Designing the Main Street Lending Program: Challenges and Options — Hutchins Center Working Paper #64
Summary
Hutchins Center Working Paper # 64, Designing the Main Street Lending Program: Challenges and Options, dated June 18, 2020, is by William B. English and J. Nellie Liang. The introduction says that, unlike the 2008 financial crisis, the current crisis reflects fundamental cash flow problems for many businesses as revenues have almost completely stopped. It says the Main Street Lending Program (MSLP) is set up to provide loans to small and mid-size firms and large below-investment-grade firms that were financially sound before the onset of the pandemic. It says the CARES Act authorizes the Federal Reserve to establish the program under its emergency authorities with capital provided by the U.S. Treasury, and calls the program a big step for the government and a difficult one.
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
- Issuer
- Source documents
- Document type
- Designing The Main Street Lending Program Challenges And Options
- Case
- Designing The Main Street Lending Program Challenges And Options
Cite this document
“Designing the Main Street Lending Program: Challenges and Options — Hutchins Center Working Paper #64.”
Original at www.brookings.edu. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/designing-the-main-street-lending-program-challenges-and-options/
, SHA-256 67863aa5d92a860b4f99715b65d50271ea6526722727c3642efacd4a322d266e.Full text
Hutchins Center Working Paper # 64
J u n e 2 0 2 0
Designing the Main Street
Lending Program:
Challenges and Options
William B. English
J. Nellie Liang*
June 18, 2020
*William B. English, Professor in the Practice of Finance, Yale School of Management,
william.b.english@yale.edu, and J. Nellie Liang, Senior Fellow, Hutchins Center for Fiscal and Monetary
Policy, Brookings Institution, jnliang@brookings.edu.
The authors thank Scott Alvarez, Lauren Anderson, Francisco Covas, Betsy Duke, Trish Mosser, Jamie
McAndrews, Bill Nelson, Simon Potter, Jeremy Stein, and David Wessel for helpful comments and suggestions,
and Manuel Alcalá Kovalski for helpful research assistance. They also thank Tim Guinnane for unstinting help
with the documents describing the German guarantee programs. All remaining errors are theirs.
The authors did not receive financial support from any firm or person for this article or from any firm or person
with a financial or political interest in this article. Neither is currently an officer, director, or board member of
any organization with an interest in this article.
THIS PAPER IS ONLINE AT
https://www.brookings.edu/research/designing-the-
main-street-lending-program-challenges-and-
options
1. Introduction
Unlike the 2008 financial crisis, the current economic crisis brought on by the COVID-19 pandemic
reflects fundamental cash flow problems for many businesses as revenues have almost completely
stopped. Businesses will need substantial financial resources—from previous saving, direct government
grants, or credit—to pay bills, survive the shutdown, and be ready to rehire workers quickly and restart
spending. Extending credit can help some businesses manage the near-term shortfall in revenues,
restructure operations, and prevent unnecessary failures at a time when bankruptcies will be costly.
The Main Street Lending Program (MSLP) is set up to provide loans to small and mid-size firms and
large below-investment-grade firms that were financially sound before the onset of the pandemic. The
CARES Act authorizes the Federal Reserve to establish the program under its emergency authorities with
capital provided by the U.S. Treasury.1 However, the MSLP is a big step for the government, and a
difficult one. Lending to risky firms is a significant challenge given U.S. aversion to government equity
stakes in private businesses and the Fed’s legal requirements to be secured to its satisfaction and to lend
only to solvent firms.
The program was announced in April and has been revised twice in response to feedback. Domestic
borrowers with fewer than 15,000 employees and less than $5 billion in 2019 revenues are eligible to
apply. Loans will be made by banks, and they will retain 5 percent of the loan and sell the remaining 95
percent to one of three Main Street facilities—the New Loan Facility, the Priority Loan Facility, and the
Expanded Loan Facility (see table 4 below for details). These facilities vary by the type of loan they will
accept, determined mainly by loan size, borrower leverage, and whether the loan is new or expands an
existing loan. All Main Street loans have a five-year maturity, defer interest payments for one year, defer
principal payments for two years, can be prepaid without penalty, and have a loan rate of LIBOR plus 3
percentage points. Borrowers commit to limits on executive compensation, shareholder distributions, and
employment.
Small-to-mid-sized businesses that could borrow from the Main Street program are an important part
of the economy. There are more than 111,000 firms with between 100 and 10,000 employees (Census
Bureau, 2017), and they employ 48 million people, about 38 percent of the aggregate. These firms are
more reliant on loans than public bond markets, and some are too large for the Payroll Protection Program
(PPP) which provides SBA-guaranteed loans with possible forgiveness, or too risky for programs
established by the Treasury and Fed to support investment-grade corporations that issue commercial
paper and corporate bonds.
The Main Street program will only succeed if it actually provides credit directly to borrowers in need.
The mere announcement that Main Street loans are available does not provide the same support to the
loan market as the Commercial Paper Funding Facility (CPFF) does for the commercial paper market and
the Primary Market Corporate Credit Facility (PMCCF) does for the corporate bond market. The CPFF
and the PMCCF offer credit at a penalty rate to high-rated firms that have recently issued commercial
paper, or that issue in the corporate bond and syndicated loan markets. By ensuring these firms access to
credit even at a higher than normal market rate, they reduce rollover risk and risk spreads in the markets,
even before any firms access the facilities. By contrast, because the potential borrowers under the MSLP
differ very widely in credit quality, the program cannot provide a guaranteed backstop, and so does not
ease credit conditions broadly.
The Main Street program purchases loan participations from banks, rather than providing low-cost
funds to banks or providing loan guarantees. It is designed to encourage banks to make loans they might
not make on their own, primarily by easing balance sheet constraints. We describe the economic
challenges in designing a loan support program and evaluate the Main Street program in terms of how it
1
See the descriptions of the programs at https://www.federalreserve.gov/monetarypolicy/mainstreetlending.htm.
2
manages significant asymmetric information, adverse selection, poor targeting, and moral hazard
problems while protecting taxpayer funds. The calibration of design features requires a difficult balance,
but banks and businesses to date are showing little interest in the existing program.2 Banks generally say
that these are loans they would make anyhow. Potential borrowers say the loans are too expensive and
burdensome.
We propose some changes to increase usage of the program and its odds of success. While we believe
that most of the recent changes to the original program go in the right direction, the current program could
be made more attractive if loan terms were more tailored to credit quality, the rigidity of fixed debt
repayments was reduced, and lenders received more compensation for taking additional risks.
• All Main Street loans have the same risk spread, regardless of borrower credit quality. We
recommend that loans to borrowers with higher credit quality should have a lower spread than
loans to borrowers that are more leveraged. Less-leveraged borrowers in the New Loan facility
should pay a lower spread than more-leveraged borrowers. Moreover, these borrowers should
have more streamlined documentation requirements.
• All loans have a maturity of five years, with payments of principal starting in the third year. We
recommend that loans have longer maturities and more delayed repayments but with incentives
for firms to repay earlier added. This change would prevent a hard debt repayment for firms that
may need more time to reorganize and survive, especially given that great uncertainty remains
about the path of the virus.
• The minimum loan size for the New Loan and Priority Loan Facilities is $250,000. If the Fed and
Treasury find that there are many small firms that either cannot access the PPP program or
require additional support, we recommend they reduce the minimum loan size further. Such a
change could also encourage smaller banks to participate in the program. The Fed has already
reduced loan size twice, from an initial $1 million.
• The bank participation share is 5 percent for all loans, regardless of borrower credit quality. We
recommend that banks in some cases be allowed to extend program loans to riskier borrowers if
they retain a larger share of the loan to demonstrate their confidence in the credit. The Federal
Reserve recently eliminated such a trade-off in the Priority Loan facility, but we believe it should
be reintroduced, creating greater flexibility in the program.
• Borrowers are required to make “commercially reasonable efforts to maintain its payroll and
retain its employees.” We recommend that the program should ease or remove the constraints on
borrowers’ behavior that would make it difficult for them to reorganize their businesses to
respond to new supplier and customer behaviors in a post-COVID-19 economy. If constraints are
required, policymakers should quickly clarify how “commercially reasonable” will be assessed,
so that this uncertainty does not limit take-up.
• Banks get a loan origination fee, servicing fees, and 5 percent of the loan spread. However, the
current fees do not appear to provide banks enough incentive to make loans they might not make
anyway, given that the risk of Main Street loans will be highly correlated with their existing
loans, which have already deteriorated as a result of the pandemic. We recommend that
policymakers consider a higher level of fees to spur banks’ willingness to use the program. In
addition, the program may need to provide additional compensation for loan workouts,
particularly if the coronavirus weighs on the economy for longer than expected, increasing the
need for banks to workout troubled loans.
2 For example, some Senators raised concerns at a hearing on implementation of the CARES Act on June 2, 2020,
that the Main Street program will not be used (see Ransom 2020). For similar concerns, see Mohsin and Scigliuzzo
(2020) and Elliott, DePillis, and Kiel (2020).
3
These changes would pose additional risk to the funds that Treasury has provided to back this
program, but we see those risks as likely justified in the current circumstances. Even with the
recommended changes, the program may have limited demand, since many businesses need equity, not
more credit, but equity cannot be provided in a lending program that would be offered by the Federal
Reserve. In addition, banks may not want to make more loans without even greater protection, given the
uncertain economic outlook. While the program could become more attractive to banks if the economy
remains very weak and loan losses mount further, putting pressure on banks’ balance sheet capacity,
banks likely will still not be anxious to take on more risk in that scenario.
However, it is critical to support businesses now. The downturn is very deep, the risk of permanent
harm to labor markets because of protracted high unemployment is high, and business activity is likely to
resume more slowly than was thought when the CARES Act was signed. While the structure of the Main
Street Program has significant benefits, particularly the use of the Fed to leverage Treasury capital,
policymakers should be prepared for the possibility that an aggressive lending program consistent with
the CARES Act may not be sufficient to help many businesses. In that case, Congress should be prepared
to authorize other types of programs as well, such as loan guarantees or forgiveness, which combined
with additional lending could better promote a more rapid recovery in employment and output and limit
long-run damage to the economy.
The remainder of this paper proceeds as follows: Section II describes characteristics of potential
MSLP borrowers and why private lending is not sufficient, Section III describes constraints for the Fed
and Treasury, and information- and incentive-related challenges when designing a government lending
program, and Section IV presents three alternative types of government programs to increase lending.
Section V outlines the MSLP program and how it addresses the main design challenges, and Sections VI
and VII present our assessment and recommendations to improve the odds of success.
