FR2028_20170731_omb

Supporting Statement for the
Survey of Terms of Lending
(FR 2028; OMB No. 7100-0061)
Summary
The Board of Governors of the Federal Reserve System (Board), under delegated
authority from the Office of Management and Budget (OMB), proposes to extend for three years,
with revision, the voluntary Survey of Terms of Lending (STL; FR 2028; OMB No. 7100-0061).
This family of reports is currently comprised of the following three reports:
 Survey of Terms of Business Lending (FR 2028A)
 Survey of Terms of Bank Lending to Farmers (FR 2028B)
 Prime Rate Supplement of the Survey of Terms of Lending (FR 2028S)
The STL collects unique information concerning price and certain nonprice terms of
loans made to businesses and farmers during the first full business week of the mid-month of
each quarter (February, May, August, and November). The FR 2028A and FR 2028B collect
detailed data on individual loans made during the survey week, and the FR 2028S collects the
prime interest rate for each day of the survey from both FR 2028A and FR 2028B respondents.
From these sample STL data, estimates of the terms of business loans and farm loans extended
during the reporting week are constructed. The aggregate estimates for business loans are
published in the quarterly E.2 statistical release, Survey of Terms of Business Lending, and
aggregate estimates for farm loans are published in the E.15 statistical release, Agricultural
Finance Databook.
The Federal Reserve proposes to (1) discontinue the FR 2028A, (2) create a new Small
Business Lending Survey (FR 2028D) that would provide focused and enhanced information on
small business lending including rates, terms, credit availability, and reasons for their changes (in
contrast to the individual loan data collected on the FR 2028A, the FR 2028D would collect
quarterly average quantitative data on terms of small business loans and qualitative information
on changes and the reasons for changes in the terms of lending), and (3) rename the STL as the
Survey of Small Business and Farm Lending (SSBFL) to more accurately describe the data
collection. No changes are proposed to the FR 2028B and FR 2028S.
The proposed final data collection for the FR 2028A would be for the May 2017 survey
week, and the proposed first data collection for the FR 2028D would be in February 2018 for the
December 31, 2017, as of date. The annual reporting burden for the proposed FR 2028 reports is
estimated to be 4,485 hours (which includes 597 hours for implementation costs associated with
the first FR 2028D submission), a decrease of 2,873 hours from the current burden of 7,358
hours. The decrease in burden would result from the cancellation of the FR 2028A survey, and
its replacement with another survey that would be less burdensome to complete.
Background and Justification
Analyses of the STL data provide estimates of the cost of important segments of nonfarm
business and agricultural credit at banks that are representative of banking institutions

nationwide. Currently, it is the Federal Reserve’s only available source of data on bank loan
pricing for individual loans of all sizes to nonfarm businesses and farmers. Since its inception in
February 1977, the STL has been periodically revised to accommodate changes in lending
practices.
The FR 2028A was designed to allow the Federal Reserve to measure the cost of nonfarm
business borrowing from banks and to assess conditions and analyze developments in nonfarm
business credit markets. It replaced the Quarterly Interest Rate Survey and portions of the Survey
of Selected Interest Rates of the Committee on Interest and Dividends (CID survey). It was
designed to provide more accurate and detailed information on nonfarm business loans,
especially concerning maturity and nonprice terms, than the previous surveys.
The survey has undergone several modifications from 1997 to 2012. Examples include
expanding the respondent panel to include U.S. branches and agencies of foreign banks, adding
questions on interest rate adjustment and maturity items, and increasing the minimum loan size
reported. The results were reported to the Federal Reserve Board and to the Federal Open Market
Committee, and the microdata are available to Federal Reserve staff for developing and
publishing research papers. The data have been the only source of individual nonfarm business
loans made to small businesses, for which banks are one of the primary sources of credit.
The FR 2028B also has undergone several modifications in recent years. In 1997, interest
rate adjustments and maturity items were added and redefined, and a risk-rating item was added
to the collection. In addition, the prime rate supplement data (FR 2028S), which previously had
been collected just from respondents to the FR 2028A, were collected from FR 2028B
respondents. In 2006, the minimum size of loans reported was increased to $3,000 from $1,000,
a level at which it had been held since the STL’s inception in 1977. The adjustment reflected
price inflation over the intervening period and the increased use of business credit cards,
developments that likely had added significantly to the burden of reporting small loan amounts.
In August 2012, the FR 2028B added an item to collect the state where the borrower is
headquartered.
Description of Information Collection
Survey of Terms of Business Lending (FR 2028A)
The survey data are used to assess conditions and to track developments in business
credit markets. For instance, during the credit market turmoil that began in the second half of
2007 and early 2008, STL data showed a smaller increase in the spread of loan rates over banks’
cost of funds than other indicators of business loan pricing suggested. Moreover, information
about the date on which commitments were finalized or renewed has been important in
understanding how loan rates evolved during the crisis, as it allowed the Federal Reserve to
study the terms on new loan commitments separately from commitments written prior to the
crisis. More broadly, the survey data have been useful for monitoring the changing role of the
prime rate as a benchmark for business loan pricing and of shifts in the mix of fixed-rate and
variable-rate lending as financial markets have changed. The STL microdata are not available to
researchers outside the Federal Reserve, but have been used in a number of research papers.

