Lessons Learned: Recognizing Structural Aspects of the Main Street Lending Program that Failed to Prevent Fraud
- Issuer
- Federal Reserve Bank of Boston
- Document type
- Report
- Date
- 2024-02-23
Summary
A lessons learned report dated February 23, 2024 in which the Special Inspector General for Pandemic Recovery (SIGPR) examines aspects of the Main Street Lending Program (MSLP) that failed to prevent fraud by borrowers. It describes MSLP loan terms, including a 5-year maturity, deferred interest and principal, and 15 percent principal amortization in the third year, and states that defaults and impairments spiked sharply as loans entered their third year. The report states that a special purpose vehicle managed by the Federal Reserve Bank of Boston bought 95 percent participation in each loan, leaving lending banks little exposure, and that lenders could rely entirely on borrower certifications. It suggests future programs tie participation levels or due diligence duties to a bank's familiarity with the borrower. SIGPR calls it the first of a series of such reports.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
Full text
Lessons Learned: Recognizing Structural Aspects
of the Main Street Lending Program that Failed to
Prevent Fraud
February 23, 2024
Introduction
In this report, the Special Inspector General for Pandemic Recovery (SIGPR) focuses
on certain aspects of the Main Street Lending Program (MSLP) that failed to prevent
fraud by borrowers. Loans that were extended under the Main Street Lending Program
(MSLP) under the CARES Act during the second half of 2020 and early 2021 required
no payment of interest during their first year, and no payment of principal until the third
of the five-year term of the loan. As those loans have entered their third year—during
which borrowers must pay not only interest but also 15 percent of the principal—the rate
of defaults and impairments for nonpayment has spiked sharply. It is certainly the case
that not all of the borrowers in default committed loan fraud. But the significantly
elevated level of defaults reflects a tradeoff embedded in the structure of the Main
Street Lending Program between ensuring the creditworthiness of borrowers and
making the MSLP loans attractive to lending banks.
The Special Purpose Vehicle (SPV) managed by the Federal Reserve Bank of Boston
(FRBB) purchased participation in all but 5 percent of each loan, and thereby left
lending banks with very little exposure in the event of a default. For this reason, and
because the program purposefully required lending banks to conduct only minimal due
diligence, the Main Street Lending Program created circumstances in which lending
banks had insufficient incentive to vet their borrowers sufficiently to prevent fraud.
In saying this, SIGPR fully recognizes that the Federal Reserve Board of Governors
consciously chose to structure the program to prioritize facilitating access to credit in the
U.S. economy, and knowingly accepted that this could create an increased risk of higher
levels of fraud. SIGPR issues this “lessons learned” report on this issue solely for the
purpose of fully highlighting for Congress and the general public the trade-offs involved
in the choices that were made, so that they can make well-informed choices in the
future. This is the first of a series of reports that SIGPR intends to issue reflecting
“lessons learned.”
Background on the Main Street Lending Program
In March 2020, Congress approved the Coronavirus Aid, Relief, and Economic Security
(CARES) Act to provide over $2 trillion to individuals and businesses negatively affected
by the coronavirus disease 2019 (COVID-19). 1 Section 4003 of the CARES Act
provided $454 billion to the Department of the Treasury to make loans, loan guarantees,
and other investments in “programs or facilities established by the Board of Governors
of the Federal Reserve System for the purpose of providing liquidity to the financial
system that supports lending to eligible businesses, States, or municipalities.” 2
The CARES Act also directed the Treasury Secretary and the Federal Reserve Board of
Governors (the “Board”) to work in conjunction to establish lending facilities. The
Secretary was to “endeavor to seek implementation of a program or facility . . . that