2. Why a government program for these businesses?
The government may want to help provide credit for small and medium-sized firms, as well as larger
firms that are not investment-grade, for the same reasons that it is providing support for small businesses
(through the Paycheck Protection Program) and larger, investment-grade businesses (through the Fed’s
commercial paper and corporate bond programs). That is, many firms with large declines in revenue
during the health emergency could fail and cut employment, which would slow the economic recovery.
Private lending may be insufficient in this case because the social benefits from a more robust
recovery and fewer inefficient bankruptcies will exceed the private benefits to banks from more loans.
Banks may be wary of lending more because they don’t want to expand their balance sheets or because
they do not want to increase risky loans. Moreover, the risk on such loans is highly correlated with that of
other similar loans already on banks’ books and this correlation is likely to be especially high, as the
progress of the coronavirus will be an important driver of defaults on a wide range of both business and
household credit.
We discuss below the employment and risk of potential Main Street borrowers, and reasons why
private lending to them might be insufficient.
Employment of small- and medium-sized businesses. Businesses that might access the MSLP
employ a significant fraction of the labor force. Data from the Census Bureau show that in aggregate
there were 128.6 million employees in almost 6 million firms in 2017 (Table 1). The vast majority of the
firms are small with fewer than 100 employees, and combined they employ 33 percent of the aggregate
workforce. But there are about 111,000 firms that have between 100 and 10,000 employees each, and
they collectively have 48 million employees, 38 percent of the aggregate. Annual payroll expenses at
these firms represent 40 percent of the aggregate. That is, the set of firms that fall between very small
4
businesses, those with less than 100 employees and the very large firms with more than 10,000
employees, represent roughly 40 percent of employment and payroll. Note that these data can only
approximate what firms are eligible for MSLP because data are not provided for firms with between
10,000 and 15,000 employees, and revenues of these firms are not available by number of employees.
While firms with between 100 and 500 employees are eligible for the PPP program, they may also be
interested in a MSLP loan. These firms had an average annual payroll expense of $9.9 million, so they
could make use of a Main Street loan, given the minimum loan size of $250,000. Their average payroll
expenses are considerably higher than the average payroll of $305,000 for firms with fewer than 100
employees. Indeed, the smaller firms may find PPP loans sufficient to their needs; the average PPP loan
made is $114,764, though 64 percent of the total number of loans were for less than $50,000.3
Table 1. Firm Employment by Size4
Firm Employment Number of Employment Share of Annual Share of Payroll per
Size Firms Employment Payroll ($Th) Payroll Firm ($th)
Total 5,996,900 128,591,812 100% 6,725,346,754 100% 1,121
<5 3,698,086 5,937,081 276,569,783 75
5 to 99 2,186,317 36,507,469 1,520,676,007 318
Subtotal < 100 5,884,403 42,444,550 33% 1,797,245,790 27% 305
100-499 92,358 18,111,531 914,291,189 9,899
500-999 10,082 7,000,139 369,884,610 36,688
1,000 to 2,499 5,887 9,042,640 520,748,082 88,457
2,500 to 9,999 3,070 14,253,746 889,750,821 289,821
Subtotal 100-
9,999 111,397 48,408,056 38% 2,694,674,702 40% 24,190
10,000 + 1,100 37,739,206 29% 2,233,426,262 33% 2,030,388
Source: Statistics of US Businesses (SUSB), Census Bureau, 2017. SUSB total excludes the self-employed,
employees of private households, employees in the railroad and agricultural production sectors, and most
government employees.
Risk of potential Main Street borrowers. Data on the riskiness of the debt of small-to-mid-sized
firms that might access the Main Street Program are fragmentary. But it is clear that these firms are much
less risky than the smallest firms, those eligible for SBA-guaranteed loans, but riskier than firms that are
rated investment-grade. Some of the larger firms that are not investment-grade that might want to access
3
While we don’t know the characteristics of firms that received PPP loans, it appears they are mainly among the
5.88 million very small firms that would have fewer than 100 employees, with an average number of employees of 7
and average payroll of $305,000. The number of PPP loans reached 4.45 million by May 23, 2020, for total funds of
$510.5 billion. In terms of the distribution of loan sizes, 64 percent of the total number of loans were less than
$50,000, another 30 percent were between $50,000 and $350,000, and only 6 percent were above $350,000.
4
Table 1 and related text were revised to correct an error in the number of firms with between 100 and 10,000
employees.
5
the MSLP may be speculative-grade risks, which tend to have higher default rates than the small-to-mid-
sized firms because they have a need and the ability to raise large amounts of debt.
Data on debt-to-asset ratios and failure rates reflect this risk profile. The Federal Reserve’s Financial
Stability Report (May 2020) shows the debt-to-asset ratios of commercial and industrial (C&I) loan
borrowers (from the banks that are required to undergo the Dodd-Frank stress tests, and hence file form
FR Y-14Q). Borrowers that are not publicly traded can be used to approximate the borrowers that might
be interested in Main Street loans (figure 1). As shown, these firms have somewhat higher debt-to-asset
ratios than publicly traded firms that have C&I loans. The debt ratios of these firms have been rising in
the past few years, though less rapidly than at publicly traded firms.
Comparing these ratios to those for all nonfinancial publicly traded companies, a debt-to-asset ratio of
34 percent in 2019 for non-public C&I loan borrowers is much lower than for speculative-grade firms,
labeled as risky firms, which have had ratios of above 40 percent in recent years. These data suggest that
many potential Main Street borrowers are less risky than speculative-grade firms but more risky than
investment-grade companies.
Figure 1. Borrower leverage for Bank C&I loans and publicly-traded nonfinancial corporations
Source: Federal Reserve Board Financial Stability Report, May 2020 and May 2019.
Loan delinquency rates also suggest smaller business loans will have higher default rates than
investment-grade bond issuers, but may have lower default rates than speculative-grade bond issuers.
PayNet reports, based on data for C&I loans provided by a sample of commercial banks, that default rates
for small businesses averaged about 3 percent over 2006-2019, higher than the 1.7 percent default rate on
all C&I loans at commercial banks, which includes loans to large publicly-traded firms.
For comparison, the average default rate on investment-grade corporate bonds was 0.12 percent from
1981-2017, but was 4.3 percent for speculative-grade corporate bonds over the same period (Standard and
Poor’s). This average default rate for speculative-grade firms is higher than that for loans to small-to-mid-
size borrowers that would access the MSLP.
In addition, data from the Census suggest there is a wide range in risk characteristics in the category
of small- to mid-size firms (table 2). The overall business failure rate peaked at 9.67 percent in 2009,
6
dominated by the nearly 5.9 million small firms, those with less than 100 employees, which had a failure
rate of 9.86 percent in 2009. But the failure rates drop off dramatically as firms get larger. The average
failure rate for businesses with between 100 and 500 employees was 0.99 percent in 2009, and it was only
0.35 percent for businesses with between 500 and 9,999 employees. These sharp distinctions by size
group in 2009 are also evident for average failure rates during the twenty-year period 1994-2014.
Table 2. Business Failure Rates
Firm Employment Number of Firms Failure rate Failure rate
Size 2014 2009 (%) 1994-2014 (%)
Total 5,060,326 9.67 8.33
< 100 4,955,252 9.86 8.48
100 to 499 84,541 0.99 0.98
500 to 9,999 19,163 0.35 0.30
Subtotal 100-9,999 103,704 0.88 0.86
Source. Bureau of Census, Longitudinal Business Database 1977-2014.
Overall, the loan default and failure data suggest that lending to small- to mid-sized businesses with
more modest needs for debt is less risky than lending to firms that issue speculative-grade bonds, though
both are riskier than lending to investment-grade companies. Moreover, the losses on defaulted loans,
which are generally more senior, tend to be lower than those on defaulted speculative-grade bonds.
Why might private lending not be sufficient? Given that these businesses are significant employers,
policymakers would like to ensure that those that are harmed temporarily but can succeed after COVID-
19 risks diminish have access to the credit they need to survive this difficult period. Unfortunately,
however, a substantial fraction of banks have recently tightened their lending standards on loans to
businesses, according to the April 2020 Senior Loan Officer Opinion Survey, suggesting that private
lenders may not be willing and able to provide the needed credit (figure 2).
7
Figure 2. Lending standards for C&I loans at commercial banks
There are four reasons why policymakers may believe that lending by private lenders to these firms is
likely to prove insufficient. The main reason is that the social benefits from a more robust economy and
fewer inefficient bankruptcies will exceed the private benefits to banks and borrowers from more loans.5
That is, there are net positive externalities which arise by promoting a more rapid recovery in output and
employment, given that levels currently are so depressed, and so encouraging a positive cycle of more
spending and more employment. In addition, banks on their own do not capture the aggregate benefits to
society if all banks lend more and a worse aggregate economic outcome is avoided. Moreover, additional
lending may avoid a negative externality from having a large number of firms go bankrupt in a short
interval. The bankruptcy system could be overwhelmed, and the result could be that some firms that could
have been reorganized in bankruptcy end up in liquidation, with a consequent social loss.6
A second reason there may be insufficient lending is that banks’ funding costs may be too high
because of dysfunction in term funding markets. Higher funding costs limit the profitability of lending,
and a government program that provides lower cost funding could be helpful. The spread of the 3-month
LIBOR to OIS was elevated following the start of the pandemic, suggesting funding pressures for some
banking institutions, although that spread has fallen back more recently.
A third reason to be concerned about access to credit is that banks may lack the risk appetite to make
additional loans. Banks no doubt already are expecting a big increase in loan losses because of the
5
We refer to all lenders as “banks” here. The Fed’s Main Street Lending Program is only open to banking
organizations. If bank balance sheet concerns are limiting banks’ willingness to lend, then allowing nonbanks to
have access to the program could boost take up. Of course, the nonbank lenders would have to be vetted to be sure
that they had the ability to do the underwriting (a bit like the PPIF evaluation process—see Treasury (nd)).
However, given the many potential complications involved in the program, the Fed may see it as helpful to operate
through entities overseen by the federal banking regulators.
6
Miller and Stiglitz (2010) have argued that in financial crises there may need to be a temporary “Super Chapter 11”
to quickly deal with a large number of bankruptcies in a way that limits harm to the economy.
8
economic slowdown due to the virus, and the risks of new loans to borrowers harmed by the virus will be
highly correlated with the existing risks. Banks may also be concerned about downside risks to loan
quality given the elevated uncertainty about the economic growth trajectory, especially in the case of
smaller and riskier firms. That uncertainty may make them even less willing to lend.