2

The FR 2028A data have limitations for assessing conditions and analyzing
developments in nonfarm business credit markets. For example, it was noted in the memorandum
for renewing the STL in June 2015 that “The STL is an important source of individual loan data
used by those concerned with lending to small businesses, for which banks are one of the
primary sources of credit.”1 However, the data were insufficient for addressing questions about
small business lending during the financial crisis, ensuing recession, or economic recovery. For
example, the data could not answer questions on whether changes in the flow of credit to small
businesses were due to supply issues, such as changes in bank lending standards or terms,
demand issues, such as changes in application rates, or both. Additionally, the FR 2028A data
could not be used to answer questions regarding changes in the credit quality of applicants or
identify potential underlying factors for observed changes in credit quality. For reasons such as
these, the June 2015 memorandum stated “The Federal Reserve is seeking alternative sources of
detailed, disaggregated data on small business loans, but there are currently none available.
Should a better source for this type of data become available, the Federal Reserve may revisit the
need for this survey.”
The Federal Reserve System has conducted a study of alternative small business loan
data sources to assess their usefulness for addressing policy questions on small business credit.
The study identified and conducted an extensive analysis of 35 existing and potential new small
business lending data collections. The data collections considered included, among others,
collections undertaken by the Board of Governors, private sector surveys such as the National
Federation of Independent Business member survey, and a Dodd-Frank Act mandated data
collection by the Consumer Financial Protection Bureau.2 The primary finding was that existing
and new data collections under consideration would not meet the policy needs for understanding
and addressing the relevant policy issues and questions.
The FR 2028D data collection is being proposed to address the gaps in existing and
planned new surveys on small business lending. In addition, other Federal Reserve reports that
have been developed in recent years provide information on large nonfarm business loans. As a
result, the information used for assessing and analyzing developments in nonfarm business credit
markets would be improved by combining the proposed FR 2028D data collection on the terms
of small business loans with the existing reports on large business loans. For these reasons, the
FR 2028A would be discontinued. The proposed final data collection for the FR 2028A would be
for the May 2017 survey week.
Survey of Terms of Bank Lending to Farmers (FR 2028B)
This survey collects basic information that the Federal Reserve uses to monitor financial
developments in the agricultural sector of the economy. For example, the data were invaluable
during the period of financial stress for many farmers and farm lenders that began early in the
1980s. When that stress began to ease, the severe drought that developed in mid-1988 renewed
concerns about the financial health of the farm sector. The STL provided a timely indicator of
1

See page 3 of the June 2015 OMB Supporting Statement for the FR 2028 at
http://www.reginfo.gov/public/do/PRAViewDocument?ref_nbr=201505-7100-002.
2
Dodd-Frank Wall Street Reform and Consumer Protection Act, Section 1071, Subtitle G – Regulatory
Improvements, Sec. 1071 – Small Business Data Collection.