1 Pub. L. No. 116-136, 134 Stat. 281 (2020).
2 Id. at § 4003 codified at 15 U.S.C. § 9042.
2
supports lending to eligible businesses, States, or municipalities.” 3 The Board was to
implement these facilities using the Federal Reserve’s emergency lending authority
under Section 13(3) of the Federal Reserve Act, 12 U.S.C. § 343(3). Since the
amendments effected under the Dodd-Frank Act, the Board can only create a section
13(3) facility with “prior approval” from the Treasury Secretary. 4
With the Secretary’s approval, the Board established the Main Street Lending Program
(“MSLP”), which was comprised of five facilities. These facilities provided credit to small
and medium for-profit businesses and nonprofit organizations that were financially
sound before the COVID-19 pandemic. 5 Three of these facilities were available to for-
profit businesses: the Main Street New Loan Facility (MSNLF), the Main Street Priority
Loan Facility (MSPLF), and the Main Street Expanded Loan Facility (MSELF). All three
facilities used the same “eligible lender” and “eligible borrower” criteria, and had many
of the same features, including the same maturity, interest rate, deferral of principal for
two years, deferral of interest for one year, and the ability of the borrower to prepay
without penalty. Other features of the loans extended in connection with each facility
differed, and the loan types also differed in how they interacted with the eligible
borrower’s existing outstanding debt, including with respect to the borrower’s level of
pre-pandemic indebtedness. The other two facilities were available to nonprofit
organizations and are not relevant for our purposes here. A set of Frequently Asked
Questions was posted to the websites of the Board and the Federal Reserve Bank of
Boston that provided more detailed information about the program. 6
General Structure of Main Street Loans
Main Street loans were to be originated by private banks but, to fulfill its purpose of
increasing access to credit, a special purpose vehicle (“SPV”) purchased 95 percent
participation in each loan. The SPV (MS Facilities LLC) is managed by FRBB. The
funds that were to be used to purchase the 95 percent participations in these loans
were to come from the Board’s emergency lending facility under Section 13(3) of the
Federal Reserve Act. The Treasury Department was the preferred equity member of the
SPV and agreed to provide it with $75,000,000,000 in funding to be used only in the
event of losses ultimately suffered by the SPV.
Loans that had the following features were eligible for participation by the SPV:
• 5-year maturity.
• Adjustable interest rate of LIBOR (1 or 3 month) plus 300 basis points (that is,
plus three percent). 7
3 Id.
4 Id.
5 Id.
6 See https://www.federalreserve.gov/monetarypolicy/mainstreetlending.htm for links to various versions
of the FAQs, which were periodically updated during the initial phases of the program.
7 The U.S. LIBOR was decommissioned as of June 30, 2023. According to FRBB officials, 96 percent of
Main Street loans have transitioned from a LIBOR-pegged rate to the Secured Overnight Financing Rate
(SOFR), and the remaining 4 percent transitioned to other comparable rates (such as the prime rate).
3
• Interest payments deferred for one year (with unpaid interest capitalized).
• Principal payments deferred for two years.
• Principal amortization of 15 percent at the end of the third year, 15 percent at the
end of the fourth year, and a balloon payment of 70 percent at maturity at the end
of the fifth year.
• Minimum loan size of $100,000.
• Subject to a maximum loan size that varied according to which type of loan was
extended, the size of the loan was limited to an amount that, when added to the
eligible borrower’s existing outstanding and undrawn available debt, does not
exceed certain multiples of the eligible borrower’s adjusted 2019 earnings before
interest, taxes, depreciation, and amortization (“EBITDA”). 8
• Prepayment permitted without penalty. 9
As a practical matter, because payment of principal was deferred for two years, and was
then backloaded (for example, 15 percent in each of the third and fourth years, and 70
percent in the fifth and final year), there were few defaults on Main Street loans during
the first year of the program, but the level of defaults has been sharply rising since
interest payments and then principal payments were first due in the second and third
years of the program, respectively.
