlined neighborhoods that were predominantly minority. Quercia and Park also documented a lack of bank CRA lending in these same neighborhoods targeted by NCRC’s approach.
CRA had focused its attention on LMI communities, as it should. However, communities of color remain underserved because of decades of redlining and discrimination. In addition to providing credit for CD activities in underserved tracts as defined via NCRC’s approach, the agencies should add these tracts as a criterion on the lending, investment and service tests. A new underserved tracts criterion on the tests would direct needed lending and investments to underserved communities. It would also help alleviate pressure on LMI tracts that are gentrifying by giving banks additional geographical areas in which to serve and receive favorable CRA consideration.
Question 4. Under the proposal, the small business and small farm revenue thresholds and the size thresholds for a small loan to a business and a small loan to a farm would increase to $2 million. Do these increases appropriately incentivize banks to engage in small business and small farm lending activities, or should other changes be made to the revenue and loan size thresholds?
Answer to Question 4: small business revenue threshold should not be increased
Per proposed § 25.03, raising the threshold for what counts as a small business from $1 million to $2 million in revenue would divert CRA’s focus away from the smallest businesses. While adjusting the loan size that counts as a small business loan to account for inflation would be reasonable, an adjustment to the revenue size of the small business would not be justified by research.
Revenue size increase would not be justified
The CFPB has found that the great majority of small businesses had revenues under $1 million. As the CFPB documented, about 20 million firms or 76 percent of all firms had annual receipts of under $100,000. An additional 5.2 million or 19 percent of all businesses had receipts between $100,000 and $999,999. Together these two categories of businesses contained 95 percent of all businesses in the United States.
The revenue size limit to define a small business must not be automatically updated to account for inflation. Instead, when considering any revenue increases to define a small business, the agencies should consult with the CFPB, Small Business Administration, and the Census for the most current revenue estimates for businesses with no employees, and those with few such as one or two employees. This would be the most accurate way of determining appropriate revenue size estimates for small businesses. The great majority of small businesses under $1 million in revenue (about 82 percent) had no employees. Since these businesses are very small, it is unlikely that their revenue size will increase at the rate of inflation. These businesses such as landscaping are likely concentrated in the services sector and thus experience small revenue growth.
The agencies simply proposed an increase in the revenue size limit to $2 million and did not present research or analysis to support such an increase. The evidence, including from other agency research, did not support their proposal.
Another instance of a proposal not explained or justified by research is the proposed increase in revenue size for a family farm. An example in the list of CRA qualifying activities was a loan to a family farm with revenues of $10 million. This example appeared in the proposed list of qualified activities without accompanying explanation regarding how to define a family farm and why a proposed $10 million revenue limit was justified. A member of the public has no means to judge the reasonableness of this proposal in terms of targeting truly small and family farms. On the contrary, it seems like a means to inflate CRA ratings by allowing for loans to larger farms, including possibly corporate farms.
In fact, according to the U.S. Department of Agriculture, only 1% of farms had sales of $5 million or more. About 76% of farms had sales of $50,000 or less. It would seem that revenues of $10 million would be much too high for classifying a farm as small.
Increase in loan size to take inflation into account might be reasonable
NCRC believes that the current definition of small business lending as non-residential loans of $1 million or less is sufficient but could be updated to take inflation into account. A GAO report found that the $1 million limit for a small business loan should be updated to $1.6 million to account for inflation. It would be reasonable to update this limit to account for inflation since the costs of equipment and other needs have increased for small businesses.
However, the data do not support a further increase beyond inflation. Using CRA loan data, NCRC conducted research to investigate dollar amounts of loans to the largest businesses with revenues above $1 million during 2016. NCRC excluded lenders with average loan amounts of $10,000 or less as these were banks that focused on smaller credit card loans. NCRC constructed the sample to identify the loans of the largest dollar amounts. The average loan amount was $70,611 and the average loan amount for the quartile of loans with the largest amounts was $343,469. These loans do not come close to the $1 million limit. Therefore, according to the publicly available data, NCRC did not observe that banks are constrained by the $1 million limit.
Increasing the loan size to account for inflation would cause less of a diversion of financing from the smallest entities than increasing the revenue size of the small business or farm. Moreover, there are likely to be instances in LMI tracts where higher loan size limits above $1 million might be needed to finance new space or equipment for smaller firms, particularly in more expensive parts of the country.
Before the agencies increase the loan size, however, they need to conduct more data analysis to determine if loan sizes near or above $1 million are occurring in LMI tracts in more expensive metropolitan areas. If the data does not indicate much of this lending activity, then the $1 million loan size limit may still be appropriate and needed to target resources towards smaller enterprises. This is another area of the NPRM in which the agencies present no data analysis to support their proposals.
Question 5. The agencies plan to publish the illustrative list on their websites and to update the list both on an ongoing basis and through a notice and comment process. Should the list instead be published as an Appendix to the final rule or be otherwise published in the Federal Register ? In addition, how often should the list be updated?
Question 6. The proposal includes a process for updating the illustrative list on an ongoing basis through submission of a form to seek agency confirmation. The agencies considered an alternative process where an agency would accept all requests from banks for confirmation that an activity is a qualifying activity, aggregate these requests, publish the list of requested items in the Federal Register for public comment and feedback, and update the list following this process once every six months. What process, including any alternative process, should the agencies adopt to update the illustrative list of qualifying activities?
Answer to Questions 5 & 6: development of a CRA qualified list of activities should only be updated through a public comment process and should occur twice a year initially and then annually
In proposed § 25.05, the agencies propose ongoing updates of the CRA list of qualified activities in addition to a process of soliciting public comments. The ongoing process would involve a form via agency websites that banks would fill out when they are requesting a determination of whether an activity would qualify for CRA consideration.
We believe this proposed approach would require the agencies to engage in notice-and-comment rulemaking each time it seeks to amend the list of qualified activities, not just every three years as the proposal contemplates. That is because the list of qualifying activities constitutes “an agency statement of general or particular applicability and future effect” that will “affect . . . individual rights and obligations.” A regulated financial institution may seek pre-investment review by “a form through the agency’s website to seek confirmation that an activity is a qualifying activity” outside of a bank’s particular examination that “would not replace a bank’s ability to discuss whether an activity qualifies with its examiners.” It thus would contain a binding determination of “qualifying activities that meet the criteria in the rule” as well as those that “do not qualify” to provide “certainty and transparency about whether an activity qualifies for CRA credit prior to a bank engaging in the activity, and to ensure consistent treatment of activities.” As such, it is subject to the rulemaking requirements of Section 553 of the APA and amendments to the list should follow from public notice available through the Federal Register and agency consideration of public comment submitted as a result.
The most transparent and fair method for updating a list of CRA qualified activities would be through a frequent request for public comment on proposed revisions and additions to the list. The agencies, especially in the early years after a new CRA regulation, should request comment twice a year (eventually the process could move to an annual one). Before each comment period, the agencies could solicit suggestions from both banks and community organizations regarding proposed activities. The agencies then would decide which activities they would propose adding to the list. They would request comment on the new additions as well as modifications to exiting activities on the list. This process provides all stakeholders with the same opportunities for influencing an important list of this nature. It could also foster collaboration where banks and community groups are suggesting proposals together.
The agencies’ proposal of a comment period once every three years would be too infrequent for a list of this nature. Stakeholders would have more frequent questions about activity eligibility, especially in the early years after a new regulation.
After a public comment period and agency revisions, a revised list could be published in a manner similar to the Interagency Question and Answer (Q&A), which is published currently in the Federal Register. It would seem that the model of the Interagency Q&A publication would be more appropriate than an appendix to the regulation since the list would be a non-exhaustive list of examples of qualified CRA activities just like the Q&As are a non-exhaustive list of guidelines for CRA.
Making a form available via agency websites to just banks (and not community groups) would deprive “interested persons an opportunity to participate in the rule making,” because it would exclude non-bank participants. It is unclear how the agencies would collaborate with each other in determining which bank proposals to accept and add to the list. This one-sided process would also invite more outlandish proposals like financing improvements to football stadiums in Opportunity Zones.
Using the process of public comment periods to inform revisions, the agencies must develop the list of qualified activities carefully and explain the list in order to avoid banks not engaging in activities that are not included in the list. The banks could develop a tendency to refrain from activities that are not on the list for fear of not receiving credit on CRA exams. This would repeat a behavior that some allege occurs today due to uncertainty as to what counts.
In addition to emphasizing that the list is not exhaustive, the agencies should not develop an inordinately long list of activities that becomes cumbersome and would seem to discourage other activities. The list in the NPRM could already be too long and unwieldy. Instead, the list must be more of a principles-based list explaining what types of activities conform to the definition of community development. A principles-based list would focus on clarifying regulatory definitions such as community development and affordable housing. To help define affordable housing, for example, the list would describe the income ranges of occupants and would also describe how pro rata procedures would apply in the case of mixed-income housing when a portion of the tenants are not LMI. In contrast, the proposed list has a number of examples indicating that affordable housing includes various percentages of units for LMI households. A bank may wonder if its project would not qualify if it has a percentage of units for LMI households not on the list. It would be more helpful for a list to describe a range of acceptable percentages or indicate if a certain percentage does not qualify and why. Instead of a detailed list of examples, the list should explain a principle or concept. The existing Interagency Q&A document is a type of principles-based list, but one that is more of a summary is needed and would be a positive outcome of a NPRM.
To supplement a principles-based list, specific examples of activities that actually received credit on CRA exams could be presented in an interactive database. An interactive database could also be creative and energizing by including links to visuals and narrative descriptions of projects describing how the project was innovative and responsive to needs. This would encourage banks to replicate activities that exams considered particularly responsive to need and would encourage healthy competition among banks to see which ones financed the most responsive and innovative activities.
Question 7. Are certain types of retail loans more valuable to LMI individuals and geographies than other types? If so, which types? Should the regulations recognize those differences? If so, how? For example, could multipliers be used to recognize those differences and provide incentives for banks to engage in activities that are scarce but highly needed?
Answer to Question 7: encourage prime and special affordable lending and do not discourage retail lending
Encourage prime lending by giving it a greater weight
The most valuable retail loans for LMI borrowers and communities are those that are affordable and sustainable. CRA exams must encourage prime lending over high-cost and subprime lending. The financial crisis revealed that high-cost lending was inherently risky. When not regulated properly, subprime lending caused massive foreclosures and unsustainable debt levels. In order to avoid an over-reliance on high cost lending for serving LMI borrowers in the future, CRA exams must encourage prime (conventional and government-insured lending) by weighting prime lending more heavily on CRA exams. Examiners now use weighting on CRA exams, so this proposal represents a new application of an existing method of analysis. Prime lending should be weighted two or three times as much as high-cost lending.
The Home Mortgage Disclosure Act (HMDA) data defined a loan as high-cost in reference to an average prime offer rate (first lien loans were defined as high cost if they were 1.5 percentage points above average prime offer rates). This definition could be used to separate prime from high-cost home loans. A similar approach could possibly be taken regrading small business lending (particularly if price information becomes available via the rulemaking associated with Section 1071 of Dodd-Frank) and with automobile lending. Subprime automobile lending had considerably higher default rates so prime automobile lending needs to be encouraged by CRA exams.
Banks were not the perpetrators of abusive loan originations that caused the crisis. Federal Reserve research has shown that bank lending was safer and sounder than that of independent mortgage companies not covered by CRA in the years leading up to the financial crisis, and that bank loans made in AAs were less likely to result in foreclosure than bank loans made outside of AAs which are not currently covered by the CRA. Any proposed change to CRA must preserve this track record and must not inadvertently encourage riskier lending. For example, inclusion of consumer lending on CRA exams must be implemented carefully so it becomes an affordable alternative to payday and other fringe non-bank lending instead of encouraging high-cost and unscrupulous lending. If the agencies do not carefully include consumer lending on CRA exams, inclusion of this lending could encourage high-interest rate and abusive credit card lending and other lending with high and hidden fees. Furthermore, if CRA exams no longer have qualitative criteria like responsiveness and innovativeness, they could provide considerable credit to entry-level credit card products that do not include pathways for LMI consumers graduating to lower cost products that are more helpful in building savings and creditworthiness.
Encourage special affordable lending
In addition to more weight for prime loans, CRA exams should encourage special affordable lending (this comment letter adopts the phrase “special affordable lending” to refer to flexible and innovative lending). The current CRA exams contain qualitative criteria that periodically measured the affordability of loans. Occasionally, CRA exams under the flexible and innovative criterion on the lending test discuss whether banks had special affordable products. The best CRA exams contain tables that include the numbers of these types of loans. Deleting the criterion of flexibility and innovation, as the NPRM would do, would represent a step backwards. Instead, the criterion should be improved and made more consistent.
The following example from a CRA exam provides insight into how performance under the flexible and innovative criterion has been evaluated:
The bank further offers a proprietary affordable mortgage product targeted toward low- and moderate-income people. This in-house program is the Community Homeownership Incentive Program (CHIP). CHIP guidelines provide for low down payments, homeownership counseling, flexible credit criteria and long-term fixed rates. During the evaluation period, the bank made several enhancements to the CHIP program to make it more attractive to a broader base of applicants. In addition, the bank began allowing existing CHIP borrowers to refinance their loans through a new option, CHIP-to-CHIP. Bank management became aware that many existing CHIP borrowers were unable to take advantage of the low interest rate environment and refinance their loans due to inadequate equity to refinance to a conventional loan. Therefore, in late 2016, bank management introduced the CHIP-to-CHIP refinance option.
The number of loans originated through these programs represents approximately 18 percent of all HMDA loans originated during the evaluation period. This represents excellent performance in innovative and flexible mortgage lending.
This narrative is robust and useful. The narrative described a bank’s own proprietary product that did not involve a government guarantee and which the bank subsequently modified in order to respond to community needs. In addition, homeownership counseling appeared to accompany the origination of a CHIP loan (though we do not know the extent of counseling or its frequency from the exam narrative). The narrative also quantified the percent of products that the examiner considered innovative and flexible. While 18% seems high in this case, CRA exams do not calculate the percentages of flexible and innovative products on a consistent enough frequency to make peer comparisons that would inform judgements.
In addition, an exam table showed about 15,000 innovative and flexible loans. Although that was commendable, about 80% of these loans were government-insured loans, not the bank’s CHIP product. Although the bank had a commendable product, the agencies could most likely do a better job motivating greater volumes of this product through more consistent methods of measuring flexible, innovative, and affordable lending. Hence, it is apparent where reform should go on this measure: the agencies and the public need more consistent data collection for this performance measure and ideally a creation of an industry-wide database so peer comparisons could be more readily made on CRA exams.
The agencies could provide motivation for increasing affordable and sustainable retail lending by including qualitative criteria on exams and making the qualitative criteria as “quantifiable” as possible by including volume measures and peer comparisons. Furthermore, the qualitative criteria could count for 20% to 30% of a component test’s rating, which would provide a powerful incentive since it could be the difference between ratings received. A greater weight for this criterion probably would not be warranted because if a bank had highly innovative and flexible products, it will tend to perform on a High Satisfactory or Outstanding manner on the distribution criteria (percent of loans to LMI borrowers or communities).
As explained in more detail below in the section of how to measure CRA activities, the agencies also need to bolster the robustness of the fair lending review and include screens against abusive lending that violate consumer protection laws. If a bank is engaged in activity that violates anti-discrimination or consumer protection laws, it is not serving community needs by providing affordable and sustainable loans.
Agencies’ counterproductive focus on balance sheet lending will decrease retail lending
The agencies are poised to dramatically decrease retail lending by penalizing banks that sell retail loans within 90 days after they originate them. The NPRM would count these loans at 25% of their dollar value in their proposed CRA evaluation measure. This would penalize banks with a business model that depends on secondary market outlets so that they could receive more capital for making more loans. The banks would respond by making fewer retail loans.
The motivation for this proposal appears to be decreasing churning, the practice of banks purchasing large volumes of loans made to LMI borrowers shortly before their CRA exams in order to inflate their ratings. A more effective approach to deterring churning would be to examine a bank’s HMDA records and toss out purchased loans that were purchased less than a year before a CRA exam. The improvements to HMDA data mandated by the Dodd Frank Wall Street Reform and Consumer Protection Act and implemented by the CFPB would enable CRA examiners to do this since the improved HMDA data requires a loan ID number. Although not publicly available, this ID number is available to the agencies so that agencies can better track loans through origination and subsequent purchases for the purposes of regulatory enforcement, which would include CRA examinations.
If the agencies want to promote portfolio lending, they should increase CRA exam consideration of special affordable lending in the manner recommended by NCRC. Special affordable lending tends to be portfolio lending because the flexibility in underwriting and product features often do not facilitate selling on the secondary market. In an effort to promote portfolio lending, a preferable approach would be to provide extra consideration for special affordable lending than to stringently penalize selling to the secondary market. Stringent penalties for selling to the secondary market needlessly interferes with the business decisions and models of banks.
Question 8. The use of multipliers is intended to incentivize banks to engage in activities that benefit LMI individuals and areas and to other areas of need; however, multipliers may cause banks to conduct a smaller dollar value of impactful activities because they will receive additional credit for those activities. Are there ways the agencies can ensure that multipliers encourage activities that benefit LMI individuals and areas while limiting or preventing the potential for decreasing the dollar volume of activities (e.g., establishing a minimum floor for activities before a multiplier would be applied)?
Answer to Question 8: The use of multipliers would decrease CRA activity
The proposal in § 25.07(b) would apply a multiplier to almost all community development activities expect purchasing Mortgage-Backed Securities (MBS) and municipal bonds. A multiplier would likely result in less financing, which even this question on the NPRM acknowledges. Under the proposal, banks would be conducting their own analysis regarding the CRA evaluation measure and presenting results to examiners. They would certainly apply multipliers in cases in which they decreased their annual levels of community development compared to their previous CRA exams, at both the overall bank level and assessment area level. The use of multipliers combined with banks conducting their own analysis of their CRA evaluation measure would be too tempting; it would likely to result in banks decreasing their community development financing.
Moreover, readers of CRA exams would not know what they would be reviewing. Would they be reviewing actual dollar amounts of community development activities or dollars that would be multiplied for many, but not all activities? It would be impossible for the public to judge the actual level of community development financing, which would frustrate the purpose of CRA of determining the extent to which banks are responding to community needs. Moreover, the agencies and examiners would increasingly need to update and refine their choice of multipliers as new circumstances or questions arise, making the system more unwieldy and confusing.
NCRC opposes the use of multipliers for these reasons, but in the event that multipliers are used, the regulators should make banks ineligible for multipliers if their actual level of community development decreased since their last CRA evaluation period.