3

the possible effects of the drought on the terms of loans for various purposes. More recently, the
STL has been able to identify turning points in agricultural lending activity. For example,
agricultural loan demand strengthened in 2011 when farm input prices rose, and again in 2014
when a steep drop in crop commodity prices boosted the need for short-term financing for
operating expenses and record high feeder livestock prices drove a sharp rise in the volume of
loans made to the livestock sector.
The FR 2028B collects information on farm financial developments that is frequently
used by the Federal Reserve in congressional testimony, meetings with legislators, and
discussions with farm groups, as well as in response to information requests from the Congress,
other government agencies, the media, and academics.
Data from the FR 2028B have been very useful in monitoring the ongoing adjustment of
rural banks to the more volatile supply of, and demand for, loan funds of recent years. In 1978,
bank regulators removed fixed ceilings on interest rates paid on certain retail deposits, which
were most prevalent at rural banks, leading these small banks to compete more vigorously for
deposits. Since that time the agricultural sector—the underlying source of prosperity in many
small communities—has experienced large swings in farm income, asset values, and rates of
resource utilization. Data from the FR 2028B on the level and distribution of loan maturities and
loan rates among smaller banks have provided a useful indicator of the degree to which these
additional sources of risk have been passed through to rural borrowers.
The FR 2028B collects unique information on lending terms for farm borrowers, and the
availability of a historical series on farm lending frequently proves useful in addressing new
questions that arise. For example, the data have been quite useful in providing a scale to measure
the amount of government subsidy that is provided to farmers through governmental or quasigovernmental agencies. Legislation passed in 1987 to assist the Farm Credit System (FCS)
forbade the FCS from offering unusually low rates of interest by basing the interest rate it
charged on farm loans on its average cost of funds rather than its marginal cost. Aggregate
estimates from the FR 2028B data provided a useful benchmark in a General Accounting Office
study of the FCS’s performance in this regard (GAO-94-39, March 1994). The aggregate
FR 2028B data also have been used in recent years by analysts at the Department of Agriculture
to monitor compliance with the North American Free Trade Agreement by estimating the
magnitude of farm subsidies arising from loans to U.S. farmers from the FCS and the Farmers’
Home Administration. In addition, the Federal Reserve has used the information on farm lending
terms and the risk ratings for individual loans to examine the degree to which commercial banks
price the riskiness of agricultural loans. This study found that only about half of the reporters for
the FR 2028B used a risk rating system, but most of the banks that did not were quite small, and
so roughly four-fifths of the survey loans carried an informative risk rating. Furthermore, after
controlling for the size and performance of the bank and as many nonprice terms of the loan as
possible, banks consistently charged higher rates of interest for farm loans that they characterized
as riskier.3 The FR 2028B data have also been used to analyze recent farm debt trends in relation

Nick Walraven and Peter Barry (2004) “Bank Risk Ratings and the Pricing of Agricultural Loans.” Agricultural
Finance Review, Vol. 64, No. 2, pp. 107-118.
3