Certifications Required for Main Street Loans
For a loan to be eligible for the Main Street program—and, in particular, for a
commitment by the SPV to purchase its 95 percent participation—the lender and
borrower were also required to execute standard certifications. 10 The instructions for
both sets of certifications specifically state that, “[f]or purposes of these certifications
and covenants, the Federal Reserve’s current Frequently Asked Questions (“FAQs”) on
the Main Street Facilities, as posted on the website of the Board or the Reserve Bank
as of the date hereof are incorporated by reference.” The instructions further noted that
the borrower and the lender “may rely on the clarifications and interpretations provided
in the FAQs, to the extent applicable.”
(See “Federal Reserve Lending Programs: Status of Monitoring and Main Street Lending Program,” at 32
n.38 (United States Government Accountability Office, December 2023) (available at
https://www.gao.gov/assets/d24106482.pdf).
8 The methodology used by the lender to calculate adjusted 2019 EBITDA was required to be the
methodology that it had previously used for adjusting EBITDA when extending credit to the borrower or to
similarly situated borrowers on or before April 24, 2020.
9 Main Street New Loan Facility Lender Transaction Specific Certifications and Covenants Instructions
and Guidance, available at https://www.bostonfed.org/supervision-and-regulation/credit/special-
facilities/main-street-lending-program/docs.aspx, at 2-5.
10 Versions of those certifications, along with the instructions that accompanied them are available at
https://www.bostonfed.org/supervision-and-regulation/credit/special-facilities/main-street-lending-
program/docs.aspx.
4
Borrower Certification Regarding Ineligible Businesses
The certifications that the borrower was obligated to execute 11 included the following:
Not an Ineligible Business. The Borrower must certify that, after reasonable, good
faith diligence, it has no reason to believe it is an Ineligible Business.
• For purposes of the Borrower Certifications and Covenants, an “Ineligible
Business” means a business of any of the types listed in 13 CFR 120.110(b)-(j),
(m)-(s), as modified and clarified by Small Business Administration (“SBA”)
regulations for purposes of the Paycheck Protection Program (“PPP”) on or
before April 24, 2020. Such modifications and clarifications include the SBA’s
recent interim final rules available at 85 Fed. Reg. 20811, 85 Fed. Reg. 21747,
and 85 Fed. Reg. 23450. The Federal Reserve may further modify the
application of these restrictions to Main Street.
• Reasonable Good Faith Diligence. For purposes of this certification, Eligible
Borrowers are expected to review the list of Ineligible Businesses in 13 CFR
120.110(b)-(j), (m)-(s), and make a reasonable, good faith effort to determine if
the Borrower’s activities or ownership would cause it to be classified within one of
the listed ineligible categories. If representatives of the Borrower have reason to
believe that the Borrower may be an Ineligible Business under the categories
listed in that regulation, Borrowers are expected to conduct further inquiry into
the SBA’s interpretations of such categories, including in the interim final rules,
and to reference the FAQs.
The referenced regulation (13 CFR § 120.110) is part of the rules usually applied to SBA
loan programs.
Lender Certifications
Like the borrower, the lender for a Main Street loan was required to have an authorized
officer or representative of the lender execute a set of standard certifications.
Lenders were required to certify that the terms of its loan to the borrower conformed
with the standard terms required for the 95 percent participation by the SPV. With
respect to the borrower’s certifications that it was a business and that it was
established prior to March 13, 2020, the lender was required to make “Due Inquiry with
Respect to Formation”— defined as receipt of supporting documentation from the
borrower certified by the appropriate governmental authority and having taken “those
steps to verify such formation as are required under the Lender’s ordinary underwriting
policies and procedures.”
With respect to all other aspects of the borrower’s certifications—including the
certification regarding ineligible businesses—the lender was only obligated to “certify
that each borrower has delivered to the Lender its own Borrower Certifications and
Covenants, signed by the persons identified as the signatories thereof.” The instructions
make clear that lenders had no obligation to investigate the accuracy of the
11 The certifications were required to be executed by the borrower’s principle executive officer and
principal financial officer “or individuals performing similar functions.”