A better approach than multipliers would be to apply less weight to activities deemed less responsive to need. Instead of using multipliers, weighting counts an activity category less when determining a rating. Exams already weight retail lending based on their volume. For example, if home purchase loans were 70% of a bank’s home loan portfolio and refinance lending was 30%, then performance on home purchase and refinance lending would contribute to 70% and 30%, respectively, of the rating or score on home lending performance. Likewise, two different investment types could be weighted differently in determining a rating or score for CD investments.
Question 9. The proposal quantifies the value of CD services based on the compensation for the type of work engaged in by the employees providing the services as reflected in the Bureau of Labor Statistics calculation of the hourly wage for that type of work. Alternatively, CD services could be valued based on a standardized compensation value for the banking industry or occupation type. For example, the median hourly compensation value for the banking industry is approximately $36, when calculated using Bureau of Labor Statistics data. Would using standardized compensation values reduce the burden associated with tracking CD services while still appropriately valuing CD services? If so, how should the agencies establish the standardized compensation values
Answer to Question 9: CD services cannot be quantified in manner suggested by agencies
NCRC rejects the premise of the question as unworkable. Community development (CD) services cannot be evaluated effectively in the manner suggested by the agencies in proposed §25.07. It would be inaccurate to multiply an hour of a CD service by a wage rate to determine its value. The value of an hour of financial counseling in which bank staff helps a LMI person improve his credit score by correcting credit bureau errors would be considerably more valuable than an hour spent by a bank staff person conducting general volunteer work in an LMI community. While general volunteer activities are valuable, they are not related to the purpose of CRA, which is to eliminate redlining and increase access to credit to consumers and LMI communities. They are outside of the statutory assessment criteria that privileges a bank’s record of “meeting the credit needs of its entire community” and lists the “convenience and needs of communities” as “the need for credit services as well as deposit services.” In contrast, improving credit scores would be directly related to increasing access to credit and must be weighted more highly.
In addition, the agencies propose to amend the definition of community development services to allow for all volunteer activity such as hammering nails or being a docent at a museum. The current definition of CD services is a service related to the provision of financial products for LMI people. It is more desirable to direct the expertise of bank staff to financial counseling than general volunteer projects in the community. If CRA is amended to permit bank staff to address the multitude of issues in a community, CRA will not be able to achieve its purposes of financial inclusion. Generating hundreds of bank staff volunteers as docents at museums, for example, does not achieve CRA’s purposes in contrast to generating hundreds of bank staff volunteers delivering financial counseling.
Recently, Federal Reserve Governor Lael Brainard further explained why CD services should not be monetized:
For example, the services and leadership provided by a small bank located in a rural community may be vital to the success of that community, even if the dollar value of those services is small compared with a branch in a large city. Because of this concern, we are inclined to propose a set of qualitative standards to evaluate retail services within the retail test, and a separate set of qualitative standards to assess community development services within the community development test.
In areas with a low density of financial services, a bank officer on the board of local community organizations could provide considerable value to the community that is not accurately reflected by monetizing volunteer hours based on their compensation.
The agencies must discard their proposal of adding CD services to the CRA evaluation measure. Instead, they must retain the existing service test and improve it. A shortcoming of the CD service part of the service test is that it is hard for banks or community groups to determine how much is enough. Some exams measured the provision of CD services in units and others used hours. A unit is a confusing measure because it is hard to know to what a unit refers. Is it one hour or some other time frame? Hours would seem to be more intuitive. If the agencies created a database of CD hours from CRA exams, they would at least be able to describe ranges of annual hours for banks of various asset sizes and would be able to develop guidelines regarding median time periods. Banks that are far above or below the medians would receive Outstanding or failing ratings on this measure.
The agencies should also standardize tables on the CD portion of the service test. Useful tables would be those that present CD services by hours across each of the categories of CD.
The measure should also have a qualitative component that should constitute 20 to 30 percent of the overall rating on this criterion. Banks should be asked to document impact. For example, if a staff person delivered 12 lectures, each an hour, to a financial education class held by a nonprofit over the course of a year, how many of the clients increased their savings or improved their credit scores. The more data on impact and how the services responded to needs in localities, the higher the score on the qualitative component of CD services.
Question 10. Should the range of retail banking services provided—such as checking accounts, savings accounts, and certificates of deposit—be considered under this proposal? If so, how could retail banking services be quantified? For example, could the types of checking and savings accounts that are offered by a bank (e.g., no fee, fixed fee, low interest-bearing, high interest-bearing) be considered in performance context?
Answer to question 10: improve the service test, do not delete it
NCRC believes that the agencies erred significantly in proposing to delete the service test that includes measuring the range of retail banking services offered to LMI consumers and communities. The agencies had been inconsistent in applying this part of the service test, but the solution is to bolster the service test through better collection of data and implementation of standard measures of performance.
Deposit accounts provide a safe and affordable place for LMI customers to store their money and are the first essential step to establishing a banking relationship. Minimizing consideration of bank accounts and services would be a major step backwards for underserved communities in that banks would deemphasize bank services in LMI communities, making them more dependent on check cashers and other high cost fringe services.
In contrast to incremental reforms to the service test suggested by NCRC, this question contemplates “quantifying” retail banking services and adding them to the proposed CRA evaluation measure. This procedure would likely under-value the provision of retail banking services since deposit accounts for LMI customers are usually of small dollar amounts and pale against the large-scale financing that would be encouraged by the proposed CRA evaluation measure. Banks would opt to provide fewer deposit accounts evaluated by the CRA evaluation measure. Instead, the agencies should revamp the service test and make it more rigorous, engaging in better data collection and analysis for holding banks accountable for providing affordable deposit accounts.
The provision of bank accounts has been a pressing issue in recent years, particularly in the context of alternative service delivery such as mobile banking. The agencies updated the Interagency Question and Answer (Q&A) document in 2016 to include a Q&A on alternative service delivery. It stated that factors such as ease of use and rate of use would be considered.
As a result, readers of CRA exams will periodically see descriptions such as this one:
The growth rate of new accounts for customers residing in LMI geographies is significantly higher than the growth rate of new accounts for customers residing in MUI geographies. The bank’s internal data also shows an increase in the usage of Alternative Delivery Systems (ADS) by customers residing in LMI geographies, and ADS usage by customers residing in LMI geographies exceeds the ADS usage by customers residing in MUI geographies. The proportion of the bank’s LMI customers using ADS is significantly greater than the conservatively estimated population of fully banked LMI consumers.
This description is a step in the right direction but is not complete in terms of a rigorous evaluation of bank service provision. The examiner actually compared the percentage of fully banked customers in the area (using the FDIC survey on unbanked and underbanked populations) to the percentage of bank customers using ADS. The examiner concluded that the percentage of LMI customers using the bank’s ADS was greater than the percentage of fully banked LMI customers in the area.
While this is encouraging, the data in this CRA exam narrative was confusing and hidden. The exam did not present actual numbers and percentages of accounts although the examiner clearly had this data. Also, it was not clear when the exam was referring to accounts in general and accounts generated via ADS. It would be an advance for the exam to contain a table comparing the percentage of LMI people in an AA to the number and percent of accounts for LMI customers via ADS and traditional branches. NCRC has advocated that data on the number and percentage of accounts for LMI customers and/or by income of census tracts be provided on exams. It would seem that this is possible per this example, but has not been implemented. If the agencies implemented regular data collection and dissemination similar to HMDA data, examiners could engage in more consistent and rigorous analysis for this important measure.
Another aspect of service that CRA examiners usually do not discuss is the cost of accounts. The Interagency Q&A encouraged the provision of low-cost deposit accounts and stated that cost of alternative service delivery should be compared against the cost of the other delivery systems of the bank. Further developing an analysis of cost of services is desirable since services that gouge the consumer, especially LMI consumers, do not truly serve community needs. For this round of CRA reform, including cost considerations in the qualitative criteria with guidelines calling for a comparison of pricing within and across banks for LMI and non-LMI customers would be an advance.
Performance context must be applied on the service test as it is on other tests. In particular, banks that make strenuous efforts to offer affordable accounts, verified via data analysis, for geographical areas with high numbers and percentages of underbanked and unbanked LMI populations as revealed by the FDIC survey of the underbanked and unbanked would score highly. Instead of retaining the service test and applying performance context in the test in a more rigorous manner, the FDIC is turning its back on the mission of increasing access to banking services. As FDIC board member Martin Gruenberg explained:
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.
Where it Counts: Updating the Geographic Scope of CRA Assessment Areas: Bank Branches and Digital/Business Footprint
Because of the effectiveness of assessment areas, the current procedure of delineating assessment areas for geographical areas containing bank branches must be retained. In order to rigorously evaluate non-traditional banks that make loans via non-branch means, assessment areas also must be established for areas in which these banks make a considerable number of loans and/or engage in a significant amount of business activity.
Most bank lending is still conducted in assessment areas. Research by Federal Reserve economist Neil Bhutta found that assessment areas captured about 70 percent of home purchase lending for large banks. Likewise, in examining the 100 largest banks, NCRC found that assessment areas captured a great majority of their lending (91 percent of their home and small business lending). NCRC’s report relied upon the percentage of loans in assessment areas reported by CRA exams, which did not consider lending by affiliates when calculating the portion of loans in an assessment area. Therefore while NCRC’s percentage might be an over-estimate, it was consistent with Bhutta’s finding that for the largest banks, current assessment area procedures captured the great majority of their lending.
Since the current procedures captured the majority of traditional bank lending, reforms should adopt an additive approach instead of implementing wholesale changes. In particular, reforms should focus on non-traditional banks that are making large volumes of loans using non-branch means including brokers and the internet. For several years, NCRC has urged the agencies to update assessment area procedures to expand the number of assessment areas to account for lending beyond branches. This is a straightforward approach that retains assessment areas where branches are located and adds assessment areas to encompass geographical areas where banks lack branches but are engaged in significant lending or other business activity
Question 11. Are the proposed methods for delineating assessment areas clear, simple, and transparent?
Answer to Question 11: the proposed assessment area reforms are not clear, simple, and transparent; agencies should use lending data to designate additional AAs
Presently, CRA exams define assessment areas as geographical areas that include bank branches and in which a substantial amount of lending activity occurred. The OCC and FDIC would retain this definition in proposed § 25.08. They would call these assessment areas (AAs) “facility-based” AAs. In addition, the FDIC and OCC would establish “deposit-based” AAs. These AAs would apply to either internet-based banks or traditional banks that conduct much of their business over the internet. Under the procedures for designating deposit based AAs, the bank would be required to establish these AAs if more than 50% of the bank’s deposits were collected beyond branch networks. Also, the bank would be required to designate additional geographical areas as AAs if 5% of more of their deposits came from these area(s). The bank would be required to use the smallest geographical area(s) (whether it be a county, metropolitan area, or state) that generated 5% or more of their deposits.
While it is commendable that the FDIC and OCC recognized the need to update AAs for banks that conduct a significant amount of their business online, the proposal is not fully developed and would exacerbate credit deserts. Firstly, banks do not currently collect data in an accurate manner concerning deposits that are generated through non-branch means. Banks now arbitrarily assign deposits collected via the internet to branches. The OCC and FDIC would need to implement a rulemaking to establish procedures for accurately collecting and disseminating this data.
The FDIC and OCC offered no data analysis estimating how many banks the new AA procedure would apply to and how many additional AAs would be created. Discussing this aspect of the rule in the NPR, the FDIC concluded, “It is difficult to accurately quantify these aspects of the proposed rule with the information currently available to the FDIC.”
The agencies stated in the NPRM that they would issue a Request for Information (RFI) regarding general data issues and deposit data collection after the NPRM and before the issuance of a final rule. The time period for agency consideration of these complex matters would be compressed and insufficient for developing a feasible approach regarding deposit data prior to issuance of a final rule. Moreover, the general public would not be privy to any additional data analysis the agencies conduced after they obtained the additional data from the banks.
Once again, the NPRM proposes a significant reform without any substantive analysis of its impact. This is an inexplicable approach towards a rulemaking of profound impact. In contrast, the agencies could have used publicly available lending data (HMDA data for home lending and CRA data for small business lending) to identify areas outside of branch networks with significant amounts of lending. Impacts of the proposal in terms of the numbers of banks affected and how many new AAs would be established could be estimated. The public could also judge whether the proposal would help address the issue of CRA hotspots and deserts by increasing attention to areas that are currently under-represented among AAs.
A further difficulty is that the public may not have access to the deposit data required for the AA reforms. Banks may claim that the data must remain confidential since it involves location of deposit customers (despite the use of HMDA data involving census tract location of borrowers not having suffered a privacy breach in over 40 years of HMDA data dissemination). If the agencies do not make this new data publicly available, interested members of the public would have no meaningful way to comment on the adequacy of bank AAs or even figure out the areas in which banks would have obligations to meet credit needs until after the release of CRA exams.
Use bank lending data to establish additional AAs
As an alternative to the use of deposit data, several CRA exams have established an important precedent regarding using lending data to establish additional AAs. The former Office of Thrift Supervision (OTS) supervised several lenders without traditional branch networks. The OTS relied upon the Interagency Question and Answer (Q&A) document allowing examination of retail lending outside of assessment areas provided the retail lending inside the assessment areas had adequately responded to needs. However, good lending performance to LMI borrowers outside of the assessment areas would not compensate for poor lending performance in the assessment areas according to the Q&A.
For example, the OTS 2009 CRA exam of Citicorp, a non-traditional thrift located in Wilmington, Delaware that made loans through 77,000 agents located throughout the country, included analyses of 10 metropolitan areas and three non-metropolitan areas with the largest percentage of lending outside of the Wilmington assessment area. Likewise, the OTS examined Capital One’s lending in 20 areas beyond its one assessment area. These 20 areas comprised 25 percent of the thrift’s lending.
A more recent exam of the Bank of the Internet further developed procedures for considering loans outside of AAs. Since AA lending in San Diego accounted for 1 percent of total lending activity, an examiner with the OCC evaluated retail lending in six states outside the San Diego assessment area. Bank of the Internet’s activity in these six states accounted for 30 percent of total deposits and 56 percent of home mortgage and small business lending. The retail lending in the states outside of the San Diego assessment area was factored into the rating for the lending test.
NCRC illustrates how AAs would work for the Lending Club, an online lender (fintech) with no branches that had made loan data by geographical area available on its webpage. In Congressional testimony, NCRC calculated that more than two thirds of Lending Club’s lending during 2012 and 2013 was in 15 states, making it feasible to designate those states as AAs. NCRC also used Texas as a case study of designating local AAs. NCRC found five metropolitan areas with more than 1,000 loans each and one area, North Texas, that could possibly be considered a rural area. The five metropolitan areas ranged in size and location across the state and included Houston, Austin, Ft. Worth, Dallas, and San Antonio.
Lending Club demonstrates that it is feasible to select AAs for states, metropolitan areas, and rural counties where a substantial amount of lending occurred for fintechs, a number of which have applied for bank charters. NCRC advocates for the agencies to use HMDA, CRA small business lending data and consumer loan data (if the agencies implement data consumer data reporting requirements) to establish AAs outside of branches in addition to branch-based AAs.
If the agencies implement deposit data collection requirements, deposit data could also be used to establish AAs outside of branch networks. For newer banks like the fintech Varo, their predominant activity is deposit taking and financial management. Thus, new AAs should establish retail service and community development obligations in AAs where a significant amount of their deposit activity occurs. NCRC, however, recommends another NPRM that would use agency data analysis to assess how AAs would be established using loan and deposit data for different types of banks. The NPRM would also discuss what the impacts would be for banks with different business models and the industry as a whole in terms of the portion of lending and deposit activity that would be covered by new AAs outside of branch networks.
Question 12. The proposal would allow banks to choose how broadly to delineate their facility-based assessment areas, but it would require banks with a significant portion, such as 50 percent or more, of their retail domestic deposits outside of their facility-based assessment areas to delineate their deposit-based assessment areas at the smallest geographic area where they receive five percent or more of their retail domestic deposits. The requirement to designate deposit-based assessment areas would impact internet banks that do not rely on branches or ATM facilities to collect deposits as well as traditional banks that, in addition to their branches and ATM facilities, collect a significant portion of their deposits online outside of their branch and ATM footprint. Do these approaches strike the right balance between allowing flexibility and ensuring that banks serve their communities? If not 50 percent, what threshold should be used to determine if a bank has a significant portion of its deposits outside of its facility-based assessment areas and why? In addition, is receiving at least five percent of domestic retail deposits from a given area the appropriate threshold for requiring a bank to delineate a deposit-based assessment in that area, or should some other threshold be implemented? If so, why?
Answer to Question 12: the proposed thresholds of 50% and 5% do not ensure that banks serve their communities
The 50% proposed threshold in § 25.08(c) would be too high of a threshold for establishing deposit-based AAs. Currently, AAs capture about 70% of bank home lending so it would stand to reason that AAs also captured a similar percentage of their deposits. Given this, only a handful of banks may need to establish AAs under a 50% threshold. Assuming a correlation between deposit-taking and lending means that approximately 30% of bank deposits in addition to loans is currently outside of AAs. Thus, the threshold should be closer to 30% for establishing deposit-based AAs. However, for large banks, a threshold such as 10% could be appropriate since this could involve hundreds of billions of dollars of deposits, which represent significant and untapped resources for reinvestments. The proposed threshold must be based on clearly described data, research and reporting of the impacts, which it is currently not.
Another issue is that 5% of a bank’s deposits would be too high of a bar. In particular, many rural counties or smaller cities would not qualify for AA status, particularly for very large banks. Thus, this threshold would likely exacerbate the problem of credit deserts for underserved communities. A better threshold would be the bank’s market share of deposits or loans, which would likely create more AAs for large banks in underserved areas (as stated above NCRC recommends using loan data as well as deposit data to establish AAs). A market share threshold would need to be established by research and data analysis.
Capturing the great majority of retail lending and deposits via AA coverage is imperative for rigorous grading. NCRC found that when the great majority of lending was not captured by assessment areas, CRA ratings were inflated. In our study of the one hundred largest banks, NCRC revealed that inflation was the largest concern when AAs covered less than 50% of retail lending but also occurred when assessment areas covered less than 75% of lending. When exams allowed banks to focus more intently on a lower percentage of their loans, it was easier to score well on the lending test. While this benefited AAs containing the minority of a bank’s loans, it resulted in less lending overall in LMI communities in areas that were not AAs as demonstrated by the Federal Reserve-sponsored studies cited above. When ratings are inflated due to inadequacies in AA procedures, local needs remain unmet. An approach such as the NPRM’s that excludes significant areas of lending and deposit-taking from designation as AAs would contribute to overall CRA rating inflation and would result in local needs being ignored and unmet.