4

to historical agricultural boom and bust cycles.4
Prime Rate Supplement to Survey of Terms of Lending (FR 2028S)
The FR 2028S is completed by banks that file the FR 2028A or the FR 2028B. The prime
rate, an administered rate, remains the base rate banks use to price a significant portion of the
loans covered by the FR 2028A and FR 2028B.5 The prime rate is by far the most common base
rate used to price variable rate business and farm loans at small and medium-sized banks. Even
for large borrowers and the largest banks, the prime rate is a pricing option frequently available
along with market-related rates. The FR 2028S imposes little burden and the information it
provides is useful in interpreting movements in rates charged on business and farm loans,
especially for small loans and for loans at smaller banks. It also provides valuable information
about variations in the prime-lending rate across banks, which can be considerable. The
FR 2028S will be renewed without revision and will be reported by FR 2028B respondents.
Information on base rates for small business loans will be included in the proposed FR 2028D.
Proposed Small Business Lending Survey (FR 2028D)
The FR 2028D would collect quantitative and qualitative information that the Federal
Reserve can use to monitor developments in the availability of credit to small businesses. Bank
lending to small businesses is critical for employment and economic growth at the local,
regional, and national levels because it is a primary source of funding for these businesses. The
FR 2028D was motivated by the inability to answer basic policy questions raised by Federal
Reserve policymakers on small business credit during the recent financial crisis and subsequent
recovery. It would also contribute to a better understanding of the role of community banks in
providing loans to small businesses and on small business access to credit in local communities.
The survey would be timed to make reports on developments in small business lending available
for the second FOMC meeting of each quarter. The data would also be available for Federal
Reserve System economists and other staff to use for research purposes. To get a complete
understanding of the availability, terms, and market conditions of bank lending to small and large
nonfarm businesses, the Federal Reserve would combine the information gathered from the
FR 2028D with other Federal Reserve data collections that gather information on large business
loans.
The FR 2028D would improve the ability to assess and analyze developments in nonfarm
small business credit markets and to answer policy questions in a timely manner. The proposed
information to be collected is not available from existing or planned surveys conducted by either
the private or public sectors. The survey would collect unique, quarterly quantitative and
qualitative information on nonfarm small business lending that improves upon the information
currently collected by the FR 2028A. The quantitative information is similar to the data in the
FR 2028A, but the FR 2028D would collect quarterly amounts or average levels of the data items
Jason Henderson and Nathan Kauffman (2013) “Farm Investment and Leverage Cycles: Will This Time Be
Different?” Federal Reserve Bank of Kansas City Economic Review, Second Quarter, pp. 89-114.
5
The FR 2028S defines the prime rate to be, “[T]he administered rate used [by the bank] for pricing business and
other credit, which [is adjusted] from time to time in response to changes in market conditions. [The] institution may
set this rate internally or may adopt as its own a published rate.”
4

5

as opposed to individual loan information from a survey week. As a result, the quantitative
information will be less costly to report and less impacted by idiosyncratic events. The
qualitative questions will provide information on changes in loan demand, credit standards and
terms, and credit quality of applicants and reasons for the changes. Information on the reasons
for denying a small business loan application will also be collected.
The FR 2028D would also improve upon current information on outstanding loans
collected on the Report of Condition and Income (Call Report; FFIEC 031, FFIEC 041, and
FFIEC 051; OMB No. 7100-0036), which collects data on loans less than a certain dollar amount
rather than on loans to small businesses. The Call Report data may result in information
distortions about the availability of credit to small businesses because not all small loans are
made to small businesses.
The FR 2028D would collect quantitative and qualitative information on loans to small
businesses from a stratified sample of 398 banking institutions. The survey would be
administered at a quarterly frequency and distributed during the second month of each quarter.
Survey responses would be based on loan activity over the previous quarter. Quantitative
information collected would include the aggregate number and dollar amount of outstanding
loans and new loans extended by banks to small businesses each quarter, as well as line-of-credit
drawdowns and the average interest rate and benchmark rate. Loans are separated into two
categories: term loans and lines of credit, with each category further separated into fixed rate and
variable rate. Additionally, quantitative information on the number and dollar amount of small
business loans with guarantees (Small Business Administration and other) would be collected, as
well as information regarding loan maturity and the use of interest rate floors. The FR 2028D
would also collect quantitative information on small business loan applications received and
applications approved during the survey quarter, including information on applications from
Low- and Moderate-Income tracts.
Qualitative information collected by the FR 2028D would include questions to gauge
changes in lending terms, loan demand, and credit standards for small business loans during the
survey period.6 Furthermore, respondents will be asked to identify possible reasons for indicated
changes in lending terms or credit standards. The survey would also include qualitative questions
on the demand for small business loans, changes in credit line usage, and changes in the credit
quality of small business loan applicants. Respondents would be asked to identify potential
factors underlying a reported change in applicant credit quality (e.g. credit scores, quality of
collateral) and to identify top reasons for denying small business loans during the survey quarter.

