1
CONSUMER FINANCIAL PROTECTION BUREAU | SEPTEMBER 2022
Data Point:
2021
Mortgage Market Activity
and Trends
2 DATA POINT: 2021 MORTGAGE MARKET ACTIVITY AND TREND
This is another in an occasional series of publications from the Consumer Financial Protection
Bureau’s Office of Research. These publications are intended to further the CFPB’s objective of
providing an evidence-based perspective on consumer financial markets, consumer behavior,
and regulations to inform the public discourse. See 12 U.S.C. §5493(d).
[1]
[1]
This report was prepared by Feng Liu, Young Jo, and Eileen Chen.
3
Table of contents
Table of contents ..............................................................................................................3
1. Introduction ...................................................................................................................4
2. Mortgage applications and originations ...................................................................7
3. Mortgage outcomes by demographic groups and loan types ............................16
3.1 Distribution of home loans ......................................................................... 16
3.2 Mortgage characteristics of home loans .....................................................21
3.3 Denial rates ................................................................................................. 29
4. Monthly mortgage trends and activities .................................................................33
5. Mortgage trends and activities by states ...............................................................50
6. Lending institutions ...................................................................................................53
7. Conclusion...................................................................................................................60
4
1. Introduction
This Data Point article provides an overview of residential mortgage lending in 2021 based on
the data collected under the Home Mortgage Disclosure Act (HMDA). HMDA is a data
collection, reporting, and disclosure statute enacted in 1975. HMDA data are used to assist in
determining whether financial institutions are serving the housing credit needs of their local
communities; facilitate public entities’ distribution of funds to local communities to attract
private investment; and help identify possible discriminatory lending patterns and enforce
antidiscrimination statutes.
1
Institutions covered by HMDA are required to collect and report
specified information about each mortgage application acted upon and mortgage purchased.
The data include the disposition of each application for mortgage credit; the type, purpose, and
characteristics of each home mortgage application or purchased loan; the census-tract
designations of the properties; loan pricing information; demographic and other information
about loan applicants, such as their race, ethnicity, sex, age, and income; and information about
loan sales.
2
The 2021 HMDA data
3
are the fourth year of data that incorporate amendments made to HMDA
by the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010 (DFA).
4
Because
of these significant changes starting with the 2018 HMDA data, the HMDA data since 2018 are
not completely comparable to the HMDA data before 2018. To maintain the consistency for
cross-year comparisons, this article uses only the HMDA data from 2018 to 2021 and focuses on
1
For a brief history of HMDA, see Federal Financial Institutions Examination Council, “History of
HMDA,” available at www.ffiec.gov/hmda/history2.htm
(last modified Sept. 06, 2018) .
2
See Filing instructions guide for HMDA data collected in 2021 (November 2020), available at
https://s3.amazonaws.com/cfpb-hmda-public/prod/help/2021-hmda-fig.pdf
for a full list of items
reported under HMDA for 2021.
3
The 2021 HMDA data, which are used for the analysis of this Data Point, cover mortgage applications
acted upon and mortgages purchased during the calendar year of 2021 and reported in 2022. Similarly,
the 2018, 2019 and 2020 HMDA data refer to applications acted upon and mortgages purchased during
the calendar years of 2018, 2019 and 2020 respectively.
4
Among the changes made by the DFA were new data points, revisions to certain existing data points, and
authorizing the CFPB to require the collection and reporting of additional data points. The CFPB issued a
final rule implementing these and other changes in October 2015 (2015 HMDA Rule). The 2015 HMDA
rule made five primary changes to the data collected starting on January 1, 2018: (1) mandated reporting
of open-end lines of credit (LOCs) by financial institutions with an annual LOC origination volume
exceeding a coverage threshold of 100 LOCs in each of the two preceding years ; (2) changed the
transactional coverage definition from loan-purpose-based to one based primarily on whether the loan
was secured by a dwelling; (3) modified definitions and values of some existing data points; (4) required
reporting of 27 new data points ; and (5) established a uniform coverage threshold of 25 closed-end loan
originations in each of the two preceding years for depository institutions (DIs) and non-depository
institutions (non-DIs), with the closed-end threshold change for DIs taking e ffect on January 1, 2017 and
the threshold change for non-DIs taking effect in 2018. This 25 loans coverage threshold was increased to
100 loans in May of 2020 by the 2020 HMDA rule, which became effective on July 1, 2020.
5
trends in mortgage applications and originations during these four years. In doing so, we have
incorporated a number of new and revised data points that were not collected prior to the 2018
HMDA data. We have also incorporated analyses of lines of credit (LOCs) and other dwelling-
secured transactions in the article. Readers who are interested in the historical trends of
mortgage applications and activities prior to 2018 can refer to the CFPB’s previous HMDA Data
Point articles.
5
Lastly, some estimates in this article are not completely consistent with those
from the previous HMDA Data Point articles because some records were excluded in previous
publications to make the post-2018 data consistent with the pre-2018 data.
6
On June 16, 2022, the CFPB published a static application-level 2021 HMDA data file that
consolidates data from individual reporters. The data file is modified to protect applicant and
borrower privacy.
7
The data file and the Data Point article reflect the data as of April 30, 2022.
Though this static file will not change, the CFPB will also provide an updated file separately to
reflect any later resubmissions or late submissions. The results using the updated file may differ
from those reported in this Data Point article, although the CFPB expects them to be largely
consistent.
The remainder of this article summarizes the 2021 HMDA data and recent trends in mortgage
applications and originations. Some of the key findings are:
8
4,332 financial institutions reported at least one closed-end record in 2021, down by 3.1
percent from 4,472 financial institutions who reported in 2020.
In total, the number of closed-end originations (excluding reverse mortgages) in 2021
increased by 2.4 percent, from 13.4 million in 2020 to 13.7 million. This is in contrast
from 2019 to 2020, where the number of originations increased by 66.8 percent and
were largely driven by the refinance boom. In contrast, most of the increase from 2020
5
See “Data Point: 2018 Mortgage Market Activity and Trends,” available at
https://www.consumerfinance.gov/data-research/research-reports/data-p oint-2018-mortgage-market-
activity-and-trends/.
6
In the first two HMDA Data Point articles published after the 2018 HMDA rule change, the CFPB
excluded all open-end LOCs, except those that are reverse mortgages, and applications for a loan purpose
other than home purchase, home improvement, or refinance. These two publications include “Data Point:
2019 Mortgage Market Activity and Trends,” available at
https://www.consumerfinance.gov/data-
research/research-reports/data-point-2019-mortgage-market-activity-and-trends/ and “Data Point: 2018
Mortgage Market Activity and Trends,” available at https://www.consumerfinance.gov/data-
research/research-reports/data-point-2018-mortgage-market-activity-and-trends/.
7
For more information concerning these modifications and the CFPB’s analyses under the balancing test
it adopted to protect applicant and borrower privacy while also fulfilling HMDA’s disclosure purposes, see
Disclosure of Loan-Level HMDA Data 84 FR 649 (Jan. 31, 2019).
8
This Data Point art icle is based on the analysis of the static consolidated application-level 2018, 2019,
2020, and 2021 HMDA data files. Some data points used in this article were modified or withheld in the
pu b lic HMDA data.
6
to 2021 was driven by an increase in the number of home purchase loans. The increase in
home purchase loans was the most prominent among jumbo loans, likely reflecting
rapidly rising house prices in 2021.
The refinance boom, especially in non-cash-out refinance that dominated the mortgage
market activities in 2019 and 2020, peaked in March 2021. The non-cash-out refinance
volume decreased precipitously throughout the remaining months of 2021. The decrease
coincided with the increase in market interest rates. On the other hand, the cash-out
refinance volume continued to increase in 2021 relative to 2020, likely because
consumers took advantage of rapid house price appreciation and converted rising home
equity into cash or liquid assets.
Continuing trends from 2020, in 2021, non-Hispanic White borrowers’ share of home
purchase loans decreased, while Asian, Black, and Hispanic White borrowers’ share
increased. The share of refinance loans declined the fastest for Asian borrowers in 2021,
just as it increased the fastest in 2019 and 2020. Black and Hispanic White borrowers’
share of refinance loans increased in 2021, accounting for 6.6 percent and 6.8 percent by
the end of 2021.