Question 13. The deposit-based assessment area delineation requirements are intended to ensure that banks serve the communities in which they operate. However, under the proposed regulation, it is possible that few banks would be required to delineate a deposit-based assessment area in less populous areas or states, despite having a significant market share in those areas (although banks with branches in those areas would be required to delineate facility-based assessment areas and banks may receive credit for qualifying activities outside of their assessment areas conducted in these areas or states). Does this framework provide sufficient incentives for banks to conduct qualifying activities in these less populous areas? Alternatively, should banks be required to delineate separate, non-overlapping assessment areas in each state, MSA, MD, or county or county equivalent in which they have at least a certain percentage of the deposit market share—regardless of what percentage of the bank’s retail domestic deposits are derived from a given area—and, if so, what should the percentage of the deposit market share be?
Answer to Question 13: 5% market share is the preferred threshold for deposit-based AAs; agencies have inadequate proposal for serving areas outside of AAs
Five percent market share of deposits is the better measure
As stated above, the 5% threshold as measured by a percentage of a bank’s total deposits would not be sufficient for establishing deposit-based AAs. Instead, a 5% market share measure would be more effective in assuring that deposit based-AAs would cover smaller metropolitan areas and rural counties.
A NCRC case study of Ames, Iowa demonstrated that a market share measure would be more appropriate. Ames is a smaller metropolitan area in central Iowa of about 66,000 people located 30 miles north of Des Moines. For very large banks, it becomes immediately clear that a measure based on a percentage of total deposits would capture, few if any additional AAs. U.S. Bank, NA had a market share of deposits in Ames of 13.2% but the deposits in Ames were only .1% of the bank’s overall deposits. Likewise, Wells Fargo had a market share of 7.9%, but the Ames deposits were even a smaller percentage of Wells’ overall deposits than U.S. Bank’s.
The agency proposal would probably fail to capture regional banks as well. Great Western Bank, based in South Dakota, had total assets of about $12.8 billion and deposits of $10.4 billion. Great Western had a deposit market share of 6.4% in Ames, but these deposits were just 1.8% of the bank’s total deposits. Thus, a measure based against a bank’s total deposits would largely or even entirely eliminate large and regional banks’ obligations in smaller metropolitan areas.
Considering banks with a market share of 5% or more of deposits, by contrast, would bring in about 76% of all of Ames’ deposits. Therefore, a market share test would more effectively ensure that banks with significant deposit share, including those without branches, had a CRA obligation in areas like Ames. It would also ensure that the great majority of deposits in a locality would be employed towards meeting local needs.
NCRC’s proposal for considering CD financing outside of AAs
The NPRM professes concern for underserved areas in Question 13 but discussed CRA activities outside of AAs in one sentence. The NPRM’s underdeveloped procedure for allowing banks to earn CRA credit outside of their AAs would encourage banks to gravitate to national level intermediaries that can put together the largest deals instead of working with locally-based CDFIs and other nonprofits on smaller deals that would be more directly responsive to the needs in their localities. In addition, banks would likely gravitate towards the easiest places to serve instead of areas outside of AAs with the greatest needs. The OCC and FDIC could respond by saying that this concern is addressed by their suggested retail lending test and established minimums of community development activity at the AA level, but under the proposal, banks would only need to reach satisfactory CRA performance in half of their assessment areas.
In comments responding to the OCC’s Advanced Notice of Proposed Rulemaking in 2018, NCRC suggested that the agencies retain current procedures for qualifying CD financing outside AAs but further facilitate outside AA financing by implementing annual data collection and increasing the number of eligible areas, called underserved areas, where banks could earn CRA credit for community development financing outside of their AAs.
Further developed in a white paper this past summer, NCRC’s approach would involve data analysis of a bank’s CD financing levels, a pre-qualification procedure, and the establishment of underserved counties as eligible areas to serve that would be outside of banks’ AAs.
The agencies would implement annual collection of CD data. This would consist of data on CD lending and investing that would correspond to the categories in the CRA regulation such as affordable housing or community facilities. NCRC proposed that this data would be collected on a census tract and county level. Precedents for this data collection include OCC databases on public welfare investments that are produced every quarter and can be downloaded.
Using CD data, an examiner could determine one year after the previous CRA exam whether a bank is meeting needs in its AAs. First, the examiner could determine whether for all AAs combined together, a bank is above or below median levels of CD financing for peer institutions. Then, the examiner could determine how current levels of CD financing in specific AAs compare to past levels on an annualized basis to see if a bank is on track to meeting needs. If the bank passes these tests, the examiner could pre-qualify the bank as eligible to pursue CD outside of its AAs. If a bank did not pass muster on these measures, it could spend the next year on improving its performance in its AAs and then pursue community development outside of its AAs. It could then ask for another agency review the second year after its previous exam.
This procedure using CD data would be more objective and certain than the current procedure, which appears to vary based on the subjective judgments of different CRA examiners. Also, it would provide an opportunity for a bank not passing muster in the first year after its previous exam to improve during the second year and become eligible for outside AA CD activities.
Create Underserved Areas
In addition to retaining the allowance for outside AA financing in statewide and regional areas, NCRC suggested that the agencies designate underserved areas that could be counties anywhere in the country that are underserved in terms of CD lending and investing. The agencies could develop measures to identify underserved counties such as the dollar amount of CD lending and investing on a per capita basis. Counties in the lowest quartile or quintile of CD financing per capita then could be candidates for designation as underserved.
The agencies could also use demographic and economic criteria for designating underserved areas much as they do now for identifying underserved and distressed rural middle-income census tracts. A combination of lending, demographic and economic criteria could be used to designate underserved counties. The counties receiving an underserved designation could be updated annually as is the case now with rural underserved and distressed tracts.
A recent NCRC white paper demonstrated via data analysis that using measures of lending levels, it would be feasible and desirable to establish underserved counties that would be areas outside of AAs that banks could serve. NCRC’s paper revealed how its designation of underserved counties would indeed target CRA resources into underserved counties that have low levels of home and small business loans, higher levels of African Americans, and higher levels of poverty and unemployment. It is targeted to underserved areas and avoids the temptation provided by the NPRM to cherry-pick the areas easiest to serve.
The designation of underserved counties would diminish the possibility of cherry-picking counties that are easier places for CD financing outside of AAs. Underserved county designation would alleviate the issue of hotspots and deserts by directing CD financing to places most in need. Also, since the designation of counties would be conducted annually, counties may come off the list if they received substantial amounts of CD financing and others may be added that have pressing needs. Banks would not need to worry that they would not receive credit for any CD financing in counties that come off the list since the agencies have established lag periods to account for CD financing in cases of rural tracts coming off lists of distressed or underserved rural tracts. This procedure would be applied to any new category of underserved tracts. The annual designation may further smooth out and more evenly distribute CD financing.
As a further inducement for CD financing in underserved counties, the pre-qualification procedures could be less stringent for a bank being allowed to engage in CD financing in an underserved county. For example, a bank may need to be at or above median levels of CD financing in its AAs to be eligible to engage in community development in statewide or in a regional area, but could be below the median if it wanted to offer CD financing in underserved counties. Perhaps, the agencies could establish a threshold such as no less than 20% below median for engaging in community development in underserved counties.
How to Measure CRA Performance: The Agencies’ Proposal is Not Objective and Would Not Accurately Measure CRA Performance
The agencies assert that their proposed evaluation framework would be more objective than the current evaluations, but the agencies’ proposal to simplify CRA evaluations would have the opposite result. The evaluation regime would become more complex and rigid while not accurately measuring whether banks met local needs. It would also skew CRA activities towards the largest-scale financing, which is not necessarily responsive to needs across a variety of local areas.
In proposed § 25.09, the agencies’ NPRM would set up two examination standards: small bank performance standards and general performance standards.
- Small bank performance standards would apply to smaller institutions with $500 million or less in assets opting to be examined under current CRA regulations for small banks.
- General performance standards would apply to institutions with over $500 million in assets.
Comments submitted to the OCC in response to last year’s Advance Notice of Proposed Rulemaking (ANPR) overwhelmingly opposed the creation of the proposed CRA evaluation measure, which would be the major determinant of a bank’s CRA rating. Yet, the NPRM continues to propose a CRA evaluation measure that would be “dominant determinant” according to the one dissenting FDIC board member, Martin Gruenberg.
Under the general performance standard, CRA exams would now have two major components for evaluations at each AA and overall at the bank level. The dollar-based component would be called the CRA evaluation measure and the other component would be called the retail test.
General Performance Standards: The proposed CRA evaluation measure
The CRA evaluation measure would continue to be the dollar amount of qualified CRA activities divided by the bank’s quarterly average for retail deposits for each AA and at the bank level. It would be the determinative factor on a CRA exam as reflected by its name, “presumptive” rating.
Agencies do not explain how empirical benchmarks for the CRA evaluation measure would increase CRA activity or make ratings more rigorous
In proposed § 25.12, the NPRM would establish specific ratios for the evaluation measure that would correspond to ratings. A ratio of 11% would correspond to Outstanding, 6% to Satisfactory, 3% to Needs to Improve and less than 3% for Substantial Noncompliance as described.
The agencies briefly described their data using the time period of 2011 through 2018 to establish these ratios but did not release their research as an appendix or accompanying paper. The agencies admitted the data was incomplete and that they had to use assumptions that are unreliable as discussed below. The OCC and the FDIC had repeatedly stated that their goal in this process was to increase CRA activity, and that moving to this approach would help accomplish that. The OCC and the FDIC must immediately release their research on the current levels of CRA activity in order for the public to be able to evaluate whether the proposed thresholds would actually motivate increases in CRA activity, would merely legitimize current rates of CRA activity, or could lead to decreases in CRA activity.
The NPRM text did not elaborate on how the ratios established for the evaluation measure would affect the CRA ratings distribution. In other words, would grading become more rigorous with fewer Outstanding ratings or higher failed ratings? Tougher grading would stimulate more CRA activity. The vague answer in the NPRM is that the agencies computed “what each bank’s average CRA evaluation measure would have been from 2011-2017 under the framework in the proposal.” Based on this description, it would appear that the proposed ratios would preserve the current ratings distribution rather than changing it. In other words, the agencies calculated the average CRA evaluation measures for all banks in the past time period and developed ranges for the ratios that would preserve the previous ratings distributions. The reader has to assume this was done because the agencies do not indicate that the ratio ranges would alter the ratings distributions.
With about 90% of banks receiving Satisfactory ratings the last several years, it would seem implausible that finer gradations in performance cannot be reflected in improved ratings made possible by enhanced scoring methodology, which would motivate increases in CRA activity. The proposed evaluation measure in the NPRM does not develop more rigorous grading.
The assumptions used in developing the CRA evaluation measure could render the empirical benchmarks unreliable. For example, data on credit card lending does not currently have information on the number and percent of LMI borrowers. Thus, the agencies had to assume that the LMI lending is captured by lending in LMI tracts. In a footnote, the agencies admitted that they would be including lending to middle- and upper-income borrowers in this approach but that they would be missing LMI borrowers who do not reside in LMI tracts. They assumed that the erroneous inclusions and exclusions would cancel each other out. However, there is no solid basis to assume this since the number of loans to LMI borrowers not residing in LMI tracts could be quite different from the number of loans to middle-and upper-income borrowers in LMI tracts. These types of assumptions make their empirical benchmarks suspect.
Further, the agencies doubled the amount of small business and small farm loans that were considered CRA eligible in calculating the empirical benchmarks for the CRA evaluation measure. They assumed that doubling was appropriate since the data now shows small business lending of amounts of under $1 million to small businesses with revenues under $1 million and small farm lending to farms with revenues below $500,000. However, this assumption rests on a theory that the distribution of lending to businesses and farms with revenues between $1 million to $2 million is the same as that for lending to businesses with revenues under $1 million and small farms under $500,000. It is likely that this is erroneous since the great majority of small businesses have revenues under $1 million as described above. Thus, another inadequately-supported assumption casts doubt on the accuracy of the CRA evaluation empirical benchmarks. To cast even more doubt on this assumption, the final NPRM published in the Federal Register replaced the word “doubled” that appeared in the original double-spaced NPRM that the FDIC board approved with the word “fraction” to describe the estimate made of CRA-qualified small business and farm lending.
Proposed CRA evaluation measure would distort activity and favor large over small financing
The proposed evaluation measure may not be “consistent with the safe and sound operation” of financial institutions as the CRA requires, because the actual ratios of CRA activity to deposits can fluctuate widely during recessions and expansions. Using a long time period that included the Great Recession as well as the ensuing recovery compromised the robustness of the agencies’ proposed benchmarks. Overall, as Governor Brainard substantiated in her recent speech, the ratios could be set too low for expansions and too high for recessions, especially if they were based on incomplete data and too long a time period. The agencies stated that they would adjust the empirical benchmarks every three years, but it is not clear why they did not use this time period to establish their initial benchmarks or how a three-year review would adjust appropriately economic cycles.
Banks may have to engage in riskier financing, such as stretching underwriting criteria too far, to meet the benchmark ratios during recessions and may have too easy a time meeting it in expansions. Also, the CRA evaluation measure in the NPRM is not adjusted or computed separately for banks of different asset classes. Thus, the NPRM benchmarks compare regional or state banks against the largest banks in the country on a determinant performance measure. This not only created competitive inequities among banks, harming smaller banks, but could also encourage the smaller banks to engage in riskier financing to achieve the ratios of their larger competitors.
The CRA evaluation measure would also distort a bank’s activities in such a way as to make the bank less responsive to their CRA obligations to serve local needs. The empirical benchmarks would encourage an over-reliance on the largest and easiest deals in order for banks to hit the benchmarks. The FDIC and OCC worsened this problem by opting against a single transaction limit that they had contemplated, such as one deal could be no more than 10% of the numerator, in order to help mitigate an over-reliance on larger deals. Without a cap on a single transaction, a bank could seek a very large deal in order to facilitate passing the empirical benchmarks for the CRA evaluation measure. For example, in order to be above the 6% needed for a Satisfactory rating, a bank could seek a large deal such as financing “essential infrastructure” in a given AA that could equal a high percentage (20% or more) of the evaluation measure’s numerator.
In practice, a locality could have pressing needs for small dollar lending to very small businesses or grants to housing counseling agencies remediating loans for borrowers on the verge of foreclosure. The banks, however, could be reluctant to meet these needs for smaller scale finance since they could be pressured to meet their ratio benchmarks.
The agencies’ promotion of large infrastructure as a regulatory category of community development would intensify the bias towards larger projects. Since the large-scale projects such as stadiums or bridges are also usually more profitable than CRA-related projects, banks would have financed them in the usual and normal course of business without CRA incentives. The reason Congress passed CRA was to ensure that banks also undertook the more labor-intensive financing often needed in underserved communities, financing that banks would overlook should the agencies enact the NPRM as proposed.
The use of a bank’s balance sheet would favor larger and longer-term financing over shorter term loans and investments. For example, short-term small business loans have less balance sheet value than a Low Income Housing Tax Credit (LIHTC) that would remain on the balance sheet for 15 years. Under the agencies’ approach, a bank would be tempted to make a few large loans and investments and let those sit on their balance sheet rather than continually seeking out new loans and investments.
The NPRM is attempting to respond to a tendency for banks to issue shorter term CD loans and investments that correspond to a three-year CRA exam cycle since a perception exists that examiners mainly focus on newer loans and investments. However, CRA exams currently consider both new and prior-period CD investments. The same procedure could apply to CD loans with decisions made regarding the exam weights accorded to new and outstanding CD financing. This would be preferable to either an over-reliance on new CD financing or only a balance sheet approach, which would bias banks towards previous CD loans and investments.
Rural and underserved areas that rely on small-dollar lending would likely experience a decrease of this lending after implementation of the proposed CRA evaluation measure. A recent Fannie Mae analysis found that smaller-dollar mortgage lending was statistically significantly higher in rural areas such as the lower Mississippi Delta, Middle Appalachia, and persistent poverty counties. The CRA evaluation measure would reduce banks’ incentive to serve these areas with low-dollar loans.
Carolina Reid’s recent white paper illustrated how the CRA evaluation measure would likely favor expensive cities over smaller towns and rural areas with depressed economies. Her analysis illustrated this in California with high-cost areas like San Francisco being favored over inland areas in California that have higher rates of unemployment and economic distress. In a particularly disturbing part of the report, Reid recounted how banks were not interested in participating in local affordable housing programs that pale against larger deals they could find in San Francisco. A city official interviewed by Reid complained that banks do not have an interest in participating in a safe and sound city affordable homeownership program.
This was disappointing but not surprising. It would seem that the proposed CRA evaluation measure would exacerbate these inequalities among geographical areas and non-responsiveness to real needs in violation of the letter and spirit of the CRA statute. In contrast, CRA reform that bolstered qualitative criteria such as innovation and responsiveness and made them “more quantitative” as discussed above would encourage more participation in local public or nonprofit sector housing and community development programs.
The CRA evaluation measure would convert CRA to a zero-sum contest for scarce resources instead of preserving the positive sum nature of the current CRA. Large-scale needs would trump smaller dollar lending. Bridges would triumph over microbusinesses. Higher-dollar home mortgage lending in expensive areas would crowd out lower-dollar mortgage lending in more affordable markets. Needs would compete against needs when in reality multiple needs in a locality must be addressed if revitalization is to succeed. An overly simplistic measure would thwart the reinvestment goals of CRA. In contrast, by having separate tests with a variety of measures, the current CRA better ensures that banks address multiple and overlapping needs.
Proposed CRA evaluation measure would short-change community development activities
The agencies must scrap the CRA evaluation measure as an over-arching measure since it cannot measure responsiveness to a variety of local needs. Instead, the agencies must retain separate retail lending tests, community development, and service tests that more effectively measure responsiveness to different needs. Furthermore, the agencies must preserve the qualitative aspects of the tests, which further enhance the tests’ abilities to measure responsiveness to needs.
The agencies could apply a ratio concept on a community development test. Community development lending and investments are now measured in a manner similar to a CRA evaluation measure. Moreover, the existing CD lending and investment tests have a ratio measure as one measure, but not the determinative one. It is more appropriate to measure community development lending and investment in a manner like this than to include also retail and service activities in one combined ratio measure. In her recent speech, Governor Brainard discussed the rigor and flexibility of a ratio comparing CD activity to deposits.
The proposed CRA evaluation measure most likely would be ineffective in stimulating increases in community development financing by large credit card lenders. NCRC examined the CRA evaluation measures of three large credit card lenders. We made an estimate of credit card lending to LMI borrowers (no data on this lending is publicly available) but we believe our estimate was conservative as explained the NCRC study. Even with a conservative estimate of credit card lending to LMI borrowers, the three large credit card lenders had ratios with just qualified credit card lending in the numerator that already exceeded the benchmark of 11% needed for an outstanding rating.
In addition, one of the three large lenders had an overall community development minimum that already exceeded 2% of deposits. In other words, this large lender likely would not have to increase its community development financing in order to keep receiving Outstanding ratings. For this lender, stagnant or diminished performance would earn it Outstanding under the proposed rule. The other two lenders would have to increase their community development financing but once they hit the 2% minimum required ratio, their performance would stagnate. The proposal probably would lead to lackluster CRA performance that does not continually and affirmatively respond to needs in the long term for these three large lenders.