Finally, banks may face balance sheet constraints that limit their willingness to lend. The constraints
could reflect capital regulation or pressures from creditors and counterparties. Banks have been operating
with capital above regulatory and supervisory requirements, with the largest banks operating with a Tier 1
common equity (CET1) ratio of 12 percent on average in 2019, but loans rose sharply in March and April
as businesses drew down existing credit lines (figure 3).7 Banks announced substantial loan loss
provisions in the first quarter, though the full effects of loan losses on capital in coming quarters are still
unknown. If a bank’s CET1 ratio were to fall by enough that it fell into its regulatory buffers, such as the
capital conservation buffer and GSIB capital buffer, it would be required to cut dividends, share
repurchases, and compensation.
Figure 3. Change in C&I loans at commercial banks
Moreover, new lending could push some of the largest banks into the next size category used to
calculate the SIFI capital surcharge since size is one of the five components determining the size of the
surcharge. Moving into the next category would raise its capital requirement by 50 basis points. This
7
The Federal Reserve estimates that business borrowers drew down significant amounts of their committed credit
lines in March and April. C&I loans increased by about $660 billion in those two months, and the Fed estimates that
only a little more than half were by firms with investment-grade ratings. Federal Reserve Financial Stability Report,
May 2020, Box “Risks Associated with Banks’ Corporate Credit Exposures through Credit Lines.”)
9
nonlinear effect may restrain banks’ willingness to lend to marginal customers. As a consequence, banks
may be unwilling to meet the demand for credit even from less risky firms.8
3. Challenges when designing a business lending program
The design of a program to support bank lending to businesses in the current health crisis can be thought
of as the solution to a constrained maximization problem. The objective is to increase loans to borrowers
that can survive the health crisis and succeed after the crisis passes, but that would not otherwise be able
to get a loan in the current environment. The constraints on the program include the legal restrictions on
the Federal Reserve’s emergency lending under section 13(3) of the Federal Reserve Act, as well as the
ability and willingness of the Federal Reserve and Treasury to absorb losses in the conduct of the
program. If policymakers are willing to take on more risk of loss, then they are more likely to succeed in
boosting lending to the desired borrowers. The problem is complicated, however, by several information-
related problems that policy makers face, including asymmetric information, adverse selection, inefficient
targeting of the support, and moral hazard.
Constraints on the Federal Reserve and Treasury. Section 13(3) of the Federal Reserve Act puts a
number of conditions on the ability of the Federal Reserve to operate an emergency lending program.
These conditions include (Federal Reserve 2020a):
• The circumstances must be “unusual and exigent;”
• The program must be approved by the Secretary of the Treasury;
• Eligibility for the program must be “broad-based,” which the Federal Reserve has judged to mean
that the program is open to at least five firms (Federal Reserve, 2015);
• Program borrowers must be “unable to secure adequate credit accommodations from other
banking institutions;”
• Borrowers under the program must be solvent, and program lending cannot assist the borrower in
avoiding bankruptcy;
• Loans under the program must be “[e]indorsed or otherwise secured to the satisfaction of the
[lending] Reserve Bank” and the Fed must take steps to ensure that the “security for emergency loans
is sufficient to protect taxpayers from losses.”
In addition, the CARES Act requires that loans made under 13(3) programs in the current crisis
cannot be forgiven. This requirement sets the 13(3) lending apart from the Paycheck Protection Program
because PPP loans are intended to be essentially grants. With no loan forgiveness allowed, the 13(3) loans
will be much less expensive for the government, but also much less attractive for the borrowers.
The first three of the 13(3) conditions are clearly satisfied in the current circumstances. The fourth is
satisfied by a Fed requirement that borrowers certify their difficulty in obtaining credit from sources other
than the Federal Reserve, and the fifth condition should be satisfied by any program aimed at firms that
can survive the crisis if they obtain credit, which is the aim of the program. The final condition poses
8
Note that balance sheet constraints could be eased, at least partly, with regulatory changes. Indeed, federal
regulators have provided some temporary changes in regulation to address such constraints. Leverage ratios for
BHCs and bank subsidiaries were just relaxed, by permitting the deduction of Treasury securities and deposits at the
Fed from exposure measures. In addition, under the new stress test capital buffer, they are not included as a
minimum regulatory requirement. Also, loans under the MMLF and PPPLF are excluded from leverage ratio
calculations. However, while regulators can take steps to ease these constraints, they cannot make banks lend.
10
greater difficulty, since business loans involve significant risk, as noted earlier, and that risk is
presumably increased in the current circumstances. However, the Congress, in the CARES Act, provided
the Treasury with $454 billion that it can use to provide security for loans made under the Federal
Reserve’s 13(3) lending programs. The decision as to how much capital to allocate to each lending
program is up to the Treasury Secretary, who has the authority to allocate the funds authorized in the
CARES Act.
The Treasury has considerable influence over an emergency lending program because the Fed must
obtain the permission of the Treasury Secretary to establish any emergency lending program under
section 13(3) and because the decision by the Treasury Secretary to allocate capital to a lending program
constrains the size of the program given the 13(3) conditions, and allows the Treasury Secretary to
condition the provision of capital on program terms that he judges acceptable, given his willingness to put
the funds allocated at risk. Thus far, Treasury Secretary Mnuchin has indicated that he has only a limited
tolerance for possible losses. In April, the Secretary noted that some of the Fed’s programs would make
money while others would lose money, but he indicated that, across the programs, “in a base case
scenario…we recover our money.” He added that “If Congress wanted me to lose all the money, that
money would have been designed as subsidies and grants as opposed to credit support” (Davidson and
Rubin 2020). At a subsequent Congressional hearing, however, the Secretary seemed somewhat more
willing to risk taking losses, stating that “our intention is that we expect to take some losses on these
facilities…That is our base case scenario” (Timiraos and Davidson 2020). Based on these comments, the
Secretary appears to be willing to commit capital to a program with a modest probability of taking losses,
but he appears averse to committing tax payer funds to back a program that is highly likely to have
significant losses. Given the riskiness of business lending, particularly to small and medium-sized firms,
this aversion to losses places significant constraints on the design of a program to support such lending.
In part, expectations for losses on 13(3) lending programs in this crisis may reflect the outcome
following the financial crisis of 2007-09. In that period, the Federal Reserve extended $600 billion of
loans under Section 13(3) of the Federal Reserve Act. All of the loans were repaid in full with interest. As
a result, the Fed actually made $9 billion on its broad-based emergency lending programs (Federal
Reserve 2010). For the Troubled Assets Relief Program (TARP), the government earned $29.5 billion
from its commitments of $269.6 billion to financial firms and the credit market programs (Liang,
McConnell, and Swagel, 2020).
However, the current situation is very different. In 2007-2009, the main problem was a loss of
confidence in the financial sector and a consequent run on virtually all forms of short-term wholesale
funding. In such a situation, the Federal Reserve could, by committing sufficient funds, slow and
ultimately reverse the runs, and as confidence returned, the loans made by the Fed would be repaid. By
contrast, the current situation is characterized by a health crisis that is forcing workers to stay home and
businesses to close. The result is a massive loss of revenue for many businesses that is likely to leave
many of them insolvent once the health crisis passes. Because the Federal Reserve can only provide
credit, and not transfers, there are many businesses that the Fed’s programs cannot help. As a
consequence, aggressive lending is likely to lead to significant losses in this case.
Information- and incentive-related challenges. In addition to the legal and financial constraints on
the Fed’s 13(3) lending, a business lending program must be designed to address four information- and
incentive-related challenges. First, there is the problem of asymmetric information. Borrowers have
private information about their prospects—for example, their likely viability once the coronavirus has
ebbed—that the Federal Reserve and Treasury do not have. For larger public companies, considerable
information can be obtained at relatively low cost, but for smaller and medium-sized firms, there is little
information that is publicly available. Banks, through their customer relationships and long experience
underwriting business loans, can make well-informed judgements about the outlook for potential
borrowers. By contrast, the Federal Reserve and the Treasury have little experience in this area, and so a
business lending program that requires the Federal Reserve and the Treasury to do the loan underwriting
11
is likely to have a high level of losses. As a consequence, it may be beneficial to design the program to
incentivize banks to participate and provide the loan underwriting needed to limit risk.
However, using banks to do the underwriting of program loans could lead to a second problem
involving the behavior of banks—adverse selection. Specifically, banks may have an incentive to keep
lower-risk loans on their own books and only pass to the program those loans that are high risk. Within
limits, such behavior is desirable, since it should help provide additional credit to the intended firms.
However, if the program design is not calibrated appropriately, banks may not have an incentive to limit
the risk of the loans that are made under the program, imposing very high losses on the Fed and the
Treasury without a compensating macroeconomic benefit. Moreover, if the spread on program loans is
very wide, then only very risky borrowers will be willing to participate, which could also result in
excessive losses.
Some program designs, perhaps responding to the possibility of adverse selection, could suffer from a
third problem—poor targeting. In this case, the borrowers receiving loans under the program may be
those that do not need the support to survive the health crisis and continue to operate. Instead, banks may
make loans that they would have been willing to make without the program, but then pass them to the
program to benefit from fees or other incentives. While such loans would likely be low risk, and so would
not impose significant costs on the Treasury, they could crowd out loans to the firms for which the
program is intended, reducing its effectiveness.
Finally, even an effective business lending program undertaken during the coronavirus crisis could
lead to subsequent moral hazard. The Federal Reserve has been very public for many years in its concern
that outsized borrowing by lower-rated nonfinancial firms and associated excesses in the syndicated loan
market were resulting in fragilities that would act as an accelerant in a downturn. If the Federal Reserve
now provides significant ex-post support to firms and lenders that it has been criticizing, managers and
investors could take the lesson that the Fed’s warnings about financial stability risks can be dismissed and
that ex-post support for imprudent risk taking will be forthcoming in future downturns (Federal Reserve
Financial Stability Report, May 2019). The result would be greater risks to financial stability once the
current crisis has passed.
Of course, even highly leveraged firms have employees and suppliers, and so addressing the moral
hazard problem by limiting lending to such firms would have macroeconomic costs. However,
policymakers may judge that the larger, more leveraged firms are likely to have management with the
skills and experience necessary to use the bankruptcy system to impose losses on creditors and continue
in operation, making credit support less critical for such firms.
In addition, the Federal Reserve and Treasury may be concerned about reputational risk. Support
provided to highly leveraged firms could be subject to considerable public criticism. Many firms
leveraged up to make payments to owners, including private equity firms, rather than to invest in their
businesses. The public may see support for such firms as undeserved and likely to encourage undesirable
behavior in the future. Such concerns might be less substantial if the problems faced by the firms reflected
illiquidity, and the lending was expected to be low risk in the end. However, that is not the case; the
additional lending would be risky and would likely result in some highly visible—and in some cases very
large—losses for the Federal Reserve and the Treasury.