6

The inclusion of qualitative questions, which are the same as those in the Senior Loan Officer Opinion Survey on
Bank Lending Practices (FR 2018; OMB No. 7100-0058), is meant to supplement the existing FR 2018 data to get a
more comprehensive view of the availability of credit to businesses. Importantly, the definitions of a small business
are different in the FR 2018 and proposed FR 2028D. The FR 2018 covers lending to both small and large firms and
defines small firms as those with annual sales of less than $50 million, which is significantly larger than the $5
million threshold in the FR 2028D. Furthermore, the FR 2018 panel only includes large institutions while the
FR 2028D panel will be a stratified sample of 398 domestic banks and include institutions of all sizes. Therefore, not
much overlap in the panels for the two data collections is expected.

6

Reporting panel
The FR 2028B panel has an authorized size of 250 domestically chartered commercial
banks. The panel of banks has been drawn from a random sample of banks stratified according to
farm loan volumes since 1989. Since that time, the authorized size of the panel has been 250
banks, with 189 banks currently reporting. The number of respondents is less than the authorized
size due to mergers among reporters and loss of respondents due to the voluntary nature of the
panel. Table 1 presents the number of reporters disaggregated by district and by bank-size group
for the current panel. If the authorized size of the panel is reduced to 189, the standard errors for
the data items would increase 17 percentage points. Moreover, the standard errors on the regional
estimates, which are based on smaller samples, likely would be greatly increased. Consequently,
the recommendation is not to change the authorized number of banks.
Table 1
Number of Reporters on FR 2028B and FR 2028S

District

Number of Small
FR 2028B and
FR 2028S
Respondents7

Number of Large
FR 2028B and
FR 2028S
Respondents8

1
2
3
4
5
6
7
8
9
10
11
12

0
0
0
13
3
5
33
22
13
34
13
1

0
2
3
8
2
5
2
5
4
7
9
5

TOTAL

137

52

The proposed authorized panel for the FR 2028D panel is 398 domestically chartered
commercial banks. The proposed size is based on obtaining survey results with a 95%
confidence level and 5% standard error, allowing for a 10% nonresponse rate. The panel of
banks would be a random sample of banks stratified according to the dollar volumes of
commercial and industrial loans with original amounts of $1,000,000 or less. Table 2 presents
the estimated number of reporters disaggregated by district and by bank size group for the panel. 9
7
8
9

Institutions with total consolidated assets less than $550 million.
Institutions with total consolidated assets of $550 million or more.
The estimations are based on current FR 2028A filers as of September 6, 2016.

7

The 398 potential FR 2028D respondents are distributed by size and district in proportion to the
current FR 2028A respondent distribution.
Table 2
Estimated Number of Reporters on FR 2028D Panel

District

Estimated
Number of Small
FR 2028D
Respondents10

Estimated
Number of Large
FR 2028D
Respondents11

1
2
3
4
5
6
7
8
9
10
11
12

4
3
2
6
7
15
20
13
10
14
14
2

16
55
22
22
12
24
37
14
7
16
27
36

TOTAL

110

288

Time Schedule for Information Collection and Publication
The FR 2028B and FR 2028S are filed every mid-quarter as of February, May, August,
and November. The proposed FR 2028D transmission period would begin the first business day
of the second month of each quarter (February, May, August, and November) and conclude 14
calendar days later. The Federal Reserve Bank of Kansas City would publish aggregated data for
the FR 2028B and FR 2028D in separate quarterly statistical releases on its public website.
Aggregate information on farm loans from the FR 2028B is published in the quarterly E.15
statistical release, Agricultural Finance Databook. The survey results are also included in
statistical compilations published both within and outside the Federal Reserve. Aggregate
information on small business loans from the FR 2028D would be published in a quarterly
statistical release approximately two months after the transmission period begins. The report also
would be available through a link on the Federal Reserve Board’s website.

10
11

Institutions with total consolidated assets less than $550 million.
Institutions with total consolidated assets of $550 million or more.