Carolina Reid’s paper also documented that a CRA evaluation measure threshold would be satisfied mostly by home and small business lending. She documented that in California, retail lending constituted about 78% of CRA lending and investing in a sample of banks. In contrast, CD lending and investing ranged from 2% to 20% in the sample of banks. It would be quite probable that in order to hit their ratio requirements, banks would therefore emphasize retail lending in the most expensive markets to the neglect of smaller dollar retail and CD financing elsewhere, particularly if they were allowed to fail in half of their assessment areas.
Proposed CRA evaluation measure would diminish qualitative considerations and local needs analysis
The proposed CRA evaluation measure would diminish important qualitative considerations such as the responsiveness of activities to community need. The agencies stated that they would retain performance context analysis that assesses a bank’s responsiveness to needs. Their brief description of performance context in the NPRM preamble and in § 25.14 was not clear in describing how this analysis would adjust ratings determined by the evaluation measure. By generously expanding upon how banks could describe their capacities and limitations in reinvestment opportunities in local areas, it appeared that NPRM performance context guidelines would be used mainly to excuse banks from not hitting CRA performance thresholds.
The NPRM’s treatment of public comments was under-developed. The NPRM briefly discusses that public comments about needs and opportunities would be a factor in performance context analysis. The NPRM, however, did not explicitly discuss whether the agencies would consider public comment on a bank’s performance on any of its metrics. The agencies also did not indicate that they would facilitate public comment. The agencies would provide banks with a form on the agency’s website for submitting comments and analyses about performance context, but no analogous form would be provided for community group comments.
The dollar metric focus of the CRA evaluation measure would make the qualitative analysis difficult because the agencies had not made clear whether they would retain the same level of detail that exists now on CRA exams. For example, would the agencies develop tables that display the dollar amount and percentages of CD financing for the major CD categories so that the percentage of CD devoted to economic development or affordable housing can be compared to socioeconomic data and public comments? The agencies have not indicated whether the exams would mainly contain ratios and little else or whether the rich detail now in exams that simulates stakeholder discussions about whether banks are meeting local needs would remain.
The agencies must considerably improve performance context analysis so that it can adequately judge whether banks are responding to needs. In a paper, How To Evaluate Community Development Financing And Services Under CRA, NCRC explained that performance context analysis should identify priority needs through analysis of economic and demographic data as well as considering community comments. If the unemployment rate, for example, is high in an area and comments confirm the difficulty of finding jobs, the agencies would expect banks to have a relatively high level of CD financing devoted to economic development or revitalization activities. If a bank’s level were low, the bank would score poorly on the qualitative part of the test, which would contribute to 20% to 30% of the rating.
Community development minimum threshold not described adequately
Under § 25.12, banks would need to achieve a minimum level of community development financing for each AA and overall. They would need to ensure that 2% of their deposits in each AA and overall are devoted to CD financing. However, would the CD minimums ensure significant levels of CD finance overall and in each AA? The agencies once again proposed a threshold without describing data analysis that revealed whether this threshold would increase levels of CD lending and investing. A recent NCRC analysis of a sample of 15 very large banks found that their median level of CD financing was already 2% of their deposits. The agencies’ proposal would not boost the CD levels of these banks. In light of the CD multiplier and the expanded list of qualifying CD activities, the proposal would likely diminish the financing of currently-qualified CD activities at those banks.
Proposed CRA evaluation measure likely to cause banks to reduce LMI retail lending
A recent NCRC analysis found that the CRA evaluation measure would likely encourage banks to decrease their retail lending to LMI borrowers and communities. After taking into account community development financing minimum requirements (and branching for a sample of very large banks), NCRC estimated that the percentage of retail lending for LMI borrowers and communities (expressed in dollars) ranged from 11% to 14% for Outstanding ratings and 5% to 6% for Satisfactory ratings for banks of various sizes. In comparison, the aggregate percentage of bank loans for LMI borrowers, census tracts, and small businesses and farms during the last four years ranged from 10% for home lending to 24% for small business lending to 68% for small farm lending. Comparing actual bank lending to CRA evaluation measure requirements suggests a large likelihood of banks relaxing their lending to LMI borrowers, communities, and small businesses.
The proposed retail test might compensate somewhat for what appears to be a relaxing effect of the CRA evaluation measure. However, the retail test would not only compare banks against a demographic benchmark but would compare banks against each other. A bank would only need to pass either a demographic or peer comparator as described immediately below. Thus if banks were induced by the CRA evaluation measure to relax their retail lending to LMI borrowers and communities, they could still pass their retail test.
Proposed retail lending distribution tests would decrease bank performance
The retail test would establish inconsistent measures for banks of different sizes and for different loan types. As a result, it would generate inconsistencies in bank performance that would not adequately respond to local needs.
The retail test in proposed § 25.11 would look at the distribution of borrowers for home and consumer lending. It would not consider home or consumer lending in LMI tracts under the general performance standards because of concerns of causing displacement (although NCRC believes there are better ways to deal with this issue as discussed above). Small banks opting to be examined under small bank performance standards would continue to be evaluated based on geographic distribution, however.  The retail test would look at the distribution of lending by borrower for small business and farm lending and also at the geographical distribution of lending for those loan types.
Under this proposal, large banks would make more of an effort to make small business and small farm loans than home loans in LMI tracts. Small banks would make more of an effort than large banks to make home loans in LMI tracts. These differences in performance would not reflect adapting to differences in local needs but would be due just to test construction.
In order to pass the borrower and geographic distribution component tests of the retail test, the bank, for all major retail product lines, would need to meet or exceed a minimum threshold associated with either the demographic or with peer comparator in that AA. These are:
- Demographic comparator: A bank’s percentage of lending for LMI borrowers or small businesses or farms would need to be at least 55% of the percentage of LMI households or small businesses or small farms in an AA. A bank’s percentage of small loans to businesses or farms in LMI census tracts would also need to be at least 55% of the percentage of businesses or farms in LMI census tracts in the AA.
- Peer comparator: A bank’s percentage of lending for LMI borrowers or small businesses or farms would need to be at least 65% of the percentage of loans to LMI borrowers or small businesses or small farms originated by all banks in the AA under the general performance standards. A bank’s percentage of small loans to businesses or farms in LMI census tracts would also need to be at least 65% of the percentage of small loans to businesses or farms in LMI tracts originated by all banks in the AA under the general performance standards.
While it is an advance to establish benchmarks for the demographic and peer comparators, the NPRM does not describe any rationale for the 55% and 65% benchmarks. If the agencies conducted data analysis and produced averages across metropolitan and rural areas, commenters do not know that. Also, these benchmarks would likely have to be computed for additional areas such as high-cost metropolitan areas, lower-cost areas, and possibly smaller metropolitan areas. In particular, the 55% demographic benchmark might be too low in lower-cost areas where it is easier to lend to LMI borrowers or too high in higher-cost areas where it is harder to lend to LMI borrowers.
The NPRM also does not explain why the agencies decided that a bank needed to exceed either the threshold on the demographic or peer comparator to pass on the geographic and borrower distribution tests. It would make more sense to propose reasonable yet rigorous thresholds for both comparators and expect a bank to pass on both in order to pass on one of the tests.
On the surface, it appears that the proposed thresholds would be too easy. As an example, consider just the peer comparator for LMI borrowers. Suppose in an AA, all lenders issued 25% of their loans to LMI borrowers. Two thirds of this would be just 16% of loans to LMI borrowers. In NCRC’s judgment based on years of data analysis, this severe of a shortfall would merit a low or failed score on this criterion. Lenders tend to be bunched up around an aggregate percentage (25% in this example), not this far behind it. The agencies would allow the poor performing outliers to earn a pass for their lackluster performance. CRA grade inflation could increase from the current pass rate of 98%. LMI households and communities would likely receive fewer loans since more banks would figure they could pass by being outliers on the low end of the distribution.
To test the proposition that the thresholds would be too easy, NCRC calculated failing rates under the proposed thresholds in twelve randomly selected core-based statistical areas (CBSAs); we chose four large CBSAs with more than 1,000 branches, four medium sized CBSAs with between 100 and 999 branches, and four small CBSAs with under 100 branches. NCRC found a wide range of failure rates depending on the loan product, size of the CBSA and whether the sample was restricted to just banks with branches in the CBSA. Generally, larger CBSAs had higher failure rates where it was possible that the high cost of housing relative to incomes for LMI households made it harder to issue loans to LMI borrowers. Moreover, banks with branches in the AAs did much better with significantly lower fail rates, particularly compared with credit card lenders in small business lending.
The wide variability of results suggests that the OCC’s and FDIC’s proposed thresholds need another round of testing, refinement and another NPRM if the agencies wish to propose a meaningful, stable and sensible rating system that does not disadvantage groups of banks, particularly smaller ones. NCRC’s results suggests that the largest banks with the highest number of AAs would be most adept at gaming a system in which failure rates of up to 50% would be allowed. Calculating their pass rates, they would simply write off large numbers of AAs, possibly the largest CBSAs, and concentrate their retail lending on medium sized and smaller CBSAs. To make matters worse, they could still receive boosts in their CRA evaluation measure at the bank level for activities in the AAs in which they concede failure in their retail tests.
Large swaths of LMI census tracts and borrowers would receive significantly fewer loans as a result, contrary to the objectives of CRA reform. Regional and community banks, in contrast, would not be able to game AAs to the same extent, and would remain in markets in which negative externalities of spotty information about borrowers and neighborhoods would re-emerge, making their retail lending riskier.
NCRC alternative thresholds for retail test would better capture gradations in performance
As an alternative to the NPRM, NCRC believes a range of thresholds corresponding to different ratings would create a rigorous retail lending test and would decrease ratings inflation. Also, the retail lending test is currently 50% of the overall rating for the large bank exam. It must retain significant weight since low levels of lending and redlining have not been overcome in many LMI communities.
On the LMI borrower test, instead of 65% of the aggregate peer, the ranges below would make more sense because they involve more gradations in comparing a bank’s performance against its peers:
Outstanding: greater than 100% because the bank would be better than its peers.
High Satisfactory: 80% to 99% because the bank would be approximately in line with its peers.
Low Satisfactory: 60% to 79% because the bank would be below its peers, but not so far below to be considered not satisfactory.
Needs to Improve: 40% to 60% because the bank would be at approximately half the level of its peers.
Substantial Noncompliance: 39% and lower because the bank would be far below the level of its peers.
In order to sum the scores within and across AAs, the ratings could be converted into numerical scores ranging from 1 for substantial noncompliance to 5 for Outstanding. Then either simple or weighted averages could be computed to derive AA ratings and overall ratings.
NCRC’s research paper used the ranges above to construct alternative ratings for the banks in the sample of twelve CBSAs. These ranges would be more effective at promoting better bank performance and increases in safe and sound lending for LMI communities and people because they would spur more competition and lending in LMI markets. In contrast, a system of pass/fail with just two thresholds would prompt many banks to decrease their performance so they would just be hitting either the 65% or 55% threshold, whichever was easier.
In contrast, a system of threshold ranges converted into ratings would reveal more gradations in performance with significant numbers of banks below average in the Low Satisfactory and failing range. This would prompt efforts by more banks to avoid exposure as poor performers. In addition, it could be possible to construct a system with low fail rates but significant numbers of banks in the lower performing levels, which would still provide powerful motivations to improve and make more loans to LMI populations. Moreover, by increased transparency regarding which geographical areas, products, and tests in which any bank is lagging, a system of ratings would more effectively promote collaborations among banks, community organizations, and public sector entities at boosting LMI lending.
NCRC’s ranges are illustrative of an alternative to the agencies’ proposal. In order to ensure rigorous grading, proposed ranges corresponding to ratings must be developed via data analysis of bank performance using HMDA and CRA small business data in a sample of metropolitan and rural areas. The NPRM does not indicate that the agencies conducted data analysis of this sort, which is the only valid way to develop a ratings methodology.
Summing Up the Ratings: Can Fail in Almost 50% of AAs and Still Pass
Under the NPRM, banks could fail in up to one half of AAs on their evaluation measure and retail distribution test, and still receive an overall Satisfactory or even Outstanding rating. This would exacerbate banking and credit deserts since banks could focus on passing in AAs where they considered it easier to conduct business. They would likely emphasize the larger areas with more population, higher employment levels, income levels, and more of a well-established infrastructure to facilitate banking activities. Cities and counties would now be competing against each other for scarce CRA dollars. The winners would likely be the larger and expensive coastal cities to the detriment of the inland parts of the country, not because they have more needs but because the banks’ CRA evaluation measures would encourage larger dollar deals. We believe the proposal would exacerbate the problem of inadequate lending in rural communities and non-MSA assessment areas.
In contrast to the NPRM, the current exams do not allow banks to escape consequences if they perform poorly in a large portion of their AAs. The current exams divide AAs into full scope and limited scope AAs. The full scope AAs are weighted more heavily in determining the final rating. However, if bank performance in the limited scope AAs is considerably worse as revealed by the quantitative performance measures, banks can experience ratings downgrades. Therefore, banks have been subject to penalties when performance across AAs was poor. The agencies have not demonstrated any reason to relax this accountability.
More robustness needed regarding the evaluation of discriminatory and illegal practices
The agencies offered a cursory explanation of their fair lending reviews under proposed § 25.15. They stated, “Specifically, in assigning a CRA rating, an agency would first evaluate a bank’s performance for the applicable time period and then make any adjustments to the presumptive rating that would be warranted based on evidence of discriminatory or other illegal credit practices, consistent with the relevant agency’s policies and procedures.” This description provided no useful information to the public as to the agencies’ intentions regarding their fair lending reviews and whether they would be improved. The agencies must improve upon the cursory fair lending reviews that now usually consist of just a few sentences affirming that examiners found no discrimination or other illegal activity.
The fair lending section should describe how the examiner confirmed compliance with a range of anti-discrimination and consumer protection laws. For example, compliance with the American with Disabilities Act (ADA) must be rigorously tested, including issues associated with physical access to a bank as well as whether people with disabilities are treated in a non-discriminatory manner when they apply for a loan, deposit account, or other banking service. The NPRM discusses an example of a proposed qualified CRA activity that would be an investment in a bond to finance sidewalk improvements to comply with ADA. However, the NPRM does not discuss banks’ compliance with ADA regarding how they treat customers.
The fair lending section must include data analysis, which would probe for any evidence of discrimination. Before the reforms to CRA regulations in the mid-1990s, CRA exams considered lending to people of color in Factors D and F on CRA exams. For example, the 1996 Federal Reserve Bank of Richmond exam of Signet Bank analyzed the percentage of applications from people of color for the bank, its affiliated mortgage company, and all lenders in assessment areas in Maryland, Virginia, and Washington DC. Denial rates of applications from people of color and whites were also analyzed. The fair lending review conducted a statistical analysis of rejected white and minority loan applications in order to assess whether denials were due to illegal discrimination. Similarly a 1995 OTS CRA exam of CenFed Bank examined applications and originations of the bank compared to demographics, concluding that while the agency could not find evidence of discrimination, the bank’s percent of applications and originations to Hispanics was considerably below the percent of the population that was Hispanic. Finally, the commonwealth of Massachusetts conducts CRA exams that include HMDA data analysis of lending to people of color.
Predatory and abusive practices must also be penalized on CRA exams. For example, in high cost areas of the country, abusive multifamily lending in LMI tracts has facilitated the displacement and eviction of LMI tenants. In response to concerns raised by NCRC members and others, banks had implemented reforms to their multifamily lending practices and state agencies had issued guidelines to ensure responsible multifamily lending. For example, New York state advised banks to conduct due diligence of landlords and property owners, assess if appraisals were accurate, and analyze loan terms and conditions to make sure that current rents would not have to increase substantially in order for property owners to repay loans. CRA examiners must monitor banks and penalize them on CRA exams if they are financing abusive activities in LMI census tracts and also disallow community development data being reported that includes predatory financing. Examiners must consult with local public agencies and community organizations to determine if regulatory, statutory, or voluntary standards exist regarding multifamily or other CRA-related financing.
Question 14. The proposed rule would define retail domestic deposits as total domestic deposits of individuals, partnerships, and corporations, as reported on Schedule RC-E, item 1, of the Call Report, excluding brokered deposits. Is there another definition—including the alternatives described above—that would better reflect a bank’s capacity to engage in CRA qualifying activities?
Answer to question 14: definition of deposits should not exclude municipal deposits
The NPRM would exclude municipal deposits from the definition of deposits. The reasoning appears to be that CRA was intended to make sure that deposits of bank customers were reinvested back into the community. Municipal deposits, however, are a form of community wealth. Cities derive much of their revenues from taxing their residents via a variety of taxes and fees. Therefore, municipal deposits in a broad sense reflect the resources of actual and potential customers of banks. They therefore should be included as an available resource from which banks can satisfy their community reinvestment obligations. The failure to include municipal deposits within this definition is arbitrary and inadequately supported by the record before the agency.
Question 15. The proposal focuses on quantifying qualifying activities that benefit LMI individuals and areas and quantifies a bank’s distribution of branches by increasing a bank’s quantified value of qualifying activities divided by retail domestic deposits (a bank’s CRA evaluation measure), expressed as a percentage, by up to one percentage point based on the percent of a bank’s branches that are in specified areas of need. Banks with no branches in these areas will not receive any CRA credit for their branch distribution under this method, even if there are very few specified areas of need in the areas they serve. Does this appropriately incentivize banks to place or retain branches in specified areas of need, including LMI areas? Does it appropriately account for the value of branches in these areas?
Answer to question 15: branches should not be quantified in the manner suggested by NPRM and this quantification would be no substitute for the service test
The numerator of the CRA evaluation measure in proposed § 25.10(b)(2) would include a measure capturing bank branches in LMI areas and other underserved areas: the percentage of all branches that are in the specified areas multiplied by .01. This branch measure is supposed to provide some weight to the importance of branches by increasing the ratio by as much as one percentage point. This provides far less weight than the current service test of the large bank exam, which now counts for 25% of the rating. This would be replaced by using branches as part of the numerator of the CRA evaluation measure, which in most cases would count for considerably less than 25% of the CRA evaluation measure. Thus, the proposed branch formula would undermine the importance of branches, contrary to the CRA finding that “the convenience and needs of communities include the need for credit services as well as deposit services.”
Under the new formula for branches, a bank with 30% of their branches in LMI census tracts, which would be a relatively high percentage of branches in LMI census tracts, would only receive a branch score of .3 percentage points in the CRA evaluative measure. This would be a mere 5% of a presumptive satisfactory rating or 2.7% of a presumptive outstanding CRA rating.