Black and Hispanic White borrowers continued to have lower median loan amounts,
lower median credit scores, and higher denial rates, in addition to paying higher median
interest rates and total loan costs compared to non-Hispanic White and Asian
borrowers.
9
The top 25 closed-end lenders held a combined market share of 43.9 percent, rising
yearly since 2018. The top 25 lenders were particularly prominent in the refinance
market, accounting for 53.0 percent of refinance loans by all reporting institutions.
9
A recent CFPB publication shows that there is heterogeneity in mortgage characteristics across Asian
American Pacific Islander subgroups. The report titled “Asian American and Pacific Islanders in the
Mortgage Market,” is available at
https://www.consumerfinance.gov/data-research/research-
reports/asian-american-and-pacific-islanders-in-the-mortgage-market/
7
2. Mortgage applications and
originations
In 2021, a total of 4,338 financial institutions—banks, savings associations, credit unions, and
non-depository mortgage lenders—reported data on approximately 23.3 million applications
and 15.0 million originations under HMDA. In contrast, in 2020, 4,475 financial institutions
reported data on 22.7 million applications and 14.5 million originations under HMDA.
Compared to 2020, the number of reporters decreased by 137, or about 3.1 percent. On the other
hand, the total number of applications and originations increased slightly in 2021. The total
number of reported applications increased by about 646,000 or 2.8 percent and the number of
originations increased by 465,000 or 3.2 percent. This slight increase in originations and
applications represents a departure from 2020, when applications increased by 50 percent and
originations increased by 56 percent in one year.
The bottom rows of Table 1 present the total number of records, including total applications,
originations, purchased loans, and requests for approvals reported each year from 2018 to 2021.
The top panels of Table 1 break down the total records by types of transactions: closed-end
excluding reverse mortgages, open-end lines of credit excluding reverse mortgages (HELOCs
10
),
or reverse mortgages. Within closed-end transactions, we further divide by property types: a
site-built one-to-four family unit, manufactured home, or multifamily transactions. We also
categorize by loan purposes: home purchase, home improvement, refinance, and other purpose.
The closed-end site-built one-to-four family originations are then disaggregated by lien status
(e.g., first lien, junior lien) and occupancy types (e.g., principal residence, second residence,
investment property). Then, the first-lien, principal-residence originations are further
disaggregated by whether they are conventional loans or not. Within the conventional loan
category, we disaggregate by whether the loan is conforming or jumbo. Within the non-
conventional loan category, we disaggregate by loans insured or guaranteed by Federal Housing
Administration (FHA), Department of Veterans Affairs (VA), or USDA Rural Housing Service or
Farm Service Agency (RHS/FSA).
For manufactured home originations, we disaggregate by
10
Open-end lines of credit secured by dwellings (excluding reverse mortgages) are commonly known as
home equity lines of credit, or HELOCs. In the rest of the article, where it is applicable, we have used the
term HELOC in lieu of open-ended lines of credit excluding reverse mortgage. Beginning with the data
collected in 2018, the reporting of HELOCs became mandatory rather than optional.
8
whether manufactured home loans are secured by land (non-chattel loans) or not secured by
land (chattel loans).
11
Lenders reported approximately 13.7 million closed-end site-built single-family originations in
2021, a 2.4 percent increase from 13.4 million originations in 2020.
12
In contrast, this figure
increased 66.8 percent from 2019 to 2020. In addition, lenders reported around 20.7 million
closed-end site-built single-family applications, which includes 4.4 million applications that the
lenders closed as incomplete or the applicant withdrew before the lender made a decision.
In 2021, about 228,000 loans secured by manufactured homes were originated, compared to
197,000 such loans in 2020. About 148,000 manufactured home loans were taken out for home
purchase, up from 137,000 in 2020. Among them, about 82,000 were secured by both home
and land, while 59,000 were secured by a manufactured home but not land. Increases also
occurred in refinance loans secured by manufactured homes. In 2021, about 75,000
manufactured home loans were taken for refinance purpose, up from 56,000 in 2020.
Unlike the previous three years, the total number of HELOCs reversed its downward trend and
increased from 869,000 in 2020 to 962,000 in 2021, a 10.7 percent increase. This number of
HELOC originations is still lower than the number of HELOC originations in 2018 and 2019.
Similar trends applied to the HELOC applications. The total number of reverse mortgage
originations also increased from 43,000 to 59,000 from 2020 to 2021, an increase of over 36
percent.
HMDA data also include information on loans purchased by reporting institutions during the
reporting year, although the purchased loans may have been originated before 2021. Table 1
shows that financial institutions purchased 2.7 million loans from other institutions in 2021, a
2.8 percent decrease from 2020.
11
Manufactured-home lending differs from lending for site-built homes. Furthermore, even among the
manufactured home loans, chattel-secured lending differs greatly from those that are not chattel secured.
Chattel-secured lending typically carries higher interest rates and shorter terms to maturity. The rest of
this article focuses almost entirely on site-built mortgage originations, which constitute most originations.
For more information on manufactured housing, see “Manufactured Housing Finance: New insights from
the Home Mortgage Disclosure Act,” available at
https://www.consumerfinance.gov/data-
research/research-reports/manufactured-housing-finance-new-insights-hmda/.
12
Throughout the rest of the report, calculations in the text are based on precise data values. Using
rounded numbers from the tables may lead to different values due to rounding errors.
13
In contrast to 2020 when the refinance loans in closed-end site-built single-family originations
increased by 149.1 percent, refinance originations slightly decreased in 2021. Refinance
originations fell from 8.4 million in 2020 to 8.3 million in 2021, which is approximately a 1.7
percent decrease. Refinance applications for site-built single-family properties also decreased
from 13.2 million in 2020 to 13.1 million in 2021.
Refinance originations decreased for all types of loans
13
except for conventional jumbo loans
and FHA loans, which increased by 28.3 percent and 5.9 percent respectively in 2021. This is
notable as these two sectors experienced some of the lowest refinance growth rates in 2020.
Percentage wise, VA loans experienced the largest decrease in refinance activity in 2021,
decreasing from 938,000 originations in 2020 to 774,000 originations in 2021, or a 17.5 percent
decline. The previous year, the refinance volume of VA loans had increased by 126.2 percent.
Given all these changes, generally, refinance origination numbers in 2021 are still notably higher
than 2019 and 2018 volumes, but not as high as the peak in 2020.
The closed-end site-built single-family mortgage originations for home purchase rose by a
similar rate in 2021 as 2020. In 2021, there were 5.1 million home purchase originations, as
compared with nearly 4.7 million in 2020. This represents a 9.2 percent growth rate from 2020,
while in 2020, the growth rate was 9.0 percent. When limited to the first-lien principal
residence, the home purchase loan originations increased by 6.7 percent, from 4.1 million in
2020 to 4.4 million in 2021. Among them, both conventional conforming and jumbo loans
increased while all nonconventional loan categories saw a decrease in volume. The conventional
jumbo market saw the largest increase of 52.8 percent in 2021, while FSA/RHS loans saw the
biggest decrease of 15.7 percent.
The change in market interest rates was likely the main driver behind the declines in refinance
applications and loans. Figure 1 plots the monthly median interest rates for 30-year fixed-rate
closed-end (excluding reverse mortgage) conventional conforming loans originated to the prime
borrowers secured by first-lien principal residence.
14
The median interest rate is computed from
originated loans in the 2018-2021 HMDA data and the month is based on the month of the
action taken date. As the figure shows, the mortgage interest rate declined substantively in 2019
and 2020, reaching historically low levels at the end of 2020. The rate began rising in 2021 but
still remained low by historical standards. Moreover, the rise in interest rates during 2021 was
smaller than the decline during 2020. Figure 1 generally aligns with Freddie Mac’s Primary
13
Mortgages secured by closed-end site-built single-family first-lien principal residence.
14
The prime borrowers are defined as those with a credit score of at least 720 and CLTV of around 80
percent (79 percent <= CLTV < 80 percent).
14
Mortgage Market Survey, which covers first-lien prime conventional conforming home purchase
mortgages with a loan-to-value of 80 percent.
