December 14, 2020
Consumer Financial Protection Bureau
1700 G Street, NW
Washington, DC 20552
Re: Comments on the CFPB Outline of Proposals Under Consideration and Alternatives Considered for Section 1071
Dear Director Kraninger:
Thank you for the opportunity to comment on the Consumer Financial Protection Bureau’s (CFPB) outline of proposals under consideration and alternatives considered for Section 1071.
The National Community Reinvestment Coalition (NCRC) consists of more than 600 community-based organizations, fighting for economic justice for almost 30 years. Our mission is to create opportunities for people and communities to build and maintain wealth. NCRC members include community reinvestment organizations, community development corporations, local and state government agencies, faith-based institutions, fair housing and civil rights groups, minority and women-owned business associations, and housing counselors from across the nation. NCRC and its members work to create wealth opportunities by eliminating discriminatory lending practices, which have historically contributed to economic inequality.
In October, the Bureau released an outline of proposals under consideration for the implementation of Section 1071 of the Dodd-Frank Act. This section amends the Equal Credit Opportunity Act (ECOA) to expressly permit and require lenders to collect information on the race, ethnicity and gender of a small business owner during the application for small business credit and to publicly report on the action taken on the application.
These proposals are the result of over ten years of deliberation, debate, research and policy analysis. They also arrive during a global economic and public crisis that has dramatically and negatively impacted small businesses and overwhelmingly harmed the financial stability and the solvency of Black businesses, Latino businesses, businesses owned by people of color, and women-owned businesses. The effects of this crisis will impact wealth creation opportunities for years to come, and coordinated federal and state action are necessary to both protect small businesses that remain open and to invest in the communities and the entrepreneurs whose wealth and livelihoods have been impacted. Data transparency alone will not achieve these goals, but it is a critical step among many to ensure that the recovery and future investment are equitable.
We believe that in many aspects, the Bureau has taken the correct approach in evaluating the efficacy of the data collection options available. However, the Bureau has also considered a number of options that will result in data collection gaps that, if adopted, will result in an incomplete picture of the small business market, limit the ability to conduct fair lending testing, and make it more difficult to ensure that lenders are meeting the credit needs of small businesses.
Our comments in response to the questions proposed in the Small Business Regulatory Enforcement Fairness Act (SBREFA) outline provide an outline for how these limitations can be addressed. To this end, we urge the Bureau to:
A) Provide a market-wide view of small business lending activity. Only a broad rule that applies to all small business lenders, covers all forms of small business credit, and includes all forms of small businesses can provide the necessary market-wide view of the small business lending activity necessary to identify lending trends and harmful practices before they become widespread.
- Broadly define financial institution and ensure that all market participants are required to report. The statutory language of Section 1071 was rightfully designed to ensure market-wide coverage and reporting from the wide variety of institutions that provide credit to small businesses, including banks, credit unions, online lenders, and other lenders. Section 1071 defines a lender as “any partnership, company, corporation, association (incorporated or unincorporated), trust, estate, cooperative organization or other entity that engages in any financial activity.” This definition would include depository institutions like banks and credit unions and non-depository institutions such as financial technology companies.
- Require lenders making 25 loans or more annually to report. We urge the Bureau to provide an activity exemption of fewer than 25 loans annually, consistent with the 2015 HMDA final rule amending Regulation C.We do not support a dollar size threshold separately or in conjunction with a loan count threshold. A dollar size threshold, including the $2.5 million threshold the Bureau offers as one of the options, would be too high.
- Report lending to small businesses with 499 or fewer employees and up to $8 million in revenue. We urge the Bureau to adopt a definition of small business as a business with fewer than 500 employees and up to $8 million in revenue, similar to the size standard alternative under consideration. We urge the Bureau, for the sake of consistency with the Annual Business Survey, to consider revising this to fewer than 500 employees (499 employees or fewer).
- Factoring agreements and merchant cash advances must be covered products under Section 1071. The Bureau is proposing to exclude merchant cash advances (MCAs) from coverage in 1071. We disagree with this proposal and urge that MCAs and factoring agreements be considered covered products. MCAs are widely used by small businesses and have a rapidly growing market share, often cause businesses to incur substantial repayment liabilities, and have other harmful terms that warrant market-wide monitoring. We also urge the Bureau to define these products as credit for the purpose of this rule, rather than rely on the current interpretation of ECOA which would exclude them.
B) Improve fair lending supervision and enforcement in the small business lending market.
- The definition of application should be consistent with the definition of application under Regulation B implementing the Equal Credit Opportunity Act. This definition is consistent with the definition of application under Regulation C implementing the Home Mortgage Disclosure Act (HMDA), and is preferable to the Bureau’s alternative proposal, which would consider an application complete when all required information is completed. The CFPB must not implement this definition because, in some cases, an incomplete application results when the lender discriminates and indirectly or explicitly discourages an applicant from completing an application.
- The proposed mandatory data points adequately implement Section 1071, with some modification. Lenders must be required to report applications and denials, as well as the reason for denial, and the race and ethnicity of the borrower or borrowers consistent with the disaggregated race and ethnicity information currently required under Regulation C.
- Additional discretionary data points should be added to improve fair lending efforts. We urge the Bureau to collect and disclose additional data on loan pricing, time in business, number of owners and employees, and the use of personal or business credit profiles in a credit decision.
C) Allow regulators, lenders, and the public to benchmark a lender’s affirmative obligation to meet the credit needs of the small businesses in the communities they serve. The Bureau has an unprecedented opportunity to improve the transparency of the small business market, identify lending patterns and trends, and enhance the ability of lenders to serve low-wealth communities and communities of color. To this end, we urge the Bureau to consider following the parameters in existing public datasets on the universe of small businesses, such as the Annual Business Survey, and collecting 1071 data consistently to provide a direct measure of small business loans per business. On an interagency basis, the agencies should determine if Section 1071 scope can become comprehensive enough to replace or be collected concurrently with CRA, CDFI, and Small Business Administration (SBA) data collection efforts.
D) Move forward with a strong rule that addresses these three key issues without delay. Taken together, these improvements can ensure that data collected under Section 1071 can achieve ECOA’s statutory objectives of preventing discrimination in credit transactions by providing publicly available data on race, gender and other demographics of small business applicants for credit. We urge the Bureau to move forward with a proposed and final rule in 2021 and an implementation of no more than 12 months.
Congress included the Section 1071 data collection requirements in Dodd-Frank to determine the ability of lenders to meet the small business credit needs of the communities they serve, identify harmful practices before they become widespread, and enforce fair lending laws. The CFPB was given the responsibility to carry out this work, and we look to the CFPB to move forward without delay, consistent with the timeline agreed to in its stipulated settlement. Further delays in carrying out this responsibility will only serve to perpetuate the disparities in small business lending that existed before the pandemic and have been so pointedly brought into focus since the pandemic began.
In the sections that follow, we describe in detail how these recommendations can be incorporated into a proposed rule in four sections, addressing the questions listed in the summary of proposals under consideration.
I) Proposals Under Consideration to Implement Section 1071 of the Dodd-Frank Act Regarding Small Business Lending Data Collection, and Alternatives Considered
In Question 1, the Bureau seeks input on other federal laws or rules that may duplicate the small business data collection proposals under consideration. Other laws and regulations require limited data disclosure of small business lending. These data are limited in scope, however, and could be replaced by data collected under Section 1071 so long as the key elements of existing disclosure requirements are preserved.
Section 1071 amends the Equal Credit Opportunity Act (ECOA) and helps achieve ECOA’s statutory objectives of preventing discrimination in credit transactions by providing publicly available data on race, gender, and other demographics of small business applicants for credit. Publicly available data on the demographics of applicants exposes racial and gender disparities and focuses the attention of the lending industry, policymakers, regulators, and community-based organizations on reducing those disparities and combating instances of discrimination.
Currently, other laws and regulations, including the Community Reinvestment Act (CRA) and call report requirements, also require lending institutions to collect and report data on small business lending. CRA requires the reporting of the location of the small business or farm (census tract location) and information on the revenue size of the small business or farm. Call report data requirements include the reporting of outstanding small business and farm dollar loan amounts. In addition, the Department of the Treasury’s Community Development Financial Institutions (CDFI) Fund requires CDFIs to report data on their small business lending activity.
On an interagency basis, the agencies should determine if Section 1071 can become comprehensive enough to replace or be collected concurrently with CRA, CDFI, and Small Business Administration (SBA) data collection efforts. We believe that this is possible after a sufficient transition period only if it is finalized and implemented in a manner that preserves the information necessary for each of the uses addressed by existing datasets. Section 1071 requires more detailed reporting than the CRA regulation and also includes the data elements that CRA now requires. Just like the Home Mortgage Disclosure Act (HMDA) data, Section 1071 data could become the data source that CRA exams use in the future. Banks and CDFIs would find it more efficient to submit data in one format as Section 1071 data than to have one or possibly two more annual data submission requirements.
In Question 2, the Bureau seeks input on the scope of its Section 1071 rulemaking authority. Specifically, Section 1071(b) states that “in the case of any application to a financial institution for credit for [a] women-owned, minority-owned, or small business, the financial institution shall—(1) inquire whether the business is a women-owned, minority-owned, or small business.” The Bureau is considering whether this statement can be read to include data collection for credit applications for all small businesses as well as for women-owned and minority-owned businesses that do not fall under the criteria of small business. The Bureau is interpreting its mandate under Dodd-Frank narrowly and is seeking to collect information consistently for the same set of small businesses and businesses owned by people of color and women. We support this interpretation of the scope of 1071, but that support is conditioned on a broad definition of small business, discussed in more detail in Question 14.
In Question 3, the Bureau seeks input on identifying businesses that would be subject to data collection requirements under Section 1071 and whether the scope proposed in Question 2 would affect that collection effort. As stated above, we agree with the proposal to limit the reporting of applicants that satisfy the Bureau’s definition of small business. Section 1071 is a statute focused on the credit needs of small businesses and we agree that the experiences of larger businesses in the credit market are different from small business. The small business lending process is “more varied and complex” as the CFPB notes. Considering the experiences of both large and small businesses may diminish the explanatory power of Section 1071 data in describing the experiences of small businesses in the credit marketplace.
1) The definition of “financial institution” rightly includes banks, credit unions, non-depositories, non-profit lenders and other market participants
In Question 5 and 10-12, the Bureau seeks comment on the definition of financial institution. The statutory language of Section 1071 was rightfully designed to ensure market-wide coverage and reporting from the wide variety of institutions that provide credit to small businesses, including banks, credit unions, online lenders and other lenders. Section 1071 defines a lender as “any partnership, company, corporation, association (incorporated or unincorporated), trust, estate, cooperative organization or other entity that engages in any financial activity.” This would include depository institutions like banks and credit unions and non-depository institutions such as financial technology companies.
Appropriately, the CFPB states that the lenders likely to be included are “online lenders/platform lenders, CDFIs (both depository institutions (DIs) and non-DIs), lenders involved in equipment and vehicle financing (captive financing companies and independent financing companies), commercial finance companies, governmental lending entities and non-profit, non-DI lenders.”
Section 1071 cannot just focus on banks but must also include non-banks that have dramatically increased their share of the small business lending marketplace. Morgan Stanley forecasts online lenders reaching $47 billion, or 16 percent of total U.S. small and medium enterprise approvals this year. The Federal Reserve Banks report that 20% of small businesses sought credit from on-line lenders. Covering online lenders is necessary because Federal Reserve surveys have revealed that small business applicants find their loan offers hard to understand, increasing the chances that the applicants will receive loans with higher interest rates than they anticipated or other undesirable terms and conditions. If online lenders reported information about their loans, including loan terms and conditions, the act of public data reporting and accountability might curb excessive pricing and onerous terms and conditions.
While credit unions have legal limits capping their small business lending, a number of credit unions have a significant presence in the small business lending marketplace. The Federal Reserve Banks found that 20 percent of business survey respondents with medium/high credit risk and with less than five years of operation sought financing from credit unions. The fact that younger businesses are more likely to use credit unions than more traditional bank financing means that their inclusion under 1071 will also help identity credit access for younger firms that would otherwise be missing. The statutory purpose of Section 1071 is to promote enforcement of fair lending laws and to identify business and community development needs. The only way to achieve these objectives is if Section 1071’s coverage of the financial industry is broad. If a sizable segment of the industry is excluded from data reporting requirements, the Section 1071 data cannot reveal the full extent to which credit needs are being met or ascertain with certainty the extent of fair lending problems in the credit marketplace.
In Question 6, the Bureau seeks information on possible exemptions for financial institutions based on size, assets or other factors. We urge the Bureau to provide an activity exemption of fewer than 25 loans annually, consistent with the 2015 HMDA final rule amending Regulation C.
We do not support a dollar size threshold separately or in conjunction with a loan count threshold. A dollar threshold, including the $2.5 million the CFPB offers as one of the options, would be too high. If a lender made 25 loans at an average of $50,000 each, the total dollar amount would be $1,250,000. This would be below the $2.5 million threshold and would thus be exempt. The CFPB would have exempted a lender that was active and important in the micro-lending sphere in its locality.
The statute provides the CFPB with the authority to exempt financial institutions from the reporting requirements. The CFPB is considering either an asset-size threshold (asset level below which an institution does not report data) or an activity-threshold (number of loans below which an institution does not report). In the case of non-depositories, the CFPB cannot readily apply an asset threshold since, unlike banks, non-depository institutions do not have stable asset sizes. Even in the case of banks, an activity threshold makes more sense since it is a clearer indication of whether a lender is substantially engaged in the business of making small business loans or whether a nominal number of loans indicates that small business lending is incidental to its lending activity.