In an extreme example, a bank with 100% of its branches in LMI tracts would receive a branch score of one percentage point in the CRA evaluation measure. If this increased the ratio from 10% to 11%, the branch score would be 9% of the CRA evaluation measure. A 9% weight for the branch score would be less than one half of the previous service test weight of 25%. This would significantly reduce the weight of branches on CRA exams since this example illustrates the greatest possible weight for branches.
Moving to this approach would greatly diminish the importance of bank branches in CRA compliance, which would likely lead to significant branch loss in LMI communities. Recent Federal Reserve research documented that the current service test that explicitly examines branch distribution across census tracts had prevented the closures of economically viable branches in LMI tracts. A test regarding consideration of branches that would be more opaque and less weighted would be less successful in preserving branches that remain a vital means of bank access to populations with limited incomes and mobility. Furthermore, as discussed above, monetizing bank deposit accounts and adding this to the CRA evaluation measure would also undervalue the importance of bank accounts, particularly when banks can seize opportunities to finance the newly eligible large and “essential” infrastructure projects.
Question 16. Under the retail lending distribution tests, the proposal would consider the borrower distribution of any consumer loan product line that is a major retail lending product line for the bank. The agencies defined a major retail lending product line as a retail lending product line that comprises at least 15 percent of the bank-level dollar volume of total retail loan originations during the evaluation period, but also considered setting the threshold between 10 and 30 percent. Should the agencies consider a different threshold? Additionally, applying the retail lending distribution test to only major retail lending product lines means that not all retail lending product lines will be evaluated for every bank. Are there any circumstances in which applying the retail lending distribution test to a consumer lending product line should be mandatory, even if it is not a major retail lending product line (e.g., if the consumer lending product line constitutes the majority of a bank’s retail lending in number of originations)? Additionally, the proposal would only apply the retail lending distribution tests in assessment areas with at least 20 loans from a major product line. Is 20 loans the appropriate threshold, or should a different threshold, such as 50 loans, be used?
Answer to Question 16: threshold for retail test should be established in reference to lending levels in the assessment area
In proposed § 25.11, the agencies contemplate evaluating a product line if it constitutes at least 15% of total retail lending of the bank. NCRC believes this would be an incorrect threshold for evaluating a retail product. A bank could be a major lender in a locality even if the product line was not 15% of its overall loan portfolio. Moreover, the 15% threshold could result in omitting products from evaluation that still constituted high volumes of loans, particularly for the largest banks, even if the product was not 15% of the bank’s total portfolio
The threshold for evaluation at the AA level should be solely how many loans of a product line the bank issued in the AA. NCRC believes that the threshold of 20 loans would be appropriate in proposed § 25.11 as 20 observations is typically the number used in analysis to generate statistically meaningful results. A higher threshold would omit from evaluation lending that was either a significant share of total lending in a smaller or rural AA, particularly underserved areas, or lending that was a sizable percentage of a smaller bank’s overall lending. Too high of a threshold would violate CRA’s mandate to evaluate a bank’s responsiveness to needs in a locality.
Question 17. Under the proposal, a bank evaluated under the general performance standards could not receive a satisfactory or an outstanding presumptive bank-level rating unless it also received that rating in a significant portion of its assessment areas and in those assessment areas where it holds a significant amount of deposit. Should 50 percent be the threshold used to determine “significant portion of a bank’s assessment area” and “significant amount of deposits” for purposes of determining whether a bank has received a rating in a significant portion of its assessment areas? Or should another threshold, such as 80 percent, be used?
Answer to Question 17: allowing banks to pass when failing in 50% of AAs would violate CRA’s purpose
The agencies proposed to allow banks to fail in up to 50% of their AAs and still pass overall at the bank level. For the reasons described above, this proposal would be an unacceptable violation of CRA’s statutory requirement that banks serve local needs. While 80% would be preferable to 50%, NCRC believes that no threshold should be established because banks must be held accountable for performance in all AAs. A threshold would tempt banks to neglect the hardest to serve AAs, which would probably be the smaller, underserved, or economically depressed areas. A threshold would therefore exacerbate the disparities among CRA hot spots and underserved deserts that the proposal purports to address.
In addition to no threshold, a more refined rating system would benefit both banks and community groups. Instead of pass/fail for the proposed retail test, the current ratings for performance in AAs must be retained and expanded. The agencies currently maintain an archaic point system of 1 to 24 that translate into CRA ratings. This point system could be converted to 1 to 100 and points could be assigned to performance in an AA. For example, 90 to 100 could correspond to Outstanding performance in an AA, 80 to 89 to High Satisfactory, 70 to 89 to Low Satisfactory, 60 to 69 to Needs to Improve and below 60 to Substantial Noncompliance. The scores for the AAs could then be averaged to provide an overall rating. A bank would not need to pass in every AA but would need to do well enough in the great majority of AAs in order to pass overall.
Also, if the final CRA rating remains one of four ratings, the point system would provide more gradation in performance overall and across AAs. The AAs with points corresponding to Low Satisfactory and below would be subject to more attention by the bank, working in collaboration with its regulator and community-based organizations (NCRC has also advocated for a point system of this nature for the component tests).
This approach would avoid the need for an arbitrary threshold and would make each and every AA an important contributor to the overall grade. At the same time, it provides some allowance for inevitable unevenness in performance as a result of recent market entry in some AAs or some other institutional reason. However, it would also focus attention where performance is poor and encourage improvements.
Question 18. Under the proposal, banks that had assets of $500 million or less in each of the previous four calendar quarters would be considered small banks and evaluated under the small bank performance standards, unless these banks opted into being evaluated under the general performance standards. Is $500 million the appropriate threshold for these banks? If not, what is the appropriate threshold? Should the threshold be $1 billion instead?
Answer to Question 18: smaller banks must not be able to opt out of new exams
The FDIC and OCC provided an option for small banks with assets under $500 million to be evaluated under a streamlined small bank exam that only had a lending test or to be evaluated under the proposed tests. The asset level would be adjusted annually to take inflation into account. As of the most recent Call Report data (December 2019), 72% of all banks or 3,725 banks would be designated as small banks. The agencies justified this by the so-called burden and costs of collecting new data required for the new tests. However, the agencies said that available data and their own analysis conclude the small banks would perform better on the new tests than their larger counterparts. This would be possible because while the small banks probably do not engage in as much community development financing as a percent of their deposits as their larger counterparts, a higher percentage of their retail lending might be for LMI borrowers, small businesses and farms.
It is not justified to provide an option for small banks to be excluded from the new exams if the agencies think they would do as well or better than their larger counterparts of the exam. The agencies would be failing in their obligation to promote banks meeting community needs. In this case, the agencies would miss an opportunity to preserve small bank retail lending while increasing their community development financing.
To make matters worse, the agencies asked whether the threshold for small banks should be increased to $1 billion. If the asset threshold is $1 billion, 84% of banks or 4,382 banks would qualify as small banks. This would surely lead to the loss of hundreds of millions of annual community development financing as documented in a NCRC study. NCRC estimated that intermediate small banks (ISB) provided $3 billion annually in CD financing. Much if not all of this would be lost if most ISB banks had an option for just a lending test instead of a lending and CD test. Most ISB banks would have this option should the threshold be raised to $1 billion. A rigorous cost-benefit analysis would not justify this threshold. While the ISB category imposes costs on smaller banks, the ISB banks have shown that these costs are manageable while they have also engaged in a significant level of community development financing.
Question 19. Under the proposal, small banks (i.e., banks with $500 million or less in assets in each of the previous four calendar quarters) may choose to exercise an opt into and a one-time opt out of the general performance standards. Should small banks that opt in to the general performance standards be permitted to opt out and be examined under the small bank performance standards for future evaluations and, if so, how frequently should this be permitted?
Answer to question 19: Intermediate small banks must not opt out of test with CD requirements
Per our response to Question 18, NCRC believes that current ISB banks must not be allowed to opt out of a test that includes CD responsibilities. The asset thresholds for qualifying for the small bank test must be the same as they are now. The agencies lack a justification for allowing more ISB banks to qualify as small banks since the ISB banks perform acceptably if not well on their current tests and provide important levels of CD financing for underserved communities. The agencies also have not produced any documentation of any onerous burden associated with the ISB test. Thus, the question posed about opting in and out of a small bank test for ISB banks would not be relevant to NCRC’s position on this matter.
Data Collection and Availability: Data Must be Robust and Publicly Available
The NPRM in § 25.19 would institute new data collection requirements for banks but would not make much of this new data publicly available. The statutory purpose of CRA is to require banks to meet local credit needs for lending and banking services. An essential means of enforcing this purpose and holding banks accountable is public data availability. It would not be sufficient for just the regulatory agencies to have this new data. The public must also have access to this data so they can determine for themselves the extent to which banks are responding to credit needs.
Furthermore, publicly-available data promotes collaboration and discussion among banks and community stakeholders. All parties have a common factual basis (the publicly available data) from which to discuss bank performance and identify strengths and weaknesses in performance. The new data would show gaps in bank performance, whether the shortfalls are retail lending or CD investing and lending, that could be addressed through bank and community partnerships. Failure to release robust data publicly frustrates the purpose of banks meeting community needs through collaboration and partnerships with community stakeholders.
The agencies recognized the benefits of data in fostering collaboration. They stated that “industry-wide reporting would enable more effective stakeholder dialogue regarding the distribution and volume of CRA activity.” Yet, the proposed data dissemination would thwart this worthy objective by needlessly constraining data dissemination.
The NPRM would require banks to collect new data on deposits by customer location and data on community development lending and investing. It refers to “certain” data being available to the public. In the proposed regulations, the agencies would provide data to the public on a county level for retail lending such as consumer lending but not data on community development financing, which would only be available at the bank level.
Public reporting of CD financing at a bank level would provide only cursory information of limited usefulness. Members of the public could gauge whether banks as a whole were providing adequate or above adequate levels of CD financing but would be unable to determine whether banks were providing sufficient levels of CD financing within particular AAs or in underserved counties. The lack of this data would completely frustrate the public’s ability to determine if a bank was responding to local community development needs with CD financing. CD data must be disseminated at least on a county level and preferably on a census tract level to more precisely determine whether various underserved neighborhoods were being served.
The agencies proposed that banks collect the data on a census tract level so no technological or institutional impediment would thwart dissemination on a census tract level. Moreover, the usual privacy considerations are moot considering that CD projects are large scale such as a multifamily dwelling, a community facility like a childcare center or a small business corridor. Data on CD projects do not present privacy sensitivities because it usually does not involve individuals as borrowers but more likely corporate or nonprofit entities. Data suppression by the agencies therefore does not appear justified.
Likewise, the new data on consumer lending would be restricted in its public dissemination. It would be disseminated only on a county level. This is in contrast to the current HMDA data for home lending which is disseminated publicly on a census tract level. The census tract level dissemination is optimal in terms of public use of the data to hold banks accountable for lending in LMI and underserved communities. Since HMDA data has not experienced a privacy breach in the more than 40 years of public dissemination, NCRC believes consumer loan data could likewise be publicly released on a census tract level.
The same level of public dissemination must be employed also for deposit data, particularly given its vital role in AA determination. Members of the public as well as agencies must be able to verify the communities in which banks would be required to serve as AAs. This would help the public to know in which communities banks would be evaluated.
The agencies proposed to delete CRA public file reporting of HMDA data since the banks would be submitting other home loan data to the agencies. NCRC does not understand why the agencies proposed to move away from HMDA data. Used over four decades, HMDA data provides an essential common factual framework for banks, regulatory agencies, and the public with which to evaluate bank performance. HMDA data facilities discussion, common understandings, and beneficial collaborations for addressing credit gaps afflicting subgroups of borrowers and communities. It does not make sense to discard this common data source.
The agencies discussed in their preamble that they intend to include construction lending as part of home mortgage lending. The agencies suggested that the inclusion of construction lending would somehow necessitate moving away from HMDA data as the CRA data source for home lending. It is unfathomable how Call Report data could replace HMDA data since it does not have information on the income of borrowers. If the agencies want to include construction loans as part of home lending, they should describe a supplemental method for banks to collect and report that data to the agencies. This supplemental reporting would only be for construction loans and HMDA data would still be the source for other home lending.
The agencies stated that banks could make the contents of their CRA public file, including data, available on their websites. For members of the public that request hard copies of the public file information, the NPRM would allow banks to charge fees. This is not appropriate. Members of the public requesting hard copies would likely be people with LMI that do not have easy access to the internet.
The agencies would make it needlessly harder for these members of the public and make them incur significant costs to access information in order to determine whether banks were serving the needs of their communities as required by law. This is contrary to the spirt of CRA. The agencies must rescind this advice and instruct banks to make the CRA public file information available at no cost in hard copy form to members of the public.
Question 20. As discussed above, the proposal would require banks to collect and report additional data to support the proposed rule. Although most of this data is already collected and maintained in some form, some additional data collection may be required. For example, banks may need to gather additional data to determine whether existing on-balance sheet loans and investments are qualifying activities. Are there impediments to acquiring this data? If so, what are they?
Answer to question 20: website form can overcome impediments to data collection
The agencies propose to develop a website form to facilitate data submission. This would help banks in understanding what data would be required and thus assist in overcoming impediments to data collection. Moreover, a useful and understandable website resource would help the public better understand the data to be collected and how the public could effectively use the data in assessing bank CRA performance.
Question 21. What burdens, if any, would be added by the proposed data collection, recordkeeping, and reporting requirements?
- What system changes would be needed to implement these requirements?
- What are the estimated costs of implementing these requirements?
Question 22. The proposal would require small banks to collect and maintain certain deposit-based assessment area data. Are there other ways the agencies can limit the recordkeeping burden associated with the designation of deposit-based assessment areas, including other ways for banks to differentiate between traditional and internet type business models?
Answers to questions 21 and 22: address burdens by not discarding HMDA data and requiring new retail home loan reporting
The agencies suggested that they might exempt small banks from the new deposit collection requirements if the banks could demonstrate via other methods whether or not they would be required to create deposit-based AAs. NCRC does not think this is appropriate. Data would be the only way to confirm whether a bank would be required to create additional AAs. As discussed above, this data must be reported publicly so the public has information about where the bank would be evaluated for its CRA obligations. The agencies could provide information on how to collect this data most efficiently and could conduct webinars and seminars instead of exempting institutions from this requirement.
A significant issue that the NPRM does not clarify is the extent to which the NPRM moves away from current databases and would require banks to submit entirely new and different data. It is not clear, for example, whether HMDA data is being abandoned altogether or would it be used on the retail test but not the CRA evaluation measure. How much new balance sheet information would banks be required to collect? A CRA reform effort should focus on improving existing and publicly available data sets instead of requiring new balance sheet data of uncertain value.
Public Accountability is a Vital Enforcement Mechanism of CRA
CRA is most effective when banks are publicly accountable for their performance and regulatory agencies are likewise publicly accountable for rigorous implementation of CRA. The NPRM proposes to consider public comment in a constrained manner and stretching out CRA exam cycles for banks with Outstanding ratings. These are proposals that would reduce accountability. In addition to responding to these misguided proposals, NCRC offers recommendations below for bolstering accountability.
Public comment must be facilitated, not constrained
The NPRM refers to public comments only a few times and only in reference to needs in AAs. The NPRM leaves open the question about whether the OCC and FDIC will continue receiving community group comments on the performance of banks in adhering to their CRA obligations as contemplated by the CRA.
Regarding performance context, the proposed regulation, § 25.14 (b)(4), stated that the OCC would consider “Any written comments about assessment area needs and opportunities submitted to the bank or the OCC.” In contrast, the current regulation stated at §25.21(b)(6) in the same section about performance context “The bank’s public file, as described in §25.43, and any written comments about the bank’s CRA performance submitted to the bank or the OCC…” The current and changed language suggested a de-emphasis on public comment about bank performance in the NPRM. The OCC and FDIC must not constrain public comment in this fashion but must instruct their examiners to consider carefully public comment on a variety of issues, including the bank’s CRA performance.
The ability of the public to comment on whether banks are meeting community needs is central to the statutory purpose of CRA. The agencies need to make increasing the ease and ability of the public to comment a central part of CRA reform. Presently, agency websites do not make it easy to comment on CRA exams. The agencies must create an easy method for the public to comment on exams and must identify public liaison staff who can work with the public to receive comments and provide transparency on exam status.
Recognition of Community Benefits Agreements
CRA examiners must recognize community benefits agreements (CBAs), and assess bank progress in implementing CBAs just as they do with conditional merger approvals. NCRC works with our members and financial institutions on a collaborative process to create CBAs where nonprofit and bank leaders discuss community needs and opportunities for CRA-related financing. CBAs commit banks to increasing CRA activity, and directing it to where it is needed most, both of which are the stated goals of the FDIC and OCC in this CRA reform process. One would be hard-pressed to think of a more ideal model of CRA implementation. Yet, the regulators do not have a process for recognizing these commitments and bank progress towards completing them.
Recognition of CBAs has gained momentum lately. CBAs negotiated with NCRC have been mentioned in four recent merger approvals as evidence of how banks are meeting the convenience and needs of community members, including in the FDIC’s approval of BB&T and SunTrust. The Treasury Department recognized CBAs as an “effective tool” to “demonstrate how [merger] application[s] would benefit the communities served.” The regulators should work with community groups and banks on the development of a process for recognizing CBAs, and for their implementation to become a factor on performance evaluations.
Automatic inclusion of affiliates on CRA exams
In its memo to the federal banking agencies in the spring of 2018, the U.S. Treasury Department stopped short of calling for mandatory inclusion of affiliates but urged the agencies to evaluate their “approach to affiliates in order to ensure that performance evaluations accurately reflect the CRA-eligible activity of the overall bank.” CRA exams allow banks to either include or exclude their mortgage company affiliates on CRA exams. And it is hard to think of a process that would be more prone to abuse. The natural tendency is for affiliates to be included on evaluations if they are responsibly lending to LMI borrowers and neighborhoods and to be excluded from exams if they are not.
An example of optional inclusion enabling abusive practices is Suntrust Mortgage Company, which Suntrust excluded from its CRA exam of 2013. The U.S. Justice Department, HUD, and the Consumer Financial Protection Bureau (CFPB) reached a $1 billion settlement with the mortgage company over widespread abuses associated with underwriting FHA mortgages and mortgage servicing that occurred in the time period covered by the CRA exam. Yet, because of the optional treatment of affiliates, Suntrust’s CRA exam did not consider the mortgage company’s lending practices and whether these practices should result in a ratings downgrade. The optional treatment is inconsistent with the interconnectedness of affiliates and their parents. Suntrust’s CRA exam states, “SunTrust Mortgage Company is the primary originator of home purchase and refinance loans for the organization.” In most cases, the affiliates’ activities are inextricably connected with the banks. The optional treatment must end.