5
certifications (other than those relating to formation). The instructions specifically
provide that, in certifying its receipt of the borrower’s certifications, “the Lender assumes
no responsibility for verifying the accuracy of such Borrower Certifications and may rely
entirely on the Borrower’s certifications….” The lender was not obligated to monitor the
borrower’s ongoing compliance with its covenants but was “expected to promptly notify
the SPV and the Reserve Bank if the Lender becomes aware of a Borrower’s material
breach of such covenants as a result of the Borrower self-reporting.” The purpose of
these instructions appears to have been to make clear that the lending institution was
not required to conduct due diligence with respect to the borrower’s certifications. The
instructions do not specify what, if any, notification obligation a lender had if it
discovered a borrower’s noncompliance by means other than the borrower’s self-
reporting. Nevertheless, the FAQs specifically state that, “[i]f an Eligible Lender
becomes aware that an Eligible Borrower made a material misstatement or otherwise
breached a covenant during the term of an MSNLF Loan, MSPLF Loan, or MSELF
Upsized Tranche, the Eligible Lender should notify the FRB Boston.”
The FAQs noted that lenders were responsible for determining the creditworthiness of a
borrower, and that the eligibility of a borrower (in terms of the eligibility rules in the SBA
regulations) did not necessarily mean that a borrower would be granted a loan. Lenders,
the FAQs noted, “are expected to conduct an assessment of each potential borrower’s
financial condition at the time of the potential borrower’s application.” In doing so,
lenders were to “apply their own underwriting standards in evaluating the financial
condition and creditworthiness of a potential borrower” and could “require additional
information and documentation in making this evaluation.” As set forth in FAQ F.3, it was
the bank—and not the Board or the SPV—that “[would] ultimately determine whether an
Eligible Borrower is approved for a Program loan in light of these considerations.”
Fees Associated with Main Street Loans
There were three types of fees associated with the origination of MSNLF, MSPLF and
MSELF loans.
First, lenders paid the SPV a transaction fee at the time of origination and were
permitted to pass this fee on to borrowers. For MSNLF and MSPLF loans, if the initial
principal amount of the loan was $250,000 or greater, the lender paid the SPV a
transaction fee of 100 basis points (that is, one percent) of the principal amount of the
loan. If MSNLF or MSPLF loans were less than $250,000, there was no transaction fee
to be paid. The transaction fee for MSELF loans was always 75 basis points (that is,
three-quarters of a percent) of the principal amount of the MSELF Upsized Tranche.
Second, borrowers also paid a loan origination fee to lenders (though lenders had
discretion over whether and when to charge borrowers this fee). For MSNLF or MSPLF
loans of $250,000 or greater, the borrower paid the lender a fee of up to 100 basis
points (that is, one percent) of the principal amount of the loan at the time of origination.
For MSNLF or MSPLF loans of less than $250,000, the borrower paid a loan origination
fee of up to 200 basis points (that is, two percent). For MSELF loans, the loan
origination fee was up to 75 basis points (that is, three-quarters of a percent) of the
principal amount of the loan at the time of upsizing.
6
Third, the SPV paid lenders a fee each year for loan servicing. For MSNLF or MSPLF
loans of $250,000 or more, the loan servicing fee was 25 basis points (that is, a quarter
of a percent) of the principal amount of the SPV’s participation. If the initial principal
amount of MSNLF or MSPLF loans were less than $250,000, the loan servicing fee was
50 basis points (that is, a half of a percent) of the principal amount of its participation
each year. The loan servicing fee for all MSELF loans was 25 basis points (that is, a
quarter of a percent) of the principal amount of the SPV’s participation.
Fee income that a lender could receive for Main Street loans could be significant. For
example, for an MSNLF loan of $25 million, the loan origination fee that the lender
would receive would be up to $250,000, and the lender would receive $296,875 in loan
servicing fees (the total amount equal to 0.25 percent of the SPV’s 95 percent
participation ($23,750,000) for each of 5 years).