We urge the Bureau to adopt an activity threshold of 25 loans annually, which is the lowest activity threshold proposed by the CFPB. Twenty-five loans as a threshold has an important precedent. The CFPB had used this threshold to determine whether a lender must report Home Mortgage Disclosure Act (HMDA) data but unwisely raised it to 100 home loans in 2020, which will likely serve to exempt many smaller lenders operating in smaller cities or rural counties and reduce the transparency of the market in these areas.
An activity threshold of 25 loans would cover the great majority of lending activity while also exempting lenders that make an incidental amount of small business loans. This threshold would also require smaller banks that were exempted from reporting CRA loan data in the mid-2000s to resume reporting data again.
Intermediate small banks (assets between $326 million and $1.3 billion) were previously required to report small business CRA data. These banks are especially important in rural communities and smaller cities. Using CRA data from 2003, one of the last years in which intermediate small banks reported data, NCRC estimated that these banks were between 15 to 20 percent of the market in the Appalachian portion of states like Maryland and Virginia.Also, the Federal Reserve survey reports that 42 percent of small businesses applying for credit applied to small banks. As of 2020, the Federal Reserve Banks report that applicants’ satisfaction rates of 79% is the highest with the small banks. Thus, Section 1071 data would be particularly useful to further probe reasons regarding whether this higher level of satisfaction is due to lower denial rates or better pricing. If these survey results were borne out via data analysis, the data would provide competitive impetus to larger banks to improve their performance.
Smaller banks are not averse to reporting data. In order for small business lending to be included on their CRA exams, some smaller banks have been regularly reporting this data, although they are not required to do so. In 2018, 157 or 22% of the 700 institutions reporting CRA small business loan data were smaller banks voluntarily reporting the data. These institutions found data reporting to be beneficial for CRA compliance purposes and/or to gauge their performance in the small business lending marketplace.
3) Evidence from HMDA suggests that compliance costs are unlikely to reduce the number of market participants
In question 8, the Bureau seeks information on compliance costs and the impact of those costs on lending. We have seen no evidence that similar data collection efforts have reduced market participants, suggesting that the benefits of disclosure outweigh the costs of compliance.
The Home Mortgage Disclosure Act (HMDA) has required data reporting for more than 40 years and has not caused any lender, to our knowledge, to cease making home loans. The Federal Reserve Board enhanced HMDA data in 1990 to include information on the action taken on applications in addition to demographic data on loan applicants. By 1998, 7,925 lending institutions reported HMDA data.
The Federal Reserve Board enhanced HMDA data again in 2004 to require reporting of price information along with additional data points such as lien status. In 2004, 8,121 lenders reported HMDA data, and by 2007 the number of reporters surged to 8,886. Over time, therefore, the number of reporters grew, which is inconsistent with the supposition that data reporting causes a decrease of lending or lenders. It was only after the subprime lending crash, which was caused by a deregulatory environment enabling risky and abusive lending, that the number of HMDA reporters plummeted. By 2012, the number of reporters had fallen to 7,632 institutions.
Likewise, CRA small business loan data reporting has been in place since 1996. We have not heard complaints from banks in our multiple conversations with them over the years that CRA small business loan data is a major compliance cost causing them to diminish their operations. Moreover, as described below, we agree with the CFPB’s analysis of the costs of data collection, which suggests modest costs and small increases in pricing of $17 to $40 per loan for a small business borrower if lenders were to pass along the costs of data collection to borrowers.
In sum, data collection should make a lender more competitive, knowledgeable about its lending operations, and better able to conduct self-audits. Data collection should make a lender more efficient; it is a modest cost that more than pays for itself in terms of enhanced competitiveness or profitability.
4) Government and non-profit lenders should be included to ensure a comprehensive view of the small business credit market
In Question 9, the Bureau seeks information on whether governmental or non-profit non-depositories should be considered covered lenders. As stated above, covered lenders should be defined broadly so that Section 1071 data covers the great majority of the lending marketplace in order to ensure that abuses or inequities in lending are not being missed by the data. Broad coverage is the only way to level the playing field for lenders and ensure that they are all competing by offering a fair deal for borrowers. Broad coverage also promotes safety and soundness by exposing any risky or abusive lending that could result in spikes in delinquencies or defaults. Stakeholders could take action to address such lending before it becomes so widespread that it threatens the health of the entire industry or even the country; this happened when the public did not have enough information on loan terms and conditions in subprime home lending.
In Question 13, the Bureau seeks input on the inclusion of online and platform lenders within the definition of a financial institution for the purposes of coverage under 1071. The Bureau says that online lenders and platform lenders are examples of non-DIs that “may be covered under the eventual 1071 rule.” We urge the Bureau to apply 1071 data collection requirements to platform lenders.
Platform lending now constitutes a sizable share of lending for small and micro-businesses. Online lenders and lending platforms have a significant presence in the small business credit market. For example, through its direct online lending and platform lending marketplace, OnDeck Capital originated $2.4, $2.1 and $2.5 billion in small business loans in 2016, 2017, and 2018, respectively. OnDeck makes loans for as little as $5,000 and never for more than $500,000. Platforms lenders’ share of the market will only continue to grow, and their exclusion would render 1071 less comprehensive every year.
Platform lenders should be required to report lending data under 1071 on behalf of the investors in those loans because lending platforms perform most of the essential functions of loan origination. While lending platforms may or may not be the sole source of capital for the loans that are procured on their websites, they are almost always involved in the key steps of loan origination. In some instances, a platform lender introduces an applicant to bidders – some corporate, some private individuals – and then steps away. In other cases, the platform lender buys all or a portion of the loan. In the latter instance, the platform may hold the loan on its balance sheet, but more often, it then sells the loan inside an asset-backed security. However, it most cases, the platform accepts the application, disburses the proceeds of the loan, provides the analytics to analyze the creditworthiness of the applicant, services the loan, authenticates the identity of the applicant, holds cash in escrow accounts on behalf of its investor members, and may be exposed to some credit risk. All of these activities reflect activities of a lender for the purpose of collecting and disclosing small business lending data.
1) The Bureau should define small business based on the number of employees and revenue, with some modification
In Question 14, The Bureau seeks feedback on various criteria for defining a small business, including annual revenue and number of employees. We urge the Bureau to adopt a definition of small business as a business with less than 500 employees and up to $8 million in revenue, similar to the size standard alternative under consideration. We urge the Bureau, for the sake of consistency with the Annual Business Survey, to consider revising this to less than 500 employees (499 employers or fewer).
When the agency issued its request for information in 2017, a common definition was a business with less than $1 million in revenues. The CFPB is considering this definition. However, the agency also says that the Small Business Administration (SBA) defines a small business as one with 500 or fewer employees and up to $8 million in annual revenues.
If the CFPB adopts the $1 million in revenue threshold, it reports that about 23% of small firms with employees would not have their applications reported in the Section 1071 data. In contrast, the SBA definition may result in an omission of just 63,000 small businesses. It is prudent to err on the side of inclusion and to adopt the more expansive definition of small business in order to maximize the accuracy of the loan database in capturing experiences of small businesses in the lending marketplace.
NCRC agrees with the CFPB’s interpretation of the authorizing language in Dodd-Frank to include data reporting requirements for applications submitted by small businesses with a maximum of 500 employees. As the CFPB notes, the types of lending and underwriting are different for large businesses as opposed to smaller businesses with 500 or fewer businesses. Moreover, a definition using more than 500 employees as a cut-off captures about 99% of women- and minority-owned small businesses.
Data fields of revenue size and number of employees can make it possible for data users to separately analyze the experiences of the smallest businesses from their larger counterparts. We discuss how to construct revenue size and employee data fields in more detail below. A guiding principle, however, is the data points must be detailed enough to allow analysts to meaningfully analyze the experiences of businesses of various sizes. The CRA data currently does not allow for robust analysis as it only allows a user to compare the experiences of small businesses below and above $1 million in revenues. The Annual Business Survey reports has a categorical data field for annual sales/receipts with the following categories: under $10,000, $10,000 to $49,000, $50,000 to $99,000, $100,000 to $249,000, $250,000 to $499,999, $500,000 to $999,999, $1 million or more. In our responses to questions 40, we offer more considerations for reporting sales/revenue size.
In Question 16 and 54, the Bureau seeks feedback on the use of NAICS code information to determine the industry of a small business loan applicant and what level of granularity is appropriate. We urge the Bureau to adopt, collect and publish the six-digit NAICS code for covered applicants for consistency and hierarchical combability with existing public data related to small business lending activity.
The use of a standard metric to meet the requirements of the statutory language in Section 1071 of the Dodd-Frank Act are clear. Specifically, the law reads that Congress enacted section 1071 for the purpose of: (1) Facilitating enforcement of fair lending laws; and (2) Enabling communities, governmental entities, and creditors to identify business and community development needs and opportunities for women-owned, minority-owned, and small businesses.
The ability to identify needs and opportunities, as discussed in previous sections, requires the ability to compare lending by sector to the total number of businesses in that sector based on other public data sources. The Annual Business Survey, which provides business characteristics, including two-digit NAICS classifications for employer firms, is one such data source. Two-digit NAICS codes are also used as part of the US Census Small Business Pulse surveys conducted to gauge the impact of the COVID-19 pandemic on small businesses. The six-digit NAICS codes are commonly used by employers and institutions, including the financial sector in order to complete annual EEOC certifications. They are also collected by the SBA as part of the process to determine the eligibility of business for SBA guaranteed loans.  The NAICS code system is maintained by the Office of Management and Budget (OMB) and is updated periodically, offering a standardized and simple method for assessing the type of businesses that are seeking, obtaining, or failing to secure credit. Taken together, this suggests that six-digit NAICS codes are already being collected by FIs, are widely accepted, and frequently used. Based on the hierarchical structure of the NAICS coding system and the progressively narrower categories of two, three, four, five and six-digit we urge the Bureau to adopt six-digit NAICS as part of this dataset to ensure comparability with existing and future public datasets that may, in conjunction with data collected under Section 1071, inform the small business market.
1) The Bureau should adopt a definition of “women-owned business,” “minority-owned business,” and “minority individual” consistent with existing, relevant public data
In Question 17, the Bureau seeks feedback back on definitions of “women-owned business,” “minority-owned business,” and “minority individual.” Race, ethnicity, gender, veteran-status and legal/ownership structure is currently collected as part of the Annual Business Survey, and the definitions used in that survey can inform how to structure the race, ethnicity, gender and other demographic information under Section 1071. We agree with the Bureau’s proposed approach to defining “minority individual” consistent with HMDA, but urge the Bureau to use the disaggregated race and ethnicity categories contained in the 2015 final rule, rather than the aggregated categories considered in the proposal (see also Q47). Likewise, existing business surveys that contain demographic information on the owner focus solely on employer firms. In non-employer firms as part of the Non-Employer Survey-Demographics project, we also urge that the Bureau consider the implications of the definition of minority individual and the disaggregated race and ethnicity definitions, and other demographics consistent with that data collection effort.
We also agree with the proposals to define ownership control as more than 50 percent of the ownership or control, with some modification. First, we urge the Bureau to collect and disclose the percent ownership by women and people of color as a continuous variable to afford the flexibility to redefine ownership thresholds as necessary.
The use of a continuous variable can help resolve two areas of ambiguity: (1) the conflicting definition of ownership control with the Annual Business Survey and (2) the challenge of determining ownership control where two partners each hold 50% ownership. We note that the ownership control definition in use by the Annual Business Survey is defined as “51 percent or more of the stock or equity in the business and is categorized by firms classifiable by sex, ethnicity, race, and veteran status and firms unclassifiable by sex, ethnicity, race, and veteran status and is not consistent with the definition of ownership control in Section 1071 of Dodd-Frank.However, this proposed change raises some issues with partnerships, which we discuss in more detail below. We also support efforts towards further simplification of this definition based on the FinCen customer due diligence rule.
In Question 18, the Bureau seeks information on legal or ownership structures of the businesses that typically apply for small business loans. We urge that the Bureau include in the definitions of “women-owned business,” “minority-owned business,” and “minority individual” all legal structures, including sole proprietorships, partnerships, and other structures. We acknowledge that a two-person partnership with others of different races or ethnicities may complicate the ownership definition discussed in Question 17. Disclosing ownership status as a continuous percent would help address this ambiguity. At a minimum, we believe that further Bureau research may be warranted to prevent undercounting firms where there is 50% ownership by women or people of color but the ownership stake does not exceed 50%.
E) The Bureau’s proposal to rely on the ECOA definition of credit and provide certain exemptions will make the small business credit market less transparent
1) The use of the ECOA definition of credit will exclude important small business lending market participants
In question 20, the Bureau seeks feedback on the use of the ECOA definition of “credit” for purposes of defining covered products under section 1071. We urge the Bureau to modify the definition of credit for the purpose of Section 1071 to include, at minimum, merchant cash advance and factoring products. Under ECOA, factoring agreements are explicitly excluded under the definition of credit. In the staff interpretation for Regulation B, lenders are instructed to exempt factoring agreements from the requirements of ECOA but apply ECOA protections for where credit is incident to a factoring agreement. The definition of credit under ECOA and the current staff interpretation, in conjunction with the exemptions the Bureau is considering in Question 21, would exempt this important and growing form of business lending as covered products. We urge the Bureau to both explicitly include factoring agreements (not just those where credit is incident to the agreement) as covered products, and to define credit more expansively than it is defined under ECOA for the purposes of this rule.