The Federal Reserve and Treasury may want to consider, or may be required to consider under the
CARES Act, some other considerations. For example, there might be social objectives, such as sustaining
employment, that need to be incorporated into the program. In addition, to limit the need for government
resources, borrowing firms could be required to limit dividends and share repurchases, and there could be
12
constraints on executive compensation as well.9 Such restrictions might make borrowers less willing to
participate in the program, and employment requirements could slow a necessary reallocation of labor
across sectors as a result of the health crisis. However, the restrictions might help with the targeting of
program resources, since firms that were facing bankruptcy would presumably be willing to abide by the
restrictions, while firms that could survive without a program loan would be less inclined to do so.
Moreover, such restrictions may mitigate moral hazard concerns of helping borrowers that had taken on
too much leverage before the pandemic.
4. The three main policy options
There are three main options for designing a government program to support business credit: subsidies,
such as provision of low-cost funding, guarantees to limit downside risk, and direct lending by the central
bank or government. All three have been used, but they differ in how they address the information and
incentive challenges, and so they may be more or less helpful in a given circumstance (summary in table
3).
Table 3. Programs to Increase Lending
Option
Subsidy Guarantee Participation
Helpful if bank lending is
constrained by:
• High funding costs Yes No No
• Limited risk Some Yes No
appetite
• Balance sheet Some No Yes
constraints
Can address challenges
caused by:
• Asymmetric Yes Partially Partially
information
• Adverse selection Yes Partially Partially
• Targeting No Requires calibration Requires calibration
• Moral Hazard Yes Requires calibration Requires calibration
Option 1: Subsidies. The central bank or other government entity can encourage business lending by
providing lenders with a subsidy for such loans. A simple way to do this is to provide low-cost funding in
return for new lending to businesses. An example of such a program is the “Funding for Lending
Scheme” that the Bank of England established in 2012. Under the program, banks, in effect, received
low-cost funding from the Bank of England, with the amount of funding and its pricing depending on the
amount of new lending to U.K. businesses and households by the bank.10 The Fed could provide a similar
incentive for business lending in the current crisis by offering longer-term, low-cost financing to banks,
with the amount of funding available to each bank, and perhaps its pricing as well, depending on the
9
Whether the program is legally required to take account of these considerations under the CARES Act is not clear
to us, and we have heard opinions on both sides.
10
The banks actually received secured loans of government securities, which they could then use to obtain funding
in the repo market. See Churm and Radia (2012) for details. Churm, et al. (2015) provide evidence on the
macroeconomic effects of the program.
13
amount of a bank’s lending to targeted firms—for example, U.S. businesses with fewer than 15,000
employees.
Such a program would help to internalize the positive externalities of such loans. For an unchanged
loan interest rate, the lower funding costs would give banks a larger spread on loans, and that would
provide extra compensation for risk-taking or increasing balance sheet size. A program along these lines
is likely to be more useful if bank term funding spreads are high, either because term funding markets are
stressed or banks are viewed as weak. In such a case, the provision of low-cost central bank funding in
return for new business loans could greatly reduce the marginal cost of such lending, and so have a
significant effect on loan rates and banks’ willingness and ability to provide loans.11 On the other hand, if
the banking system is generally seen as healthy, and banks can obtain wholesale term funding at low cost,
then the program presumably would be less effective. Nonetheless, the central bank could still encourage
lending by providing longer-term funding at a low rate—perhaps at the policy rate, as the Bank of
England is doing (see below). A reliable source of stable, low cost funding should increase banks’
willingness to provide longer-term loans to businesses.
Under such a program, banks would continue to do the underwriting for their loans, and the loans
would remain on their books, with the banks responsible for any losses. As a consequence, the central
bank would not have to worry about asymmetric information or adverse selection. In addition, unless the
subsidy were very large, it seems unlikely to lead banks to provide loans to firms with very high leverage
and risk, mitigating potential moral hazard problems. The central bank could also decide on the size of the
program and so could control the cost of the subsidy.
However, a funding for lending program may not target the assistance very well. For example, if the
low-cost funding is available for all SME loans, banks may choose to lend more at lower rates to
relatively safe firms that would have gotten loans in any case. The firms that were not getting credit, and
which need loans to survive the crisis, might still be seen as highly risky, particularly given the banks’
existing exposures to coronavirus risk, or as generating an undesirable increase in balance sheet size. One
way to improve targeting would be to base the amount of low-cost funding on a more narrowly targeted
set of firms, for which credit is critical for firm survival. However, efforts to target the program in such a
way would run into the asymmetric information problem—the central bank does not have the information
needed to identify such firms. As discussed below, the Bank of England and the ECB have provided low
cost funding to banks to increase lending to SMEs, but have not tried to target their programs more
narrowly, presumably on the view that an overall increase in SME lending should lead to some increase
in lending to firms that would not have received loans otherwise.
Option 2: Guarantees. Business lending could be encouraged by providing, in return for a below-
market fee, a guarantee that limits the downside risk to the lender. For example, the Treasury could
guarantee repayment of at least 80 percent of the loan amount, leaving the lender responsible for only the
first 20 percent of losses.12 Such a guarantee would encourage banks to make loans that are riskier than
those they would otherwise make. In the current crisis, banks may be particularly concerned about tail
risks, in which the virus has larger and more protracted economic effects. By limiting the maximum loss
on additional loans, the guarantee would make banks more-willing lenders. However, if bank lending
were limited by balance sheet constraints, guarantees would be less helpful, since the loans would remain
on banks’ books.13
11
For evidence on these effects, see Havrylchyk (2016).
12
The Federal Reserve would find it difficult to provide such a guarantee given the legal constraints on Section
13(3) lending.
13
In the case of a partial guarantee, the remaining portion of the loan would still be included in the calculation of the
risk-weighted capital ratios. And even with a 100 percent guarantee, the loan would count toward the leverage ratio,
unless the banking agencies provided a temporary exception.
14
Guarantees could be structured to address the information problems raised by lending programs,
though the problems would remain. So long as the first-loss position retained by the banks was
sufficiently large, banks would still need to do loan underwriting. As a result, a guarantee program would
help address the asymmetric information problem, since banks would have an incentive to use the
information they have on the riskiness of borrowers when approving loans. Similarly, banks with a first-
loss position would have an incentive to avoid loans that had a high probability of default, limiting the
adverse selection problem. However, there could still be some adverse selection, since banks might make
loans that had a high risk of very adverse outcomes, since their downside risk would be capped by the
guarantee.
Given the limited information available to the central bank, targeting a guarantee program correctly
would be a challenge. If the guarantee fee were set too low, then banks would have an incentive to make
too many risky loans to firms that are unlikely to survive the crisis. The result would be very high
expenses for the government, without a commensurate macroeconomic benefit. In addition, providing
low-cost guarantees for loans to risky, leveraged firms could encourage moral hazard after the recovery
gets underway. However, excessively high fees would also be a potential problem, since they would
discourage use of the program, and so not help provide credit to the firms that need it.
Option 3: Participations. The final option is for the central bank to take some or all of the business
loans onto its own books through purchases of loan participations. Under this arrangement, banks would
make the loans and sell a portion (perhaps even all) of each loan to the central bank, and in return banks
would receive fees for originating and servicing the loans. This approach would allow the program to
provide credit to businesses while harnessing the banks’ information and underwriting advantages. It
would also remove the portion of the loan purchased by the program from the bank’s balance sheet,
directly addressing limits on lending due to balance sheet constraints. However, by leaving a pari passu
share of the loan on the banks’ books, participations may not address limits on lending that reflect banks’
concerns about risk since the portion of the loan that the bank retains would have the same risk
characteristics as the entire loan and would have to meet the bank’s underwriting criteria. Of course, the
program fees—either up front origination fees or servicing fees paid over the life of the loan—would
encourage risk-taking and increase lending. Up-front fees would, in effect, provide additional capital to
support the lending, while fees paid over the life of the loan would boost the effective risk spread on the
loan.
A participation structure would help address some of the information problems noted earlier. So long
as the lending banks retain a significant enough portion of the loans, they will have an incentive to do
careful underwriting, reducing the asymmetric information problem. Indeed, because they are taking a
vertical slice of the loan, their underwriting incentives are not affected by the program except as a result
of the fees provided. And, so long as the sale of the participation to the program is made at the time of the
loan, this structure should also help limit the adverse selection problem, since banks will retain the risk on
the part of the loans they keep. Note, however, that if banks sell the entire loans to the program, as
suggested recently by Hubbard and Scott (2020) in the case of the Main Street Lending Program, then the
bank would have no incentive to underwrite the loans to limit risk. As a consequence, the program would
have to provide detailed underwriting criteria, which would make the asymmetric information problem a
particular concern. Moreover, since banks could choose which loans to make and hold, and which to
make and sell to the program, there would also be an adverse selection problem. Specifically, banks could
make weak loans in order to obtain fee income, in the knowledge that the resulting loan losses would be
absorbed by the program.
Even if banks retained a portion of the loans, policymakers might be concerned that the program
would not provide a sufficient incentive for banks to perform appropriate underwriting and limit adverse
selection. If so, the program could impose a minimum set of underwriting standards for eligible loans.
However, such standards would require borrowers to satisfy two sets of underwriting criteria, one set by
the program and the other by lending banks. The two sets of underwriting criteria could excessively limit
15
the set of eligible business loans, suggesting that the standards set by the program be kept relatively
simple, so as not to overly constrain the program.
Stein (2020) suggests an additional structural element to help address the asymmetric information
and adverse selection problems: extending program loans in several tranches over time. The program
would start by making relatively small loans, and then provide additional credit contingent on how the
borrower’s situation evolves, as well as incoming information on the coronavirus. So firms that had
particularly large losses might not get additional credit, since they would be unlikely to emerge from the
crisis as a viable firm unless they received transfers from the government. This tranched approach would
allow the government to set easier ex ante underwriting standards while still limiting the government’s
exposure to loss. Stein also proposes the use of less-senior claims, such as preferred stock, to reduce
potential debt overhang post crisis, with the program taking warrants to gain some upside to compensate
for the increased risk. That approach could have benefits, but the Federal Reserve can only lend, not
purchase stock. Moreover, the resulting equity stakes would be difficult for the government to exit in the
case of small-to-mid-sized businesses, which generally do not have preferred stock.14
Approaches employed in other countries. Many countries are facing similar issues related to the
effects of the coronavirus on firms’ revenues and the consequent need to ensure the availability of credit.
While a complete catalog of the approaches that have been taken would be well beyond the scope of this
paper, it is interesting to consider some of the approaches that have been considered in other jurisdictions.