15
FIGURE 1: Monthly Median Interest Rate of 30-Year Fixed Rate Conventional
Conforming Loans
NOTE: Monthly median interest rate of 30-year fixed-r a te , conventional conforming, home pu rch ase , closed-
end loans secured by site-built single-family homes to prime borrowers between January 2018 and December
2021. Limited to borrowers with credit score>=720 and 79%<=LTV <=80%, 1st lien, principal residence.
15
See https://www.freddiemac.com/pmms. According to the Primary Mortgage Market Survey (PMMS),
the average rate on 30-year fixed-rate conventional conforming mortgage loans started at a historically
low 2.65 percent at the beginning of 2021 and increased to 3.11 percent by the end of the year. In contrast,
the interest rates decreased from 3.72 percent at the beginning of 2019 to 2.67 percent by the end of 2020.
The reported interest rates in the HMDA data follow a consistent pattern as that observed in PMMS.
15
On the other hand, fast-rising house prices in 2021
16
likely explain the increase in home
purchase loan origination volume. In particular, the large increase of jumbo loan volume, for
both home purchase and refinance loans, could be largely attributable to rising house prices.
Similarly, as house values appreciated and existing homeowners’ home equity increased, a
higher share (about 35.0 percent) of closed-end site-built single-family refinance loans were
cash-out refinances versus non-cash-out refinances. This is higher than 2020 when cash-out
refinance loans accounted for 25.1 percent of all refinance loans.
16
See S&P CoreLogic Case-Shiller U.S. National Home Price Index for a commonly cited monthly measure
of the value of single-family housing, available at https://fred.stlouisfed.org/graph/?g=RtST. According
to the index, nationwide the house prices (not-seasonally adjusted) increased by 18.9 percent from
December 2020 to December 2021.
16
3. Mortgage outcomes by
demographic groups and loan
types
The HMDA data are a key resource for policymakers and the public to understand the
distribution of mortgage credit across demographic groups. Tables 2 through 4 provide
information on loan shares, product usage, certain mortgage/borrower characteristics, pricing
information, and denial rates by applicant income, neighborhood income, and applicant race
and ethnicity. Tables 2 through 4 focus on closed-end first-lien home purchase and refinance
loans secured by site-built one-to-four-family, principal residence properties, which accounted
for approximately 80.0 percent of all HMDA originations excluding purchased loans in 2021.
3.1 Distribution of home loans
Table 2 presents different groups’ shares of closed-end (excluding reverse mortgage) site-built
one-to-four-family, first lien, principal residence home purchase and refinance loans and how
these shares have changed since 2018. Continuing the trend, Black borrowers’ share of home
purchase loans increased from 6.8 percent in 2018 to 7.9 percent in 2021, whereas the share for
Hispanic White borrowers was at 9.2 percent in 2021, holding at a steady level since 2019. For
non-Hispanic White borrowers, their share of home purchase loans was 55.6 percent in 2021,
down from 62.0 percent in 2018. On the other hand, Asian borrowers’ share of home purchase
loans increased from 5.5 percent in 2020 to 7.1 percent in 2021.
Non-Hispanic White borrowers accounted for about 58.3 percent of all refinance loans in 2021,
down slightly from 61.0 percent in 2020. Asian borrowers still took up a larger share among
refinance loans in 2021 at 6.0 percent compared to their shares in 2018 and 2019, but their
share was down from a high of 6.7 percent in 2020. The shares of Black and Hispanic White
borrowers refinance loans returned to 2019 levels after a low in 2020. The Black borrowers’
share of refinance loans increased from 4.2 percent in 2020 to 5.4 percent in 2021, and the
Hispanic White borrowers’ share of refinance loans increased from 5.3 percent in 2020 to 6.1
percent in 2021.
17
We note that such observations are set against the backdrop in which origination volumes of
refinance loans have slightly deceased from peak volumes after the refinance boom of 2019 and
2020.
The shares of home purchase and refinance loans exhibit opposite trends for low- or moderate-
income (LMI) borrowers compared with high-income borrowers.
17
The LMI borrower share of
home purchase loans decreased from 30.4 percent in 2020 to 28.7 percent in 2021, whereas
high-income borrowers’ share increased from 41.2 percent to 43.2 percent. The LMI borrower
share of refinance loans increased from 18.9 percent in 2020 to 23.9 in 2021, while high-income
borrowers’ share decreased from 45.1 percent to 42.0 percent. For both home purchase and
refinance loans, these trends are the opposite of the ones seen from 2019 to 2020 and shares
generally returned to 2019 levels during 2021.
Like 2020, the trends in shares of LMI and high-income neighborhoods mirror those of the
borrowers for refinance loans but not for home purchase loans.
18
The LMI neighborhoods’ share
of refinance loans increased, whereas high-income neighborhoods’ share of refinance loans
decreased. On the other hand, the share of home purchase loans in LMI neighborhoods
increased slightly, while the share in high-income neighborhoods decreased slightly between
2020 and 2021.
The share of refinance loans for most racial/ethnic groups, LMI borrowers, and LMI
neighborhoods has fluctuated in the past few years, while the number of refinance loans has
increased for all groups between 2018 and 2021.
19
For example, the share of refinance loans for
Hispanic White borrowers decreased from 6.8 percent in 2018 to 6.1 percent in 2021 but the
number increased by about 354,000. The increase in the number of refinance loans over the
past few years is especially large for Asian borrowers, high-income borrowers, and high-income
17
In accordance with the definitions used by the federal bank supervisory agencies to enforce the
Community Reinvestment Act, LMI borrowers are defined as those with incomes less than 80 percent of
the estimated current area median family income (AMFI). Middle-income borrowers have incomes of at
least 80 percent and less than 120 percent of AMFI, and high-income borrowers have incomes of at least
120 percent of AMFI. AMFI is estimated based on the incomes of residents of the metropolitan area or
nonmetropolitan portion of the state in which the loan-securing property is located. For AMFI estimates,
see Fe deral Financial Institutions Examination Council (2020), “FFIEC Median Family Income Report,”
available at https://www.ffiec.gov/Medianincome.htm
. A very small percentage of records had income
reported as zero or negative. They are included in the LMI group.
18
Definitions for LMI, middle-income, and high-income neighborhoods are identical to those for LMI,
middle-income, and high-income borrowers, but are based on the ratio of census-tract median family
income to AMFI measured from the census data.
19
The bottom of Table 2 provides the total loan counts for each year, and thus the number of loans to a
given group in a given year can be easily computed. For example, the number of home purchase loans to
Asians in 2021 was approximately 312,000, calculated b y multiplying 4.4 million loans by 7.1 percent.
18
neighborhoods. The same can be said of home purchase loans. The numbers of originations for
home purchase loans were higher in 2021 than 2018 across all groups in Table 2.
19
TABLE 2: DISTRIBUTION OF HOME PURCHASE AND REFINANCE LOANS, BY
BORROWER AND NEIGHBORHOOD CHARACTERISTICS (PERCENT EXCEPT
AS NOTED)
Home Purchase
Refinance
2018
2019
2020
2021
2018
2019
2020
2021
A. Borrower race and
ethnicity
(1)
Asian
5.9
5.7
5.5
7.1
3.7
5.5
6.7
6.0
Black or African American
6.8
7.0
7.3
7.9
6.2
5.3
4.2
5.4
Hispanic White
8.9
9.2
9.1
9.2
6.8
6.2
5.3
6.1
Non-Hispanic White
62.0
60.3
59.1
55.6
63.1
60.9
61.0
58.3
Other minority
(2)
0.8
0.8
0.9
1.0
0.9
0.8
0.7
0.8
Joint
3.6
3.7
3.9
4.1
2.9
3.3
3.5
3.4
Missing
12.0
13.3
14.1
15.1
16.3
18.0
18.5
19.9
All
100
100
100
100
100
100
100
100
B. Borrower income
(3)
Low or moderate
28.0
28.6
30.4
28.7
29.0
23.1
18.9
23.9
Middle
26.7
27.2
27.4
27.0
25.2
22.1
21.8
23.2
High
43.9
43.1
41.2
43.2
41.3
43.5
45.1
42.0
Income not used or not
applicable
1.3
1.2
1.1
1.0
4.4
11.2
14.2
10.9
All
100
100
100
100
100
100
100
100
C. Neighborhood income
(4)
Low or moderate
16.5
16.5
16.1
17.1
16.3
14.0
11.7
13.6
Middle
44.2
44.3
44.3
44.3
45.6
43.0
41.1
42.8
High
38.8
38.9
39.3
38.4
37.7
42.7
47.0
43.5
All
100
100
100
100
100
100
100
100
Total (in thousands)
3,594
3,736
4,101
4,378
1,605
3,052
7,849
7,605
NOTE: Closed-end (excluding reverse mortgage), first-lien home purchase or refinance mortgages secured by site-
built, one- to four-family homes used for a principal residence.
