Community Lending Fact Data Edition book cover

We created the Fact Book and Data Portal to help a range of users monitor changes to key financial indicators in neighborhood markets. We present data from the community up to the regional level on mortgage lending, community demographics, foreclosure trends, changes in the income levels of local homebuyers, business lending, and the activity of top lenders.

User Guide

The guide is organized by geographic area. Each geographic area presents the same five data sections:

1. Neighborhood Overview

Each community area and suburban region profile begins with a summary of basic Census Bureau data on housing and demographic characteristics. This includes an estimate of the household income ranges, mortgage status, units in structure, and data on the number and percentage of owner-occupied units, renter-occupied units, vacant units, and total housing units.

Also included are data on the racial and ethnic composition of the population and homeownership rates by race and ethnicity. These data are from the 2019 Five-Year American Community Survey (ACS 5-year). Note that data on ethnicity (Hispanic/Latino origin) are reported separately from race in Census data. Therefore, household share figures in this section may not add up to 100 percent.

This section also includes data on the total number of housing units, from the ACS five-year, and bank branches in the area. Bank branch information comes from the Federal Deposit Insurance Corporation (FDIC). Figures include only physical branches and exclude non-deposit-taking locations such as corporate headquarters.

2. Top Lenders (by number of loan originations)

Data for the top 20 lenders is further broken down by the amount of originations, number of purchase or refinance originations, and number of originations to minority borrowers or low- and moderate-income borrowers. Note that many areas have more lenders originating loans than are listed on the page. The Fact Book lists only those lenders originating the largest number of loans. The source is HMDA data.

3. Mortgage Lending Data

The housing lending data are listed for two lender types:Banks, Thrifts, and Direct Operating Subsidies Subject to CRA Requirements

  • Non-Bank Mortgage Companies and other Lenders Not Subject to CRA Requirements
  • The first rows show applications, originations and denials by race, ethnicity and low- and moderate-income borrowers.

The balance of the table presents figures on the number and aggregated value of all loan originations. These include the number of conventional loans, loan purpose, type of property, and loan characteristics, such as number of mortgages with a high loan-to-value or debt-to-income ratio, or the number of originations with high closing costs.

The number of applications that result in either loan originations or denials might not equal the total number of applications. An applicant may withdraw an application, an application may be incomplete, or a loan may be offered but not taken by the applicant. This section has been revised to incorporate several new variables that were not previously available in the source, HMDA data.

4. Small Business Lending

The Fact Book lists the total number of loans made under $100,000 and under $1 million to businesses in each area, plus the total number and amount of loans made to businesses with less than $1 million in gross annual revenue. Small business loans are presented for each geographic area as a whole as well as for all low- or moderate-income census tracts in that geographic area.

Data on the number of business addresses come from U.S. Postal Service data aggregated by the Department of Housing and Urban Development. Data on the small loans (loans less than or equal to $1 million) to businesses come from reports that large financial institutions must submit (and some voluntarily submit) to the Federal Financial Institutions Examination Council (FFIEC) under the CRA. For the purposes of reporting, banks generally aggregate any extension of credit, including traditional loans, lines of credit, and credit cards.

5. Foreclosures

The final chart shows the number of foreclosure filings in each year.


Data Quality

HMDA Data

The mortgage lending data contained in the Fact Book come from the 2019 HMDA data obtained from the CFPB. HMDA data comes from reports that mortgage lenders submit to federal regulators.

On March 1 of each year, lenders are required to submit the Loan Application Register (LAR) to their respective regulatory agency. The LAR contains data on all the mortgage applications the lender received in the previous calendar year. Although regulators are charged with ensuring a high level of data quality, inaccuracies may exist due to faulty reporting, inaccurate compilation, or insufficient data cleaning.

Problems in HMDA LAR files might include invalid census tract numbers or incorrect loan or income amounts. For the Fact Book, applications with missing census tract information were not used in tabulations based on location (i.e., community area, city, or county subregion pages).

Beginning in 2017, the CFPB made changes to how it made HMDA data available to the public. In previous years, the FFIEC made static HMDA loan-level data available to the public in the fall of each year. For data reported since 2017, the CFPB updated the data it made available to the public on an ongoing basis to reflect late submissions and resubmissions. The Fact Book uses the CFPB dataset for the 2019 HMDA data, updated to August 2020.

Changes to reporting requirements implemented in 2017 mean that post-2017 HMDA data cover a smaller proportion of lenders than in previous years. Changes to HMDA’s Regulation C exempted depository institutions from reporting data if they originated fewer than 25 loans in each of the two preceding calendar years. In 2017, the number of reporting institutions declined by 13% nationally and in 2018 declined even further by 2.9% due in part to this reporting threshold change. For 2019, the number reporting was down another 3.0%.

The CFPB made significant changes to some variables within HMDA datasets for the 2018 data going forward. In comparing data from 2017 to the data for 2018 and in this Fact Book, several of the new variable options were consolidated to approximate the 2017 categories more closely.