For example, the approaches in the United Kingdom and Germany are broadly similar, but differ
substantially from those taken in the United States. In both of the European countries, the focus has been
on providing loan guarantees and low-cost funding, with relatively little use of participations.
In the United Kingdom, the British Business Bank, a government development bank, is providing
guarantees of new loans to small and medium-sized businesses (see table A1). There is a 100 percent
guarantee of smaller loans (based on 2019 turnover, but limited to £50,000), and an 80 percent guarantee
of larger loans. Borrowers must be British firms, and there are limits on loan maturity as well as on the
payment of interest and fees in the first year. The British Business Bank also has a program offering
convertible loans that is aimed at newer, faster-growing firms. That program will match private funds
raised now, and the loan converts to equity when the firm next raises equity finance.
In Germany, the bulk of the support is in the form of loan guarantees provided by the government
development bank, KfW, under several separate guarantee programs (see table A2). In addition, both the
federal government and state governments have other assistance programs. The KfW provides guarantees
to lending banks for 100 percent of the loan amount for smaller loans (limited based on 2019 turnover,
but as much as €800,000) and generally for 90 or 80 percent for larger loans. Borrowers must satisfy
some broad criteria—e.g., operating in Germany and not having been already troubled prior to the
coronavirus crisis. Some loan terms are set by the programs (e.g., maximum maturity and interest rates),
and there are significant restrictions on the borrowing firms, including limits on executive compensation,
limits on the ability of firm to pay dividends, and limits on capital distributions. In addition to the
guarantee programs, the KfW also has a program under which it will purchase participations in large,
syndicated loans.
While these programs have not be undertaken by the central bank, in both jurisdictions the central
bank has also taken steps to support lending to businesses. In the United Kingdom, the Bank of England
and the Treasury have opened a COVID Corporate Financing Facility under which the Bank will
purchase eligible commercial paper in the primary and secondary market, including from middle-market
firms that have not previously issued commercial paper, on terms comparable to those prior to the crisis,
14
A possible way to benefit from something like a warrant in the case of small firms would be to require the firm to
agree to a higher income tax rate starting a few years after the health crisis ends. The firms that survive and are
operating at that time would, in effect, be paying a share of their income to the Treasury.
16
and with maturities of up to a year (Bank of England 2020a). In addition, the Bank of England is
providing low-cost funding to banks through its “Term Funding Scheme with additional incentives for
SMEs.” The program offers banks low-cost, four-year funding based on the amount of new lending
provided to U.K. businesses, with the amount of lending to SMEs given extra weight in the calculation of
the amount of and rate on the central bank funding (Bank of England 2020b). Given the very low level of
market rates, the Bank of England concluded that some banks might find it difficult to reduce deposit
rates in line with market rates, and so term funding at low rates would help banks to offer lower-cost term
loans to their customers.
In Germany, banks are eligible for the European Central Bank’s Targeted Long-term Refinancing
Operations (the TLTRO III program). Under this program, banks can receive low-cost funding (with rates
as much as 50 basis points below the ECB’s deposit rate, so -100 basis points) depending on their lending
behavior. In addition, the ECB has eased its collateral rules to help ensure that banks have the collateral
that they need to take advantage of the TRTROs. The collateral easing measures include allowing banks
to use SME loans as collateral, among other assets, thereby providing low-cost funding banks can use to
finance additional SME lending. In addition, the ECB has established a program of Pandemic Emergency
Longer-term Refinancing Operations as a bridge to the TLTRO III operations (Lane 2020).
5. What was done: The Main Street Lending Program
In contrast to the British and German programs, the Main Street Lending Program established by the
Federal Reserve and Treasury supports bank lending to businesses through the purchase of participations
in bank business loans. Since these loans will involve potentially significant risk, the Treasury has
provided $75 billion of capital to support up to $600 billion of loans purchased by the MSLP. The MSLP
consists of three facilities—the New Loan Facility, the Priority Loan Facility, and the Expanded Loan
Facility—open to borrowers that are U.S. businesses with 15,000 or fewer employees, or annual revenue
of $5 billion or less. 15
The three facilities are structured to address the funding needs of different types of firms, and they
offer different loan sizes and leverage limits (see table 4 for a summary). The New Loan Facility is aimed
primarily at small and medium-sized businesses that do not already have high levels of debt. The
minimum loan size in this facility was lowered recently to $250,000 from $500,000 in April, and from $1
million in March, and the maximum leverage is 4 times 2019 earnings before interest, taxes, depreciation,
and amortization (EBITDA). The Expanded Loan Facility is aimed at larger firms that may already have
significant debt outstanding in the form of syndicated loans. The program can allow very substantial loans
(raised recently to $300 million) to be made to “upsize” an existing loan agreement. Given the relatively
high debt levels of such firms, the maximum level of leverage in the facility is 6 times EBITDA. The
Priority Loan Facility is aimed at small and medium-sized firms that may have larger debt than those
using the New Loan Facility, and the loans can be used to refinance existing credit to another lender as
well as to increase borrowing. The maximum loan sizes are $50 million, higher than $35 million in the
New Loan Facility, and the maximum leverage is higher (again 6 times EBITDA).
Somewhat surprisingly, the terms on the loans under the three facilities are quite similar, despite the
very different types of borrowers and loans that are covered. Notably, in all three cases, the borrower pays
a rate of LIBOR (1 month or 3 months) plus 300 basis points, and the maturity of the loan is five years,
despite large differences in loan size and maximum debt coverage ratio requirements. Interest payments
are deferred for the first year, principal amortization is deferred for the first two years, and repayments of
principal starting in year three are in similar proportions across the facilities. The origination fee for the
15
On June 15, the Federal Reserve announced plans to open similar programs for not-for-profit entities. Those
programs are not considered here.
17
New and Priority Loan programs is 100 basis points, but it is only 75 basis points for the Expanded Loan
Facility, presumably reflecting the larger size of the loans under that program.
The facilities are designed to address the four challenges noted earlier. Because the banks keep a 5
percent share of each loan on their books, they have an incentive to do careful underwriting, limiting the
asymmetric information problem faced by the Federal Reserve and the Treasury. The retained share also
limits adverse selection, since banks cannot avoid a portion of any losses on loans under the MSLP.
However, the share of credit risk banks must retain does not vary by loan risk, suggesting there is no more
protection to the Fed and Treasury from this source where it is needed more. Under the Expanded Loan
Facility, which allows the largest and most leveraged loans, there is a limit of 35 percent of the total credit
of which the loan is a part, which provides an additional significant incentive for the bank to take careful
account of risk.
Adverse selection is also addressed by the minimum underwriting criteria that the MSLP imposes.
Borrowers must have been healthy at the end of 2019, as judged by the lender’s internal rating of existing
credits at that time, and they must certify that they believe that they are able to meet their financial
obligations and not file for bankruptcy for at least 90 days. Borrowers must commit to not using the loan
to voluntarily repay other debts before the MSLP loan is repaid, and that they will not seek to cancel other
credit lines over that period. Borrowers also must limit executive compensation, dividends, and capital
distributions while the loan is outstanding. In addition, lenders under the program must be a regulated
U.S. depository institution or a U.S. branch or agency of a foreign banking institution. In either case, the
participating lenders would be subject to supervision by the U.S. banking agencies, and so their
operations and incentives should be relatively well understood. Moreover, the lending banks cannot
request early repayment of other debts during the life of an MSLP loan, and they cannot cancel or reduce
any existing loan commitment or line of credit over the same period.
These program features also help to address some of the constraints on the Fed’s 13(3) lending
authority. In particular, the incentives for bank underwriting and the minimum underwriting criteria help
ensure that the borrowers are solvent. In addition, the Fed is secured by the SPV collateral and the
Treasury capital, presumably to its satisfaction, and the Treasury backing should help protect taxpayers
from losses.
In addition, the underwriting requirements help to limit the risk of loss to the Treasury. As noted
earlier, the extent to which the Treasury is willing to take on such risk depends on the Treasury
Secretary’s risk tolerance, and the Secretary has suggested only a limited willingness to take losses on the
CARES funds. The actual outcome will depend in part on the calibration of the MSLP. Under the
program, the Treasury will receive fees from the banks for purchasing the 95 percent participations, as
well as 95 percent of the spread income on the loans (less the servicing fees paid to the bank) over the life
of the loan. This income will help to offset losses on the loan portfolio. If the fees and spreads received by
the Treasury are too low relative to the losses on the loans, the Treasury could lose money. On the other
hand, if fees or spreads are too high, the Treasury would earn money on the program. However, the gains
and losses are not the measure of program success, which must take account of the aid provided to the
intended firms.
In that regard, the program terms should also help target the credit to businesses that are neither so
strong that they do not need the assistance nor so weak that they are likely to fail. As we note below,
strong firms will not find the pricing attractive, since they are likely able to borrow at a spread less than
300 basis points. Such firms also will not want to accept the constraints the program imposes on
dividends and capital distributions. In addition, the MSLP requires that borrowing firms must make
“commercially reasonable efforts” to maintain employment and payroll during the life of the loan, a
constraint that stronger firms would likely choose to avoid, given uncertainty about what compliance
would entail and the potential need to restructure some of their operations in response to the economic
fallout from the virus. Conversely, very weak firms will likely not satisfy the requirements of the
18
program (such as the limits on leverage), helping to limit the program to borrowers most likely to benefit
from a loan.
Finally, the program is designed to limit moral hazard by excluding loans to firms that had excessive
levels of leverage prior to the pandemic. As noted earlier, the Federal Reserve has emphasized for some
time that excessive leverage in the nonfinancial business sector is a potential risk to the economy because
a business downturn caused by other factors could be reinforced by a wave of failures and bankruptcies
by highly leverage firms. In the interagency guidance on leveraged lending in 2013, the federal banking
agencies noted that “a leverage level after planned asset sales…in excess of 6X Total Debt/EBITDA
raises concerns for most industries” (Federal Reserve 2013). Thus, it seems appropriate that borrowers
under the MSLP be required to have a debt to EBITDA ratio of no more than 6, even after receiving the
program loan. The program’s limits on compensation, dividends, and share repurchases for borrowers
also may serve as an offset to future incentives to increase leverage.
Table 4. Main Street Lending Program Terms
Facility Intention Terms
Main Street New Supporting new loans to • 5-year maturity, rate of LIBOR+300 bps
Loan Facility businesses that are small (LIBOR is 1 or 3 months). Origination fee of up
or medium-sized and to 100 bps.
which have low levels of • Size limited to between $250,000 and $35
debt. (Thus, the loan sizes million, with a limit on total leverage of 4x
are relatively modest, and 2019 EBITDA. Any outstanding loan has an
leverage is limited.) internal risk rating of “pass.”