(1) Applications are placed in one category for race and ethnicity. The application is designated as “joint” if one
applicant was reported as White and the other was reported as one or more minority races or if the application is
designated as White with one Hispanic applicant and one non-Hispanic applicant. If there are two applicants and
each reports a different minority race, the application is designated as two or more minority races. If an applicant
reports two races and one is White, that applicant is categorized under the minority race. Otherwise, the applicant is
categorized under the first race reported. "Missing" refers to applications in which the race of the applicant(s) has
not been reported or is not applicable or the application is categorized as White, but ethnicity has not been
reported.
(2) Consists of applications by American Indians or Alaska Natives, Native Hawaiians or other Pacific Islanders, and
borrowers reporting two or more minority races.
(3) The categories for the borrower-income group are as follows: Low- or moderate-income (or LMI) borrowers
have income that is less than 80 percent of estimated current area median family income (AMFI), middle-income
20
borrowers have income that is at least 80 percent and less than 120 percent of AMFI, and high-income borrowers
have income that is at least 120 percent of AMFI.
(4) The categories for the neighborhood-income group are based on the ratio of census-tract median family income
to area median family income published by FFIEC (available at https://www.ffiec.gov/Medianincome.htm), and the
three categories have the same cutoffs as the borrower-income groups (see note 3).
21
3.2 Mortgage characteristics of home loans
The median characteristics of mortgage loans and borrowers may differ substantially by race
and ethnicity, borrower income and neighborhoods, and loan types. Table 3A shows the median
loan amount, credit scores of borrowers, interest rates, and total loan costs of home purchase
loans for different racial/ethnic groups, borrower income, neighborhood income, and enhanced
loan types over time.
20
The sample is limited to closed-end first-lien mortgages for site-built
one-to-four-family, principal residences with loan purpose being home purchase. Table 3B
presents similar information for refinance loans.
The median loan amounts for home purchase loans have risen consistently across all categories,
likely reflecting the rise in home prices.
21
Among different racial/ethnic groups, in 2021, Asian
borrowers continued to take out home purchase loans with the largest median loan amount,
with a median of about $414,000, an increase from a median loan amount of $362,000 in 2020.
In contrast, Black borrowers continued to take out loans with the smallest median loan amount
for home purchase, at approximately $264,000, also an increase from 2020. The median loan
amount of home purchase loans for Hispanic White borrowers was $272,000 in 2021, while the
median loan amount of home purchase loans for non-Hispanic White borrowers was $274,000.
In 2021, the median loan amount of home purchase loans for high-income borrowers was
$394,000. As in 2020, high-income borrowers had a median loan amount which was almost
twice that of LMI borrowers. Similarly, the median loan amount of home purchase loans secured
by properties in high-income areas was $355,000, compared to $235,000 in LMI
neighborhoods.
Loan amounts vary by loan type. The home purchase jumbo loans have a median loan amount of
$948,000 in 2021. In comparison, the median loan amounts of home purchase loans were
$172,000 for RHS/FSA loans and $241,000 for FHA loans.
20
We report the medians instead of averages in Tables 3A and 3B because medians are more stable and
less subject to outliers than averages. In general, the patterns of averages looked similar to the patterns in
medians. Combining the transaction types (closed-end, open-end, reverse mortgage), loan types reported
under HMDA (conventional, FHA, VA, RHS/FSA), conforming loan status based on the loan amount
reported and the conforming loan limits published by the Federal Housing Finance Agency (FHFA), all
single family LARs can be grouped into seven categories: 1) Conventional Conforming; 2) Conventional
Non-conforming or Jumbo; 3) FHA; 4) VA; 5) RHS/FSA; 6) HELOC; and 7) Reverse Mortgage.
These
categories are referred to as the “Enhanced Loan Types”.
21
All dollar amounts are reported in nominal terms.
22
Similar to the home purchase loans, in 2021, the median loan amount of refinance loans for
Asian borrowers was also higher than other racial/ethnic groups, at approximately $355,000 in
2021, compared to $221,000 for Black borrowers, $243,000 for Hispanic White borrowers, and
$232,000 for non-Hispanic White borrowers. For all racial/ethnic groups, the median refinance
loan amount decreased from 2020 to 2021, just as the refinance loan origination volume
decreased. This is in contrast with home purchase loans, where the median loan amount
increased for all racial groups and origination volumes increased slightly.
Credit scores are widely used in mortgage underwriting and pricing. Where applicable, credit
scores have been collected and reported in HMDA since 2018.
22
Table 3A shows that the median
credit scores for closed-end home purchase mortgage loans secured by first-lien principal-
residence site-built single-family properties have been increasing from 2018 through 2021. But
the variations of median credit scores among different groups remain. Black and Hispanic White
borrowers continued to have lower median credit scores than other racial/ethnic groups. The
median credit score of Black borrowers who took out home purchase loans was 691 in 2021, and
the median credit score of Hispanic White home purchase loan borrowers was 716. In
comparison, the median credit scores of non-Hispanic White and Asian home purchase loan
borrowers in 2021 were 750 and 764, respectively.
High-income home purchase borrowers have higher median credit scores than middle and LMI
borrowers, and borrowers taking out home purchase loans secured by properties in high-income
neighborhoods have higher median credits scores than borrowers in middle and LMI
neighborhoods as well. In terms of loan types, the median credit score for home purchase jumbo
loan borrowers was 776 in 2021 and the median credit score of conventional conforming home
purchase loan borrowers was 759, while the median credit score of FHA home purchase loan
borrowers was 664.
The median credit scores of refinance loan borrowers are higher than those of home purchase
loan borrowers, both overall and across most groups. The median credit score of Black
borrowers who refinanced in 2020 was 719, and the median credit score of Hispanic White
refinance borrowers was 737. In comparison, the median credit scores of non-Hispanic White
and Asian borrowers who refinanced in 2021 were 763 and 776, respectively. The median credit
score for refinance jumbo loan borrowers was 776 in 2021, the median credit score of
conventional conforming refinance loan borrowers was 764, and the median credit score of FHA
refinance loan borrowers was 706.
22
To protect applicant and borrower privacy, credit score is excluded from the application-level HMDA
data made available to the public.
23
The mortgage interest rate increased during 2021 from its historical low at the end of 2020.
However, the rate of increase during 2021 was slower than the rate of decrease during 2020 as
illustrated in Figure 1. As a result, the median interest rate for home purchase loans secured by
first-lien principal-residence site-built single-family properties was 3.000 in 2021, compared to
3.125 percent in 2020.
The variations in interest rates across different racial/ethnic groups and loan types still
remained in 2021. At 3.125 percent, Black and Hispanic White borrowers continued to pay
higher median interest rates than all other racial/ethnic groups for home purchase loans. The
median interest rate for Asian borrowers was 2.875 percent and that for non-Hispanic White
borrowers was 3.000 percent. For home purchase loans, FHA loans had higher median interest
rates than any other enhanced loan types, at 3.125 percent. On the other hand, VA loans had the
lowest median interest rate at 2.750 percent.
Total loan costs, a data point collected and reported under HMDA since 2018, represent the sum
of origination fees that the lender charges, charges for the services that borrowers cannot shop
for (e.g., appraisal fees or credit report fees), and charges for services borrowers can shop for
such as settlement agent or title insurance fees.
23
The median total loan costs for home purchase loans in 2021 was $4,889, up by 3.2 percent
from $4,736 in 2020. Hispanic White borrowers for home purchase loans paid $6,282 in
median total loan costs, the highest among all racial/ethnic groups. The next highest was Black
borrowers. The median total loan costs for Black home purchase loan borrowers were $6,186, up
from $5,980 in 2020. In comparison, the median total loan costs of home purchase loans were
$4,462 for non-Hispanic White borrowers and $5,095 for Asian borrowers.