Alternatively, regulators could adjust the CRA rating of the bank if the affiliate is engaged in activity that is at wide variance from the bank such as abusive lending or not lending to modest income populations while the bank is offering a higher percentage of their loans to these populations. Disparate patterns like this could very well reflect gaming exams.
Under the NPRM, banks would be conducting their own CRA exams, which is contrary to the statute
Contrary to the statutory requirement that “the appropriate Federal financial supervisory agency shall . . . assess the institution’s record” and “prepare a written evaluation” under the CRA, under the proposal, banks would be calculating their CRA evaluation measures and results of the retail tests and then presenting results to the CRA examiners. CRA examiners become checkers and auditors rather than conducting their own analysis. The exam would become largely bank-driven with little judgment applied by the examiner. This is almost like students evaluating their performance and handing results to professors, who then rely on the students’ data to confirm results.
Instead, examination procedure must remain, as it is currently, in which the examiner uses data that has been verified by the regulatory agency to calculate results of performance measures. In addition, the agencies must instruct examiners to apply weighting, qualitative factors, and performance context. The way to make exams more rigorous and consistent is to make examination procedures and instructions uniform across agencies and to establish more thresholds for the quantitative measures. In contrast, under the NPRM, exams would become less rigorous self-graded evaluations with examiners mostly double-checking the work of the banks and applying little judgment of their own.
Proposed stretch-out for Outstanding ratings is unnecessary and would reduce bank performance
The NPRM states that banks receiving Outstanding ratings would be subject to CRA exams once every five years, as opposed to the current schedule of once every two to three years. Five years is too long a time period between exams; it fails to hold banks to the statutory requirement of “continuing” and “affirmative” obligation to meet community needs since banks would relax their commitment to CRA, especially in the first year or two at the beginning of a five-year time cycle. Also, CRA ratings are an integral part of merger reviews, which will be compromised by stale exams that are less likely to reflect recent past CRA performance.
Timely production and release of CRA exams
NCRC’s research of a sample of large bank CRA exams found that exams of very large banks had been delayed partly because of violations of consumer protection law and fair lending laws that involved lengthy investigations. The OCC had proposed releasing CRA exams 90 days after completion if a fair lending investigation was not resolved. If examiners subsequently found violations, then ratings on future exams could be lowered. NCRC has recommended instead that ratings on the most recent exam be retroactively downgraded.
The ratings downgrade must be applied as closely as possible to the years in which the transgressions occurred so the bank experienced consequences as close as possible to the time period of their misdeeds. Future penalties would be less certain since examiners could be inclined to be lenient towards transgressions that occurred in the past.
Ongoing fair lending investigations must be completed and any downward adjustment to CRA ratings must occur before banks are allowed to merge or become financial holding companies. It violates the spirit and intent of bank merger law and CRA to allow mergers or approvals of other applications when fair lending investigations are pending that could impact the decisions on applications. For example, bank holding companies are not allowed to become financial holding companies and thereby engage in non-banking financial activities if not all of the subsidiary insured depository institutions have at least Satisfactory CRA ratings. Pending fair lending investigations can impact whether the subsidiaries have Satisfactory CRA ratings.
Poorly conceived reform proposals are not needed in order to improve timeliness of exam release. Instead, improvements regarding how to handle fair lending investigations that could impact ratings are needed.
Summary of NCRC Recommendations
Immediately below is a summary and a list of NCRC’s major recommendations regarding the major sections of the NPRM:
- Regulatory definition of community development: NCRC opposes the proposal to delete the criteria of “economic development” and “revitalize and stabilize” and adding “essential infrastructure” as criterion of the regulatory definition of community development. These changes would divert banks’ attention from holistic community development that depends on a variety of affordable housing, economic development, community facilities and revitalization activities occurring simultaneously. At the same time, declaring essential infrastructure as community development would divert substantial resources from development of LMI neighborhoods into major road, bridge and other infrastructure projects.
- Athletic stadiums and other projects in Opportunity Zones: The proposed list of CRA eligible activities must not include athletic stadiums, which would consume an inordinate amount of bank financing and divert it from the statutory focus of CRA, that is, the revitalization and reinvestment of LMI communities. Activities must not count automatically if they are located in Opportunity Zones. Only projects that conform to the current regulatory definition of community development should count if they are located in Opportunity Zones.
- Including middle-income housing in high cost areas as affordable housing must be limited or eliminated: NCRC’s research found that this provision would divert financing away from LMI affordable housing in the highest cost counties, which comprise 13% of the population of the United States. These are the counties in greatest need of affordable housing for LMI households and families. If this provision remains in a final rule, it must be limited via a cap or some other mechanism, ensuring that the great majority of bank financed affordable housing is for LMI households.
- Naturally Occurring Affordable Housing (NOAH) must have safeguards: As proposed, an automatic qualification for NOAH if rent levels of housing developments are affordable to LMI tenants would likely result in much housing being occupied by middle- and upper-income tenants. The agencies must propose methods for assuring that this housing will be occupied by LMI tenants.
- Financial education must remain targeted to LMI people: As proposed, lifting income restrictions on financial education programs would result in significant numbers of beneficiaries being middle- and upper-income. The statutory purpose of CRA directs agencies to rectify redlining, which created the most need for LMI people to be educated about banks since banking had been relatively scarce in redlined communities.
- Pro rata credit must be awarded carefully: The agencies proposed a radical departure from the primary purpose or primary benefits standard generally requiring more than 50% of a community development activity to benefit LMI people or communities. Partial or pro rata credit for CD dollars when projects benefit less than a majority of LMI people must only be applied when the calculation can be conducted with precision. For example, pro rata or partial credit for a major bridge would be difficult or impossible to calculate unless a bank knew the percentage of LMI people in cars crossing the bridge. In contrast, pro rata credit can be more readily calculated for, e.g., a light rail line that has 40% of its stations in LMI tracts.
- Home lending in LMI communities must be reinstated as a criterion on the retail lending test: The agencies proposed removing this as a criterion on the lending test out of concerns about providing CRA credit for lending to affluent households in LMI tracts that were undergoing gentrification. However, other methods exist for avoiding loans to affluent households causing displacement of LMI people. Both academic researchers and NCRC have developed a method for identifying LMI tracts that are gentrifying. In these tracts, the agencies can limit or not allow credit for loans issued to middle- and upper-income borrowers. Eliminating lending in LMI tracts as a criterion on the lending test would make it harder for economically struggling LMI communities to revitalize since banks would have reduced incentives to lend in these communities.
- The proposed definition of underserved areas must be changed: The agencies presented no data analysis in the NPRM to support their definition of underserved areas. NCRC had developed another definition that focused on low levels of lending. NCRC found that this definition more effectively targets retail lending and CD financing to census tracts with the lowest levels of lending, high percentages of people of color, higher poverty rates, higher rates of unemployment and lower home values. Adding the criterion of underserved tracts as NCRC defines them would also capture some middle income areas in need of more lending and would relieve pressure on gentrifying neighborhoods.
A CRA reform effort must include a concerted effort to include communities of color more explicitly on CRA exams since a large body of research shows continuing and stark racial disparities in lending. NCRC’s proposal on underserved tracts is grounded in data analysis and addresses the significant racial disparities we found.
- The small farm and small business revenue thresholds must not be raised: The agencies present no justification for raising these thresholds, which would result in lending being diverted away from small businesses and farms. Other federal agencies’ research showed that the vast majority of the small entities had revenues of $1 million or less, which is the current revenue size threshold.
- The list of CRA eligible activities must be amended: A transparent and evenhanded process is needed for updates to the list via a public request for comments and not via a website form that only one party, the banks, fill out. Moreover, the list must be a principles-based list instead of a list of several examples, which banks would be most likely to interpret as exhaustive and thus would curtail their CRA activities. The list could be accompanied by an interactive database of examples from CRA exams.
- In terms of what should count more, prime lending must be provided more weight than high-cost or subprime lending. Moreover, innovative and flexible lending must be encouraged via retention and strengthening of the innovative and flexible criterion on CRA exams.
- The use of multipliers for CD activities would decrease the amount of bank-financed CD. If the agencies believe that some CD activities are more important than others, the favored ones can be weighted more on CRA exams. Weighting is an existing procedure and would not reduce activity like multipliers would.
- The service test must be retained and not deleted as proposed by the NPRM: CD services cannot be monetized as proposed as that would inadequately measure their value. The proposal would also greatly diminish the value of branches in the CRA exam. Branches remain vital for extending lending and banking services to LMI consumers and communities. In addition, the service test needs to collect better data on bank accounts for LMI customers and communities.
Where it Counts
- The NPRM recognition that AAs must include areas beyond bank branches is an important acknowledgment that bank activities beyond bank branches must be evaluated by CRA exams. Research has documented that CRA evaluations increase safe and sound bank lending, investing and branching for LMI people and communities. In order for CRA exams to be most effective in boosting activities for LMI people and communities, exams must capture the great majority of activity via AA reform.
- The NPRM approach of using deposits to designate AAs renders data analysis of the impacts impossible: The current deposit data collected by the FDIC is not detailed enough for the purpose of designating AAs. The agencies should have first proposed this data collection in a NPRM before issuing the CRA NPRM. Furthermore, they should have also used HMDA and CRA small business and farm data to identify areas outside of branches with high levels of activity that would be designated as AAs.
- The NPRM’s thresholds of 50% and 5% are not adequate for designating AAs. The proposed 50% threshold would be too high and would exclude areas from being designated as AAs when they have substantial volumes of lending or deposits. The overall objective is to ensure that AAs cover the great majority of lending and other bank activity. When AAs cover less than a majority, NCRC had found that ratings were inflated. The 5% threshold for designation of an AA cannot be based on percent of a bank’s total deposits but rather the bank’s market share of deposits or loans in a geographical area in order to evaluate a bank where it has significant market share and is important to an area.
- The NRPM’s proposal that activities anywhere outside of AAs would count must be rescinded. This would result in banks, particularly very large banks, searching nationwide for the easiest and largest deals. Instead, NCRC proposes using data analysis to develop a list of underserved counties that would be updated annually. In addition to allowing statewide and regional CD activities provided a bank satisfies needs in AAs, a bank can also provide CD financing in underserved counties. The NCRC proposal would be more effective in targeting CRA deserts for CD activity in contrast to the agency proposal, which would exacerbate the disparities between CRA deserts and hotspots.
- Annual data collection and reporting of CD data on a county level: The NPRM proposed annual data collection but not dissemination of data on a county level. County level data is needed to measure whether a bank is meeting CD needs in its AA before venturing outside of its AA. Also, annual data on a county level would enable more effective targeting of CRA deserts.
How it Counts
- Any reform to CRA evaluations must be grounded in data analysis: The agencies did not adequately describe the data analysis they performed in order to determine empirical benchmarks for the CRA evaluation measure. The agencies most likely could not perform rigorous data analysis. They made a number of assumptions and asked banks for additional data in a Request for Information (RFI) during the comment period of the NPRM. Because of the lack of rigorous agency data analysis and transparency regarding the data analysis, the general public cannot meaningfully comment on the NPRM. For example, the public does not have enough information to know whether the distribution of CRA ratings would change or whether CRA loans, investments and services would increase (though NCRC believes based on our own research that CRA activity for LMI people and communities would decrease).
- CRA evaluation measures cannot consist of a single metric that drives the overall rating: The CRA evaluation measure will distort CRA activity and will encourage banks to pursue the largest and easiest deals, regardless of whether localities are most in need of large-scale financing. Smaller dollar home and small business lending, which is needed in several communities, would likely decrease. Instead of using a ratio measure as the presumptive rating, a ratio such as CD financing divided by deposits (or assets or Tier 1 capital) could continue to be used on a community development (CD) test as one measure on that test, not the determinative measure. Qualitative measures must continue to be an important part of the CD or investment test since they take into account the extent to which banks are responding to local needs across AAs.
- The retail lending test must retain considerable weight: A pass/fail retail lending test as proposed in the NPRM would likely result in decreased retail lending in LMI and other underserved communities, the opposite outcome desired in a CRA reform effort. The current lending test counts for 50% of the large bank exam. A substantial weight for retail lending must remain on CRA exams since progress needs to continue in combating decades of redlining and disparities in lending. In addition, any empirical benchmarks for the retail test need to be grounded in data analysis in contrast to the proposed benchmarks which would likely result in even more grade inflation.
- All AAs must count on CRA exams: The agencies must discard the allowance for banks to fail in either 50% or some other large portion of AAs. Bank performance across all AAs must be averaged on CRA exams to determine a final rating.
- Thresholds for retail lending are too high: The NPRM’s proposal for inclusion of a retail lending product line if it constitutes 15% of the bank’s portfolio is too high and would result in lending in some communities not being evaluated although the lending could be of high volume. The threshold should be a number of loans in an AA; a threshold of 20 loans is reasonable.
- Smaller banks must not be allowed to opt out of exams which consider CD financing: Currently, intermediate small banks (ISBs) are an important source of CD financing. A reform proposal must retain the CD requirement for the ISB banks and must not exempt a subset of these banks from CD responsibilities. Since ISB banks tend to serve rural areas and small towns, cutting back on their CD responsibilities would reduce the amount of CD financing in smaller towns and rural areas.
- Fair lending exams must be more robust and include data analysis of lending to people and communities of color. Illegal and abusive lending must be penalized via lower ratings.
Accountability must be increased
- No stretch-out in exam time cycles for Outstanding ratings: This proposal would only reduce CRA-related lending, investing and services as stated above. The agencies should propose other incentives for achieving Outstanding ratings. Perhaps, banks could receive reduced premiums for FDIC insurance much as they do for high scores on their safety and soundness exams.
- Data on a county level of CD financing and HMDA-like for consumer lending: The effectiveness of CRA hinges on public data dissemination. It is not sufficient for agencies to propose additional data collection requirements and then not disseminate the new data. Data on CD financing must be released to the public on a county level so that members of the public as well as CRA examiners hold banks accountable for serving all communities, including current CRA deserts. Likewise, new consumer loan data should be disclosed like HMDA data, which in more than 40 years has not experienced a single privacy breach (at least no agency had documented that in any report to Congress).
- Affiliates of banks must be automatically on CRA exams, not at the option of the bank. Mortgage company affiliates of banks not included on CRA exams were significant abusive lenders in the years before the financial crisis. Affiliates must be encouraged to make safe and sound loans and investments by being included with banks on CRA exams. The distinction between banks and their affiliates has been blurred in recent years as affiliates often make loans on behalf of their banks.
- Public comments: The role of public comments must be elevated, not diminished. The agencies did not discuss the vital role of public input in any detail except to state that public comments on performance context would be considered. This reference could be interpreted as limiting public comment to a narrow range of issues. Public comment on bank performance and whether banks are serving needs must be an integral part of CRA exam analysis. The agencies must also make commenting on CRA exams and merger applications easier, including providing easy access to public liaison staff that can guide the public in making comments. The central point of CRA is ensuring that banks meet local needs. For agencies to ascertain that, they must listen carefully to the public.
- Recognition of CBAs: CBAs are an effective model of CRA implementation for accomplishing the NPRM’s stated goals of increasing CRA activity and directing it to where it is needed most. The regulators should work with community groups and banks on a process for recognizing CBAs, and make implementation of CBAs a factor in performance evaluations.
- Timely release of exams: Delays in the release of the largest bank CRA exams have been caused by lengthy investigations of fair lending violations. CRA exams should be released even if fair lending investigations are ongoing. Any adjustments to ratings due to fair lending violations should occur retroactively instead of on future exams.
The FDIC and OCC must rescind this NPRM and work with the Federal Reserve on an incremental approach towards strengthening CRA. The FDIC and OCC are correct in that AA reform and better data are needed to make exams more rigorous. However, the public cannot objectively assess the NPRM’s AA reform because it is based on data that does not currently exist. In addition, the new data collection required by the NPRM is not accompanied by transparency in data dissemination.
The core of the reform proposal regarding what counts on CRA exams and how to measure CRA performance would divert banks from the statutory purpose of CRA. The statute and Congressional record clearly indicate that CRA is intended to combat redlining by requiring that banks serve credit and deposit needs of communities with a focus on LMI people and communities. The FDIC and OCC proposed to broaden what counts to include an array of activities that benefits the community writ large; this would result in substantial shifts of CRA-qualifying lending and investing away from LMI borrowers and communities.
The CRA evaluation measure, a single metric or ratio of CRA activities divided by deposits, would exacerbate the shift away from LMI people and communities by encouraging the large deal over smaller financing needed in many LMI communities. Allowing banks to fail in up to one half of their AAs and garner credit anywhere in the country for CRA activities would further divert CRA lending and investing away from underserved and distressed communities, the opposite outcome of a well-thought-out CRA reform effort.
The NPRM’s lack of rigorous data analysis assessing the proposal’s impact and lack of a clear description of the agency data analysis are fatal flaws. The current proposal lacks justification and the public can only guess as to its impact on CRA ratings and the level of CRA activity to LMI people and communities. CRA is too important to our nation’s communities, efforts to fight redlining and inequality for this proposal to move forward. We urge the agencies to withdraw it and work with the Federal Reserve Board on a new proposal that adheres to the intent and purpose of the CRA statute.
Please ask us or Josh Silver, Senior Advisor, if you have any questions. Thank you for the opportunity to comment on this important matter.
Jesse Van Tol
President and Founder, NCRC
 Office of the Comptroller of the Currency (OCC) and Federal Deposit Insurance Corporation (FDIC), Joint Notice of Proposed Rulemaking (NPRM), Community Reinvestment Act (CRA) regulations, Docket ID OCC-2018-0008 & RIN 3064-AF22, https://www.regulations.gov/document?D=OCC-2018-0008-1515, p. 1204.
 NRPM, p. 1205.
 NPRM, pp. 1205-1206.
 NPRM, p. 1205.
 NPRM, p. 1207.
 This comment letter will use “CRA evaluation measure” and “proposed CRA evaluation measure” interchangeably.
 Statement by Martin J. Gruenberg, Member, FDIC Board of Directors, Notice of Proposed Rulemaking, Community Reinvestment Act Regulations, December 12, 2019, p. 1, https://www.fdic.gov/news/news/speeches/spdec1219d.pdf
 NCRC forecast: Weakening the Community Reinvestment Act would reduce lending by hundreds of billions of dollars, September, 2018, https://ncrc.org/ncrc-forecast-weakening-the-community-reinvestment-act-would-reduce-lending-by-hundreds-of-billions-of-dollars/
 NPRM, p. 1206.
 NPRM, p. 1207.
 NPRM, p. 1207.
 The SBA calculates small businesses account for 60 percent of net new jobs created from mid-2009 to mid-2012. See U.S. Small Business Administration, Frequently Asked Questions, March 2014, available via http://www.sba.gov/sites/default/files/FAQ_March_2014_0.pdf.
 NPRM, p. 1207.
 NPRM, p. 1227.