• No interest or principal payments in the first
year. Required minimum amortization of 15
percent, 15 percent, and 70 percent in years 3 to
5, respectively.
• Limits on executive compensation, dividends,
distributions of capital, and employment.
• Loan not used to voluntarily repay other debt.
• Borrower must make “commercially reasonable
efforts” to maintain employment and payroll.
• Bank participation 5 percent.
• Bank pays SPV 100 bps of principal amount,
and is paid 25 bps of principal per annum for
servicing.
Main Street Supporting increases in Same as the New Loan Facility, except:
Expanded Loan loans to businesses that • Origination fee of up to 75 bps.
Facility are larger and already • Size limited to between $10 million and $300
have significant debt. million, with a limit on total leverage of 6x
(Thus, loan sizes are 2019 EBITDA.
considerably larger, and • Size also limited to 35 percent of the borrower’s
allowable leverage is debt that is pari passu with the participation.
higher than in the New • Minimum amortization of 15 percent, 15
Loan Facility.) percent, and 70 percent in years 3-5,
respectively.
• Bank participation 5 percent
• Bank pays SPV 75 bps of principal, and is paid
25 bp of principal amount per annum for
servicing.
19
Main Street Supporting new loans to Same as the New Loan Facility, except:
Priority Loan businesses that are small • Size limited to between $250,000 and $50
Facility or medium-sized but may million
already have significant • Limit on total leverage is 6x 2019 EBITDA.
debt. (Thus, the loan • The loan can be used to refinance existing debt
sizes are relatively owed to another lender.
modest, but allowable
leverage is higher than in
the New Loan Facility.)
6. Will the MSLP be successful?
The success of the MSLP will be measured ultimately by whether more loans are made to borrowers in
need, with future losses consistent with Federal Reserve and Treasury constraints and preferences. Thus, a
program with very few loans would not be successful since its terms were presumably too restrictive to
generate the social benefits of preserving firms. However, a program with lots of loans but large losses for
the Treasury and the Federal Reserve would also be a failure, given that the Treasury has indicated a
desire to limit losses.
Will borrowers be helped? The MSLP makes loans, not government transfers. Not all firms that have
lost business because of the pandemic will be helped with a loan that increases its total debt burden,
whereas all firms that receive transfers to cover some of their lost revenue will be helped. But loans can
be structured so the additional debt is not overly burdensome. In particular, loans with longer maturity,
lower rates, and deferred payments would make them more equity-like. The MSLP loans have a few of
these characteristics.
For all three programs, the loans are five-year maturity, with interest payments deferred for the first
year. Principal payments are deferred for two years, and the bulk of the repayments—70 percent—come
only in the final year. These terms are helpful because they allow the firm more time for recovery. Indeed,
the Federal Reserve recognized potential benefits to borrowers of more “equity-like” features when it
lengthened maturities and pushed back amortization in changes announced on June 8 (Federal Reserve
2020c).
In addition, the loan rate of 3-month LIBOR + 300 basis points is the same for the three programs,
despite minimum loan sizes that vary from $250,000 for the New Loan Facility and a maximum loan ratio
of 4 times EBITDA, and a minimum of $10 million for the Expanded Loan Facility and a maximum
leverage ratio of 6 times EBITDA.
However, spread data on C&I loans by commercial banks and on institutional leveraged loans
available for transactions in mid-2017 (the latest data available from the Fed’s Survey of Terms of
Business Lending, or STBL) suggest substantial variation across size and credit rating (table 5). For C&I
loans by large domestic banks, with an average size of $593,000 and a risk rating of 3.1, the average
spread is 2.13 percent (a rate of 3.43 percent).16 The spread for low risk firms is lower than the average,
1.61 percent, and for loans above $1 million, 2.07 percent. All of these spreads are lower than the 300
basis points for the Main Street program. At their recent peak in 2010, following the financial crisis, C&I
loan spreads peaked at 3.17 percent, and were above 300 basis points for moderate risk loans and smaller
16
The category "moderate risk" represents the average loan under average economic conditions at the typical lender.
The risk rating is calculated by assigning a value of 1 to minimal risk loans, 2 to low risk loans, 3 to moderate risk
loans, 4 to acceptable risk loans, and 5 to special mention and classified loans.
20
loans.17 These data suggest both wide variation in spreads by size and risk, and that a 300 bp spread is
much closer to a peak recession spread than a spread in more normal times. Such a spread is high for low-
risk firms and even for moderate risk firms except at their cyclical peak, suggesting the program will not
attract higher-quality borrowers. This pricing is unlike the more favorable terms offered to investment-
grade corporations, where the PMCCF will purchase new issues at market rates, and the SMCCF
purchases bond ETFs or single-name bonds in order to reduce risk spreads on outstanding bonds, which
then should reduce the rate on bonds purchased in the PMCCF.
Table 5. Spreads on C&I loans and institutional leveraged loans, by risk and size, 2010 and 2017
2010:Q4 2017:Q2
Average C&I loan 3.17 2.13
2010 loan size $447,000, Risk rating 3.2
2017 loan size $593,000, Risk rating 3.1
Risk
Low 2.47 1.61
Moderate 3.24 2.18
Size ($Th)
100 to 1,000 3.60 2.50
1,000 to 10,000 2.95 2.07
Institutional leveraged loans
BB/BB- rating 4.00 2.50
B+/B rating 5.50 4.00
Source: For C&I loans, large domestic banks, Survey of Terms of Bank Lending, Federal Reserve Board. For
institutional leveraged loans, S&P Global, Leveraged Commentary & Data, as reported in the Federal Reserve
Board Financial Stability Reports. Spreads are relative to 3-month LIBOR.
Note that spreads on institutional leveraged loans for below-investment-grade borrowers ranged on
average from 250 basis points to 400 basis points, and this range would encompass the 300 basis point
spread for borrowers under the Expanded or Priority facilities. But again, the range is fairly large,
suggesting that a common spread would be a disincentive for higher quality borrowers within this group
to borrow from the program.
In short, the LIBOR+300 basis points spread on MSLP loans would not be attractive to less-leveraged
middle-market firms that access credit mainly through banks. Rather, it appears attractive mainly for
higher-risk borrowers and smaller loans. We believe that a bank with an existing customer that met the
conservative debt limits under the New Loan Facility would find this spread to be high, and so would be
willing to make a loan at this spread and keep it on its own books, suggesting that a narrower loan spread
17
The STBL was discontinued in May 2017. A new survey was started for small business loans—Federal Reserve
Bank of Kansas City (2020). However, the loans covered in that survey are much smaller than those that had been
included in the STBL. For example, the average C&I loan offered by domestic commercial banks in the STBL in the
May 2017 survey was $491,000, while the average loan size in 2017:Q4 in the new survey for a new fixed-term loan
was $93,000 and a new variable-term loan was $160,000. Loan spreads on new fixed-term loans ranged from 3.2 to
3.9 percent between 2017:Q4 and 2019:Q4, reflecting their smaller size. No information is available in the new
survey on the risk of the loans.
21
may be called for to support business borrowing and the continued operation of small-to-mid-sized
businesses.
Some have argued that the leverage ratios in the MSLP are too low, and that there are firms with
higher than 6 times EBITDA that would benefit from a Main Street loan. The leverage ratios serve to
reduce risk to the program, as well as to mitigate moral hazard. That said, there may be situations where a
borrower in an industry that is characterized by high leverage is a significant employer, and its
bankruptcy would be especially inefficient. In such cases, the Fed could establish a process to allow banks
to make exceptions to the general underwriting guidance and lend to such firms under the program, so
long as the bank agreed to retain a larger share of the loan, such as 15 or 20 percent, in order to ensure
that the bank was comfortable with the credit. Indeed, when the Priority Loan facility was introduced in
April, it allowed higher leverage than the New Loan facility in return for a higher bank participation
share. This tradeoff was eliminated in the revised terms issued in June. But policymakers should monitor
closely whether the program could be more effective in serving the needs of borrowers and the economy
if higher leverage could be accommodated. If so, the higher leverage should be accompanied by an
increase in bank retention share to limit risk, and moral hazard concerns could be mitigated by the limits
on dividends, buybacks, and compensation.
Some prospective borrowers may also be concerned about the stigma associated with taking a
government-supported loan when it is not designed to be used broadly and offered on favorable terms. In
addition, the restriction on retaining employees may hinder the ability of a firm to make necessary
adjustments to its business practices to remain viable in a post-COVID environment, and thus limit
demand. Moreover, lack of clarity about how compliance with this requirement will be measured would
make borrowers less willing to use the program.
Will banks participate given the terms that have been established? We outlined above three ways to
incentivize banks to lend more, by a funding for lending subsidy, a guarantee, and a participation such as
in the MSLP. The ability to sell 95 percent of the loan to the SPV should help if the main impediment to
lending to the intended beneficiaries is balance sheet capacity. If the main impediment to bank lending is
limited risk appetite, however, rather than balance sheet pressures, it isn’t clear that the MSLP will get
much use because the bank needs to retain some of the loan, and even if only 5 percent, it will apply its
own underwriting standards to the loan.
Which is it? Banks’ responses to the April 2020 Senior Loan Officer Opinion Survey, (based on
responses in March) show that that 40 percent of domestic banks surveyed, on net, indicated that they had
tightened lending standards for firms of all sizes (figure 2 above). Moreover, they indicated that reduced
tolerance for risk was a more important factor than bank’s current or expected capital position. In
particular, the survey shows that banks attributed the changes to increased uncertainty about the outlook,
a worsening of industry-specific problems, and a reduced tolerance for risk. Only a few banks indicated
that a deterioration in their bank’s current or expected capital position had played a significant role.18
Thus, the current program, which is more effective in addressing balance sheet constraints than
concerns about loan risk, may not give enough encouragement to banks to participate. However, while
banks may want to avoid lending to riskier firms on their own, the fees on MSLP loans should make them
more willing to do so. For example, under the New and Priority Loan Facilities, banks extend only 5
percent of the loan, but they receive an origination fee of up to 100 basis points of the loan, and then a
servicing fee of 25 basis points of the remaining 95 percent of the loan each year. Thus, the total fees paid
to the banks are more than 40 percent of the portion of the loan that the banks retain—a significant
18
Of course, banks may be wary of reporting concerns about their capital positions on a Federal Reserve survey –
even in the global financial crisis, when many banks faced severe capital pressures, survey respondents pointed
primarily to higher risk and decreased risk tolerance as the reasons for tighter lending standards.
22
incentive to lend to firms that may benefit from the credit but are not strong enough to obtain loans
directly from the bank.