The median total loan costs of home purchase loans were higher for high-income borrowers
($5,148) and middle-income borrowers ($4,980) than LMI borrowers ($4,425)
24
. On the other
hand, the median total loan costs of home purchase loans were $4,952 for loans secured by
properties in LMI neighborhoods, $4,767 for loans in middle-income neighborhood and $4,995
for loans in high-income neighborhoods.
23
The total loan costs collected under HMDA only applies to originated loans that are subject to specified
requirements in Regulation Z. It is limited to “buyer-paid” portions of the total loan costs on the TILA-
RESPA Integrated Disclosure Rule (TRID) Closing Disclosure of applicable loans. In other words, under
the HMDA reporting requirements, it includes the charges by the lenders as well as the charges by the
third party service providers in connection with obtaining the loan to the extent those are paid by a
consumer rather than by a seller or other third party.
24
Just as other statistics presented in this table, the discussion here does not control for differences in
loan amount that may vary across income groups.
24
Among various enhanced loan types for home purchase loans, FHA loans had the highest
median total loan costs, at $8,427, likely reflecting the required upfront mortgage insurance
premium which could be a significant part of the total loan costs of FHA loans
25
. The median
total loan costs were $6,830 for VA home purchase loans. To the extent that Black and Hispanic
White borrowers are more likely to take out FHA loans than Asian and non-Hispanic White
borrowers, the high total loan costs of FHA loans could contribute to higher median total loan
costs for Black and Hispanic White borrowers as observed above.
The median total loan costs for refinance loans in 2021 was $3,336, which is lower than that for
home purchase loans, and only up very slightly from $3,310 in 2020. The Hispanic White
borrowers who refinanced in 2021 paid $3,879 in median total loan costs, also the highest
among all racial/ethnic groups for refinance loans. The next highest was Black borrowers. The
median total loan costs for Black refinance loan borrowers were $3,857. The median total loan
costs of refinance loans were $3,214 for non-Hispanic White borrowers and $3,030 for Asian
borrowers.
The median total loan costs of refinance loans were slightly lower for low-income ($3,286) and
middle-income borrowers ($3,313) in comparison with high-income borrowers ($3,387).
Conversely, the median total loan costs of refinance loans were higher for loans secured by
properties in LMI neighborhoods ($3,454) than those in middle-income ($3,337) or high-
income neighborhoods ($3,299). The pattern of median total loan costs for refinance loans by
enhanced loan types was consistent with the pattern for home purchase loans, with FHA
refinance loans having the highest loan costs and VA refinance loans having the lowest.
25
In the time period examined in this report, FHA charged an upfront mortgage insurance premium of
175 Basis Points (bps) (1.75%) of base loan amounts with the exception of a fe w limited programs and
products, such a s Hawaiian Home Lands and Indian Lands Programs. (See FHA Single Family Housing
Policy Handbook 4000.1 for more details.) Such upfront mortgage insurance premium is included in the
Closing Disclosure as one of the services that borrowers cannot shop for. By definition, if the upfront
mortgage insurance premium is “borrower-paid”, it should be included as part of the Total Loan Costs
that institutions report to HMDA.
29
3.3 Denial rates
The overall denial rate for home purchase applications for all applicants was 8.3 percent in
2021, lower than that in 2020 (9.3 percent) and in 2019 (8.9 percent).
26
About 12.4 percent of
FHA applications (excluding withdrawn or incomplete applications) for home purchase loans
were denied in 2021, followed by applications for FSA/RHS loans which had a denial rate of
10.0 percent. The conventional conforming home purchase applications had a denial rate of 7.1
percent in 2021, the lowest among all enhanced loan types.
The denial rates for refinance applications were higher than those for home purchase loans. The
overall denial rate on applications for refinance loans was 14.2 percent in 2021, up from 13.2
percent in 2020, but still much lower than the denial rate in 2018 (29.1 percent) and 2019 (19.2
percent). Consistent with home purchase applications, applications for FHA refinance loans
were more likely to be denied (23.6 percent) than all other enhanced loan types, while
applications for conventional conforming refinance loans were the least likely to be denied (13.0
percent).
As in past years, Black and Hispanic White borrowers had notably higher denial rates in 2021
than non-Hispanic White and Asian borrowers. Among home purchase applications, the denial
rates were 15.3 percent for Black applicants and 10.6 percent for Hispanic White applicants in
2021, both of which were lower than 2020 denial rates. In contrast, the denial rates of home
purchase applications were 7.9 percent for Asian applicants and 6.3 percent for non-Hispanic
White applicants.
Within each enhanced loan type except FHA and FSA/RHS loans, Black and Hispanic White
applicants for home purchase loans had higher denial rates than non-Hispanic White or Asian
applicants. For example, the denial rate of home purchase loan applications for conventional
conforming loans was 15.7 percent for Black applicants and 9.7 percent for Hispanic White
applicants. In contrast, the denial rate of home purchase loan applications for conventional
conforming loans was 5.4 percent for non-Hispanic White applicants. On the other hand, for
FHA home purchase applications, the denial rate of Asian applicants was higher than that of
Hispanic White applicants but lower than that of Black applicants.
26
Denial rates are calculated as the number of denied loan applications divided by the total number of
applications, excluding withdrawn applications and application files closed for incompleteness.
30
Consistent with denials for home purchase loans, Black and Hispanic White applicants were also
more likely to be denied for refinance applications, with denial rates almost twice as high for
Black applicants as compared with non-Hispanic White applicants. In 2021, about 23.6 percent
of Black applicants and 17.6 percent of Hispanic White applicants applying for refinance loans
were denied, compared to denial rates of 12.3 percent and 11.8 percent for Asian and non-
Hispanic White loan applicants, respectively. The disparities in refinance denial rate were
smaller among FHA and VA loan applications compared with conventional conforming and
jumbo loan applications. For example, the denial rates for conventional conforming refinance
applications stood at 23.9 percent, 17.1 percent, 11.9 percent, and 10.5 percent for Black,
Hispanic White, Asian, and non-Hispanic White applicants in 2021, respectively. In contrast,
the denial rates for FHA refinance applications were 25.3 percent, 21.9 percent, 23.8 percent,
and 21.9 percent for Black, Hispanic White, Asian, and non-Hispanic White applicants,
respectively.
33
4. Monthly mortgage trends and
activities
The discussion and tables in previous chapters provide historical comparisons on an annual
basis. In this chapter, we show the monthly trends of mortgage activities from 2018 to 2021. In
doing so, we rely on the action taken date that was collected and reported under HMDA. We
note that the action taken date is a data field not available in the publicly released version of the
modified HMDA data.
Figure 2 plots the number of originations for closed-end mortgage loans secured by site-built
single-family homes in each month between January 2018 and December 2021, separated by
three loan purposes: home purchase, cash-out refinance, and non-cash-out refinance. The home
purchase origination volume clearly displays a strong seasonal pattern, as home sales typically
fluctuate as the numbers of home sellers and buyers entering the market change with the
seasons. Figure 2A plots the year-over-year change in closed-end loan origination volumes by
loan purpose between January 2018 and December 2021. The home purchase loan origination
volume rose year-over-year starting in the second half of 2019 until April 2020 when the
pandemic and nationwide shutdown hit. Then starting in July 2020, the home purchase volume
recovered and increased significantly despite the typical low-home-sale seasons near the end of
the year. The year-over-year rise of home purchase loan volume continued well into the first half
of 2021, reaching 520,000 originations in June 2021. Then it mostly plateaued or dipped into
negative territories in terms of year-over-year change.
On the other hand, the refinance volume, especially in non-cash-out refinance that dominated
the mortgage market activities in 2019 and 2020, clearly peaked in March 2021, as shown in
Figure 2. Then the non-cash-out refinance volume dropped precipitously through the rest of
2021. At its peak, over 763,000 non-cash-out loans were originated in a month. By December
2021, the monthly non-cash-out refinance volume fell to 236,000, below the volume of cash-out
refinance for the first time since April of 2019. This coincided with the increase in market
interest rates, as noted in Figure 1. For most of 2021, the cash-out refinance volume continued
to grow, staying above 200,00 originations per month, likely because consumers took advantage
of rapid house price appreciation and converted some of their rising home equity into cash or
liquid assets.