Current Level of Defaults and Charge-offs in the Main Street
Lending Program
As part of SIGPR’s assessment of how the structure of the Main Street Lending
Program increased the risk of fraudulent borrowing, SIGPR reviewed the SPV’s “loan
loss allowance”—the “estimate of uncollectible amounts used to reduce the book value
of loans…to the amount that” the SPV expects to collect, as well as the “actual credit
losses” that the SPV has reported up through the point at which MSLP borrowers were
required to make their first payments of principal. SIGPR compared those loss levels to
those that banks with similarly sized non-MSLP loan portfolios have sustained and
found that the MSLP losses were substantially larger.
In February 2021, after the last of the MSLP loans had been made, the SPV reported
that the total principal outstanding in the Main Street Lending Program at the end of
January 2021 was $16,448,448,031. 12 In March 2021, the SPV reported a loan loss
allowance of $2,400,000,000, which was equal to 14.6 percent of the outstanding
principal.
By the end of 2022, due primarily (if not completely) to early repayment of some loans,
the outstanding principal had decreased to $10,409,503,261. In January 2023, the SPV
reported to Congress that its loan loss allowance stood at $1,400,000,000, which is
equal to 8.5 percent of the original outstanding principal.
Calendar year 2023, however, was the period during which the first principal payments
(of 15 percent) became due on MSLP loans. As of January 2024, the SPV reported total
principal outstanding of $7,313,165,772, and a loan loss allowance of $820,000,000.
This loan loss allowance represented 4.9 percent of the original outstanding principal.
12 The data about principal outstanding, loan loss allowances and actual losses that is included in this
section of SIGPR’s report was obtained from the Periodic Report that the SPV sends to Congress each
month and from audited financial reports of the SPV issued annually by KPMG. Charts setting forth some
of this data are attached to this report as Appendix A.
7
The level of actual losses that the SPV reported is even more telling about the trend line
after borrowers became obligated to repay a portion of the principal. As of the end of
August 2022 (before any principal payments were due because it was before any loans
reached Year 3), the SPV reported actual losses of $42,000,000. By the end of 2022,
reported actual losses jumped to a total of $95,000,000. And by the end of 2023, total
actual losses spiked up to $564,000,000. These figures reveal that the SPV’s actual
losses just during calendar year 2023 amounted to $469,000,000.
The significance of the percentage represented by those loss allowances and the size
of those actual losses becomes apparent when they are compared to analogous figures
at a number of commercial banks that had similarly sized portfolios of non-MSLP
loans. 13 During 2023, these banks had loan portfolios ranging from approximately
$16,005,325,000 to $17,587,254,000. The percentage of these banks’ portfolios
represented by loan loss allowances (representing anticipated losses) during 2023
ranged from 0.95 percent to 1.35 percent. By the end of 2023, the MSLP’s loan loss
allowance had dropped to $820,000,000 (from a peak of $2,400,000,000). Yet, this loan
loss allowance still represented 4.99 percent of the original outstanding principal—
nearly five times the rate of roughly comparable non-MSLP bank portfolios. 14
Likewise, the actual losses that these banks suffered during 2023 in similarly sized
portfolios of non-MSLP loans were substantially smaller than those that the MSLP
program suffered during that calendar year. The net charge offs that the banks suffered
(representing charge offs against which recoveries were set off) ranged from
$7,182,000 to $62,773,000 (from 0.04 to 0.37 percent of the portfolio). The MSLP’s
2023 actual loss of $469,000,000 is substantially larger in actual size and expressed as
a percentage (2.85 percent) of the original outstanding principal.
While defaults or losses certainly do not necessarily indicate fraud, the existence of
significantly higher levels of loss suggests the high likelihood of significantly higher
levels of borrower fraud. At a minimum, these comparisons strongly suggest that
lending banks failed to take sufficient steps to ensure the creditworthiness of many
MSLP borrowers.