There is an important public policy reason for this expansive definition. Factoring and MCA agreements are widely used by small businesses, particularly very small businesses, who are more likely to face heightened challenges accessing business credit. From 2013 to 2016, non-bank providers supplied an average of $94 billion in receivables-based financing to small businesses.
In question 21, the Bureau seeks information on leases, trade credit, factoring, or merchant cash advances (MCAs) and whether they should be considered covered products under the 1071 rule. The Bureau is proposing to exclude merchant cash advances (MCAs) from coverage in 1071. We disagree with this proposal and urge that MCA and factoring agreements be considered covered products as described in our response to Question 20. MCAs provide a business with an up-front lump sum payment (the advance) in return for a percentage of that business’s credit and debit card sales and should be reportable under Section 1071. Despite often advertising themselves as loans, MCA providers have been strongly opposed to labeling their products as loans.
a) Factoring agreements should be reported under Section 1071
The Bureau is considering excluding factoring agreements from the definition of covered credit, consistent with the staff interpretation of the definition of credit under ECOA. We urge the Bureau to require the reporting of these agreements regardless of whether there is a credit agreement incident to the factoring agreement. In its 2019 survey of credit sources used by firms with less than $1 million in annual revenues, the Federal Reserve found that five percent of firms who applied for credit sought financing through factoring. Small businesses use factoring as a source of working capital. Many small businesses may not have the credit profile to qualify for other forms of financing, and as a result, they smooth their cash flows by selling the proceeds of their receivables. In spite of that, known challenges exist in this product sector. Factoring is an expensive form of credit, and additionally, some borrowers find it difficult to determine the true cost of borrowing under these agreements. The structure of a factoring agreement differs significantly from traditional term loans and open-end lines of credit, and accordingly, we outline two categorical data points that the Bureau should require in Question 51.
b) MCAs are an important and growing source of small business credit
Similar to factoring, MCAs should be included under the definition of covered product. These loans have become a staple of the small business lending industry, and there is evidence that most small businesses perceive these transactions as a form of credit.  MCAs grew rapidly following the 2008 financial crisis, as traditional lenders tightened lending standards and fled from small business lending markets. The MCA industry was estimated to have provided $19.2 billion in small business funding by the end of 2019.
A CFPB white paper estimates that the number of factoring and merchant cash advances is about 8 million (7 million for factoring and one million for MCAs), which exceeds the 6 million loan term accounts. While the dollar amount of factoring and merchant cash advance is smaller than term loans, measuring factoring and merchant cash advance by the number of accounts illustrates that these types of credit are widespread. These were estimates for the year 2013, and recent filings from major MCA lenders suggest that the number of MCAs have significantly increased since then.
According to its latest annual report, Square Capital has by itself issued approximately 1 million MCAs, for a total amount of more than $6.3 billion, since 2014. Shopify Capital extended $153 million in MCAs in the second quarter of 2020, a year-over-year increase of 65 percent. In its corresponding presentation to investors, Shopify’s management noted that COVID created new demand for its credit offerings, in part because the kinds of smaller firms that it primarily serves encountered greater-than-normal challenges to accessing capital through traditional bank financing during the pandemic. Future growth will be fueled by the pandemic’s effect on our economy: For example, 60 percent of the 80,000 small businesses that received their PPP loan via Square had no prior relationship with the company.
c) MCAs have the potential to causes businesses to incur repayment liability
MCAs have the potential to cause businesses to incur repayment liability even after default. MCA often include personal guarantees and collateral requirements that allow lenders to seize a business owner’s assets. The FTC has found that provisions sometimes go unadvertised to small business consumers, suggesting that MCA providers are engaging in deceptive practices. For example, the FTC found RCG Advances LLC, a New York-based MCA provider, advertised its products as having “no personal guaranty of collateral from business owners.” Yet RCG’s lending contracts include provisions that if a borrower is deemed to be unable to make a repayment, any guarantor on the MCA contract “will be jointly and severally liable to RCG for all of RCG’s losses and damages.” RCG and other providers were also found to have told borrowers that delays in repayment or failure to repay due to bankruptcy were not a violation or default on the lending contract, yet proceeded to hold borrowers in default and pursue repayment under such circumstances. In pursuing this repayment, the FTC found that MCA providers went as far as to threaten physical violence against borrowers who failed to make repayment.
d) Collecting and disclosing data on MCAs is necessary to monitor the market and identify the growth of higher-cost products
Exemptions for the MCA industry under ECOA and state usury laws does not justify exempting the industry from 1071. MCA providers label their products as factors to avoid state usury laws and licensing requirements, allowing them to charge often high interest and annual percentage rates (APR). A study by Opportunity Fund found that MCAs have an average APR of 94 percent, almost three times higher than some of the laxest state usury limits. In some of the most egregious cases, MCAs have rates that exceed 1,000 percent.
These high rates can force borrowers to take out another MCA to make repayments. This creates a cycle of debt as the borrower now owes the original amount borrowed several times over. For example, Antelope Valley Community Clinic, a healthcare services non-profit that helped underserved communities, originally received $250,000 through an MCA but ended up owing $4.3 million in cumulative MCA debt. With concerns over predatory and abusive pricing, including the MCA’s industry under the reporting requirements of 1071, especially if pricing information is collected, could provide more information on the industry’s pricing practices. We describe how to determine pricing disclosures for MCAs in more detail in response to question 50.
e) MCA often have other harmful terms that warrant market-wide monitoring under Section 1071
MCA providers often require that small business borrower’s sign “Confessions of Judgement” (COJs) agreement within their lending contracts. COJs waive a borrower’s procedural due process rights and force them to accept liability and automatically lose any legal disputes with lenders. MCA providers claim COJs are efficient and prevent the lengthy and costly parts of the litigation process. However, in practice, COJs allow MCA providers to claim that a business owner has breached the lending contract without providing sufficient evidence. This can result in the seizure of the borrower’s business accounts and even personal assets without the borrower even being aware that the provider has alleged that their contract has been breached. MCA providers can claim the lending contract has been breached and freeze a borrower’s bank account if a single daily payment is missed. MCA providers have been found to have altered or even forged COJ agreements. Similar concerns and abuses about COJs led the Federal Trade Commission (FTC) to outlaw them in consumer credit transactions in 1985, considering it to be an abusive practice. However, this reform did not apply to commercial small business transactions.
Despite these previous concerns, COJs have become an increasingly common practice within MCA lending. Bloomberg found the number of MCA cases ending with a confession in favor of a merchant cash advance company in New York state rose from 14 cases in 2014 to over 3,500 cases in 2018. These COJ cases are estimated to have won the MCA industry an estimated $500 million.
Federal regulators have begun to crack down on the industry. In August, the FTC filed a complaint against one of the MCA industry’s pioneers, Yellowstone Capital, the same company that Antelope Valley received financing from. The FTC alleged that Yellowstone illegally withdrew millions of dollars in excess payments from borrowers and relied on deceptive marketing, including misrepresenting collateral requirements, personal guarantees, and financing amounts available under their products.
f) Two states have already taken steps to define MCAs as credit
California (in rulemaking) and New York (passed but not yet signed into law) each passed state small business truth in lending laws, and both include merchant cash advances within their scope of coverage. The bills identify MCAs as a form of credit, establish a system for collecting data on their issuance, create disclosure requirements, and arrive at methods for calculating an APR. Additionally, the California Department of Financial Protection and Innovation recently filed a consent order with the MCA provider Allup LLC in which the Department found that Allup’s products placed the risk of loss on the merchant similar to a lending transaction. As a result of this finding, Allup agreed to cease lending in California until the company applied for a license under California Financing Law and refund any fees or payments collected above California’s usury limit.
In question 23, the Bureau seeks feedback on the definition of application. We urge the Bureau to consider a definition of application consistent with HMDA.
The CFPB proposes to define an application as “an oral or written request for an extension of credit that is made in accordance with procedures used by a creditor for the type of credit requested.” According to the CFPB, the definition accords with that in Regulation B, which implements the Equal Credit Opportunity Act. This is sensible and accords with the definition of application contained in the Home Mortgage Disclosure Act (HMDA). The HMDA definition is “Application means an oral or written request for a covered loan that is made in accordance with procedures used by a financial institution for the type of credit requested.” Using a definition for application under Section 1071 that is consistent with HMDA would minimize costs for lenders that need to comply with both statutes. This definition also accurately reflects the lending process in that the definition of application and associated fair lending obligations is triggered when a customer says that he or she desires credit.
This definition is also much preferable to an alternative definition, which the CFPB is also considering. The alternative definition involves receiving all the information that the creditor needs in order to render a decision. The CFPB must not implement this definition because applicants sometimes do not submit all the needed information. In some cases, an incomplete application results when the lender discriminates and indirectly or explicitly discourages an applicant from completing an application. The CFPB would miss recording incomplete applications in its database if it defines an application as occurring only when an applicant submits all the necessary information necessary for rendering a decision.
II) Both mandatory and discretionary data points need to be improved to implement the proposed definitions of small business and covered product, and to fulfill the public policy purpose of Section 1071, consistent with other public small business data sources
In Question 25, the Bureau seeks feedback on mandatory data points required under Section 1071 of Dodd-Frank. For the mandatory data points of loan and credit type, the CFPB contemplates requiring three data fields: type of loan product, type of guarantee, and loan term. These data fields are appropriate choices because each of them is necessary separately and in combination to help determine whether lenders are responding to the needs for credit by offering affordable and sustainable products to traditionally underserved small businesses.
Under loan product type, the CFPB appropriately proposes to include not only term loans but also credit card lending. Credit cards are widely used by small businesses, often with smaller principal balances and higher interest rates than term loans. It is thus important to assess whether the smallest businesses or women- and minority-owned businesses have equitable access to term loans or are served disproportionately by credit card loans or other small business credit products. If an imbalance exists, stakeholders can then determine the factors behind the imbalance and take steps to narrow the disparities in access. The SBA reported that Hispanic and African-American owners are more likely to rely upon credit cards than other businesses. Section 1071 data would inform us if this continues to be the case.
A significant limitation in the CRA small business data is a lack of separate reporting for different types of loans. The CRA small business data combines credit card lending and non-credit card lending without the ability to separate credit card and non-credit card lending in analyses. Section 1071 must eliminate this limitation.
The CFPB also contemplates, including lines of credit that meet important credit needs to help businesses weather fluctuations in revenues. Lines of credit will help inform stakeholders whether minority- and/or women-owned businesses are able to access this important credit type or whether they experience disproportionate amount of denials.
The lines between consumer and small business lending can get blurred, especially for the smallest businesses. Studies find that about one-quarter of new small businesses are funded by the personal loans received by their owners. In the case of personal loans, the CFPB should consider asking lending institutions for data on borrowers using their personal credit cards or non-credit card loans to finance their businesses.
The CFPB lists in its outline a loan purpose of “refinancing existing debt.” However, in a summary document accompanying the outline, the CFPB contemplates excluding from the definition of an application whether the request is for a renewal of the loan. This is contradictory unless the CFPB is defining refinancing existing debt as debt that does not include a previous loan from the particular lender the borrower is currently using. In any case, all refinances, and renewals should be included.
The CFPB is considering requiring lenders to report data on an application when a borrower requests additional loan dollars. We agree with this approach, but the database should be made more complete by including requests for refinances or renewals (extension of existing principal past the original due date) of the original loan as well as when a borrower requests additional dollars.
The CFPB contemplates requiring lenders to ask if the loan involves a federal or non-federal government guarantee or a personal guarantee by the borrower. These are important data points because they allow stakeholders to further determine whether traditionally underserved small business receive more costly or onerous products. In home lending, people of color are disproportionately represented as recipients of Federal Housing Administration (FHA) loans or other government guaranteed-loans. While providing a critical source of credit to people of color, government-guaranteed loans are more expensive. It is therefore important to see if lenders can work with other stakeholders to increase conventional home lending to people of color. Likewise, federal and non-federal government guarantees would be valuable data points in the Section 1071 database to determine the access of women- and minority-owned businesses to government guarantees and conventional loans.
The CFPB also proposes to require an indication of whether the borrower guaranteed the loan. This data point would allow an analysis of whether they are disparities in the frequency of minority-owned and women-owned businesses being required to guarantee loans compared to non-minority and male-owned businesses. Such disparities could be significant obstacles to credit. Related to personal guarantees is whether a business is required to pledge collateral.
The CFPB should consider additional data points describing collateral requirements. According to businesses involved in Federal Reserve focus groups and meetings, collateral asset values held by their businesses dropped during the Great Recession. As a result, several lenders required additional collateral that small businesses need for loans, making it harder for small businesses to be approved for loans. Collateral information will, therefore, shed light on underwriting changes and approaches during various economic conditions and will help stakeholders better understand fluctuations in access to credit. In addition, the CFPB should consider asking lending institutions to report the value of any personal collateral pledged for loans in addition to collateral owned by the business. According to Robb, start-ups tend to pledge personal assets when securing financing. In addition, personal guarantees were offered on 41 percent of small business loans, according to a study using data collected by the Federal Reserve Board.
Loan term is a necessary component of the total cost of borrowing and should be included as proposed.