Figure 2A plots the year-over-year change in closed-end loan origination volumes by loan
purpose between January 2018 and December 2021. The home purchase loan origination
volume rose year-over-year starting in the second half of 2019 until April 2020 when the
34
pandemic and nationwide shutdown hit. Then starting in July 2020, the home purchase volume
recovered and increased significantly despite the typical low-home-sale seasons near the end of
the year. The year-over-year rise of home purchase loan volume continued well into the first half
of 2021, reaching 520,000 originations in June 2021. Then it mostly plateaued or dipped into
negative territories in terms of year-over-year change.
35
FIGURE 2: Number of Closed-end Mortgage Originations by Loan Purpose
NOTE: Monthl y originations of closed-end loans secured by site-bu ilt si ngl e-family homes between January
2018 and December 2021.
36
FIGURE 2A: Year-over-Year Percentage Change of Number of Closed-end
Mortgage Originations by Loan Purpose
NOTE: Year-over-year percentage changes of monthly originations of closed-end loans secured by site-built
single-family homes between January 2019 and December 2021.
Figures 3 to 7 show the monthly loan origination volume of home purchase, cash-out refinance,
and non-cash-out refinance loans by enhanced loan types. The trends in the conventional
conforming market (Figure 3) were largely similar to the overall trends of the entire mortgage
market presented in Figure 2, as conventional conforming loans are the most common enhanced
loan type. The rise of home purchase loans in 2021 was the most prominent among jumbo loans
(Figure 4), with monthly origination volume of jumbo loans substantively higher than the
previous three years and reaching 33,700 originations in a month by the end of 2021. This partly
reflected rapidly rising house prices in 2021.
37
FIGURE 3: Number of Closed-end Conventional Conforming Originations by Loan
Purpose
NOTE: Monthly originations of closed-end conventional conforming loans secured by site-built single-family
homes between January 2018 and December 2021.
38
FIGURE 4: Number of Closed-end Jumbo Originations by Loan Purpose
NOTE: Monthly originations of closed-end jumbo loans secured by site-built single-famil y homes between
January 2018 and December 2021.
39
FIGURE 5: Number of Closed-end FHA Originations by Loan Purpose
NOTE: Monthly originations of closed-end FHA loans secured by site-built single-family homes between
January 2018 and December 2021.
40
FIGURE 6: Number of Closed-end VA Originations by Loan Purpose
NOTE: Monthly originations of closed-end VA loans secured by site-built single-family homes between
January 2018 and December 2021.
41
FIGURE 7: Number of Closed-end RHS/FSA Originations by Loan Purpose
NOTE: Monthly originations of closed-end RHS/FSA loans secured by site-bu ilt si n gl e-family homes between
January 2018 and December 2021.
Figure 8 shows the monthly origination volume of closed-end home purchase loans broken
down by race/ethnicity. Compared to the same period in 2020, Asian, Black, Hispanic White,
and non-Hispanic White borrowers all saw a significant increase in home purchase origination
volumes in the second half of 2021. Because the share of non-Hispanic White borrowers far
exceeds the shares of Asian, Black, and Hispanic White borrowers, to aid the visual comparison
of the lending volumes and shares of minority borrowers over time, we included Figures 8A, 9A,
10A, and 11A where non-Hispanic White borrowers are excluded, and the vertical axis is rescaled
accordingly. Figure 9 and 9A show the relative shares of monthly home purchase originations by
racial/ethnic group. In 2021, non-Hispanic White borrowers’ share among home purchase loans
continued a downward trend that began in 2020 (Figure 9). Meanwhile, the share of home
purchase originations for Asian borrowers rose sharply beginning in May 2020. In addition, the
share of home purchase originations for Black borrowers increased in 2021 compared to the
previous years (Figure 9A).
42
FIGURE 8: Number of Closed-end Home purchase Loan Originations by Race and
Ethnicity
NOTE: Monthly originations of closed-end home p urchase loans secured by si te-built single-family homes
between January 2018 and December 2021.
43
FIGURE 8A: Number of Closed-end Home purchase Loan Originations by Race
and Ethnicity (Minorities Only)
NOTE: Monthly originations of closed-end home p urchase loans secured by site-built single-family homes
between January 2018 and December 2021.
44
FIGURE 9: Share of Closed-end Home purchase Loan Originations by Race and
Ethnicity
NOTE: Monthly originations of closed-end home p urchase loans secured by si te-built single-family homes
between January 2018 and December 2021.
45
FIGURE 9A: Share of Closed-end Home purchase Loan Originations by Race and
Ethnicity (Minority Only)
NOTE: Monthly originations of closed-end home p urchase loans secured by si te-built single-family homes
between January 2018 and December 2021.
Plotting the monthly origination volume of closed-end refinance loans broken down by
race/ethnicity, Figure 10 shows the refinance boom during 2019 and 2020 largely ended in
2021. The refinance volume increased for Asian, Black, Hispanic White, and non-Hispanic
White borrowers during 2019 and 2020. Figure 10A shows that just as the refinance volume
increased the fastest for Asian borrowers in 2019 and 2020, it also declined the fastest in 2021.
Figure 11 and 11A show the relative shares of monthly refinance originations by racial/ethnic
group. Figure 11A shows a significant decline in Asian borrowers’ share of refinance loans in
2021. At its peak, Asian borrowers accounted for close to 7.7 percent of all refinance loans in
September 2021, far exceeding the shares of Black and Hispanic White borrowers. By the end of
2021, the share of refinance loans by Asian borrowers dropped to about 4.8 percent, dipping
below the shares of Black and Hispanic White borrowers. In comparison, the Black and
Hispanic White borrowers’ shares have steadily increased in 2021, accounting for about 6.6 and
6.8 percent of refinance loans by the end of 2021.
46
FIGURE 10: Number of Closed-end Refinance Loan Originations by Race and
Ethnicity
NOTE: Monthly originations of closed-end refinance loans secured by site-built single-family homes between
January 2018 and December 2021.
47
FIGURE 10A: Number of Closed-end Refinance Loan Originations by Race and
Ethnicity (Minority Only)
NOTE: Monthly originations of closed-end refinance loans secured by site-built single-family homes between
January 2018 and December 2021.
48
FIGURE 11: Share of Closed-end Refinance Loan Originations by Race and
Ethnicity
NOTE: Monthly originations of closed-end refinance loans secured by site-built single-family homes between
January 2018 and December 2021.
49
FIGURE 11A: Share of Closed-end Refinance Loan Originations by Race and
Ethnicity (Minorities Only)
NOTE: Monthly originations of closed-end refinance l oans secured by site-built single-family homes between
January 2018 and December 2021.
50
5. Mortgage trends and
activities by states
The mortgage activities not only changed across time, but they also varied across geography. In
this chapter, we show an annual growth rate of home purchase and refinance loans from 2020
to 2021 by state. All figures in this chapter are restricted to closed-end originations secured by
site-built single-family first-lien principal residences, excluding reverse mortgages.
As Figure 12 illustrates, seven states (Idaho, Utah, Arizona, Montana, New Hampshire,
Nebraska, and Tennessee) had a negative annual growth rate in home purchase originations.
The home purchase originations increased by less than five percent in sixteen states (South
Dakota, Texas, Oklahoma, Minnesota, Wisconsin, Georgia, Indiana, Vermont, Missouri, Rhode
Island, Kansas, Maine, Ohio, Arkansas, Iowa, and Kentucky). In contrast to year 2020 when 29
states experienced an annual growth rate of over 10 percent, only 13 states experienced an
annual growth rate over 10 percent in 2021.
27
27
While not shown on the map, the annual growth rates of home purchase loans for Puerto Rico and
Washington DC from 2020 to 2021 were about 35.9 percent and 11.5 percent, respectively.
51
FIGURE 12: Annual Growth Rate of Closed-end Home purchase Loan Originations
by State
NOTE: Annual growth rate of closed-end home purchase originations secured by site-b uil t si ng le-family
homes from 2020 to 2021.
For refinance loans, twenty-eight states saw a decline in 2021 compared to 2020, as shown in
Figure 13.
28
Eleven of them - South Dakota, Utah, Colorado, North Dakota, Washington,
Nebraska, Hawaii, Louisiana, Minnesota, Wisconsin, and Oregon – experienced an annual
decline of over 10 percent. This is in sharp contrast with 2020 when all states experienced a
yearly growth rate of over 100 percent.
28
While not shown on the map, the annual growth rates of refinance loans for Puerto Rico and
Washington DC from 2020 to 2021 were about - 12.8 percent and 39.8 percent, respectively.