For example the property type variable went from having only three values in 2017 – “single-family,” “manufactured housing,” and “multifamily” – to having the four values of “single family,” “multifamily,” “single-family manufactured,” and “multifamily manufactured.”

Similarly, where 2017 HMDA data only had loan purpose values of “home purchase,” “home improvement,” and “refinancing,” the new HMDA data had additional values “cash-out refinancing” and “other purpose”. In interpreting the 2019 HMDA data for this Fact Book, the 2019 value “single-family manufactured” was considered parallel to the 2017 value “manufactured housing,” and both 2019 “multifamily” and “multifamily manufactured” values were combined into the category of “multifamily.” Also, in this Fact Book, the loan purpose value of “cash-out refinancing” was consolidated into the overall category of “refinancing” and the value “other purpose” was excluded from the analysis.

The CFPB also changed the variables for race, ethnicity, and loan amount in 2018, and those changes still apply to the current HMDA data. The variable for race now includes codes for subgroups within the race categories for the 2017 and earlier HMDA data. To preserve comparability with earlier Fact Book data, the subgroups were re-aggregated under the broader group racial designation. The same change applied to the ethnicity variable, with more detailed data on subgroups.

As with race, the subgroups were re-aggregated to the broader category to preserve comparability with earlier data. The reported data for the applicant is used to categorize the race or ethnicity of the borrower for purposes of the Fact Book analysis. In previous HMDA datasets, loan amounts were reported within $1,000 dollar ranges. As of the 2018 HMDA data, the reported loan amount data are grouped in $5,000 dollar ranges. The lowest range value is $5,000 and all reported loan amounts below this value were automatically rounded up by the CFPB to $5,000.

Unreported race, ethnicity, and income data remain a significant concern when working with HMDA data. Nearly 13% of HMDA data entries for the seven-county Chicago region lacked data on the race of the applicant. These missing data are attributed to non-reporting by applicants on mail, telephone, and Internet applications.

In analyzing data found in the Fact Book, it is important to consider the differing sizes of lending institutions. The size of an institution will affect the type and/or amount of lending it does and the relationship it has to a particular community.

It should also be noted that top lender data in the Fact Book is only for mortgage lending. Some institutions specialize in commercial lending and may be actively providing business loans in a community. (The Fact Book includes data on the total number of business loans, but not data for individual business lending institutions).

Although the data reported in the Fact Book represent the majority of mortgage lending activity of for-profit lenders, the Fact Book does not include mortgage lending by non-profit lenders, which are not required to report data under HMDA. There are numerous community development financial institutions (CDFIs) and other non-bank lenders who receive some funding from financial institutions and other sources to make housing loans for community development purposes. These loans are often riskier than those that banks would normally make and typically reach segments of the housing or business lending market that may have difficulty accessing credit from mainstream financial institutions.

CRA Small Business Lending Data

CRA data on small business loans are for loans under $1,000,000 to any size business and cover some but not all of lending activity to businesses. In 2019, lending institutions were required to report small business lending data only if they had assets totaling over $1.284 billion as of December 31 each of the two prior calendar years. Some smaller institutions under this threshold chose to report data voluntarily. Analysis of data from Bank Call and Thrift Financial Reports indicates that CRA reporters accounted for 89.4% of all small business loans and 74.4% of the dollar value of all small business loans made by depository institutions. CRA data do not include loans made by institutions not covered by the CRA, such as commercial finance companies.

CRA small business data are less comprehensive than HMDA data. Unlike HMDA data which include all applications, CRA small business data include only census tract-level totals of the number and amount of originated loans. Apart from the number of loans made to businesses with less than $1 million in gross annual revenue, the data do not contain any additional demographic information about the business or the applicant. Banks are not required to collect revenue data on businesses to which they make loans. Therefore, it is possible the data underreport the number of loans made to businesses with less than $1 million in revenue.

Loan amounts are reported in categories:

  • less than or equal to $100,000;
  • greater than $100,000 and less than or equal to $250,000; and
  • greater than $250,000 and less than or equal to $1 million.

Banks generally aggregate any extension of credit—credit cards, lines of credit, and term loans— in reporting. Lenders report the census tract of the loan, but interpreting the distribution of loans can be complicated. For example, if the proceeds of a small business loan are used in more than one location, the lending institution can record the loan location as either the address of the borrower’s business headquarters or the location where the greatest proportion of the proceeds are applied.


Changes in the 2019 Data

Woodstock Institute made several changes to Fact Book data this year in response to user feedback and to incorporate some of the new mortgage lending variables now available in the HMDA dataset. The goal is to provide users with more detailed information and to lay the foundation for future editions of the Fact Book.

Because of the changes, Woodstock urges users to exercise caution in comparing some of the data in this edition with the data from earlier editions. These notes highlight the changes and where the data may differ substantially from earlier Fact Books.