The CFPB is appropriately considering requiring the reporting of a variety of loan purposes. This is important because it gauges whether small businesses can obtain credit for various needs relating to loans for start-up, working capital, expansion, buying equipment or acquiring commercial real estate. CRA requires federal bank agency examiners to determine if banks are responding to a variety of credit needs. The degree of a bank’s responsiveness to priority needs is also an important criterion on CRA exams. These proposed data points would therefore be critical to help analyze whether the smallest business in low- and moderate-income census tracts are receiving loans that meet their needs for a variety of purposes, including start-up and expansion. This data would provide important motivations for banks to improve their CRA performance because it would measure, for the first time, banks’ records of making loans to meet these purposes. We also urge the Bureau to consider allowing a general loan purpose value for small business credit cards where it may be difficult to determine the loan purpose.
In addition, it would help identify priority needs for various geographical areas. For example, if loans for working capital were especially scarce in one particular geographical area, CRA examiners could reward banks with higher ratings on the small business part of the CRA exam if the banks were performing better than their peers in providing loans for working capital. This in turn, may encourage other banks to increase their lending for working capital.
The amount of credit is a data element that is important for the purposes of Section 1071, which includes enforcing fair lending laws and assessing whether credit needs are met. The Minority Business Development Administration (MBDA) finds that businesses owned by people of color received lower loan amounts than white-owned firms. This finding remains even after controlling for sales level of firms. The average loan received by high-sales firms owned by people of color was $363,000 compared to $592,000 for white-owned firms. Disparities in loan amounts received can impede the growth of businesses owned by people of color and women-owned businesses and thus impair overall economic growth. The data on loan amount will enable stakeholders to investigate these disparities and attempt to devise strategies for narrowing or eliminating the disparities. Disparities in loan amounts can manifest themselves in both credit card and non-credit card lending.
We agree with the CFPB’s proposal to require lenders to report the initial amount requested at the time of application. We also agree with the proposal to require lenders to report the originated loan amount and the credit limit approved in the case of open-end lending.
Section 1071 requires lending institutions to report “the type of action taken with respect to such application.” Currently, the largest dataset available to the public, the CRA small business loan data, has information on loan approvals only. In contrast, some surveys, such as the Kauffman Firm Survey has information on loan denials as well as approvals.
Laderman and Reid state that their paper has limitations because the CRA data does not have other action categories. According to Laderman and Reid:
Our paper cannot tease out how changes in small business lending are driven by changes in supply or changes in demand. The lack of data on small business loan applications severely limits our ability to understand where there is unmet demand for credit and whether or not small businesses in LMI neighborhoods are being denied small business credit by banks. While we do have some information on patterns of small business lending by banks with a CRA obligation, we do not have the ability to assess to what degree those loans reflect a true commitment to meeting the credit needs of the banks’ local communities. Access to more comprehensive data on small business lending – especially across the different types of credit markets – would greatly facilitate our understanding of credit barriers and gaps in LMI neighborhoods.
As Laderman and Reid maintain, it is not possible to determine with the CRA small business loan data whether differences in the percentages of loans by demographic category of a neighborhood are due to differences in demand or lender action on applications. In order to assess whether credit needs are being met, data on applications and denials, in addition to approvals, are needed. Moreover, data on denials enable regulators and the public at large to determine if there are fair lending concerns arising from disparate denial rates experienced by different races or genders. Research based on survey data reveals statistically significant differences in denials by race and gender, which suggests that data for individual lending institutions is needed for enforcement purposes to investigate the possibilities of discrimination on the part of certain lenders. A recent Federal Reserve paper reconfirmed previous findings that even after controlling for firm characteristics such as profitability and creditworthiness, minority-owned firms are still more likely to be denied than white-owned firms.
Analysis of the application and denial rates by geographical area can also shed light on whether the needs for credit are being met in a uniform manner or if some regions have unmet credit needs. In a report conducted for the Appalachian Regional Commission, NCRC obtained special tabulations from the Kauffman Firm Survey aggregated to the county level in order to assess potential differences in application and denial rates in Appalachia compared to the United States as a whole. This research found that application rates are similar in Appalachia and the nation, but that denial rates are considerably higher in Appalachia. This suggests to policymakers that unmet credit needs are present in Appalachia compared to the nation and that either fair lending enforcement should be increased or the creditworthiness of a segment of small businesses in Appalachia should be improved so that they can receive the loans they desire.
In questions 34-37, the Bureau seeks feedback on the type of action taken. The action categories in HMDA are a good precedent that can inform the action categories in the small business data. We support the CFPB proposal to use action categories that resemble those in HMDA. In addition to approvals, HMDA reports whether the application was approved but not accepted, if the application was denied, if the loan application was incomplete, if the loan was approved but not accepted by the borrower, or if the borrower withdrew the application. All of these categories are important for fair lending enforcement. Disparities in incomplete or withdrawn applications can be due to discouragement and discriminatory treatment applied to women or minorities. Moreover, disparities in approvals not accepted by the applicant can be indicative of less favorable pricing or loan terms and conditions applied to women to people of color.
8) Denial codes must be required and should be granular to ensure that applicants receive useful and actionable information
In question 35, the Bureau seeks feedback on the reporting of denial codes. We urge that denial codes be required, based on the denial codes currently in Regulation C. We also discuss other considerations when implementing denial codes, such as additional reasons and appropriate granularity to provide the applicant with useful and actionable information. HMDA also requires mandatory reporting of reasons for loan denial for certain lenders. All lending institutions reporting small business data must be required to report reasons for the denial. This data can help stakeholders determine if fair lending concerns are present, if underwriting factors need to be addressed, or if borrower creditworthiness should be improved. In its survey, the Kauffman Foundation asks respondents to indicate if they were denied loans because of insufficient collateral, business credit history, personal credit history, loan too large, inadequate documentation, or business too new. Another denial reason that could be applicable is the debt service coverage ratio, which is an important measure of whether the loan is sustainable for the business.
In response to the Bureau’s recent Request for Information for possible updates to ECOA, we also discuss the need for more detailed detail codes, given the widespread use of alternative underwriting criteria and the use of data outside traditional credit reports. Improved and more granular denial reason codes are necessary to provide useful and actionable information to the small business applicant.
In Question 35 and 39, the Bureau seeks feedback on the use of census tracts for determining the location of a business. The CFPB prefers that a lender reports an address where the loan proceeds will be principally applied. This would be geocoded into a census tract. If this address is unknown, the CFPB asks for the address of the business headquarters or the address associated with the application. The CFPB is correct in prioritizing the address where the proceeds would be used. The geographical location of the business is critical for regulatory enforcement of the Community Reinvestment Act (CRA) and fair lending laws. Designed to combat redlining, CRA exams measure whether banks are making loans in low- and moderate-income census tracts and thus whether they are meeting credit needs in these tracts. While progress has been made in combating redlining, fair lending issues still remain, according to a paper by Robb and Bates. Robb and Bates’ econometric analysis shows that minority-owned firms are more likely to receive loans in census tracts with higher levels of poverty. They suggest that banks are attuned to their CRA responsibilities and will make more efforts to lend to creditworthy or marginally creditworthy minority-owned firms in lower-income census tracts.
Likewise, Robb and Bates show that minority-owned firms are more likely to be approved for loans if they are located in minority census tracts, which shows banks’ attention to fair lending and redlining concerns. However, fair lending enforcement has not been completely successful since businesses owned by people of color are less likely to be approved when they are located in predominately white neighborhoods. Data must capture the location of businesses in order to measure the effectiveness of CRA and fair lending enforcement and to identify areas where enforcement needs to be improved.
Improved data on business location is needed to continually track lending in neighborhoods, including throughout business cycle fluctuations. Employing the CRA small business loan data, Reid and Laderman’s analysis illustrates that more steps need to be made to increase access to all neighborhoods. During the economic boom years from 2004 and 2007, their analysis reveals that low- and moderate-income (LMI) and African-American neighborhoods did not benefit as much as middle- and upper-income neighborhoods (MUI) from the increase in small business lending. Moreover, the contraction in lending was greater in LMI than MUI neighborhoods between 2008 and 2009. By 2009, there was one loan per 22.6 small businesses in MUI neighborhoods but just one loan per 28.4 small businesses in LMI neighborhoods.
The Brookings Metropolitan Policy Program and Gallup found that firms in communities of color earn lower Yelp ratings and fewer consumer reviews than firms in predominantly white neighborhoods. This damages the growth potential of firms in communities of color and points to the need for accurate census tract data so that these disadvantages can be further researched and addressed.
Since location is a significant part of CRA and fair lending enforcement, lending institutions must be instructed to provide the census tract location of small business loans. Currently, the interagency CRA Question and Answer (Q&A) guidance advises banks that “prudent banking practices and Bank Secrecy Act regulations dictate that institutions know the location of their customers and loan collateral.” The agencies further advise banks to obtain addresses beyond a post office box. In cases where only rural route numbers are provided, the agencies state that the banks still usually know the census tract location of the businesses. In cases that census tracts cannot be geocoded, the bank is to report the census tract or county as “unknown.”
This guidance for banks reporting under CRA appears to strike a reasonable balance in exhorting banks to make strenuous efforts to obtain census tract locations for loans. In fact, comparing CRA small business lending reporting to HMDA reporting, NCRC found tracts unknown for about four percent of the small business loans and five percent for conventional home purchase loans (reported under HMDA). This suggests the CRA regulation may have tighter reporting requirements than HMDA for census tract location.
Currently, the FFIEC publicly reports loans by individual census tracts for the aggregate (all banks, as a group) on a county level. For individual banks, the FFIEC publicly reports data by income category of census tracts. For example, for a fictitious bank such as “Main Street Bank,” the data disclosure would report how many loans were in low- or moderate-income tracts. The data will not report how many loans were issued by Main Street Bank in each of the census tracts in the county. In contrast, HMDA data reports data by each census tract for individual banks. Members of the public can more effectively hold banks accountable for making loans in various neighborhoods if the data is reported for each census tract. Below, suggestions are offered for overcoming an issue of borrower privacy that has made the agencies reluctant to release data for each tract for individual banks.
10) Gross annual revenue should be collected to disaggregate lending to the smallest firms, consistent with existing small business public data
In Question 40, the Bureau seeks feedback on the use of gross annual revenue. Section 1071 requires lending institutions to report “the gross annual revenue of the business in the last fiscal year of the women-owned, minority-owned, or small business loan applicant.” We urge that revenue size to be reported as a continuous variable as that is the most accurate reporting method. However, if the CFPB adopts another reporting method, it should consider a carefully developed categorical variable or a modified version of continuous reporting that involves selecting the mid-point within $10,000 increments for reporting.
The gross annual revenue size of the small business is a critical data element since research shows that smaller businesses are less likely to receive loans. In a report conducted for the Appalachian Regional Commission (ARC), NCRC obtained Pepperdine survey data (see above for a description of the Pepperdine survey) that revealed approval rates for businesses of various revenue sizes. As revealed by survey data for the first quarter of 2012, just 18 percent of the small businesses with revenues less than $500,000 who sought loans obtained them. In contrast, 35 percent of the businesses with revenues between $500,000 and $1 million and 55 percent of the businesses with revenues between $1 million and $5 million received loans.
Revenue categories should be more detailed than those in the CRA small business loan data. In the CRA small business data, the revenue categories are small businesses with revenue below $1 million dollars and small businesses with revenues above $1 million. Just the one example above from the Pepperdine survey reveals the inadequacies with this classification since businesses with revenues below $500,000 had markedly less access to loans than businesses with revenues above this amount.
In order to capture microenterprises, revenue categories need to be expanded. The Association of Enterprise Opportunity (AEO) has a further breakdown for the smallest microbusinesses as those with gross sales and receipts below $50,000. The CDFI Fund also uses $50,000 as a category in the data it makes publicly available. At a minimum, the revenue categories should include businesses with annual sales/receipts below $10,000, $10,000 to $49,999, $50,000 to $99,999, $100,000 to $249,999, $250,000 to $499,999, $500,000 to $999,999, and $1 million or more. In 2017, the vast majority of small businesses had receipts under $100,000. The Census Bureau data supports AEO’s classification in that a classification system that does not have separate categories below $500,000 will fail to adequately capture the experience in the lending markets for large numbers of small businesses.
The CFPB should further investigate the body of research to determine whether additional revenue categories are needed beyond the Census categories. According to JP Morgan Chase, “the median Black-owned firm earned $39,000 in revenues during its first year, 59 percent less than the $94,000 in first-year revenues of a typical White-owned firm. Small businesses founded by Hispanic owners earned $74,000 in revenues, or 21 percent less than the median for White-owned firms.” Revenue categories of up to $50,000 and then $50,000 to $100,000 would work for capturing differences between white and Black owned-firms since the median revenue amounts for these two racial categories of businesses are in different Census revenue categories. However, the Census categories would not work for capturing differences in median revenue amounts between white and Hispanic businesses since the difference of $20,000 would be within one revenue category. For businesses with revenues under $100,000, the revenue categories would probably need to be in $10,000 increments.
In addition, the need for care in revenue categories is reinforced by the size distribution of women-owned small businesses. According to an American Express report, 88% of all women-owned businesses have revenues less than $100,000. Just as with minority-owned firms, a reporting method needs to be developed to ensure that the data accurately describes the credit experiences of women-owned businesses. Within the $100,000 or less in revenue category, the experiences of women-owned businesses could be quite different within themselves and/or different compared to male-owned businesses.
An alternative to a categorical variable for revenue size is adopting the HMDA procedure for reporting loan amount. Lenders are instructed to report the dollar amount for loan size that represents the mid-point within a $10,000 increment in which the actual dollar amount falls. For example, if the loan amount is $112,000, this loan amount falls within the $110,000 to $120,000 interval and would be reported as $115,000. The CFPB adopted this procedure to safeguard borrower privacy.