52
FIGURE 13: Annual Growth Rate of Closed-end Refinance Loan Originations by
State
NOTE: Annual growth rate of closed-end re fi nance originations secured by site-built singl e-family homes
from 2020 to 2021.
53
6. Lending institutions
In 2021, 4,332 financial institutions reported closed-end applications and originations excluding
reverse mortgages
29
, down from 4,472 in 2020. This represents a continuation of the downward
trend of institutions reporting closed-end applications and originations from 2019, where there
were 5,500 institutions. Some of the decrease in the number of institutions reporting closed-end
transactions is likely driven by the increased reporting threshold that took effect in July 2020
due to the 2020 HMDA rule. As discussed in Section 2, the overall market volume of closed-end
transactions excluding reverse mortgages, however, has continued to increase over the past few
years, from 13.0 million applications in 2019 to 20.9 million in 2020 and 21.5 million in 2021.
The financial institutions are broadly categorized into depository institutions (DIs) and non-
depository institutions (non-DIs). Table 5A shows the number of DIs and non-DIs among the
closed-end reporting institutions in 2021. DIs included 1,937 banks and thrifts (hereafter,
banks) of which 1,299 were small (assets less than $1 billion), and 1,355 credit unions. Non-DIs
included 68 mortgage companies affiliated with DIs and 972 independent mortgage
companies.
30
Among DIs, banks collectively originated most of the reported closed-end loans, whereas,
among non-DIs, independent mortgage companies originated significantly more loans than
mortgage companies affiliated with DIs. For example, banks collectively originated 3.7 million
loans, accounting for 26.3 percent of all reported closed-end originations in 2021. Credit unions
originated 1.0 million loans, accounting for only 7.5 percent. Independent mortgage companies
originated 8.9 million loans, accounting for 63.3 percent of all reported loans, whereas affiliates
of DIs originated only 400,000 loans or 2.9 percent.
In 2021, independent mortgage companies originated 64.1 percent of all closed-end (excluding
reverse mortgage) first-lien, owner-occupied, site built, one-to-four-family, home purchase
loans, which compares to 60.7 percent in 2020. They also originated 65.8 of all closed-end,
29
Reverse mortgages can be structured as either closed- or open-end transactions. Tables 5a and 5b
exclude reverse mortgages. There is no separate reporting threshold for reverse mortgages.
30
Data on bank assets were drawn from the Federal Deposit Insurance Corporation’s Reports of
Condition and Income. The $1 billion threshold is based on the combined assets of all banks within a
given b anking organization. Fed. Fin. Inst. Examination Council, HMDA Data Publication,
https://ffiec.cfpb.gov/data-publication/
(data from the HMDA Reporter Panel can be used to help
identify the various types of institutions).
54
first-lien, owner-occupied, one-to-four family site-built refinance loans, which compares to 63.1
percent in 2020.
31
Only a small percentage of institutions (15.1 percent, 652 out of 4,332) reported fewer than 100
closed-end originations in 2021, accounting for about 35,000 total originations or about 0.2
percent of all originations. About 3 percent of institutions (140 out of 4,332), originated fewer
than 25 loans, totaling just over 1,000 originations.
32
A number of differences exist between DIs and non-DIs with respect to the closed-end activity
reported in 2021. First, DIs originated a significantly higher percentage of conventional home
purchase loans than non-DIs. Second, overall, smaller shares of loans originated by DIs went to
minority borrowers, LMI borrowers, and in LMI neighborhoods than non-DIs. Third, non-DIs
sold more of their originated loans compared to DIs. The HMDA data provide information on
whether lenders sold originated loans within the same calendar year that they were originated,
as well as the type of institution to which the lenders sold the loans, such as one of the GSEs or a
banking institution.
33
Table 5A shows that non-DIs, particularly affiliated mortgage companies,
sold almost all of their loans in the same calendar year that they originated them.
A distinct pattern emerges within a specific institution type. For example, among DIs, credit
unions were the least likely to sell originated loans compared to banks. Small banks and large
banks sold 81.1 percent and 73.8 percent, respectively, of their refinance loans within the same
calendar year of the originations. In contrast, credit unions sold 40.3 percent of their refinance
loans during the same period.
31
See “2020 Mortgage Market Activity and Trends,” available at https://www.consumerfinance.gov/data-
research/research-reports/2020-mortgage-market-activity-and-trends/.
32
These results include all originated dwelling-secured, closed-end loans, excluding reverse mortgage for
all reporters. Effective July 1, 2020, the reporting threshold of 100 closed-end originations applies to
originations in each of the previous two years.
33
Because loan sales are recorded in the HMDA data only if the loans are originated and sold in the same
calendar year, loans originated toward the end of the year are less likely to be reported as sold. For that
reason, statistics on loan sales are computed using only loans originated during the first three quarters of
the year.
55
TABLE 5A: LENDING ACTIVITIES BY INSTITUTION TYPES: CLOSED-END EXCLUDING
REVERSE MORTGAGES
Type of Institution
Small
Bank
Large
Bank
Credit
Union
Affiliated
Mortgage
Company
Inde-
pedent
Mortgage
Company
All
Number of institutions
1,299
638
1,355
68
972
4,332
Applications (thousands)
867
4,512
1,655
565
13,867
21,465
Originations (thousands)
652
3,032
1,042
400
8,859
13,986
Purchases (thousands)
8
1,100
25
7
1,520
2,660
SIZE DISTRIBUTION
Institutions with originations <25
Number of institutions
9
2
74
3
52
140
Originations (thousands)
<1
<1
1
<1
<1
1
Institutions with >= 25 originations < 100
Number of institutions
137
22
302
<1
51
512
Originations (thousands)
10
2
19
<1
3
33
Institutions with >= 100 originations < 1000
Number of institutions
1,068
307
762
21
351
2,509
Originations (thousands)
340
151
270
10
163
933
Institutions with >= 1000 originations
Number of institutions
85
307
217
44
518
1,171
Originations (thousands)
303
2,879
753
391
8,693
13,018
Home purchase loans (thousands)
207
906
252
205
2,808
4,378
Conventional (percent)
78.7
85.2
88.9
63.5
64.4
70.7
LMI borrower (percent)
31.1
25.1
28.4
29.4
29.7
28.7
LMI neighborhood (percent)
14.9
14.6
16.3
15.2
18.3
17.1
Non-Hispanic White (percent)
70.7
60.6
61.5
53.5
52.5
55.6
Minority borrower (percent)
16.2
21.8
19.0
27.1
27.3
25.2
Sold (percent)
77.7
67.3
45.5
99.2
93.9
85.0
Refinance loans (thousands)
283
1,603
576
138
5,004
7,605
Conventional (percent)
89.7
95.9
94.7
91.7
76.8
83.0
LMI borrower (percent)
23.5
22.3
26.0
27.9
24.1
23.9
LMI neighborhood (percent)
10.3
11.2
14.0
12.9
14.5
13.6
Non-Hispanic White (percent)
74.9
67.1
63.9
66.5
53.7
58.3
Minority borrower (percent)
10.0
15.9
15.8
14.5
20.0
18.4
Sold (percent)
81.1
73.8
40.3
99.5
95.7
86.6
(1) Small banks consist of those banks with assets (including the assets of all other banks in the same banking
organization) of less than $1 billion at the end of 2021. Affiliated mortgage companies are nondepository mortgage
companies owned by or affiliated with a banking organization or credit union.
(2) Closed-end first-lien mortgages for site-built single-family, principal-residence homes.
(3) See table 2, note 3.
(4) See table 2, note 4.
56
(5) See table 2, note 1. "Minority borrower" refers to non-White borrowers (excluding joint or missing) or Hispanic
White applicants.
(6) Excludes originations made in the last quarter of the year because the incidence of loan sales tends to decline
for loans originated toward the end of the year, as lenders report a loan as sold only if the sale occurs within the
same year as origination.
SOURCE: FFIEC HMDA data; bank asset data drawn from Federal Deposit Insurance Corporation Reports of
Condition and Income (https://www.fdic.gov).
In 2021, 936 financial institutions reported open-end applications and originations excluding
reverse mortgages (Table 5B), very similar to the number of institutions reporting in 2020 and
2019 (938 and 936 respectively). As discussed in Section 2, the overall market volume of open-
end transactions excluding reverse mortgages increased slightly from approximately 1.7 million
applications in 2020 to 1.8 million in 2021.