13 The data about these banks was available through the Board of Governors of the Federal Reserve
System based on reports as of September 30, 2023, as well as reports available through Federal
Financial Institutions Examination Council as of December 31, 2023. A chart setting forth the identity of
the banks and some of the relevant data is attached to this report at Appendix B. Two of the banks—City
National Bank of Florida and Arvest Bank—included one or more MSLP loans in their loan portfolios. The
small amount of the portfolios represented by those MSLP loans does not affect the analysis in the text
above and, in any case, makes the comparison even more significant.
14 The conclusions here are similar to those set forth by SIGPR’s auditors in their interim report dated May
12, 2023, which is available at https://www.sigpr.gov/sites/sigpr/files/2023-05/Audit_of_the_Effects_the_
MSLPs_Loan_Losses_Have_on_Treasurys_Investment_in_the_Program--FINAL.pdf.
8
Lessons Learned: Structural Aspects of the Main Street
Lending Program that Increased the Risk of Fraud
SIGPR believes that certain structural aspects of the Main Street Lending Program
created a set of circumstances that increased the risk of fraud committed by MSLP
borrowers. The specific features of the MSLP program that heightened these risks are
(1) the very limited exposure that lenders retained after the SPV’s purchase of 95
percent participation; (2) the limited obligations placed on lender banks in terms of
vetting borrowers for the MSLP program; and (3) the extent to which the SPV played no
role whatsoever in assessing the creditworthiness of MSLP borrowers. While none of
these program features taken alone posed heightened risks of fraud, the combination
and interplay among them appears to have done so.
As the examples below demonstrate, lending banks would recover a sufficient amount
from fees and interest such that they could break even—that is, suffer no losses
whatsoever—even if a large number of their MSLP loans went into default. In terms of
the eligibility of borrowers—relating, for example, to the existence of the company, its
solvency, and the type of business in which it was engaged—lending banks had minimal
obligations. Indeed, lending banks had no obligation to ensure that their MSLP
borrowers complied with eligibility rules. And while the lending bank was obligated to
perform its customary underwriting as to the creditworthiness of its MSLP borrowers, the
SPV did little to ensure that lending banks adequately performed this function.
Moreover, other than the lenders relying on borrower certifications, and the SPV relying
on lender certifications, there was little oversight by the SPV, FRBB or the Board. To put
it simply, because the lending banks had very little to lose, they had substantially less
incentive to be diligent about ensuring that borrowers were not likely to default on MSLP
loans, and there was no entity charged with ensuring that lending banks did, in fact,
guard against fraud. While the borrower and lender certifications were good tools for
pursuing those who were later determined to have committed fraud, they may not have
been sufficient safeguards against those intending to commit fraud at the time when the
loans were made.
For example, if a lending bank made a $1,000,000 MSLP loan, it sold $950,000 of the
loan to the SPV and retained $50,000 of the loan on its own books. The lending bank
would receive an origination fee of 100 basis points (that is, one percent), which would
total $10,000. If divided equally over the period of five years, that $10,000 one-time loan
origination fee represents a return of 4 percent per year.
The lending bank would receive servicing fees paid annually totaling $11,875 assuming
the loan did not default and was fully paid at maturity. (The servicing fee was equal to
one-quarter percent of the purchased participation – that is 0.0025 multiplied by
$950,000.) This servicing fee represented an additional return of 4.75 percent per year
with regard to the portion of the loan retained by the lending bank.
9
In addition, assuming LIBOR remained at one percent, 15 the lending bank would also
receive four percent interest on the five percent of the loan it retained. That interest
would be calculated as follows:
Interest Paid
Year to Lender Calculation of Interest Paid
No interest was due in Year 1. Rather, interest of $2,000 (4
percent of $50,000) would be capitalized—that is, added to the
Year 1 $0
principal due. That portion of the principal due to the lending bank
at the end of Year 1 would be $52,000 .