 Josh Silver, An Evaluation Of Assessment Areas And Community Development Financing: Implications for CRA reform, July 2019, https://ncrc.org/an-evaluation-of-assessment-areas-and-community-development-financing-implications-for-cra-reform/
 NPRM, p. 1207.
 NPRM, p. 1207.
 Silver, An Evaluation Of Assessment Areas, and Josh Silver, The Community Reinvestment Act and Geography, NCRC, May 2017, https://ncrc.org/the-community-reinvestment-act-and-geography/ .
 5 U.S.C. § 706(2)(A) & (D).
 Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983).
 State Farm, 463 U.S. 29, 43 (1983).
 FCC v. Fox Television Stations, 556 U.S. 502, 515 (2009).
 Department of Commerce v. New York, 139 S. Ct. 2551, 2575 (2019).
 12 U.S.C. § 2903(a)(1); 12 U.S.C. § 2901(b).
 12 U.S.C. § 2901(a)(3).
 12 U.S.C. §?2906.
 12 U.S.C. § 2903(a)(1).
 As discussed more below in response to question 17, the Proposed Rule seeks comment on whether “another threshold, such as 80 percent, be used” to define “significant portion.” 85 Fed. Reg. 1226. An 80 percent threshold would similarly fail to account for a bank’s “entire community.” 12 U.S.C. § 2903(a)(1).
 85 Fed. Reg. 1217, 1218.
 Banking Committee Hearings on S. 406, March 1977, page 6.
 Banking Committee Hearings on S. 406, March 1977, page 7
 Banking Committee Hearings on S. 406, March 1977, Testimony of the American Bankers Association, p. 315 and Senator Garn, page 324.
 85 Fed. Reg. 1218.
 12 U.S.C. § 2901(a)(3), (b).
 In the Banking Committee Hearings in March of 1977, a loan-to-deposit evaluation was considered but critiqued by lenders who felt this would not be the best way to evaluate whether a bank is meeting the credit needs of the community. Banks feared that for a given level of deposits, there might not be a proportionate loan demand by credit worthy borrowers. Thus banks, which would be unable to predict future market needs, could establish branches in areas where there was not a loan-to-deposit ratio deemed acceptable by the regulatory agencies. In an effort to meet their CRA requirements, these banks may then make loans that were not in accord with safe and sound lending policies; this could lead to necessary government bank bailouts. Hearings Before the Committee on Banking, Housing, and Urban Affairs, U.S. Senate 95-1 (March 23, 24, and 25, 1977) at 151-52.
 In making his statements before the Committee on Banking, Housing, and Urban Affairs, Henry Schechter, Director of the Department of Urban Affairs, AFL-CIO, suggested that to determine if banks have met the credit needs of their communities, the use of a single metric ratio would not be detailed enough to allow the regulatory agency to decide if credit needs have been met. Therefore, the regulatory agency should consider adopting a plan that uses more than one ratio to make this determination. The AFL-CIO supported the CRA legislation but not the use of a determinative one ratio. Ibid at 151. Also, A.A. Milligan, testifying on behalf of the American Bankers Association, seconded the concerns of the Federal Reserve Chairman Burns about inadvertently restricting lending from where it was needed most. Mr. Milligan stated: “Banks in urban areas such as Milwaukee, Chicago, or Minneapolis, that are providing the necessary funds for rural community development in Wisconsin would not be considered to be meeting the needs of their own communities. They would be labeled derelict in their responsibilities to their own communities even if their communities had no current need,” Ibid at 315. See more analysis of the CRA hearings at https://ncrc.org/the-purpose-and-design-of-the-community-reinvestment-act-cra-an-examination-of-the-1977-hearings-and-passage-of-the-cra/
 Ibid at 2.
 60 Fed. Reg. I (1995), 22168. “Performance criteria. The 1994 proposal provided that to determine whether a small institution’s CRA record is satisfactory, the agencies would consider the institution’s loan-to-deposit ratio, adjusted for seasonal variation and, as appropriate, other lending-related activities, such as loan originations for sale to the secondary markets, community development loans, or qualified investments. This provision of the 1994 proposal responded to concerns following the 1993 proposal that institutions that package and sell their loans would be disadvantaged when compared to portfolio lenders by a strict loan-to-deposit ratio test. This provision of the 1994 proposal was retained in the final rule. Evaluations would also take into account the institution’s size, financial condition, and the credit needs of its assessment area. The final rule also required consideration of the proportion of the institution’s total lending made to borrowers in its assessment area. The agencies would take into account local lending and investment opportunities in assessing this criterion.”
 12 U.S.C. §?2906.
 12 U.S.C. §?2906.
 5 U.S.C. § 553(c), 706.
 NPRM, p. 1222.
 Federal Register, Vol. 85, No. 7, Friday, January 10, 2020, Proposed Rules, p. 1286.
 While the OCC promised banks confidentiality for their specific data, data aggregations could be made publicly available.
 NPRM, p. 1237.
 Only 15 of 40 recent rules during 2016 through 2018 had cost-benefit analysis. Moreover, the agency did not engage its economists early in the rulemaking process in order to promote rigor in cost-benefit analysis. An article about the report stated, “The FDIC was also hit for not requiring its chief economist to review cost-benefit analyses before publication to ensure “accurate, sufficient, logical, unbiased” and data-based conclusions.” Brendan Pederson, FDIC not consistent in assessing impact of rules: Watchdog, American Banker, February 5, 2020, https://www.americanbanker.com/news/fdic-not-consistent-in-assessing-impact-of-rules-watchdog; See also, FDIC Office of Inspector General, Cost Benefit Analysis Process for Rulemaking, February 2020, https://www.fdicoig.gov/sites/default/files/publications/20-003EVAL.pdf
 5 U.S.C. § 553(c).
 Todd Garvey, A Brief Overview of Rulemaking and Judicial Review, Congressional Research Service, March 2017, p. 2, https://fas.org/sgp/crs/misc/R41546.pdf
 The NCRC letter can be found here: NCRC, More Than 500 Community Organizations Urge FDIC To Not Join Flawed Notice Of Proposed Rulemaking For Community Reinvestment Act, December 11, 2019, https://ncrc.org/more-than-500-community-organizations-urge-fdic-to-not-join-flawed-notice-of-proposed-rulemaking-for-community-reinvestment-act/
 Relatedly, we reiterate our concern, previously expressed in a letter from our outside counsel, that OCC has failed to include detailed summaries of all conversations it has had with industry stakeholders. See Letter for Joseph M. Otting from Nitin Shah & Jeffrey Dubner, Democracy Forward Foundation, March 17, 2020 (submitted to public docket). Without detailed descriptions of these calls in the record, the public is unable to review the entire basis for the agency’s proposed rule and meaningfully evaluate and respond to the proposal.
 OCC, Summary of Proposal to Modernize Community Reinvestment Act Rules https://www.occ.gov/topics/consumers-and-communities/cra/summary-of-cra-proposed-rule.pdf. The version of the document on OCC’s website does not include the matrix.
 See two videos: the first one is produced by the OCC and the second one by John Hope Bryant interviewing senior OCC staff: https://www.youtube.com/watch?v=rdEclUWxbpw and https://www.youtube.com/watch?v=JXl57u8sK3k&t=552s.
 12 U.S.C. §?2905.
 12 U.S.C. §?2901(b).
 Id. § 2903(a)(1).
 Lei Ding and Leonard Nakamura, Don’t Know What You Got Till It’s Gone: The Effects of the Community Reinvestment Act (CRA) on Mortgage Lending in the Philadelphia Market, Working Paper No. 17-15, June 19, 2017, https://www.philadelphiafed.org/-/media/research-and-data/publications/working-papers/2017/wp17-15.pdf and Lei Ding, Raphael Bostic, and Hyojung Lee, Effects of the CRA on Small Business Lending, Federal Reserve Bank of Philadelphia, WP 18-27, December 2018, https://www.philadelphiafed.org/-/media/research-and-data/publications/working-papers/2018/wp18-27.pdf.
 123rd Cong. Rec. 17630 (1977) (statement of Sen. Proxmire).
 Robert E. Litan, Nicolas P. Retsinas, Eric S. Belsky, Susan White Haag, The Community Reinvestment Act After Financial Modernization: A Baseline Report, U.S. Treasury Dept., April 2000.
 Ibid; Joint Explanatory Statement of the Managers of the Committee of Conference, Public Law 95-128, Subcommittee on Housing and Community Development of the Committee on Banking, Finance and Urban Affairs, House of Representatives, 95th Cong., 1st Sess. Washington, D.C.: U.S. Government Printing Office (October 1977), p. 92.
 Ibid. at note 7 U.S Treasury Baseline Report.
 Federal Register, May 4, 1995, Volume 60, Number 86 https://www.fdic.gov/regulations/community/community/crapreamb.txt
 Interagency Q&A, Federal Register, July 2016, Q&A, §_.12(h)—8:, p. 48530.
 Interagency Q&A, Federal Register, July 2016, §__.12(g)—2, p. 48525.
 Interagency Q&A, Federal Register, July 2016, §__.12(g)(2) and §__.12(g)(3)—1:, p. 48526.
 NPRM, p. 1213.
 NPRM, p. 1233.
 Interagency Q&A, Federal Register, July 2016 §__.12(g)(3)—1:, p. 48526.
 Federal Reserve Bank of Dallas and Kansas City, Engaging Workforce Development: A Framework for Meeting CRA Obligations, https://www.kansascityfed.org/community/workforce/engaging-workforce-cra-framework.
 NPRM, see list of proposed CRA activities that correspond to §§ 25.04(c)(9) and 345.04(c)(9), p. 1233.
 NPRM, p. 1210.
 NPRM, p. 1211.
 Interagency Q&A, Federal Register, July 2016, Q&A, §_. 12(h)—8, p. 48530. Generally, a majority of dollars were devoted to community development, which had been focused on LMI people and communities.
 NPRM, p. 1230.
 Harry Jaffe, The Insane Highway Plan That Would Have Bulldozed DC’s Most Charming Neighborhoods, in the Washingtonian. October 2015, https://www.washingtonian.com/2015/10/21/the-insane-highway-plan-that-would-have-bulldozed-washington-dcs-most-charming-neighborhoods/.
 NPRM, p. 1234.
 Noah Buhayar and Jesse Hamilton, Financing for Sports Stadiums Could Count as Helping the Poor, Bloomberg, December 16, 2019, https://www.bloomberg.com/news/articles/2019-12-16/banks-may-call-nfl-stadium-financing-aid-to-poor-in-rule-change. Also, see Jimmy Atkinson, These 18 NFL stadium neighborhoods are eligible for the Opportunity Zones tax break, Opportunity Zones Database, February 2019, https://opportunitydb.com/2019/02/nfl-stadiums-eligible-for-opportunity-zones-tax-break/.
 Column in Sports Illustrated, Development of Sports Facilities Usually Regarded as a “Net Negative” for Local Municipalities, Multi-Use Arenas Can Be the Exception, https://www.si.com/johnwallstreet/sports-business/community-reinvestment-act-stadiums; Scott A. Wolla, The Economics of Subsidizing Sports Stadiums, Federal Reserve Bank of St. Louis, May 2017, https://research.stlouisfed.org/publications/page1-econ/2017-05-01/the-economics-of-subsidizing-sports-stadiums/; Andrew Zimbalist and Roger G. Noll, Sports, Jobs, & Taxes: Are New Stadiums Worth the Cost?, Brookings Institution, 1997, https://www.brookings.edu/articles/sports-jobs-taxes-are-new-stadiums-worth-the-cost/.
 NCRC, CRA Reform Proposals Would Count Sports Stadiums As Community Development!, January 2020, https://ncrc.org/cra-reform-proposals-would-count-sports-stadiums-as-community-development/.
 Christopher Hong, City Council unanimously approves $90 million improvement to EverBank Field, the Florida-Times Union, December 2015, https://www.jacksonville.com/news/metro/2015-12-08/story/city-council-unanimously-approves-90-million-improvement-everbank-field.
 Market shares computed by the FDIC Deposit Market Share reports, https://www7.fdic.gov/sod/sodMarketBank.asp?barItem=2.
 Statement of Joseph M. Otting, Comptroller of the Currency, before the Committee on Financial Services, United States House of Representatives, January 29, 2020, pp. 9-10, https://www.occ.gov/news-issuances/congressional-testimony/2020/ct-occ-2020-9-written.pdf.
 NPRM, p. 1211.
 Interagency Q&A, Federal Register, July 2016, Q&A, §_.12(h)—8:, p. 48530.
 Middle-income housing was discussed on pages 1210, 1211, 1231, 1241, 1242 of the NPRM, and p. 1241 defines high-cost area.
 NPRM, p. 1231.
 NPRM, p. 1210.
 NPRM, p. 1211.
 NPRM, p. 1241.
 NCRC, For The Trump Administration: Affordable Housing Means Middle-Income Housing For Counties With 43 Million People, January 2020, https://ncrc.org/for-the-trump-administration-affordable-housing-means-middle-income-housing-for-counties-with-43-million-people/.
 Poverty rates of Humboldt, Tulare, are above 20%; see Census look-up tables https://www.census.gov/quickfacts/tularecountycalifornia; https://www.census.gov/quickfacts/humboldtcountycalifornia; https://www.census.gov/quickfacts/maderacountycalifornia.
 NPRM, p. 1230.
 Comment letter in response to the CRA ANPR of National Association of Affordable Housing Lenders, p. 11, see https://naahl.org/wp-content/uploads/2018/12/NAAHL-CRA-Comment-FINAL.pdf.
 FDIC National Survey of Unbanked and Underbanked Households, 2017, p. 19, https://economicinclusion.gov/downloads/2017_FDIC_Unbanked_HH_Survey_Report.pdf.
 Rakesh Kochhar and Anthony Cilluffo, How wealth inequality has changed in the U.S. since the Great Recession, by race, ethnicity and income, Pew Research Center, November 2017, people of color in the lower income ranges had the lowest levels of wealth and homeownership rates among all population groups, https://www.pewresearch.org/fact-tank/2017/11/01/how-wealth-inequality-has-changed-in-the-u-s-since-the-great-recession-by-race-ethnicity-and-income/.
 Jason Richardson, Bruce Mitchell, PhD, Juan Franco, Shifting Neighborhoods: Gentrification and Cultural Displacement in American Cities, NCRC, March 2019, https://ncrc.org/gentrification/. This report describes a methodology of how to categorize tracts as gentrified.
 Richard Florida, America’s Biggest Problem Is Concentrated Poverty, Not Inequality, August 2015, in Citylab, https://www.citylab.com/equity/2015/08/americas-biggest-problem-is-concentrated-poverty-not-inequality/400892/.
 Ding and Bostic, Effects of the CRA..
 Richardson and Mitchell, Shifting Neighborhoods.
 Elizabeth Kneebone, Brookings Institution, The Growth and Spread of Concentrated Poverty, 2000 to 2008-2012, July 2014, https://www.brookings.edu/interactives/the-growth-and-spread-of-concentrated-poverty-2000-to-2008-2012/ and Daniel Friedman, The Rise of Income Segregation in Post-Recession America, August 2017, in Harvard Political Review, https://harvardpolitics.com/united-states/the-rise-of-income-segregation-in-post-recession-america/.
 New Research on Mobility: Studies by Profs. Chetty, Hendren, and Katz, May 4, 2015 See http://economics.harvard.edu/news/new-research-mobility-studies-profs-chetty-hendren-and-katz
 NPRM, p. 1242.
 Regulatory Background, distressed and underserved tracts, FFIEC, see https://www.ffiec.gov/cra/pdf/Regulatory%20Background%20-%20Distressed%20and%20Underserved%20Tracts%20FINAL.pdf.
 Bruce Mitchell, PhD and Josh Silver, Adding Underserved Census Tracts As Criterion On CRA Exams, NCRC, January 2020, https://ncrc.org/adding-underserved-census-tracts-as-criterion-on-cra-exams/.
 Aaron Glantz and Emmanuel Martinez, For People of Color, Banks are Shutting the Door on Homeownership, February 15, 2018, https://www.revealnews.org/article/for-people-of-color-banks-are-shutting-the-door-to-homeownership/; NCRC, Foreclosure in the Nation’s Capital: How Unfair and Reckless Lending Undermines Homeownership, April 2010, https://ncrc.org/foreclosure-in-the-nations-capital-how-unfair-and-reckless-lending-undermines-homeownership/; Even among LMI borrowers there was substantial disparity in several key data points among races. In 2018, NCRC observed that the rate spread, the difference in the interest rate charged to the borrower from the average prime offer rate (APOR), was much higher for LMI Black and Latino borrowers compared with LMI White and Asian borrowers. LMI Black and Latino borrowers had rate spreads that were 50% to 100% higher than their White and Asian counterparts. LMI Black and Latino borrowers were also less likely to have their application result in a home purchase. While 70% of LMI white applicants and 69% of LMI Asian applicants closed on their loans, those figures fell to just 57% and 63% for LMI Black and Latino applicants, respectively.
 Kevin A. Park and Roberto G. Quercia, Who Lends Beyond the Red Line? The Community Reinvestment Act and the Legacy of Redlining, a Penn Institute for Urban Research working paper, September 2019, https://penniur.upenn.edu/uploads/media/Park_Quercia.pdf.
 NPRM, p. 1211.
 Consumer Financial Protection Bureau (CFPB), Key Dimensions of the Small Business Lending Landscape, p. 10, May 2017, https://www.consumerfinance.gov/data-research/research-reports/key-dimensions-small-business-lending-landscape/.
 CFPB, Key Dimensions, p. 10.
 Ibid., p. 8.
 The Small Business Administration (SBA) estimates that small businesses employ about 85% of all workers in the service sector, see Office of Advocacy, SBA, 2018 Small Business Profile, p. 3, https://templatelab.com/small-business-profiles-us/
 NPRM, p. 1233.
 This article, https://en.wikipedia.org/wiki/Corporate_farming, discusses that some family farms are the largest corporate farms in the country. The article states, “‘Family farm’ and ‘corporate farm’ are often defined as mutually exclusive terms, with the two having different interests. This mostly stems from the widespread assumption that family farms are small farms while corporate farms are large-scale operations. While it is true that the majority of small farms are family owned, many large farms are also family businesses, including some of the largest farms in the US.”
 2017 Census of Agriculture Highlights: Farm Economics – Value of Production, number of farms, and income down slightly, April 2019, https://www.nass.usda.gov/Publications/Highlights/2019/2017Census_Farm_Economics.pdf.
 NCRC comment on the Advance Notice of Proposed Rulemaking, November 2018, see section of comment titled, Evaluations Of Small Business Lending Must Be Focused On The Smallest Businesses, https://ncrc.org/ncrc-comments-regarding-advance-notice-of-proposed-rulemaking-docket-id-occ-2018-0008-reforming-the-community-reinvestment-act-regulatory-framework/.