HELOCs are dominated by DIs. In 2021, 809 DIs, including 271 banks, of which 44 were small
(assets less than $1 billion), and 538 credit unions, originated, in total, 934,000 HELOCs
reported under HMDA, accounting for 97.1 percent of all HELOC originations reported. Only
127 non-DIs, including 2 mortgage companies affiliated with DIs and 125 independent mortgage
companies reported HELOC originations, accounting for 2.9 percent of the market. Only a very
small percentage (3.3 percent) of HELOCs were sold to other institutions.
57
TABLE 5B: LENDING ACTIVITIES BY INSTITUTION TYPES: OPEN-END EXCLUDING
REVERSE MORTGAGES
Type of Institution
Small Bank
Large
Bank
Credit
Union
Affiliated
Mortgage
Company
Inde-
pedent
Mortgage
Company
All
Number of institutions
44
227
538
2
125
936
Applications (thousands)
2
1132
566
<1
62
1762
Originations (thousands)
1
572
360
<1
28
962
Purchases (thousands)
<1
1
3
<1
<1
5
SIZE DISTRIBUTION
Institutions with originations <100
Number of institutions
37
25
226
1
114
403
Originations (thousands)
<1
<1
4
<1
1
5
Institutions with >= 100 originations < 100
Number of institutions
7
31
102
1
3
144
Originations (thousands)
1
8
18
<1
1
27
Institutions with >= 500 originations
Number of institutions
<1
171
210
<1
8
389
Originations (thousands)
<1
565
338
<1
27
929
LMI borrower (percent)
19.9
19.2
18.3
10.8
15.7
18.8
LMI neighborhood (percent)
10.9
10.5
12.5
17.5
16.1
11.4
Minority borrower (percent)
80.3
70.2
67.3
62.5
54.7
68.7
Non-Hispanic White (percent)
6.7
11.9
13.5
12.5
19.6
12.7
Sold (percent)
0.8
0.3
3.7
69.2
60.4
3.3
(1) See table 5A, note 1.
(2) See table 2, note 3.
(3) See table 2, note 4.
(4) See table 5A, note 5.
(5) See table 5A, note 6.
58
Table 6A lists the top 25 closed-end reporting institutions by the total number of closed-end
originations and their lending characteristics
34
, limited to first-lien mortgages for site-built
single-family, principal-residence homes excluding reverse mortgages. Together the top 25
closed-end lenders originated 6.1 million loans in 2021, accounting for 43.9 percent of the
national market, the highest share of the national market in the last four years. Since 2018, the
top 25 closed-end lenders have increased their share of total origination volume each year, from
33.8 percent in 2018, to 37.2 percent in 2019, to 38.9 percent of all closed-end originations in
2020.
35
The top 25 lenders were particularly prominent in the refinance space. In 2021, they
originated 4.0 million refinance loans, accounting for 53.0 percent of refinance loans by all
reporting institutions. The top 25 lenders originated 1.5 million home purchase loans, or 34.4
percent of home purchase loans.
With about 1.2 million originated loans, Rocket Mortgage (formerly known as Quicken Loans)
continued to be the highest volume closed-end lender, with a market share of about 8.8
percent.
36
Nineteen of the 25 top financial institutions were independent mortgage companies, 5
were large banks, and 1 was a credit union. Two institutions , Better Mortgage Corporation and
Navy Federal Credit Union, were new to the list of top 25 closed-end reporting institutions in
2021. Both institutions had a market share of 1.0 percent or less. Two institutions, Finance of
America Mortgage and Primelending, A PlainsCapital Company fell off the list.
Table 6B lists the top 25 HELOC reporting institutions by origination volume in 2021. In total,
the top 25 HELOC reporters accounted for 423,607 HELOC originations or 44.0 percent, of all
34
Some institutions may be part of a larger organization; however, the data in Tables 6A and 6B are at the
reporter level. Because affiliate activity has declined markedly since the housing boom in the mid-2000s,
a top 25 list at the organization level is not likely to be significantly different from Tables 6A and 6B.
35
For the top 25 closed-end lenders over the past four years, see “2020 Mortgage Market Activity and
Trends,” available at
https://www.consumerfinance.gov/data-research/research-reports/2020-mortgage-
market-activity-and-trends/, “Data Point: 2019 Mortgage Market Activity and Trends,” available at
https://www.consumerfinance.gov/data-research/research-reports/data-p oint-2019-mortgage-market-
activity-and-trends/, and “Data Point: 2018 Mortgage Market Activity and Trends,” available at
https://www.consumerfinance.gov/data-research/research-reports/data-point-2018-mortgage-market-
activity-and-trends/.
36
Notably, loan counts and market shares derived from the HMDA data can differ from some other
industry sources, such as the market shares compiled by Inside Mortgage Finance
(https://www.insidemortgagefinance.com/
). For HMDA reporting purposes, institutions report only
mortgage applications for which they make the credit decision. Under HMDA, if an application was
approved by a third party (such as a correspondent) rather than the lending institution, then that third
party reports the loan as its own origination, and the lending institution reports the loan as a purchased
loan. Alternatively, if a third party forwards an application to the lending institution for approval, then the
lending institution reports the application under HMDA (and the third party does not report anything). In
contrast, Inside Mortgage Finance considers loans to have been originated by the acquiring institution
even if a third party makes the credit decision. Thus, many of the larger lending organizations that work
with sizable networks of correspondents report considerable volumes of purchased loans in the HMDA
data, while Inside Mortgage Finance considers many of these purchased loans to be originations.
59
HELOC originations reported under HMDA in 2021. All but one of the top 25 HELOC lenders
were DIs and none of the DIs were small banks. The two largest HELOC reporters in 2021 were
Citizens Bank and PNC Bank, accounting for 5.1 and 4.2 percent of all HELOC originations
reported in 2021. Last year, Bank of America had the highest market share of HELOC
origination at 5.6 percent, while this year the institution was the fifth largest with a market share
of 3.3 percent.
Table 6C lists the top 10 reverse mortgage lenders by origination volume in 2021.
37
In total, the
top 10 reverse mortgage lenders accounted for a little under 55,000 reverse mortgage
originations, or approximately 93.1 percent, of all reverse mortgage originations reported under
HMDA in 2021. American Advisors Group continued to be the largest reverse mortgage lender
that reported HMDA data in 2021, accounting for approximately 31.3 percent of all reverse
mortgage originations reported. It was followed by Finance of America Reverse LLC with an
annual market share of 18.0 percent.
37
Applicants generally do not need to provide income information when applying for reverse mortgage
loans. Therefore, the values for applicant income of reverse mortgages in the HMDA data are mostly “Not
Applicable” and we excluded the LMI borrower column in Table 6C.
66
7. Conclusion
The 2021 HMDA data are the fourth year of data that reflect changes implemented by the 2015
HMDA rule. The 2015 HMDA rule made changes to the institution and transaction reporting
criteria and revised the data points that institutions covered under HMDA must report.
The number of closed-end originations (excluding reverse mortgages) increased modestly from
13.4 million in 2020 to 13.7 million in 2021 or by 2.4 percent. Most of the increase was driven by
an increase in the number of home purchase loans. The increase in home purchase loans was the
most prominent among jumbo loans, likely reflecting rapidly rising house prices. In 2021, the
share of home purchase loans by non-Hispanic White borrowers continued a downward trend
while the share by Black borrowers continued an upward trend that began in 2018. The share of
home purchase loans by Asian borrowers rose in 2021 after being relatively stable in the past
three years.
The number of refinance loans decreased year-over-year by 1.7 percent. The refinance boom,
especially in non-cash-out refinance that dominated mortgage market activities in 2019 and
2020, peaked in March 2021. The non-cash-out refinance volume dropped precipitously
throughout the remainder of 2021. The downward trend coincided with the increase in market
interest rates.
Black and Hispanic White borrowers continued to have lower median loan amounts, lower
median credit scores, paid higher median interest rates and total loan costs, and were more
likely to be denied than non-Hispanic White and Asian borrowers. The market was more
concentrated among the largest lenders in 2021, with the top 25 closed-end lenders accounting
for 43.9 percent of the national market and 53.0 percent of all refinance loans.