8

Legal Status
The Board’s Legal Division has determined that these surveys are authorized by section
11(a)(2) of the Federal Reserve Act (12 U.S.C. § 248(a)(2)) which authorizes the Board to
require any depository institution to make such reports of its assets and liabilities as the Board
may determine to be necessary or desirable to enable the Board to discharge its responsibility to
monitor and control monetary and credit aggregates. The reports are voluntary. Individual
responses reported on the FR 2028A, FR 2028B, FR 2028D, and FR 2028S are regarded as
confidential under the Freedom of Information Act (5 U.S.C. § 552(b)(4)).
Consultation Outside the Agency
From August to October 2016, the Federal Reserve Bank of Kansas City had discussions
about the FR 2028D survey instrument with senior officers at six banks. The results of these
comments were incorporated into the current proposal. On April 21, 2017, the Board published a
notice in the Federal Register (82 FR 18759) requesting public comment for 60 days on the
extension, with revision, of the FR 2028. The comment period for this notice expired on
June 20, 2017. The Board did not receive any comments. On July 10, 2017, the Board published
a final notice in the Federal Register (82 FR 31779) and the information collection will be
revised as proposed.
Estimate of Respondent Burden
As presented in the table below, the current annual reporting burden for the STL reports
is estimated to be 7,358 hours. The estimated average hours per response for the FR 2028B is
unchanged at 1.4 hours and the proposed annual burden would be 1,400 hours. Based on
feedback received from senior officers at the six banks that provided comments on the FR 2028D
survey instrument, the estimated average hours per response for the FR 2028D would be 1.5
hours. With the exception of the first survey year, the proposed annual burden would be 2,388
hours. The estimated burden for the first survey of the data collection is 3.0 hours. Allowing the
additional 1.5 hours for first-time respondent implementation cost would increase the proposed
annual burden for the first survey year by 597 hours. As a result, the net reporting burden would
decline 2,873 hours, from the current burden of 7,358 hours to 4,485. These reporting
requirements represent less than 1 percent of the total Federal Reserve System paperwork
burden.

9

FR 2028
Current
FR 2028A
FR 2028B
FR 2028S

Estimated
Estimated
average hours annual burden
per response
hours

Number of
respondents12

Annual
frequency

398
250
567

4
4
4

3.6
1.4
0.1

5,731
1,400
227
7,358

250
250
398
398

4
4
4
1

1.4
0.1
1.5
1.5

1,400
100
2,388
597
4,485

Total
Proposed
FR 2028B
FR 2028S
FR 2028D
FR 2028D (First Time only)
Total
Change

(2,873)

Based on the proposed revisions, the total cost to the public is estimated to decrease from
$403,954 to $246,227 for the FR 2028 survey.13
Sensitive Questions
These reports contain no questions of a sensitive nature, as defined by OMB guidelines.
Estimate of Cost to the Federal Reserve System
The proposed cost to the Federal Reserve System for collecting and processing the
FR 2028 is estimated to be $322,163 per year, an increase of $2,820 from the current cost of
$319,343. The one-time cost to implement the revised report is estimated to be $103,700.

12

Of the actual respondents, 137 for the FR 2028B and 110 for the FR 2028D are small entities as defined by the
Small Business Administration (i.e., entities with less than $550 million in total assets)
www.sba.gov/contracting/getting-started-contractor/make-sure-you-meet-sba-size-standards/table-small-businesssize-standards.
13
Total cost to the public was estimated using the following formula: percent of staff time, multiplied by annual
burden hours, multiplied by hourly rates (30% Office & Administrative Support at $18, 45% Financial Managers at
$67, 15% Lawyers at $67, and 10% Chief Executives at $93). Hourly rates for each occupational group are the
(rounded) mean hourly wages from the Bureau of Labor and Statistics (BLS), Occupational Employment and Wages
May 2016, published March 31, 2017 www.bls.gov/news.release/ocwage.t01.htm. Occupations are defined using the
BLS Occupational Classification System, www.bls.gov/soc/.

10