An important issue in the CRA small business loan data is that for a sizable number of loans, the revenue size of the small business is unknown because the bank did not consider revenue size in its underwriting. This could be the case, for example, in credit card lending. However, the publicly available data provided by the FFIEC does not indicate for which loans the revenue size is unknown. When analysts seek to calculate the percentage of loans for businesses of various revenue categories, the percentage can be incorrect since it is not possible to subtract from the denominator the number of loans for which revenue size is unknown. At the very least, reporting institutions should be required to indicate the loans for which revenue size is unknown so that the data’s accuracy can be improved. The CFPB, however, must instruct lenders to make strenuous efforts to collect revenue size since it is a key variable for assessing whether lenders are meeting the credit needs of businesses of various revenue sizes.
The inaccuracy due to loans for which revenue is unknown impairs regulatory enforcement. For example, the CRA exam for East Boston Savings Bank reports that the bank issues a higher percentage of its loans than the aggregate (all other banks, as a group) to businesses with revenues of less than $1 million. However, the CRA exam acknowledges that it is unknown to what extent the bank exceeds the aggregate. The exam states:
The disparity between the aggregate and Bank performance can, in part, be explained by a significant volume of business credit card lending by larger, nationwide institutions, which often do not perform underwriting that considers GAR (Gross Annual Revenue) information. Therefore, it is reasonable to assume that a significant portion of the aggregate data comprises loans to businesses upon which GAR information was not relied on, rather than to businesses with GARs over $1 million.
In sum, this inaccuracy diminishes the ability of even regulatory agencies to adhere to their CRA responsibilities in assessing whether banks are meeting credit needs of small businesses.
A final point for the CFPB to consider is whether the revenue variable would be reported on a cash flow or accrual basis.
11) Collect and report disaggregated race and ethnicity data consistent with the 2015 HMDA final rule
In question 46, the Bureau seeks feedback on the potential challenges, costs, and benefits of collecting and reporting the race, sex, and ethnicity of principal owners using aggregate categories. The 2015 HMDA final rule, revised in 2017, currently requires that lenders collect and report disaggregated race and ethnicity data for Latino and Asian borrowers. Our research has found that, in 2019, borrowers are willing to report these data, with almost 60% of borrowers that were eligible to report disaggregated race and ethnicity data choosing to do so. Likewise, analyses of these disaggregated data for Asian and Latino borrowers have provided additional, useful information on the ability of Asian and Latino borrowers of different countries of origin to access credit and the terms in which that credit is accessed. We urge that the Bureau collect similar data under Section 1071, in a manner consistent with the 2015 HMDA final rule.
In Question 47, the Bureau considers the option of reporting a principal owner’s race, sex, or ethnicity based on visual observation or surname and concludes that the use of visual observation and surname should not be used. We urge the Bureau to consider the use of visual observation and surname if the applicant declines to provide this information, consistent with its application under Regulation C. If the Bureau does not adopt this requirement, we urge, at minimum, that the Bureau to clarify this method is permissible as part of the institution’s internal small business fair lending review procedures.
B) Additional discretionary data points are necessary to apply the definitions of small business and to fulfill the public policy purpose of Section 1071
In addition to the mandatory data fields listed in the statute and described in the prior section, we urge the Bureau to collect and disclose additional discretionary data fields, including loan pricing, time in business, number of employees, timing considerations, and source of credit report used for the any determination of creditworthiness. We discuss the efficacy and cost of collecting each of these data points in response to the Bureau’s questions 48 and 49 and in further detail in the following questions.
1) Collecting and disclosing loan pricing data is necessary and consistent with other lending disclosure requirements
In Question 49-51, the Bureau seeks feedback on potential challenges and costs for collecting pricing information, including any challenges with collecting and disclosing pricing information for various products, including term-loans, open-end credit, merchant cash advance and factoring agreements. We urge the Bureau to collect interest rate (expressed as a rate spread, similar to pricing information collected under Regulation C), total fees, the total cost of borrowing and the annual percentage rate (consistent with the APR required under the Truth in Lending Act). In some cases, one or more of components of the total loan price may not be required under TILA. We describe how the Bureau could address some of these omissions and ensure that pricing disclosure for certain products is consistently applied in the sections below.
a) Collecting and disclosing pricing information in necessary to monitor the market for emerging high-cost products and fair lending concerns
Loan pricing data is a critical fair lending tool and affords regulators, advocates and industry the opportunity to conduct fair lending reviews and monitor the market for emerging high-cost products. Pricing information was only included in Home Mortgage Disclosure Act data starting in 2004 and even then, it provided only rate spread information above a certain threshold. Additional pricing data points were added in 2015 and provided more detailed information, including information on total loan costs, total points and fees, origination charges, discount points, and lender credits. These data have been critical to identifying disparate pricing among protected classes, including higher prices charged to Asian borrowers, Latino borrowers and same-sex couples. While some pricing data were required by the HMDA amendments in Dodd-Frank, other data points were added under the Bureau’s discretionary authority. The challenges that women-owned businesses and businesses owned by people of color have when seeking to access credit are well-documented in prior sections, and analyses of pricing disparities in the small business market must be included in efforts to address these challenges. To this end, we urge the Bureau to use its similar discretionary authority under Section 1071 to ensure this information is collected and disclosed.
b) Methods for collecting pricing data as an annual percentage rate are well-established for closed-end small business loans and could be modified for open-end loans
In Question 51, the Bureau seeks feedback on the benefits of using various pricing metrics. Assessing the total cost of credit expressed annually is a critical tool in comparing loan products, including loans used to start, sustain or expand a small business. Credit extended primarily for business, commercial, or agricultural purposes are exempt from the disclosure requirements of Regulation Z, however. Methods for calculating loan APRs are well-established for closed-end consumer loans under Regulation Z, and with some modification, could also be applied to open-end loans.
First, the Bureau should consider using its discretionary authority under Section 1071 to collect the data points necessary to calculate an annual percentage rate, including interest rate, additional fees, term (proposed as part of the loan/credit type data point in question 29), payment frequency, payment amount for closed-end loans. For open-end loans, the amount of the line of credit (discussed in prior sections) and the minimum payment should also be included.
The second option is to require the disclosure of the annual percentage rate as currently required for consumer closed-end loans, and to open-end loans using the methods described above. Under this second option, we urge the Bureau to still collect and disclose the interest rate expressed as a rate spread and any additional fees.
Additionally, since 1071 data are reported yearly and interest rates fluctuate throughout the year, we urge that the Bureau collect and disclose the contractual interest rate as well as a rate spread expressed in basis points over a determined index, similar to the rate spread reported under HMDA.
c) States have led the way in determining how to calculate and disclose pricing for merchant cash advances
We urge the Bureau to require pricing disclosures for MCAs, using the procedures adopted in New York and California. Admittedly, the Bureau will have to address certain complexities associated with MCAs, as their structures differ from traditional lines of credit and term loans. We urge the Bureau to price MCAs using an internal rate of return calculation, based on the use of estimates derived from lenders for future cash flow, with the ability to also capture fees, and with a system for retroactively identifying cases where lenders are falsely deflating projected cash flows that include enforcement provisions.
To capture an annual percentage rate for an MCA, the Bureau must adopt a procedure for determining the term of the MCA. This presents some challenges since, under an MCA, repayment periods are determined by future cash flows. Generally, a lender collects a percentage of debit and credit card sales. A hypothetical example: a lender advances the merchant $10,000 with the expectation that a business will return $14,000, with repayments made on a monthly basis for 10 percent of card-related sales, and over a period of no more than three years. If the borrower can generate card sales of $3,889 per month, the lender will satisfy the debt after 36 months. In this instance, the APR using a term of 36 months would be 23.3 percent (assuming no fees). However, if the business recorded monthly average card sales of $5,833, leading to repayment in 24 months, the APR is 34.6 percent. The challenge is that card sales cannot be predicted with certainty, and if a borrower repays earlier, the interest rate is higher. This uncertainty requires a pricing calculation that uses an estimate of future sales. Moreover, unless the regulator implements retroactive procedures to verify the consistent accuracy of estimates, a lender could inaccurately disclose the APR by underestimating predicted card sales.
The formulas used by California and New York address the uncertainty of the loan term by estimating the future cash flows of the small business. Other facts, such as the amount disbursed, the amount that a business is obligated to repay (including any fees), and the portion of cash flows that the lender will capture to satisfy the debt, are readily available from stated terms of the MCAs.
The California and New York methods impute the loan term by using one of two methods to estimate future revenues. One system (the “historical” method) uses recent sales records; the other (the “opt-in”) allows the lender to submit an estimate of likely future cash flows. The latter approach solves for the problem of applications made by startup businesses. The Bureau would not have to create the estimates, now would it have to decide which method applied, as lenders develop these estimates internally during the course of underwriting. The Bureau could ask the lenders to indicate the type of source (historical or opt-in) and the expected amount.
The Bureau can determine the APRs using a standard internal rate of return calculation. In some cases, a contract may also call for the payment of certain fees independent of charges associated with card sales. The IRR calculation has the flexibility to accommodate that condition, even if the fees are not charged in a consistent cadence.
Given the possibility that an unscrupulous lender could manipulate the system by intentionally underestimating cash flows in the opt-in approach – with the effect of extending the loan term and therein lowering the APR – both states adopted review processes to verify the integrity of a lender’s reporting system. Enforcement provisions have been added when evidence shows that a lender consistently underestimates cash flows.
d) States have led the way in determining how to calculate and disclose pricing for factoring agreements
We believe that the Bureau can collect pricing data for factoring agreements in ways that are neither burdensome nor ambiguous. The critical data points to determine pricing for factoring agreements are the value of the underlying receivables, the advance rate, the number of days between the distribution of the advance and the due date of the receivables, the periodic factoring rate, and any administrative fees. Factoring companies have these data points, and they are often provided in the contract. Broadly speaking, collected data should include the specific points needed to calculate the cost of credit as well as certain variables that provide clarity on how receivables default risk is assigned.
Consider a hypothetical example where a factoring company advanced $16,000 to a small business at an advance rate of 80 percent on a $20,000 portfolio of receivables due in 60 days. The factoring fee is 2 percent per month (or 0.07 percent per day). The contract charges a monthly maintenance fee of $50, to be paid separately on day 30. On a daily accrual basis, the small business pays a credit expense of $13.15 per day, but for the purposes of an internal rate of return calculation, the cost of the credit is experienced when the factoring company debits the factoring fees against the $4,000 held in the outstanding escrow deposit at the end of day 60. Thus, the small business receives $16,000 on day 0 and pays back the $16,000 in principal and $789 in factoring fees on day 60. The IRR before administrative fees is 29.3 percent, and the APR after accounting for the maintenance fee paid on day 30 is 31.2 percent.
We urge the Bureau to collect data points on all administrative fees, such as fees to transfer funds, mail documents, or for establishing the account. Additionally, the data should indicate who bears the risk for defaulted receivables and if the receivables are for a portfolio or a single (“spot”) invoice, as such indicators will add important context to understanding the risk-reward tradeoff. Additionally, as factoring uses risk-based pricing, data set users need additional data points. To convey the default risk on the collection of receivables, the data should indicate if the contract has recourse or non-recourse structure. To convey diversification risk, the data should report the number of invoices in the portfolio, as factoring companies charge more when a portfolio includes fewer invoices.
2) Collecting time in business is necessary to ensure fair access to credit for start-ups and younger firms
In Question 53, the Bureau seeks information on a firm’s time in business. We urge that these data be collected given the well-documented challenges that start-ups and younger firms encounter in accessing credit compared to more established businesses. In a recent survey of microbusinesses, CFED (now Prosperity Now) finds that younger businesses are twice as likely to indicate trouble accessing credit than more established businesses. In addition, businesses under one-year-old used an average of 3.9 financial products in contrast to 6.6 used by businesses over ten years old. The Federal Reserve Banks report that their 2016 survey revealed that start-up firms have greater difficulty accessing credit than mature firms and that they are less likely than mature firms to receive the full amount of credit requested. Therefore, years in business is an important control variable for explorations of gender and racial disparities in access to credit.
3) Number of employees is a necessary field to apply an employee-based definition of small business and provide insight into credit availability for the smallest firms
In Question 55, the Bureau seeks input on the ability to collect the number of employees during a small business application. In question 14, we urge the Bureau to adopt a definition of small business as any firm with 499 employees or fewer, necessitating the collection of the number of employees as a discretionary data point. Likewise, we urge the collection of total number of employees and number of owners separately.
Given the large percentage of non-employer firms and previous challenges collecting accurate employee counts under an aggregated reporting structure, we believe this is a necessary step to prevent errors and confusion. Data collected under the Paycheck Protection Program provides an illustrative example. After the release publicly released dataset, analysis of these data suggested some confusion on behalf of both the lender and the borrower about how to report the number of employees for a non-employer firm with some non-employers clearly reporting the number of employees as zero, with other non-employers reporting one. We urge the Bureau to require lenders to report owner and employees separately to reduce the need for further verification and to facilitate verification at later stages of origination or after origination as needed. Regardless of whether employees and owners are collected separately or in aggregate, user testing and guidance are necessary to ensure that the number of employees is reported accurately and consistently.