Lender would receive 4 percent interest on $52,000 principal
Year 2 $ 2,080
balance at the end of Year 1.
Assuming the 15 percent principal payment was made on the last
Year 3 $ 2,080 day of Year 3, lender would receive 4 percent interest on $52,000
principal due to the lending bank at the end of Year 2.
Assuming the 15 percent principal payment was made on the last
Year 4 $ 1,765 day of Year 4, lender would receive 4 percent interest on $44,125
principal due to the lending bank at the end of Year 3.
Assuming the 70 percent principal payment was made on the last
Year 5 $ 1,450 day of Year 5, lender would receive 4 percent interest on $36,250
principal due to the lending bank at the end of Year 3.
Total
Interest
$ 7,375
to
Lender:
That would, of course, result in an additional 4 percent return per year. 16
Accordingly, in this example, the economic return to a lending bank on its retained
portion of a $1,000,000 loan that was repaid in full on schedule would total $29,250.
This represents a 12.75 percent total return (4 percent from the origination fee, 4.75
percent from the servicing agreement, and 4 percent from interest (plus any increase in
LIBOR above one percent). 17
15 Between October and December 2020, when most MSLP loans were made, LIBOR was between .081
and .083 percent. (See https://www.global-rates.com/en/interest-rates/libor/american-dollar/2023.aspx.)
Of course, to the extent that the basis for the applicable interest rate (whether LIBOR, SOFR or
otherwise) has exceeded the one percent rate we assumed in the example in the text for most of the life
of MSLP loans, the example actually underestimates the return to a lending bank. SOFR is currently over
5 percent. (See https://www.newyorkfed.org/markets/reference-rates/sofr.)
16Interest that accrues during the first year of an MSLP loan was not due and payable, but instead was
capitalized—that is, added to the outstanding balance of the loan. As a result, by the end of the first year,
the outstanding balance included the interest that accrued in that first year. However, the interest on
MSLP loans did not otherwise compound.
17 The return on such a loan would be smaller if a borrower prepaid some or all of the principal because it
would reduce the amount of interest due (if the outstanding principal were smaller) or could retire the loan
as fully repaid before the full 5-year life of the loan. On the other hand, to the extent that the basis for the
10
In contrast, if no payments ever were made on the loan (and no collateral ever was
collected), then a lending bank ultimately would lose $28,125 of its investment of
$50,000 (after collecting $10,000 in the loan origination fee from the borrower, and
$11,875 in the loan servicing fee from the FRBB). 18
These figures are relatively rough approximations but are nevertheless instructive. A
lending bank that made one MSLP loan that defaulted with no payments of interest or
principal and one MSLP loan that was fully repaid at maturity would come out ahead: It
would lose $28,125 on the defaulted loan but would make $29,250 on the compliant
loan. In short, in order to make a profit on its portfolio of MSLP loans, a lending bank
needed only half of those loans not to default. This unquestionably had the potential to
make lending banks more willing to make MSLP loans to borrowers who were less
creditworthy than the banks might have otherwise required. 19
SIGPR believes that it is no coincidence that the higher default rates (which may reflect
higher rates of fraud) occurred in a program structured the way the MSLP is. Indeed, it
is clear that the SPV anticipated higher rates of default than those found in commercial
lending; that explains why the loan loss allowance set at the very beginning of the
program represented 14.6 percent of the outstanding principal—many multiples of the
rate found in non-MSLP portfolios. The critical question is why that was the case.