 NPRM, p. 1213.
 5 U.S.C. § 551(4).
 Chrysler Corp. v. Brown, 441 U.S. 281, 301 (1979).
 85 Fed. Reg. 1213.
 85 Fed. Reg. 1213.
 85 Fed. Reg. 1214.
 NPRM, p. 1213.
 5 U.S.C. § 553(c).
 NPRM, p. 1230.
 CFPB, What is a “higher-priced mortgage loan?”, September 2017, https://www.consumerfinance.gov/ask-cfpb/what-is-a-higher-priced-mortgage-loan-en-1797/.
 Jason P. Brown, Kansas City Federal Reserve Bank, Auto Loan Delinquency Rates Are Rising, but Mostly among Subprime Borrowers, The Macro Bulletin, August 2018. https://www.kansascityfed.org/en/publications/research/mb/articles/2018/auto-loan-delinquency-rates-rising.
 Elizabeth Laderman and Carolina Reid, Federal Reserve Bank of San Francisco, “CRA Lending during the Subprime Meltdown” in Revisiting the CRA: Perspectives on the Future of the Community Reinvestment Act, a Joint Publication of the Federal Reserve Banks of Boston and San Francisco, February 2009, p. 122, and Neil Bhutta and Daniel Ringo, Assessing the Community Reinvestment Act’s Role in the Financial Crisis, Feds Notes, May 2015, https://www.federalreserve.gov/econresdata/notes/feds-notes/2015/assessing-the-community-reinvestment-acts-role-in-the-financial-crisis-20150526.html.
 This term is similar to the term used in the following Federal Reserve report. That report found that this lending was safe and sound. The Performance and Profitability of CRA-Related Lending: Report by the Board of Governors of the Federal Reserve System, submitted to the Congress pursuant to section 713 of the Gramm-Leach-Bliley Act of 1999, July 17, 2000, http://www.federalreserve.gov/boarddocs/surveys/craloansurvey/cratext.pdf, p. 66.
 BB&T CRA exam, January 2017, https://www5.fdic.gov/CRAPES/2017/09846_170117.PDF, pp. 24-25.
 See a discussion on the innovative and flexible criterion in Josh Silver, Do CRA Exams Measure Retail Lending Well? December 4, 2019, NCRC, https://ncrc.org/do-cra-exams-measure-retail-lending-well/.
 NPRM, p. 1214.
 NPRM, p. 1215.
 NPRM, p. 1220.
 12 U.S.C. §§?2903(a)(1), 2901; see also §?2903(a)(1) (the Section 2903 exam shall be included in “a written evaluation of the institution’s record of meeting the credit needs of its entire community.”).
 NPRM, pp. 1212-1213.
 Federal Reserve Governor Lael Brainard, Strengthening the Community Reinvestment Act by Staying True to Its Core Purpose, speech at the Urban Institute, January 8, 2019, https://www.federalreserve.gov/newsevents/speech/brainard20200108a.htm.
 US Bank exam uses hours for CD services, see page 34, U.S. Bank CRA exam, October 2017, https://www.occ.gov/static/cra/craeval/sep19/24.pdf and BB&T uses units, see page 36, BB&T CRA exam, January 2017, https://www7.fdic.gov/CRAPES/2017/09846_170117.PDF.
 See the table on page 36 from BB&T’s CRA exam. It should use hours instead of units but has a good breakdown of CD service across the categories of CD.
 Community Reinvestment Act; Interagency Questions and Answers Regarding Community Reinvestment Act Guidance, OCC, Board of Governors of the Federal Reserve System, §__.24(d)(3)—1, Fed. Reg. 81, 142, p. 48542 https://www.gpo.gov/fdsys/pkg/FR-2016-07-25/pdf/2016-16693.pdf
 Capital One CRA exam 2017, p. 28, https://www.occ.gov/static/cra/craeval/jul18/13688.pdf.
 Interagency Q&A, op cit., §__.24(d)(3)—1, Fed. Reg. 81, 142, p. 48542
 Gruenberg speech of December 12, 2019, p. 6.
 Neil Bhutta, Jack Popper, and Daniel R. Ringo, The 2014 Home Mortgage Disclosure Act Data, in the Federal Reserve Bulletin, https://www.federalreserve.gov/pubs/bulletin/2015/articles/hmda/2014-hmda-data.htm, Figure 13 and accompanying narrative.
 Silver, The Community Reinvestment Act and Geography p. 9.
 See the CRA regulation, https://www.ffiec.gov/cra/regulation.htm, §25.22 lending test pertaining to consideration of affiliate lending and that affiliate lending is not considered in portion of loans in assessment areas.
 NPRM, p. 1216.
 NPRM, p. 1237.
 NPRM, p. 1222.
 The consideration of lending outside of assessment areas is described in Q&A § __ .22(b)(2) & (3)—4, see Federal Register, Vol. 81, No. 142, Monday, July 25, 2016, page 48538. The OCC reiterated this procedure in a recent bulletin, https://www.occ.gov/news-issuances/bulletins/2018/bulletin-2018-17.html#ft6.
 OTS 2009 Citicorp CRA exam, see http://www.occ.gov/static/cra/craeval/OTS/CRAE_14470_20091109_64.pdf.
 OTS 2005 Capital One CRA exam, see http://www.occ.gov/static/cra/craeval/OTS/CRAE_13181_20050718_64.pdf.
 OCC 2016 CRA exam of Bank of the Internet, https://www.occ.gov/static/cra/craeval/nov16/716456.pdf.
 Bank of Internet exam, p. 11.
 Bank of Internet exam, p. 7.
 Bank of Internet exam, p. 1.
 NCRC Congressional Testimony on Fintech Oversight, February, 2018, https://ncrc.org/ncrcs-congressional-testimony-fintech-oversight/.
 Bhutta and Popper, The 2014 Home Mortgage Disclosure Act; Laurie Goodman, Jun Zhu, and John Walsh, The Community Reinvestment Act: Lending Data Highlights, November 2018, Urban Institute, https://www.urban.org/research/publication/community-reinvestment-act-lending-data-highlights.
 Silver, The Community Reinvestment Act and Geography, p. 9.
 Ding and Nakamura, Don’t Know What You Got and Ding and Bostic, Effects of CRA on Small Business Lending.
 The data for this part of the comment letter is derived from the FDIC’s Deposit Market Share reports calculator; see Source: https://www7.fdic.gov/sod/sodMarketBank.asp?barItem=2, Data from June 2019, the most recent data is currently used by the calculator.
 NPRM, p. 1216.
 NCRC Comments Regarding Advance Notice Of Proposed Rulemaking (Docket ID OCC–2018-0008) Reforming The Community Reinvestment Act Regulatory Framework, https://ncrc.org/ncrc-comments-regarding-advance-notice-of-proposed-rulemaking-docket-id-occ-2018-0008-reforming-the-community-reinvestment-act-regulatory-framework/.
 Silver, An Evaluation Of Assessment Areas.
 Josh Silver and Bruce Mitchell, PhD, How To Consider Community Development Financing Outside Of Assessment Areas By Designating Underserved Counties, NCRC, January 2019, https://ncrc.org/how-to-consider-community-development-financing-outside-of-assessment-areas-by-designating-underserved-counties/
 The FFIEC web page has a section on CRA that discusses a one-year lag time period for underserved and distressed rural tracts, see https://www.ffiec.gov/cra/pdf/Regulatory%20Background%20-%20Distressed%20and%20Underserved%20Tracts%20FINAL.pdf.
 Gruenberg speech of December 12, 2019, p. 3.
 NPRM, p. 1220.
 NPRM, p. 1221.
 NPRM, p. 1221.
 NPRM, p.1221-1222, see footnote 46.
 NPRM, p. 60 of the original double-spaced NPRM used the word “doubled” to describe their adjustment. However, the final NPRM on p. 1221 changed the wording to “a fraction” to describe their adjustment. The public does not have a clear sense of how they made the estimation of dollar amount of loans that would qualify.
 12 U.S.C. §§ 2901(b), 2903(a)(1).
 Governor Brainard’s speech of January 8, 2019. The Governor stated, “a uniform ratio that does not adjust with the local business cycle could provide too little incentive to make good loans during an expansion and incentives to make unsound loans during a downturn, which could be inconsistent with the safe and sound practices mandated by the CRA statute. Industry commenters also expressed concern that discretionary adjustments to the uniform metric are likely to lag behind the economic cycle and undermine the certainty a metric purports to provide.”
 NPRM, p. 1222.
 Nuno Mota, An Analysis of Small Balance Loan Origination in Rural and High-Needs Rural Areas, Fannie Mae, December 2019, p. 12, https://www.fanniemae.com/resources/file/research/housingsurvey/pdf/analysis-of-small-balance-loan-origination.pdf.
 Carolina Reid, Quantitative Performance Metrics for CRA: How Much “Reinvestment” is Enough?, in Penn Institute for Urban Research, September 2019, pp. 11-13, https://penniur.upenn.edu/uploads/media/Quantitative_Performance.pdf.
 The CRA itself recognizes that the convenience and needs of communities include both “the need for credit services as well as deposit services.” 12 U.S.C. §?2901(a)(2).
 Governor Brainard’s speech of January 8, 2019. The Governor stated, “Our analysis suggests there are a set of metrics that can be compared to appropriately tailored benchmarks to provide greater certainty regarding community development lending and investment. The proposed metric would aggregate loan and investment dollars that are originated or purchased during the evaluation period with the book value of all other community development loans and investments that are held on the bank’s balance sheet.” She continued, “The proposed test would compare the combined measure of a bank’s community development financing relative to deposits in its local assessment area to a national average, set differently for rural and urban areas, and a local average in the bank’s assessment area.”
 Bruce Mitchell, PhD and Josh Silver, NCRC Research Memo – Impact Of Proposed Rule-Making On Major Credit-Card Lending Banks, February 2019, https://ncrc.org/ncrc-research-memo-impact-of-proposed-rule-making-on-major-credit-card-lending-banks/
 Reid, Quantitative Performance Metrics, p. 11.
 NPRM, pp. 1222-1223.
 NPRM, p. 1222.
 Josh Silver, How To Evaluate Community Development Financing And Services Under CRA, December, 2019, NCRC, https://ncrc.org/how-to-evaluate-community-development-financing-and-services-under-cra/.
 NPRM, p. 1222.
 Josh Silver, The CRA Evaluation Measures Would Allow Banks To Relax Their Retail Lending To LMI Borrowers And Communities, NCRC, February, 2019, https://ncrc.org/the-cra-evaluation-measures-would-allow-banks-to-relax-their-retail-lending-to-lmi-borrowers-and-communities/.
 NPRM, p. 1219.
 NPRM, p. 1220.
 NPRM, p. 1219.
 Bruce Mitchell, PhD, Proposed OCC and FDIC Geographic Analysis of Home and Small Business Lending: Permission to Decrease Lending for the Largest Banks?, March 2020, https://ncrc.org/proposed-occ-and-fdic-geographic-analysis-of-home-mortgage-and-small-business-lending-permission-to-decrease-lending-for-the-largest-banks/.
 NPRM, pp. 51-52. The NPRM says that a bank could not receive a Satisfactory or Outstanding rating unless it also has this rating in more than 50% of its AAs. The NPRM does not offer a rationale or provide an exact percentage of AAs in which a bank needs to have a passing rating. Question 17 on pages 76 and 77 asks if the percentage should be 50% or higher at 80%.
 Interagency Large Institution Examination Procedures, 2014, p. 15, https://www.federalreserve.gov/supervisionreg/caletters/CA_14-2_attachment_1_Revised_Large_Institution_CRA_Examination_Procedures.pdf and also see FDIC compliance manual, pp. 10.3 -10.5, https://www.fdic.gov/regulations/compliance/manual/11/xi-10.1.pdf.
 NPRM, p. 1223.
 NPRM, proposed CRA activity conforming to §§ 25.04(c)(6)(ii) and 345.04(c)(6)(ii), p. 1233.
 Federal Reserve Bank of Richmond, CRA Exam of Signet Bank, January 1996, pp. 18-20, https://www.federalreserve.gov/dcca/cra/1996/460024.pdf.
 Office of Thrift Supervision CRA Exam of CenFed Bank, November 1995, p. 9, https://www.occ.gov/static/cra/craeval/OTS/CRAE_01788_19951127_60.pdf.
 See https://www.mass.gov/archive/community-reinvestment-act-compliance for a description of the Massachusetts CRA. An example of this analysis is the CRA exam of Abington Bank, https://www.mass.gov/doc/abington-bank-cra-public-evaluation/download.
 See Blog Posts of the Association of Housing and Neighborhood Development, https://anhd.org/blog/bad-boy-carveoutand https://anhd.org/blog/new-york-state-lenders-you-are-accountable-multifamily-displacement-lending. Also see Kevin Stein, Banks Should do More to Prevent Renters from Being Displaced, American Banker, August 14, 2018.
 New York State Department of Financial Services, DFS Advises State Chartered Banks of Their Responsibilities in Lending to Landlords of Rent-Stabilized or Rent Regulated Multifamily Residential Buildings, September 25, 2018, https://www.dfs.ny.gov/about/press/pr1809251.htm.
 NPRM, p. 1218.
 12 U.S.C. §?2901(a)(2).
 NCRC in The CRA Evaluation Measures Would Allow Banks to Relax Their Retail Lending To LMI Borrowers and Communities shows that the median percent of branches in LMI and other underserved tracts was 30% for a sample of very large banks and 24% for the top 100 banks in terms of dollar amount of deposits.
 Lei Ding and Carolina Reid, The Community Reinvestment Act (CRA) and Bank Branching Patterns, Federal Reserve Bank of Philadelphia, Working Papers, WP 19-36, September 2019, https://www.philadelphiafed.org/-/media/research-and-data/publications/working-papers/2019/wp19-36.pdf.
 NCRC comments on the CRA ANPR, the section with the A 100 Point Scale Would Better Differentiate Between The Four Assigned Ratings discusses this issue, November 2018, https://ncrc.org/ncrc-comments-regarding-advance-notice-of-proposed-rulemaking-docket-id-occ-2018-0008-reforming-the-community-reinvestment-act-regulatory-framework/.
 Call Report data is available from the FDIC website via: https://www7.fdic.gov/sdi/main.asp?formname=customddownload.
 NPRM, p. 1224.
 FDIC Call Report data.
 NCRC, Intermediate Small Banks: The Forgotten But Significant Resource For Affordable Housing And Community Development, November 2017, https://ncrc.org/intermediate-small-banks-forgotten-significant-resource-affordable-housing-community-development/
 Under the proposed rule, banks would not be required to make public the data required by the rule. Instead, the agencies themselves would need to disclose data as part of the statutory CRA written evaluations procedure. The CRA requires the agencies to produce written evaluations “of the institution’s record of meeting the credit needs of its entire community,” including separate public and confidential sections. 12 U.S.C. §?2906(a). The public section of a report shall contain an agency’s “conclusions for each assessment factor,” “the facts and data supporting such conclusions,” and “the institution’s rating and a statement describing the basis for the rating,” §?2906(b). The confidential section is limited to customer, employee, or witness identity information as well as “statements obtained or made by the appropriate Federal financial supervisory agency in the course of an examination which, in the judgment of the agency, are too sensitive or speculative in nature to disclose,” §?2906(c). Data used to support an assessment factor conclusion and a rating would still need to be publicly available consistent with these requirements.
 NPRM, p. 1209.
 NPRM, p. 1227.
 NPRM, pp. 1250-1251.
 NPRM, pp. 1226-1227.
 NPRM, p. 1211.
 NPRM, p. 1228.
 NPRM, p. 1227.
 NPRM, p. 1228.
 NPRM, 1223, 1227.
 See, e.g., 12 U.S.C. § 2906(c) (providing for confidentiality for “any person or organization that has provided information in confidence to a Federal or State financial supervisory agency”).
 NPRM, p. 1247.
 OCC CRA exam of Sterling Bank, January 2017, pp. 2-3, https://www.occ.gov/static/cra/craeval/jul17/25075.pdf.
 FDIC Approval of BB&T- SunTrust Merger, November 2019, pgs. 9-10, 15, https://www.fdic.gov/news/news/press/2019/pr19111a.pdf. Federal Reserve Approval of KeyBank-First Niagara Merger, July 2016, pgs. 17-18, 27, https://www.federalreserve.gov/newsevents/pressreleases/files/orders20160712a1.pdf/ Federal Reserve Approval of Huntington-FirstMerit Merger, July 2016, pg. 31, https://www.federalreserve.gov/newsevents/pressreleases/files/orders20160729a1.pdf, Federal Reserve Approval of Fifth Third-MB Financial Merger, March 2019, pgs. 14 and 20, https://www.federalreserve.gov/newsevents/pressreleases/files/orders20190306a1.pdf.
 Memorandum for the Office of the Comptroller of the Currency, the Board of Governors of the Federal Reserve System, the Federal Deposit Insurance Corporation from the Department of Treasury, Community Reinvestment Act – Findings and Recommendations, April 2018, p. 22, https://home.treasury.gov/sites/default/files/2018-04/4-3-18%20CRA%20memo.pdf.
 Treasury Memo, p. 24. Available online at https://home.treasury.gov/sites/default/files/2018-04/4-3-18%20CRA%20memo.pdf.
 Department of Justice, Federal Government and State Attorneys General Reach Nearly $1 Billion Agreement with SunTrust to Address Mortgage Loan Origination as Well as Servicing and Foreclosure Abuses, June 2014, https://www.justice.gov/opa/pr/federal-government-and-state-attorneys-general-reach-nearly-1-billion-agreement-suntrust.
CRA Exam of Suntrust Bank, Federal Reserve Bank of Atlanta, March 2013, https://www.frbatlanta.org/-/media/Documents/banking/cra_pes/2013/675332.pdf, p. 2.
 The Interagency Q&A recognizes possible gaming and cherry-picking using affiliate loans and imposes a prohibition on such behavior, see §_.22(c)(2)(ii)—1 and –2, pgs. 48539 and 48540, Community Reinvestment Act, Interagency Question and Answers Regarding Community Reinvestment Guidance, Federal Register, Vol. 81, No. 142, Monday, July 25, 2016, Rules and Regulations https://www.gpo.gov/fdsys/pkg/FR-2016-07-25/pdf/2016-16693.pdf.
 12 U.S.C. § 2903(a)(1), 2906.
 NPRM, p. 1220.
 12 U.S.C. §?2901(a)(3).
 Silver, An Evaluation Of Assessment Areas and The Community Reinvestment Act And Geography.
 Community Reinvestment Act: Supervisory Policy and Processes for Community Reinvestment Act Performance Evaluations, OCC Bulletin 2018-17, June 15, 2018, https://www.occ.gov/news-issuances/bulletins/2018/bulletin-2018-17.html.
 12 U.S.C. §?2903(c)(1).