4) Section 1071 disclosure requirements should be triggered when the applicant requests a loan application
In Question 56, the Bureau seeks information on timing considerations for certain data. We urge the Bureau to collect data when an applicant requests an application. The CFPB proposes that lenders do not need to collect Section 1071 data during any specified time period during the application process. The CFPB states it does not want to impose costs on lenders by interfering with the timing of their data collection activities during applications. However, the agency acknowledges that a lack of standardizing when to collect Section 1071 may lead to inconsistencies among lenders or delays in obtaining Section 1071 data if the data is collected “late in the process when applicants may be less motivated to supply their demographic information.”
Instead, the CFPB should require the collection of Section 1071 data when a customer has requested an application, that is when the customer requests credit. The CFPB mentioned it was considering this option and should adopt this option. An additional benefit of collecting data earlier in the process at the time of application is that it promotes standardization of treatment of applicants. The possibility of disparities in treatment by race and gender may diminish since collecting Section 1071 data intended to reduce the incidence of discriminatory treatment would occur early in the process and impress upon the lenders’ and applicants’ the importance of fair and non-discriminatory treatment. The CFPB should also consider conducting testing and focus groups on determining the optimal time for collecting Section 1071 data.
5) The Bureau should collect information on the use of personal or business credit reports or collateral for business loans
The Bureau should collect and disclose whether a financial institution relied on a personal or business credit profile, credit score, or other determinants of credit history under Section 1071. Small business owners frequently rely on their personal credit reports to apply for credit. In fact, 86% of employer firms stated that they relied on their personal credit score, while only 13% relied solely on the credit score of the business. Other types of personal collateral are also used frequently, with 58% of employer firms stating they used a personal guarantee and 31% using personal assets to secure credit.
G) The Bureau can take steps to address the privacy considerations included in the 1071 proposals under consideration
In Questions 69 through 73, the Bureau seeks feedback on privacy-related issues. The CFPB is charged with balancing the benefits of data disclosure against the risks of privacy invasions and identification of specific borrowers using Section 1071. While not minimizing the seriousness of this issue, we note that in the more than 40 years of HMDA data reporting, the federal regulatory agencies have not reported any instances of privacy invasions in either Congressional testimony or other official reports or documents. Likewise, we believe that Section 1071 can include robust data disclosure without risking privacy invasions. Privacy concerns, particularly preventing the re-identification of applicants based solely on publicly disclosed data, can be sufficiently prevented by preventing the collection of personally identifiable information (PII), increasing the number of covered institutions and covered products, and implementing masking techniques to prevent re-identification. Collectively, and in addition to other measures that might be necessary, we believe that the Bureau can minimize privacy concerns while ensuring that the fair lending and community reinvestment purposes of Section 1071 are met.
Neither HMDA data nor the Section 1071 data pose significant privacy concerns because the data does not have personally identifiable information that can be used to target specific borrowers. In contrast, the breach of Equifax data imperils the privacy of up to 143 million consumers precisely because the Equifax data contains personally identifiable information such as Social Security numbers, street addresses, and birth dates.
Concerns have been expressed that lending data enables predatory marketing. In this scenario, bad actors could use the publicly available data to identify vulnerable borrowers or applicants that have been denied loans and then aggressively market abusive products to them. In the HMDA context, these claims have been far-fetched. A lack of data on the extent of abusive home lending was a major reason predatory lending was able to proliferate. If the Dodd-Frank HMDA enhancements had been implemented sooner, the public and regulatory agencies could have used the data to identify surges in problematic lending and taken steps to deter abusive lending before it reached crisis proportions. Thus, data disclosure is a tool to deter predatory practices, not enable them. Like HMDA data, Section 1071 data could be used to identify problematic pricing or loan terms and conditions before these proliferate into crisis proportions. Also, just as with HMDA data, we believe that the CFPB can take steps to protect against any possibilities of predatory marketing resulting from privacy invasions. The techniques described below would effectively protect borrower privacy.
2) Increasing the number of reported applications and loans will make it more difficult to identify borrowers in a specific geography
Increasing the universe of lenders reporting small business data will help alleviate the concern about privacy. HMDA data has generally involved a higher volume of loan data than CRA small business data. The CRA small business data loan volume would be boosted considerably if smaller banks, on-line lenders, credit unions, merchant cash advance, and factoring firms would be required to report. For example, the large banks that currently report CRA small business loan data made 70 percent of the small business loan dollars in 2015. Smaller banks that are not required to report small business loan data issued 30 percent of the loan dollars, illustrating that substantially more loans would be captured if data reporting requirements are expanded to more banks and non-banks.
If the number of observations can be increased by one third to one half from currently reported levels, the small business lending levels in publicly available databases would be comparable to HMDA. HMDA has loan-level reporting on a census tract level, in part, because higher loan volumes make it difficult to identify borrowers using HMDA data. If small business loan volumes mirror HMDA loan volumes, robust data disclosure on a census tract level is possible without fear of significant privacy invasions.
It is not only the number of reporters that accounts for fewer small business loan observations than HMDA observations. HMDA has more action categories, including applications and denials than small business data, which only includes originations. Since action categories, including applications and denials, would be mandatory data reporting items, including them in the small business database would significantly increase the number of observations, making it much harder for predatory actors to identify small business applications using Section 1071 data.
3) Implement masking techniques to prevent re-identification and preserve the usefulness of Section 1071 data in aggregate
In addition to increasing the number of reporters and action categories, there are masking techniques that make it more difficult to identify specific borrowers. In the case of census tracts with low levels of lending to a specific type of small businesses (for example, either a minority-owned or a microenterprise with one employee), the data associated with those particular small businesses could be moved to a contiguous or nearby census tract.
In particular, the CFPB should consider using data swapping to protect the privacy of loan applicants whose demographic data points are part of a small group. Following the methodology used by the U.S. Census in its Public Use Microdata Sample, the CFPB could switch records for similarly-situated loan applicants between nearby census tracts, making it nearly impossible to reconnect individual loan applicants with public records while maintaining the utility of Section 1071 data, including for users doing analysis at the neighborhood level. The CFPB should explore which of the data swapping techniques the U.S. Census uses would best balance protecting loan applicants’ privacy and allowing users to do a meaningful analysis of Section 1071 data at the census tract level. This would not significantly compromise the integrity of the database and could more effectively mask the identity of the small business.
Additional masking techniques include making some variables categorical variables instead of reporting specific values. For example, the revenue size of the small business could be reported in categories rather than specific revenue amounts. If some census tracts have low levels of loans below a specific threshold, sensitive variables could be reported in percentages or aggregate numbers rather than loan-level reporting. For example, for a particular census tract, data disclosure could report the total number of minority-owned businesses that were approved and denied rather than loan-level reporting that occurs in HMDA data. While not ideal, this would be an improvement over the level of detail in the CRA small business data and would be more useful for analytical purposes in determining the extent to which credit needs are being met for different types of small businesses or neighborhoods.
4) Increased transparency will create a more, not less, competitive small business credit market
Finally, some reporting institutions may claim their competitors can determine their business or pricing strategies by using detailed, publicly available data. When disclosed carefully, data would not significantly impede the competitive position of reporting institutions. For example, a pricing disclosure can reveal overall costs to borrowers without providing detailed aspects of how a particular lender may adjust fees and interest rates or stagger payments in order to entice borrowers by easing repayment stress for borrowers. Moreover, publicly available data also provides insights to lenders about how to become more competitive. Based on conversations with industry stakeholders, when HMDA data becomes available in early April each year for individual lenders, lenders often use data from their competitors so that they can glean insights into their market positions as compared to their peers. Overall, data prevents monopolization and exploitation and promotes competition through transparency and accountability.
5) The Bureau must consider the three primary use cases when publishing loan-level small business loan data
In Question 74, the Bureau seeks feedback on the public disclosure of data collected under Section 1071. We urge the Bureau to address the public disclosure of these data in three forms. The Bureau should make loan-level data publicly available as a machine-readable source file in commonly available formats, along with any necessary input code for commonly available statistical packages. The Bureau should also develop a robust and user-friendly web query interface that allows the public to quickly access summary statistics by borrower demographic and geography using the previously available HMDA Explorer as a starting point. Given the wide range of possible queries that could be incorporated into such an interface, we urge the Bureau to convene a user advisory committee to solicit periodic feedback, as well as other ongoing public feedback mechanisms. We also urge that the Bureau publish its own annual report of trends in small business lending, similar to the reports published with the annual release of HMDA data.
H) The Bureau should issue a strong final rule without delay, with an implementation period of no more than 12 months
In Question 76, the Bureau seeks feedback on the implementation period. Dodd-Frank was passed ten years ago, and since then, the mandate to collect small business data has gone unfulfilled. In fact, during this time, the Bureau has taken on numerous efforts to make lending data less available by raising reporting thresholds for HMDA and then by reducing the ability of novice data users to retrieve HMDA data in concise summary tables. We urge that the Bureau complete a proposed and final rule in 2021 consistent with the recommendations described above. We also urge an implementation period of no more than 12 months.
IV) The impact on small entities can be minimized while ensuring market transparency and necessary fair lending functions
In question 78, the Bureau seeks feedback on the cost associated with an eventual 1071 final rule. Data disclosure entails benefits and costs, and, in this case, we believe that the benefits of data disclosure significantly outweigh the costs. Data disclosure promotes more efficient and equitable markets, enabling more small businesses to acquire loans, expand their size, and create more jobs. While there are costs, bank behavior towards data disclosure illustrates a pragmatic acknowledgment of the benefits of data disclosure. We also believe that Section 1071 data can replace existing disclosure requirements for many lenders that are currently reporting data under other statutory or regulatory requirements. Given the important public purpose of these data, we believe that the costs associated with Section 1071 data collection are reasonable given the vital purpose of small business market transparency. Some lenders opt-in to disclosing lending data because it provides a vital and competitive function.
As mentioned above, about 22% of banks reporting CRA small business data in 2018 had assets below the threshold that would require them to submit the data. These smaller banks voluntarily reported because they wished to be evaluated on the large bank CRA exam, or they saw some other benefit to data disclosure. Perhaps, they wanted to compare their performance with peers in order to figure out how to be more competitive and/or to find some overlooked and profitable lending opportunities. Perhaps, they believed their performance was commendable, and they wanted to disclose their data as a means of bolstering their reputations as responsive to the needs in their communities.
Despite the claims of some, it appears there is a consensus among community organizations and significant numbers of lending institutions (specifically those that are voluntarily reporting) that data disclosure is a vital part of well-functioning markets in that it identifies and reduces burden and obstacles to lending.
B) The costs associated with Section 1071 data collection are reasonable given the vital purpose of small business market transparency
In the SBREFA outline, the CFPB describes a detailed analysis of costs for smaller banks of $600 million in assets or less. The CFPB assumes that these small banks are divided into two categories: Type A receive an average of 75 applications per year, and Type B receive about 300 applications per year. The CFPB describes data entry costs, costs associated with preparing the data for reporting and transmitting the data, and costs associated with auditing and compliance for Type A and Type B institutions.
For type A institutions, the per-application costs range from $26 to $56, while the net income ranges from $37,000 to $45,000. For type B, the per-application costs range from $62 to $145, while net income ranges from $12,000 to $13,000. The costs are small compared to the net income generated from each loan for the two categories of lenders. The CFPB, therefore, predicts that the low variable costs would range from $17 to $40 per loan. The agency concludes, “Even if the variable cost were passed on in full to small business borrowers in the form of higher interest rates or fees associated with a loan or line of credit (or even applicants in the form of application fees), the Bureau expects that this would comprise a small portion of the total cost of the average loan to the small business borrower.”
Based on the CFPB’s analysis of costs and how much of them would be passed onto the borrower, costs are quite low compared to the benefits of more precision in estimating community needs and improved fair lending enforcement. Moreover, the Board estimates that the costs of adding discretionary data points in addition to the mandatory ones would add only $4 to $5 per application.
C) All applicants should be provided with an anti-discrimination notice, not just applicants where demographic information is accessible to underwriters
In question 59, the CFPB considers methods to shield demographic information from underwriters Section 1071 requires that underwriters not have information on the demographic information of applicants. We recognize that, for very small institutions, it may not be feasible to shield this information from underwriters. The statute accommodates this by requiring all applicants should receive an anti-discrimination notice. We urge the Bureau to provide this notice to all applicants, not only applicants whose information cannot be shielded in a “firewall” from underwriters. We do not believe that small business status should be shielded from underwriters.
We appreciate the opportunity to comment on the Bureau’s proposals to implement Section 1071 of Dodd-Frank. For more information, please do not hesitate to contact me or Tom Feltner, Director of Policy at firstname.lastname@example.org or Josh Silver, Senior Policy Advisor at email@example.com.
Jesse van Tol, CEO
National Community Reinvestment Coalition
 Fairlie, R. W. (2020). The Impact of Covid-19 on Small Business Owners: Evidence of Early-Stage Losses from the April 2020 Current Population Survey (No. w27309). National Bureau of Economic Research. https://doi.org/10.3386/w27309
 12 CFR Part 1003 (as amended October 15, 2015)
 Q1) Are there any relevant Federal laws or rules which may duplicate, overlap, or conflict with the Bureau’s proposals under consideration beyond those discussed in Appendix C? How might the Bureau’s proposals under consideration for implementing section 1071 impact other aspects of ECOA/Regulation B compliance?
 12 CFR § 25.42
 Q2) Please provide feedback and information on the approach the Bureau is considering regarding the scope of its section 1071 rulemaking particularly the proposal to limit reporting to applicants that satisfy the Bureau’s definition of a “small business.” Are there any alternative approaches the Bureau should consider?
 Q3) How often does your FI make loans to businesses that are not “small”? Would you anticipate any specific complexities or costs in identifying women-owned and/or minority-owned applicants that are not small businesses, and collecting 1071 data about their applications for credit?