Banks also may be less willing to participate if there is uncertainty about the costs associated with
workouts of loans to defaulting borrowers. The program requires that the banks, which have considerable
experience in managing troubled loans, will be responsible for any workouts. Thus, in evaluating
participation in the program, banks will need to take account of the likely costs they will incur in the case
of loan defaults. One problem with this approach, however, is that banks may view the workouts as too
burdensome, or be concerned that in situations where loan losses are high, they will face large workout
costs that they will find hard to manage. In such situations, banks also may face funding pressures and so
may be inclined to limit workout costs by moving quickly to close defaulting firms. By contrast, the Fed
and Treasury do not face funding pressures, and they can take account of the social benefits to keeping
firms going during the recovery and avoiding liquidations—hence they would be inclined to be more
flexible in workouts. To encourage the banks in this regard, the Fed and Treasury could pay higher
servicing fees for delinquent loans, particularly in periods of generally high loan delinquencies, which
would encourage more loan originations and more efficient workouts.19
7. How to make the MSLP more effective
This paper has laid out the key elements of the MSLP program as a solution to the problem of increasing
bank lending to firms that need to obtain credit to survive the health crisis, subject to the constraints of the
Fed and the Treasury. We define success as an increase in loans to such borrowers relative to the level
that banks would have provided on their own, subject to the Fed’s legal constraints to lend only to solvent
borrowers and be secured to its satisfaction, and with losses at a level acceptable to the Treasury. Since
many of the firms that could benefit from the program are inherently risky, if there are no losses to the
Treasury, the program probably did not take enough risk. Moreover, unlike in 2008, when the intention of
Fed lending programs was primarily to counter a run on the financial system, the intention of the MSLP is
for the government to share the risk of losses caused by the shutdown and the associated changes in
economic behavior.
The MSLP uses a loan participation structure between the bank and government, with the program
terms designed to address the asymmetric information, adverse selection, poor targeting, and moral
hazard problems that arise as a result of the differing information and incentives of the borrowers, banks,
and the government. In particular, the various terms and conditions are calibrated to encourage banks to
participate and to encourage use of the program by borrowers that can benefit from it. The calibration of
the parameters of the MSLP is difficult, given that it is a new program being introduced in an entirely
novel situation. The two sets of adjustments that the Fed has made already to the original proposal based
on more than 2,200 comment letters and other feedback from potential borrowers and lenders illustrate
the difficulty of designing a program that can be effective. This paper, which highlights why private
lending may not be sufficient in the current situation and the challenges in setting up a government
program to support such lending, provides a framework to understand adjustments of the parameters of
the program and the tradeoffs between borrowers, lenders, and the government.
19
An alternative approach would be to use the SPV or a separate “resolution” SPV to manage the troubled loans on
behalf of the Federal Reserve and Treasury and enter into loss-sharing arrangements with the banks. This approach
would help to align the incentives of those conducting the workout with the interests of the Federal Reserve and
Treasury, and so could lead to better outcomes. Allowing for troubled loans to be pooled across the program also
could increase returns, particularly if the troubled loans were ultimately securitized and sold. Hiring asset managers
with skills in working out troubled credits could free the Fed and Treasury from those responsibilities, which require
specialized skills and could be politically fraught. And, while it would take some time to contract with appropriate
asset managers, there likely would be time because program loans do not require payments in the near term.
23
One reason the program may not be effective is because targeted firms do not see the program terms
as attractive enough. Since the set of firms that could benefit from the MSLP varies widely, from small
private borrowers that have limited debt to very large, complex companies with significant debt, the terms
of the program should be better tailored to the risks and needs of the borrowers. In particular, we think
that the terms for the New Loan facility should be eased relative to those for the Priority and Extended
facilities, since the smaller firms that would likely borrow under the New Loan facility are, because of
their lower leverage, likely simpler and less risky, and have fewer alternatives to avoid liquidation or
inefficient bankruptcies. As a consequence, we propose, first, that loan spreads be recalibrated to reflect
differences in credit quality across the three loan facilities, and that the New Loan facility have a lower
spread, given the better credit quality of the borrowers. In addition, the paperwork required for loans in
the New Loan facility should be streamlined, if possible. Currently the borrower and lender certification
requirements appear to reflect the participation by large banks and large, complicated firms, which have
the legal staff needed to comply with them. However, the smaller banks and simpler borrowers that might
use the New Loan facility will likely find these certifications daunting.
Second, we fully agree with the recent changes to lengthen loan maturities from four to five years,
and the pushback in repayment schedules and required amortization. These changes give borrowing firms
more time to recover, and they may be especially helpful in the case of a protracted post-coronavirus
recession. For comparison, investment-grade and fallen angel firms borrowing from the PMCCF can issue
loans or bonds for a five-year term with repayment only at maturity, so five-year loans with some
amortization under the Main Street program seem less favorable to the borrowers. Another possible
comparison is to SBA 7(a) loans, which can be extended for ten years (maximum statutory maturity). If it
appears that borrowers would find that longer maturity loans would be helpful, we would support going
beyond five years. The program could offer a longer-term loan where the interest rate begins to increase
after five years, which gives borrowers an incentive to repay the loan after five years but allows for
greater flexibility.
While maximum debt-to-EBITDA ratios have been set for each of the facilities, which helps to
address adverse selection, better targeting, and moral hazard concerns, there may be situations where
borrowers would exceed the limits, but banks believe that such lending would benefit the borrower and
the economy. The current program does not allow for this type of flexibility, but if policymakers wanted
to permit it in some situations, they could require the banks to retain a larger share of the loan in
exchange, so the Federal Reserve and Treasury would still be adequately protected.
In addition, we believe it is important for the Fed and Treasury to provide greater clarity regarding
what is meant by “commercially reasonable efforts” to maintain employment. Providing more clarity on
what adjustments would be deemed reasonable would reduce uncertainty for businesses considering a
MSLP loan and encourage take-up. In practice, we also believe that firms should be given significant
flexibility to respond appropriately in order to stay commercially viable because COVID-19 may require
firms to make some important changes to their operations and business models to respond to changes in
the behavior of their customers and suppliers.
Another reason the MSLP could get limited use is because banks do not find it attractive. The
participation structure of the current program helps banks to manage balance sheet constraints they may
face, but does less to reduce loan risk than a loan guarantee program would. However, commercial
bankers have attributed their current tighter lending standards more to heightened uncertainty and reduced
risk tolerance rather than balance sheet constraints. That being said, balance sheet constraints could
become more pronounced if the economy remains very weak, and the MSLP could become more
attractive to banks over time with needed balance sheet relief and fee income. We believe, though, that it
is important for the program to aim to boost credit sooner rather than later. That increase would reduce
the likelihood that the sharp decline in current activity leads to significant permanent reductions because
businesses could not survive the immediate crisis.
24
To encourage bank participation in the program if they lack the risk appetite to make the desired
loans, we recommend, first, that the government increase fees, giving banks a larger part of the spread on
the entire loan, to provide greater compensation for the risk that banks retain. The government could
even pay a higher fee to banks that do more lending under the program; this additional incentive for
greater program use would be similar in spirit to the approach taken to reward lenders for more SME
lending in the Bank of England’s term funding scheme. Second, given that banks may be concerned
about the costs associated with a higher than usual level of loan workouts, the program could pay higher
servicing fees for delinquent loans, particularly in periods of generally high loan delinquencies, in order
to encourage banks to do loan workouts that are more efficient than those that only minimize the banks’
short-term costs.
Of course, some of these changes could worsen the adverse selection and moral hazard problems the
program is designed to address, and they would also require the government to accept an increased risk of
loss. Since some of the other 13(3) programs (PDCF, CPFF, MMLF) are likely to make money, the
Treasury could, in effect, mutualize the programs using gains on some programs to offset the losses on
others. Policymakers should prepare for the possibility that interest in the program remains weak and be
ready to make additional adjustments as needed. The Fed and the Treasury have been aggressive thus far
in their response to the coronavirus, and they may need to be even more so to make the MSLP a success.
However, policymakers also should be prepared for the possibility that even an aggressive MSLP
program consistent with the CARES Act may not be sufficient to meet the financial needs of many
businesses. In that case, they should quickly try an alternative approach. One possibility would be for the
Federal Reserve to introduce a funding for lending program to reduce bank funding costs that might be
limiting their willingness to lend. More broadly, the Federal Reserve and the Treasury could go back to
Congress and ask for changes to the CARES Act that would allow for a more effective support program.
For example, Congress could authorize a loan guarantee program that should increase banks’ willingness
to lend despite a limited risk appetite. Such a program could be structured like an expanded SBA 7(a)
program, and it would be similar to the programs that have been introduced in some other countries.
Alternatively, Congress could allow the Federal Reserve or the Treasury to purchase preferred stock, or
even common equity, which would avoid the problems associated with debt overhang, and so allow
support to be provided to many more firms. Finally, Congress could lift the prohibition on loan
forgiveness contained in the CARES Act, allowing the Fed and Treasury to forgive loans if the effects of
the coronavirus on borrowers turn out to be significantly larger than anticipated. Such forgiveness would
effectively provide insurance against future shutdowns caused by the coronavirus, and so could support
both banks’ willingness to lend and firms’ willingness to borrow.20
20
Metrick (2020) outlines a possible structure for such a program.
25
References
Bank of England (2020a). “Covid Corporate Financing Facility (CCFF): information for those seeking to
participate in the scheme.” https://www.bankofengland.co.uk/markets/covid-corporate-financing-
facility
Bank of England (2020b). “Term Funding Scheme with additional incentives for SMEs (TFSME),”
Market Notice, March 11. (https://www.bankofengland.co.uk/markets/market-notices/2020/term-
funding-scheme-market-notice-mar-2020)
British Business Bank (2020a). “Bounce Bank Loan Scheme.” (https://www.british-business-
bank.co.uk/ourpartners/coronavirus-business-interruption-loan-schemes/bounce-back-loans/)
British Business Bank (2020b). “Coronavirus Business Interruption Loan Schemes and Future Fund.”
(https://www.british-business-bank.co.uk/ourpartners/coronavirus-business-interruption-loan-
schemes/)
Census Bureau (2017). “Statistics of U.S. Businesses.” (https://www.census.gov/programs-
surveys/susb.html)
Census Bureau (20xx), Longitudinal Business Database 1977-2014.
(https://www.census.gov/library/working-papers/2002/adrm/ces-wp-02-17.html)
Churm, Rohan and Amar Radia (2012). “The Funding for Lending Scheme,” Bank of England Quarterly
Bulletin. Fourth Quarter. (https://www.bankofengland.co.uk/-/media/boe/files/quarterly-
bulletin/2012/the-funding-for-lending-
scheme.pdf?la=en&hash=57B13A5C9E94B5D02A451670C8B013DD22CF16EA).