During its investigations, SIGPR discovered a number of factual scenarios that further
reflected the ways in which the structure of the Main Street program failed to protect
adequately against borrower fraud. For example, in one instance, a lending bank denied
an MSLP loan to a borrower because its EBITDA 20—which the Main Street program
used to determine the amount of permissible loans—was too low, but then granted the
loan to that same borrower based on a different EBITDA calculation. In another
investigation, SIGPR determined that a lending bank had cut ties with two of its
directors after learning that the two directors controlled an entity that purchased a loan
made by that lending bank, but the lending bank still gave an MSLP loan to an entity it
knew was controlled by two other directors of the lending bank. In yet another instance,
a lending bank made an MSLP loan to one borrower with knowledge that the borrower
would use that money to repay a separate commercial loan that that lending bank had
made to a different entity controlled by the same individuals as the MSLP borrower. The
applicable interest rate indeed substantially exceeded 1 percent during the life of most MSLP loans, that
fact has resulted in increased returns to the lending banks for those loans.
18 It is our understanding that, in the event of a default, a lending bank would continue to collect its
servicing fee each year as it assisted with efforts to collect the loan. The statement that a lending bank
would collect all of the servicing fees even in the event of a default assumes that those efforts continued
up to the full maturity date of the loan.
19 Similar analysis of MSLP loans under $250,000 yields even more extreme conclusions due to the
higher fees (on a percentage basis) that lending banks received for those loans. If a lending bank made 8
MSLP loans of $200,000 that defaulted, it would still break approximately even if it made a single MSLP
loan of $200,000 that was fully paid at maturity. As it happens, only 12 MSLP loans were made in
amounts less than $250,000, and no lender made more than two of them. Nevertheless, this analysis
corroborates the way in which the fee structure gave lending banks significantly diminished incentives to
take greater care with MSLP loans.
20 EBITDA stands for earnings before interest, taxes, depreciation, and amortization, and is an alternative
measure of profitability to net income.
11
bank did so despite language in a number of publications associated with the loan
program that stated that MSLP loans were not intended to shift the risk of existing loans
onto the FRBB or the Treasury.
SIGPR believes that a higher risk of fraud was inevitable in a system in which lending
banks had little “skin in the game” and few obligations to conduct genuine due diligence
of the type they likely perform with respect to non-MSLP loans. The very small portion of
the loans that MSLP lending banks retained (5 percent), together with the fees the
banks earned for origination and servicing, meant that a lending bank could make a
good profit on a successful MSLP loan but would lose very little if such loans went into
default. This gave banks insufficient incentive to look critically at the background and
creditworthiness of MSLP borrowers. SIGPR’s investigations have revealed that, in
many cases, lending banks extended MSLP loans to borrowers who were not existing
customers. In those case, the banks knew very little about their borrowers. While the
MSLP rules required lending banks to undertake the same “Know Your Customer” and
underwriting that they customarily performed, there was little or no oversight to enforce
compliance with those obligations. And it appears that, in many cases, compliance with
these obligations was incomplete or entirely absent.
It is important to note that SIGPR does not claim to be able to identify a specific
percentage of participation that would eliminate the higher risk of fraud we discuss in
this report. The point is, however, that 5 percent appears to have been too little.
One possibility would be to link the level of the SPV’s participation to a lending bank’s
existing familiarity with the prospective borrower. It could be that a bank would not be
permitted to make a loan in a future program to a borrower with whom the lending bank
does not have pre-existing familiarity. An alternative could be that the level of
participation could be adjusted based on whether or not the borrower is already known
to the lending bank. A different alternative could be to impose greater due diligence
obligations on lending banks in the case of borrowers with whom the lending bank has
no pre-existing familiarity. Any of these changes would serve to increase the incentive of
lending banks to make loans backed up by taxpayer money only to borrowers that the
lending banks have appropriately determined to be eligible and creditworthy. Such a
change would certainly serve to reduce waste, fraud and abuse in future loan and
assistance programs by the federal government.
12
File and source
- File
- REPORT_SIGPR_lessons-learned-mslp-structural-aspects-failed-prevent-fraud_2024-02-01.pdf
- Size
- 462,685 bytes
- SHA-256
- 1aa8dee2aa16fcb743fedae763a54f3773606b53bfa9ac8a57098900d5ebd161
- Original
- www.sigpr.gov