 CFPB. (2020). Small Business Advisory Review Panel for Consumer Financial Protection Bureau, Small Business Lending Data Collection Rulemaking, Outline of Proposals Under Consideration and Alternatives Considered. p. 9, https://files.consumerfinance.gov/f/documents/cfpb_1071-sbrefa_outline-of-proposals-under-consideration_2020-09.pdf
 CFPB (2020). Small Business Advisory Review Panel for Consumer Financial Protection Bureau, Small Business Lending Data Collection Rulemaking, Outline of Proposals Under Consideration and Alternatives Considered. p. 9, https://files.consumerfinance.gov/f/documents/cfpb_1071-sbrefa_outline-of-proposals-under-consideration_2020-09.pdf
 Q5) Please provide feedback and information on the approach the Bureau is considering regarding the general definition of “financial institution,” along with any alternative approaches the Bureau should consider.
 CFPB (2020). High-Level Summary of Outline of Proposals Under Consideration for SBREFA: Small Business Lending Data Collection Rulemaking. https://files.consumerfinance.gov/f/documents/cfpb_1071-sbrefa_high-level-summary-of-outline-of-proposals_2020-09.pdf. pp. 2-3.
 Karen Gordon Mills and Brayden McCarthy (2016). The State of Small Business Lending: Innovations and Technology and the Implications for Regulation. Harvard Business School. p. 48.
 Federal Reserve Banks (2020). 2020 Report on Employer Firms, Small Business Credit Survey. p. 8. https://www.fedsmallbusiness.org/medialibrary/FedSmallBusiness/files/2020/2020-sbcs-employer-firms-report.pdf
 Barbara J. Lipman, Federal Reserve Board Division of Consumer and Community Affairs, and Ann Marie Wiersch, Federal Reserve Bank of Cleveland Community Development Department (November 2019). Searching for Small Business Credit Online: What Prospective Borrowers Encounter on Fintech Lender Websites. Consumer and Community Context, Vol. 1, No. 2, a Federal Reserve System publication, pp. 3 and 10.
 Federal Reserve Banks (August 2017). Small Business Credit Survey: Report on Startup Firms. p. 15.
 Q6) Please provide feedback and information on the approach the Bureau is considering regarding the possible exemptions for FIs based on size and/or activity, along with any alternative approaches the Bureau should consider.
 12 CFR Part 1003 (as amended October 15, 2015)
 Access to Capital and Credit for Small Businesses in Appalachia. (2007). National Community Reinvestment Association. https://ncrc.org/wp-content/uploads/2007/05/ncrc%20study%20for%20arc.pdf
 Federal Reserve Banks, Small Business Credit Survey, April 2017, 14
 NCRC analysis of 2018 CRA small business data
 Federal Reserve Banks, 2020 Report, p. 20.
 Q8) What compliance costs would cause your FI to stop or decrease your small business lending?
 Federal Financial Institutions Examination Council, History of HMDA, https://www.ffiec.gov/hmda/history2.htm
 Federal Financial institutions Examination Council, Background and Purpose of HMDA, https://www.ffiec.gov/hmda/history.htm
 Q9) Are there certain types of FIs, such as governmental lending entities or non-profit non-DI lenders, that the Bureau should consider not including within 1071’s data collection and reporting requirements? If so, why?
 Q13) Please provide feedback and information on the approach the Bureau is considering regarding treatment of FIs that are not the lender of record, along with any alternative approaches the Bureau should consider.
 OnDeck Capital. (2019, March 1) Annual Report for the Year Ending December 31st, 2018. Submitted to the Securities and Exchange Commission. Retrieved at https://www.sec.gov/Archives/edgar/data/1420811/000142081119000045/ondk-20181231x10k.htm
 Q14) Please provide feedback and information on the approach the Bureau is considering regarding the definition of “small business,” along with any alternative approaches the Bureau should consider. For example, should the Bureau include or exclude applications from particular types of borrowers from the scope of its eventual 1071 rule in addition to or differently than as described herein?
 CFPB.Outline. p. 61.
 CFPB. Outline. p. 9.
 U.S. Census Bureau, (n.d.). Annual Business Survey Methodology. The United States Census Bureau. Retrieved November 30, 2020, from https://www.census.gov/programs-surveys/abs/technical-documentation/methodology.html
 See Q16) Are you familiar with the SBA’s six-digit NAICS code-based size standards, and does your FI currently use them for any purpose? What would the cost be to implement a small business definition based on the SBA’s size standards? And Q54)
 15 U.S.C. 1691c-2
 Classification Development Branch, E. S. M. D. (n.d.). North American Industry Classification System (NAICS) Frequently Asked Questions (FAQs). Retrieved November 24, 2020, from https://www.census.gov/eos/www/naics/faqs/faqs.html#q4
 Table of size standards. (2019). Small Business Administration. https://www.sba.gov/document/support–table-size-standards
 Classification Development Branch, E. S. M. D. (n.d.). North American Industry Classification System (NAICS) Main Page. United States Census Bureau. Retrieved December 3, 2020, from https://www.census.gov/eos/www/naics/
 Classification Development Branch, E. S. M. D. (n.d.). North American Industry Classification System (NAICS) Frequently Asked Questions (FAQs). Retrieved November 24, 2020, from https://www.census.gov/eos/www/naics/faqs/faqs.html#q4
 Q17) Please provide feedback and information on the approach the Bureau is considering regarding the definitions of “women-owned business,” “minority-owned business,” and “minority individual,” along with any alternative approaches the Bureau should consider.
 Small Business Advisory Review Panel for Consumer Financial Protection Bureau Small Business Lending Data Collection Rulemaking: Outline of Proposals Under Consideration and Alternatives Considered. (2020). Consumer Financial Protection Bureau. https://files.consumerfinance.gov/f/documents/cfpb_1071-sbrefa_outline-of-proposals-under-consideration_2020-09.pdf (p. 18)
 Luque, A. et al (2019). Nonemployer Statistics by Demographics (NES-D): Using Administrative and Census Records Data in Business Statistics. CES 19-01, 142.
 U.S. Census Bureau, Annual Business Survey Methodology.
 15 U.S.C. § 1691c–2(h)(5)
 Q18) What are the legal or ownership structures of the businesses that typically apply for small business loans from your FI (i.e., sole proprietorship, partnership, limited liability company, “S” corporation, etc.)? Do those businesses typically have an indirect ownership structure (i.e., ownership interests are held by other entities)? What persons or group of persons are typically responsible for the operations of such business (i.e., whether a managing member, two or more partners, a CEO, or some other person or group of persons)?
 Q20) Please provide feedback and information on the approach the Bureau is considering regarding covered products and use of the ECOA definition of “credit” for purposes of defining covered products under section 1071, along with any alternative approaches the Bureau should consider. Are there any products that should or should not be covered by the Bureau’s eventual 1071 rule, and if so why?
 Board of Governors of the Federal Reserve System. (September 2017). Report to the Congress on the Availability of Credit to Small Businesses. Retrieved at https://www.federalreserve.gov/publications/files/sbfreport2017.pdf
 Q21) What challenges would you anticipate if leases, trade credit, factoring, or MCAs or some subset(s) thereof, were included as covered products under the 1071 rule? Do you have suggestions on how to mitigate or resolve those challenges? If a subset of any of these products were included, do you have suggestions on how to define such a subset, what to include, and why (for example, including only capital leases as a covered product or only including a subset of MCAs)?
Heskin, R. S. (2019, June 26). ”Crushed by Confessions of Judgement: The Small Business Story” Testimony Before the United States House of Representative Committee on Small Business. https://smallbusiness.house.gov/uploadedfiles/06-26-19_mr._heskin_testimony.pdf
 Federal Reserve Banks. (2019) Small Business Credit Survey: 2019 Report on Employer Firms. Retrieved at https://www.fedsmallbusiness.org/medialibrary/fedsmallbusiness/files/2019/sbcs-employer-firms-report.pdf
 Pryor, Charles, Nicholas Lynch, and Stephen S. Gray. (December 2020). ”When factoring receivables can help SMEs improve cash flow.” Journal of Accountancy. Association of International Certified Professional Accountants. Retrieved at https://www.journalofaccountancy.com/issues/2020/dec/factoring-receivables-can-help-improve-cash-flow.html
 Lipman, Barbara and Ann Marie Wiersch. (June 2018). ”Browsing to Borrow: ’Mom and Pop’ Small Business Perspectives on Online Lenders.“ Board of Governors of the Federal Reserve System. Retrieved at https://www.federalreserve.gov/publications/files/2018-small-business-lending.pdf
 Sweeney, P. (2019, August 19). ”Gold Rush: Merchant Cash Advances Are Still Hot.” Debanked. https://debanked.com/2019/08/gold-rush-merchant-cash-advances-are-still-hot/
 CFPB, Key Dimensions of the Small Business Lending Landscape, 21-22, May 2017, https://www.consumerfinance.gov/data-research/research-reports/key-dimensions-small-business-lending-landscape/
 Square. 2019 10-K, (2020, February 26). Annual Report filed to the Securities and Exchange Commission. Retrieved at https://s21.q4cdn.com/114365585/files/doc_financials/2019/q4/Square-2019-10-K.pdf In its filing, Square provides additional commentary on its lending model: “Generally, for loans to Square sellers, loan repayment occurs automatically through a fixed percentage of every card transaction a seller takes. Loans are sized to be less than 20% of a seller’s expected annual GPV and, by simply running their business, sellers repay their loan in eight to nine months on average. https://s21.q4cdn.com/114365585/files/doc_financials/2019/q4/Square-2019-10-K.pdf
 Credit Suisse Equity Research. (2020, September 27th). “Square: Updated Model with New Seller & Cash App Segmentation; Updated GPV Exit Rates & App Download Data.” A research note to investors.
Federal Trade Commission v. RCG Advances LLC, Ram Capital Funding LLC, Robert L. Giardina, Jonthan Braun, Tzvi Reich. Case No. 20-CV-4432. https://www.ftc.gov/system/files/documents/cases/192_3252_rcg_advances_-_complaint.pdf
 Louis, T., Weaver, E., Brown G., McShane, C., (2016, May.) Unaffordable and Unsustainable: The New Business Lending on Main Street. Opportunity Fund. https://www.opportunityfund.org/blog/unaffordable-and-unsustainable-new-opportunity-fund-report/
 New York Attorney General (2020, June 10) Press Release: Attorney General James Sues Predatory Lender That Threatened Violence and Kidnapping, and Illegally Collected Millions from Small Businesses. https://ag.ny.gov/press-release/2020/attorney-general-james-sues-predatory-lender-threatened-violence-and-kidnapping
 McLean, B. (2018, December 8). ‘We’re coming after you’: Inside the merchant cash advance industry. Yahoo! Finance. https://finance.yahoo.com/news/merchant-cash-advances-salvation-small-businesses-payday-lending-reincarnate-161835117.html?guccounter=1
 Lewis, K. M. (2019, January 3). ”Agreeing in Advance to Lose? Legal Considerations in Regulating Confessions of Judgment.” Congressional Research Service. https://fas.org/sgp/crs/misc/LSB10239.pdf
 Mider, Z.R., Faux, Z. (2018, November 20) Sign Here to Lose Everything Part 1: ”I Hereby Confess Judgement.” Bloomberg. https://www.bloomberg.com/graphics/2018-confessions-of-judgment/?srnd=confessions-of-judgment
 16 CFR § 444.2
 Mider, Z.R., Faux, Z. (2018, November 27) Sign Here to Lose Everything Part 2: The $1.7 Million Man. Bloomberg. https://www.bloomberg.com/graphics/2018-confessions-of-judgment-millionaire-marshal/
 Federal Trade Commission. (2020, August 3). FTC Alleges Merchant Cash Advance Provider Overcharged Small Business Millions. https://www.ftc.gov/news-events/press-releases/2020/08/ftc-alleges-merchant-cash-advance-provider-overcharged-small
 State of California. SB 1235 An act to add Section 22780.1 to, and to add Division 9.5 (commencing with Section 22800) to, the Financial Code, relating to commercial financing. Retrieved at https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=201720180SB1235
 State of New York. 2019. S5470 The Truth in Lending Act. Retrieved at https://www.nysenate.gov/legislation/bills/2019/s5470
 Commission of Financial Protection and Innovation v. Allup Finance LLC. CFL File No: 60DBO-77076. https://dfpi.ca.gov/wp-content/uploads/sites/337/2020/11/Consent-Order-Allup-Finance-LLC.pdf
 Q23) Please provide feedback and information on the approach the Bureau is considering regarding the definition of “application,” along with any alternative approaches the Bureau should consider.
 CFPB, Outline. p. 22.
 HMDA, § 1003.2 Definitions (b), https://www.consumerfinance.gov/policy-compliance/rulemaking/regulations/1003/2/
 CFPB. Outline. p. 23.
 Q25) Please provide feedback and information on the approach the Bureau is considering for each mandatory data point, along with any alternative approaches the Bureau should consider.
 Small Business Administration, Office of Advocacy (September 2011) Frequently Asked Questions about Small Business Finance. http://www.sba.gov/sites/default/files/files/Finance%20FAQ%208-25-11%20FINAL%20for%20web.pdf.
 Scott Shane (May 2014). How Personal Credit Affects Small Business Borrowing. http://smallbiztrends.com/2014/05/personal-credit-affects-small-business-borrowing.html/print/
 CFPB. Outline. p. 28.