Housing and Demographic Data

Population, Homeownership Rate, Occupancy, Income, Mortgage Status, and Units in Structure data are from the 2019 ACS 5-year estimates. The new data provide more detail about neighborhood characteristics, including a more detailed breakdown of income and the nature of the housing stock, than the older version of the Fact Book.

While most of the data are directly comparable to the older version of the Fact Book, the data on the total number of housing units in the current version is a change from the older version, which had an estimate of the number of Mortgageable Properties, defined as:

Units in one-unit attached and detached structures and units in mobile homes are counted as 1 mortgageable property each; units in 2-unit structures are divided by 2 to get the number of mortgageable properties; and units in 3- and 4-unit structures are divided by 3.5 to get the number of mortgageable properties. The number of condominiums is estimated by taking the number of owner-occupied units in structures with 5 or more units.

Woodstock made the change because it decided that the estimate of Mortgageable Properties was a complex construct which did not substantially improve the description of the potential number of properties that could be eligible for a mortgage. For example, the estimate for units in two-unit structures was the total number of units in those structures divided by two.

That method of estimating assumes that one household owns both units and excludes half of the units in two-unit structures that have been converted to condominiums, a common practice in some parts of the region. The estimate also excludes all renter-occupied units in larger buildings which are condominiums. Those units are potentially mortgageable in sales to other investors or new owner-occupants. The change to total housing units avoids those issues.

Bank branch information comes from the FDIC. Data include only physical branches and exclude non-deposit taking facilities, such as corporate headquarters. Data are as of December 16, 2019.

Mortgage Lending Data

The data in this section contain most, but not all, of the data included in earlier editions of the Fact Book, plus many new variables now available in the HMDA data. Three of the new variables are presented in an X of Y format. This format is used because of the reporting rules for those data. The CFPB, which governs HMDA data reporting requirements, added loan-to-value, debt-to-income, and total loan cost variables, among many others, to the data it required lenders to report.

At the same time, the CFPB exempted smaller institutions from the new reporting requirements. What this means is that all institutions report some of the data, including the older variables required before the CFPB expanded the data to be reported, but only larger lenders report the new variables. The X of Y format allows the user to know how many loan records included the new variables.

The borrower income data in the HMDA dataset include some negative values, which would suggest that the borrower had less than $0 income. One possible explanation for this is simply that it is a data entry error. To avoid including negative income values, all income values below $0 are omitted from the calculations.

Small Loans to Businesses

The data on small loans to businesses include the data reported in earlier editions of the Fact Book, but now the data are also displayed for loans to active business addresses in low- or moderate-income census tracts.

Top Lenders (by number of loan originations)

The section on top lenders has been extensively revised. In earlier editions, the data were displayed separately for the top 30 home purchase and refinance lenders, and the top eight home improvement and multifamily lenders. The data included only the number of loans originated by each lender. For this year, the top lenders were determined by the total number of originations and, if the number for two lenders was the same, by the total amount of loans originated. In addition, only the top 20 lenders are displayed. For those lenders, the data now include the number of purchase and refinance originations, as well as the number of originations to minority or low- and moderate-income borrowers.

Trend – Past Five Years

Data on the number and amount of loan originations are taken from the 2015, 2016, 2017, 2018, and 2019 HMDA data. Conventional loans are defined as any loan other than Federal Housing Administration (FHA), Veterans Administration (VA), Farm Service Agency (FSA), or Rural Housing Service (RHS) loans. Note that FHA/VA filings may include a small number of FSA or RHS loans, depending on the geography. The data in this section are for one year less than in the earlier editions of the Fact Book, but they now include data on the amount of the loans as well as the number.

The owner-occupied home purchase lending by borrower income level data are based on the income levels for median family income in the HMDA data. Those data vary, depending on the Metropolitan Region in which the loan was made. Cook, DuPage, Kendall, McHenry, and Will Counties are in the Chicago-Arlington Heights, IL Metropolitan Division. Lake County is in the Lake County-Kenosha County, IL-WI Metropolitan Division. Kane County is in the Elgin, IL Metropolitan Division. Median family income figures are different for each metropolitan division. Therefore, a borrower income that falls into an income level category in one metropolitan division may not fall into the same income level category in another division.

Foreclosure data are from reports acquired from Record Information Services.


Future Changes & Your Feedback

While Woodstock hopes that the data will remain consistent from year-to-year in the future, we recognize changes may be warranted. For example, the range of data available may change, and we may take advantage of any additions to the data to add new information. As always, we continue to ask our users for feedback on the Fact Book that we incorporate into the design of the Fact Book in print and online and our technical assistance.

If you have questions, comments or suggestions related to the Fact Book, or would like technical assistance assessing or enhancing access to credit for your community, please contact Director of Research Spencer Cowan by email at or by phone at 312-368-0310.