 CFPB. High-Level Summary. p. 4.
 CFPB. Outline. p. 23.
 Addressing the Financing Needs of Small Businesses: Summary of Key Themes from the Federal Reserve System’s Small Business Meeting Series (2010). http://www.federalreserve.gov/newsevents/conferences/sbc_small_business_summary.pdf.
 Alicia Robb (April 2013). Access to Capital among Young Firms, Minority-Owned Firms, Women Owned Firms and High-Tech Firms, for the SBA Office of Advocacy. p. 14.
 This is data from a 1998 study so the percentages may have changed over the years, but personal guarantees probably remain important, especially for sole proprietors and microbusiness. John Moon (Winter 2009/2010). Small Business Finance and Personal Assets. Federal Reserve Board of Governors, Community Investments, Volume 21, Issue 3. http://www.frbsf.org/community-development/files/moon_john.pdf.
 CFPB. Outline. p. 28.
 Minority Business Development Agency (January 2010). Disparities in Capital Access between Minority and Non-Minority Owned Businesses: The Troubling Reality of Capital Limitations Faced by MBEs. p. 5.
 CFPB. Outline. p. 28.
 Elizabeth Laderman and Carolina Reid (October 2010). The Community Reinvestment Act and Small Business Lending in Low- and Moderate-Income Neighborhoods during the Financial Crisis. Federal Reserve Bank of San Francisco. pp. 9-10. http://www.frbsf.org/community-development/files/wp2010-05.pdf.
 Cavaluzzo, K.S., L.C. Cavaluzzo and J.D. Wolken (January 2002). Competition, Small Business Financing, and Discrimination: Evidence From a New Survey. Journal of Business, vol. 75, no.4. Alicia Robb, Marin Consulting, LLC (April 2013). Access to Capital among Young Firms, Minority Owned Firms, Women Firms, and High Tech Firms. SBA Office of Advocacy.
 Mels de Zeeuw, Federal Reserve Bank of Atlanta Community and Economic Development Department, and Brett Barkley, Federal Reserve Bank of Cleveland Supervision and Regulation Department (November 2019). Mind the Gap: Minority-Owned Small Businesses’ Financing Experiences in 2018. Consumer and Community Context, a Federal Reserve System publication, Vol. 1, No. 2, p. 16. https://www.federalreserve.gov/publications/2019-november-consumer-community-context.htm
 Josh Silver, Archana Pradhan (NCRC), Spencer Cowan (Woodstock Institute) (July 2013). Access to Capital and Credit in Appalachia and the Impact of the Financial Crisis and Recession on Commercial Lending and Finance in the Region. Appalachian Regional Commission (ARC). pp. 102-103. https://www.ncrc.org/access-to-capital-and-credit-in-appalachia-and-the-impact-of-the-financial-crisis-and-recession-on-commercial-lending-and-finance-in-the-region/
 CFPB. Outline. p. 30.
 Q35) Would FIs prefer reporting denial reasons to help explain the decision on an application? If so, should those reasons be voluntary or mandatory fields?
 Alicia Robb and Joseph Farhat (June 2013). An Overview of the Kauffman Firm Survey: Results from the 2011 Business Activities.
 See Q38) Does your FI currently geocode addresses for a reporting requirement, such as HMDA, and what geocoder do you use? Would that geocoder be viable for purposes of 1071 data reporting? What are the costs to geocode addresses? and Q39) How often and in what circumstances does your FI know the address where the borrower’s loan proceeds will be used?
 CFPB. Outline. p. 31.
 Timothy Bates and Alicia Robb (May 2014). Has the Community Reinvestment Act Increased Loan Availability Among Small Businesses Operating in Minority Neighborhoods, Urban Studies Journal.
 Laderman and Reid. Federal Reserve Bank of San Francisco, op. cit., p. 6.
 Andre Perry, Jonathan Rothwell, and David Harshbarger (February 2020). Five-star reviews, one-star profits: The devaluation of businesses in Black communities. Brookings Metropolitan Policy Program and Gallup. p. 2. https://www.brookings.edu/research/five-star-reviews-one-star-profits-the-devaluation-of-businesses-in-black-communities/
 NCRC calculations using 2012 data reported on the FFIEC webpage; table 2-1 CRA National Aggregate for small business data and Table 7-2, Conventional Home Purchase Loans.
 Q40) Does your FI collect gross annual revenue from applicants? If so, for which types of lending products? Are there any products for which your FI does not collect gross annual revenue? Does your FI verify the gross annual revenue provided by applicants? Are there any situations in which you do not verify the gross annual revenue provided by applicants?
 CFPB,. Outline. p. 31.
 Josh Silver, Archana Pradhan (NCRC), Spencer Cowan (Woodstock Institute), report funded by ARC, op. cit., p. 86.
 Association of Enterprise Opportunity. Bigger than You Think – The Economic Impact of Microbusiness in the United States.
 See 2017 SUSB Annual Data Tables by Establishment Industry, United States Census, March 2020, https://www.census.gov/data/tables/2017/econ/susb/2017-susb-annual.html and United States Census Bureau (n.d.). Annual Business Survey Methodology. The United States Census Bureau. Retrieved November 30, 2020, from https://www.census.gov/programs-surveys/abs/technical-documentation/methodology.html
 JP Morgan Chase & Co. Institute (July 2020) Small Business Owner Race, Liquidity, and Survival. p. 17. https://www.jpmorganchase.com/institute/research/small-business/report-small-business-owner-race-liquidity-survival
 American Express. The 2018 State of Women-Owned Businesses Report. p. 9.
 CFPB (December 2018). Executive Summary of the HMDA Data Disclosure Policy Guidance. p. 4. https://files.consumerfinance.gov/f/documents/HMDA_Data_Disclosure_Policy_Guidance.Executive_Summary.FINAL.12212018.pdf
 Community Reinvestment Act: Interagency Questions and Answers, March 2010, p. 11670, Sections _.42(a)(4)-2, 42(a)(4)-3, 42(a)(4)-4 discuss revenue of small business and that lending institutions are not required to ask for revenue size. Lending institutions report “revenues not known” on loans for which they do not collect revenue information. However, the FFIEC does not include a code in the publicly available data that indicates when revenue is unknown. See http://www.ffiec.gov/cra/pdf/2010-4903.pdf.
 2011 CRA exam of East Boston Savings Bank, http://www2.fdic.gov/crapes/2011/33510_111117.PDF.
 Q46) What are the potential challenges, costs, and benefits of collecting and reporting the race, sex, and ethnicity of principal owners using aggregate categories? Although the Bureau is not considering proposing that FIs use disaggregated race and ethnicity categories when collecting and reporting the race and ethnicity of principal owners, what would be the potential challenges, costs, and benefits of such a requirement?
 12 U.S.C. 2803(b)(4); 12 CFR 1003.4(a)(10). Revised §?1003.4(a)(10)(i); revised comment 4(a)(10)(i)-1; revised appendix B to part 1003.
 Agnani, S., & Richardson, J. (2020). Mortgage Lending in the Asian American and Pacific Islander Community. National Community Reinvestment Association. https://ncrc.org/mortgage-lending-in-the-asian-american-and-pacific-islander-community/ and Richardson, J., & So, A. (2020). Hispanic Mortgage Lending 2019 Analysis. National Community Reinvestment Association and UnidosUS. https://ncrc.org/hispanic-mortgage-lending-2019-analysis/
 Appendix B to Part 1003 – Form and Instructions for Data Collection on Ethnicity, Race and Sex, see point 10, https://www.consumerfinance.gov/rules-policy/regulations/1003/B/#9-v
 See Q48) Please provide feedback and information on the approach the Bureau is considering for each discretionary data point, along with any alternative approaches the Bureau should consider and Q49) What would the potential challenges and costs be for collecting, checking, and reporting each discretionary data point?
 2002 Amendments to Regulation C, effective January 1, 2004 required lenders to report pricing information for loan originations in which the annual percentage rate (APR) exceeds the yield for comparable Treasury securities by a specified amount or threshold — the thresholds are a spread of 3 percentage points for first-lien loans and 5 percentage points for subordinate-lien loans.
 12 CFR § 1003.4
 See Richardson, J., & So, A. (2020). Hispanic Mortgage Lending 2019 Analysis. National Community Reinvestment Association and UnidosUS. https://ncrc.org/hispanic-mortgage-lending-2019-analysis/, Agnani, S., & Richardson, J. (2020). Mortgage Lending in the Asian American and Pacific Islander Community. National Community Reinvestment Associatio n. https://ncrc.org/mortgage-lending-in-the-asian-american-and-pacific-islander-community/ and Richardson, J., & Kali, K. S. (n.d.). Same-Sex Couples and Mortgage Lending. NCRC. Retrieved December 10, 2020, from https://www.ncrc.org/same-sex-couples-and-mortgage-lending/.
 Q51) What are the potential costs and benefits associated with collecting and reporting pricing using each of these metrics (i.e., APR, TCC, interest rate and total fees)? Could the costs and benefits vary depending on the type of small business credit product about which pricing is being reported? Is there another metric that would be preferable in order to lower reporting burden?
 12 CFR § 1026.3
 See 12 CFR § 1003.4(a)(12)(i)
 The APR is the period rate, annualized. The APR of a factoring transaction, calculated according to TILA/ Reg Z Appendix J, using the “general equation” as simplified for “single advance, single payment transactions,” just as described in NY S5470 § 806(c).
 Q53) Does your FI currently collect information about the time in business of small business credit applicants? In what format (years / months / years and months / date established) does your FI request that applicants provide the information? Does your FI obtain or verify this information from a third party such as a business credit bureau? Does your FI separate small businesses by time in business for determining risk in underwriting or eligibility? If so, what time parameters are used? Would including a time in business data point help avoid misinterpretation of the 1071 dataset, when a denied application might be explained by relative lack of experience in the business?
 Lauren Williams and Kasey Wiedrich, In Search of Solid Ground: Understanding the Financial Vulnerabilities of Microbusiness Owners, CFED, April 2014, p. 13, https://prosperitynow.org/resources/search-solid-ground-understanding-financial-vulnerabilities-microbusiness-owners-full
 Federal Reserve Bank of New York (August 2017). Small Business Credit Survey: Report on Startup Firms. p. iv.
 Q55) Does your FI currently collect number of employees from any small business applicants? Does your FI take any steps to verify this information? What do you anticipate the potential costs and burdens would be if your FI was required to collect number of employees from small business applicants?
 Q56) Please provide feedback and information on the approach the Bureau is considering with respect to the timing for collection of data points provided by applicants, along with any alternative approaches the Bureau should consider.
 CFPB. Outline. p. 36.
 Small Business Credit Survey: Report on Employer Firms. (2019). Federal Reserve System. https://www.smefinanceforum.org/sites/default/files/blogs/SBCS-Employer-Firms-Report.pdf
 Washington Post (September 2017). Equifax Breach Hits Credit Data of Millions. https://www.washingtonpost.com/news/the-switch/wp/2017/09/08/after-data-breach-equifax-asks-consumers-for-social-security-numbers-to-see-if-theyve-been-affected/?hpid=hp_hp-more-top-stories_equifax-1145am%3Ahomepage%2Fstory&utm_term=.bf6cc31f537d
 Silver, J. (2014). Small Business Loan Data: Recommendations to the Consumer Financial Protection Bureau for Implementing Section 1071 of the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010. National Community Reinvestment Coalition. https://ncrc.org/wp-content/uploads/2014/08/recommendations-to-cfpb-on-small-business-loan-data.pdf
 Federal Financial Institutions Examination Council (August 2016). Reports – Findings from Analysis of Nationwide Summary Statistics for 2015 Community Reinvestment Act Data Fact Sheet. https://www.ffiec.gov/hmcrpr/cra_fs16.htm
 The CFPB reports that in 2015, the HMDA database contained 14.4 million records (see Highlights of the 2015 HMDA data via https://www.consumerfinance.gov/data-research/hmda/learn-more). In contrast, the small business database contained about 6 million records (see national aggregates via https://www.ffiec.gov/craadweb/national.aspx). If the number of records can be increased by one third or one half in the small business database, then privacy concerns are ameliorated, enabling robust census tract level disclosure.
 Privacy in Statistical Databases. https://link.springer.com/book/10.1007%2Fb97945
 Please provide feedback and information on the approach the Bureau is considering regarding public disclosure of 1071 data by the Bureau on behalf of FIs, along with any alternative approaches the Bureau should consider.
 Silver, J. (2019, April 29). The CFPB Needs to Keep Easy Access to HMDA Data. NCRC. https://ncrc.org/the-cfpb-needs-to-keep-easy-access-to-hmda-data/
 For example, Data Point: 2018 Mortgage Market Activity and Trends A First Look at the 2018 HMDA Data. (August 2019). Consumer Financial Protection Bureau. https://files.consumerfinance.gov/f/documents/cfpb_2018-mortgage-market-activity-trends_report.pdf and accompanying tables.
 Q76) Please provide feedback and information on the approach the Bureau is considering regarding an implementation period, along with any alternative approaches the Bureau should consider.
 Q78) The Bureau’s overall methodological approach to measuring one-time and ongoing costs of the eventual 1071 rule, along with any alternative approaches the Bureau should consider.
 CFPB. Outline. p. 48.
 CFPB. Outline. p. 57.
 CFPB. Outline. p. 64.
 Q59) Please provide feedback and information on the approach the Bureau is considering regarding the firewall under section 1071(d)(1), along with any alternative approaches the Bureau should consider.
 15 U.S.C § 1691c–2(d)(2)