Churm, Rohan, Michael Joyce, George Kapetanios, and Konstantinos Theodoridis (2015).
“Unconventional monetary policies and the macroeconomy: the impact of the United Kingdom’s
QE2 and Funding for Lending Scheme,” Bank of England Staff Working Paper No. 542. August.
(https://www.bankofengland.co.uk/-/media/boe/files/working-paper/2015/unconventional-
monetary-policies-and-the-macroeconomy-the-impact-of-the-uk-qe2.pdf)
Davidson, Kate and Richard Rubin (2020). “Steven Mnuchin Says U.S. Aims to Get Back Its Money
From Fed Programs,” Wall Street Journal, April 29. (https://www.wsj.com/articles/mnuchin-
says-u-s-not-aiming-to-lose-money-on-fed-lending-facilities-11588178749)
Elliott, Justin, Lydia DePillis and Paul Kiel (2020). “The Trump Administration Says a New Bailout
Program Will Help 35 Million Americans. It Probably Won’t,” ProPublica, May 14.
(https://www.propublica.org/article/the-trump-administration-says-a-new-bailout-program-will-
help-35-million-americans-it-probably-
wont?utm_source=pardot&utm_medium=email&utm_campaign=majorinvestigations&utm_cont
ent=river)
Federal Reserve (2013). SR Letter 13-3.
(https://www.federalreserve.gov/supervisionreg/srletters/sr1303a1.pdf)
Federal Reserve (2015). “Federal Reserve Board approves final rule specifying its procedures for
emergency lending under Section 13(3) of the Federal Reserve Act,” Press Release, November
30. (https://www.federalreserve.gov/newsevents/pressreleases/bcreg20151130a.htm)
26
Federal Reserve (2017). Survey of Terms of Business Lending.
(https://www.federalreserve.gov/releases/e2/)
Federal Reserve (2020a). “Section 13: Powers of Federal Reserve Banks.”
(https://www.federalreserve.gov/aboutthefed/section13.htm)
Federal Reserve (2020b). Senior Loan Officer Opinion Survey on Bank Lending Practices, April.
(https://www.federalreserve.gov/data/sloos.htm)
Federal Reserve (2020c). “Federal Reserve Board expands its Main Street Lending Program to allow
more small and medium-sized businesses to be able to receive support,” Press Release, June 8.
(https://www.federalreserve.gov/newsevents/pressreleases/monetary20200608a.htm)
Federal Reserve (various). Financial Stability Report.
(https://www.federalreserve.gov/publications/financial-stability-report.htm)
Federal Reserve, Office of the Inspector General (2010). “The Federal Reserve’s Section 13(3) Lending
Facilities to Support Overall Market Liquidity: Function, Status, and Risk Management,”
November. (https://oig.federalreserve.gov/reports/FRS_Lending_Facilities_Report_final-11-23-
10_web.pdf)
Federal Reserve Bank of Kansas City (2020). “Small Business Lending Survey.”
(https://www.kansascityfed.org/research/indicatorsdata/smallbusinesslendingsurvey)
Havrylchyk, Olena (2016). “Incentivising Lending to SMEs with the Funding for Lending Scheme, Some
Evidence from Bank-level Data in the United Kingdom” OECD Working Paper No. 1365,
December. (https://www.oecd-ilibrary.org/docserver/a4352865-
en.pdf?expires=1590771330&id=id&accname=guest&checksum=432DAA4CB7BCC04DFF038
37647DAD93F)
Hubbard, Glenn and Hal Scott (2020). “Main Street Needs More Fed Help,” Wall Street Journal, April
16, 2020. (https://www.wsj.com/articles/main-street-needs-more-fed-help-11587055459)
Juenemann, Michael, Johannes Wirtz, and Pascal Leitmann (2020). “COVID-19: Germany Establishes
Economic Stabilisation Funds” Bird and Bird report, April.
(https://www.twobirds.com/en/news/articles/2020/germany/covid-19-germany-establishes-
economic-stabilisation-funds)
KfW (2020). “KfW coronavirus aid: loans for companies.”
(https://www.kfw.de/inlandsfoerderung/Companies/KfW-Corona-Hilfe/; Note that the links on
the German-language page provide more detail on the programs than are available on the Englis-
language pages.)
Lane, Philip (2020). “The monetary policy response to the pandemic emergency,” Blog Post, European
Central Bank, May 1.
(https://www.ecb.europa.eu/press/blog/date/2020/html/ecb.blog200501~a2d8f514a0.en.html)
Liang, J. Nellie, Meg McConnell, and Phillip Swagel, 2020. “Evidence on the Outcomes from the
Financial Crisis Response,” in First Responders: Inside the U.S. Strategy for Fighting the 2007-
2009 Global Financial Crisis, Bernanke, Ben S., Timothy F. Geithner, Henry M. Paulson, and J.
Nellie Liang, Eds. New Haven, CT: Yale University Press, 2020.
Metrick, Andrew (2020). “Proposal for Dual-Trigger Insurance + Loan (DTIL) Program,” mimeo, Yale
University, June 6.
27
Miller, Marcus and Joseph Stiglitz, 2010. "Leverage and Asset Bubbles: Averting Armageddon with
Chapter 11?" Economic Journal, vol. 120, pp 500-518.
Mohsin, Saleha and Davide Scigliuzzo (2020). “Wall Street Doubts Fed’s $600 Billion Lending Plan Can
Succeed,” Bloomberg, June 6. (https://www.bloomberg.com/news/articles/2020-06-06/wall-
street-doubts-fed-s-600-billion-lending-plan-can-succeed)
PayNet. Small Business Delinquency Index. (https://sbinsights.paynetonline.com/loan-performance/)
Ransom, Diana (2020). “Why the Fed's New $600 Billion Lending Program May Get a Cool Reception,”
Inc Magazine, June 4. (https://www.inc.com/diana-ransom/main-street-lending-program-credit-
markets.html)
S&P Global. Leveraged Commentary & Data, as reported in Federal Reserve Board Financial Stability
Report (various).
Standard and Poor’s. 2018 Annual Global Corporate Default and Rating Transition Study.
Stein, Jeremy (2020). “An Evaluation of the Fed-Treasury Credit Programs,” Bendheim Center webinar
(https://bcf.princeton.edu/event-directory/covid19_15/)
Timiraos, Nick and Kate Davidson (2020). “Powell, Mnuchin Outline Contrasting Perils Facing
Economy,” Wall Street Journal, May 19. (https://www.wsj.com/articles/powell-mnuchin-set-to-
face-lawmakers-over-crisis-lending-programs-11589880602)
Treasury (n.d). “Legacy Loans Program: Summary of Terms.” (https://www.treasury.gov/press-
center/press-releases/Documents/legacy_loans_terms.pdf)
28
Table A1: Summary of U.K. COVID-19 Business Credit Support Programs
Source of Support Type of Support Eligibility Terms
British Business Bank Guarantee SMEs 100 percent guarantee
Bounce-back Loans of loans of £2000 to the
(SMEs are defined as a lesser of 25 percent of
business with less than business turnover or
250 employees and £50,000.
either (a) a turnover of Loan term of 6 years,
less than £44.45m or loan interest rate of 2.5
(b) a balance sheet of percent.
less than £38.22m.) No fees and no
payments for a year.
Government pays
interest for the first
year.
British Business Bank Guarantee SMEs 80 percent guarantee of
Business Interruption loans over £50,000 and
Loans (SMEs are defined as a up to £5 million.
business with less than No interest and fees for
250 employees and the first year.
either (a) a turnover of
less than £44.45m or
(b) a balance sheet of
less than £38.22m.)
British Business Bank Guarantee Larger firms with 80 percent guarantee of
Large Business turnover greater than loans up to £25 million
Interruption Loans £44.45 million. for firms with turnover
from £ 44.45 million to
£250 million and up to
£50 million for those
with turnover above
£250 million.
Term of up to 3 years
No interest and fees for
the first year.
British Government Convertible loan Newer, faster growing Bridge financing of
Future Fund firms. between £125,000 and
£5,000,000, matching
private funds raised.
Interest rate of 8
percent for a maximum
of three years.
Converts to equity at
the next funding round
of the firm with a 20
percent discount.
Bank of England Primary and secondary Firms that were Terms comparable to
COVID Corporate market purchases of investment grade prior those in the market
Finance Facility commercial paper to the crisis. prior to the crisis.
29
(Joint program of the Maturity of up to a
BOE and Treasury) year.
Minimum purchase is
£1 million. Maximum
total purchase is £1
billion for large, highly
rated firms.
Bank of England Low-cost funding British depository Floating rate funding
Term Funding Scheme institutions with a 4-year term.
with additional Priced at Bank Rate,
incentives for SMEs unless lending to
(TFSME) British businesses and
households declines, in
which case it increases
by as much as 25 bps.
Quantity of low-cost
funding depends on the
level of lending to
businesses and
households in the UK,
plus increases in such
lending over a base
period. Increases in
lending to SMEs get a
multiple of 5 in the
calculation.
Sources: British Business Bank (2020a), British Business Bank (2020b), Bank of England (2020a), Bank of
England (2020b).
30
Table A2: Summary of German COVID-19 Business Credit Support Programs
Source of Support Type of Support Eligibility Terms
KfW Instant Loan Guarantee of loan for German firms with 100 percent guarantee
(KfW backed by the investment or operating more than 10 of a loan of up to lesser
German Government) costs. employees and positive of €500,000 and 25
(Loan can’t be used to profit in 2019 or on percent of 2019
payoff other loans or to average over 2017- turnover for firms with
invest in other firms.) 2019. up to 50 employees,
Firms not in troubled and lesser of €800,000
condition at the end of and 25 percent of 2019
2019. turnover for firms with
more than 50
employees.
Term is up to 10 years.
Fixed rate, currently 3.0
percent.
No underwriting by the
bank.
Limits on executive
compensation,
dividends, and
distributions of capital.
KfW Startup Loan -- Guarantee of loan for German firms in Guarantee of 90 percent
Universal investment or operating existence for 3- 5 years. of loans for SMEs and
costs. Firms not in troubled 80 percent of loans for
(Loan can’t be used to condition at the end of larger firms.
payoff other loans or to 2019. Maximum loan size of
invest in other firms.) €1 billion (with other
(SMEs are firms with limits based on annual
fewer than 250 turnover, wage costs,
employees and turnover debt, and financing
of less than €50 million needs).
OR assets of less than Term of up to 10 years.
€43 million.) Fixed Rate, currently
2.0 to 2.12 percent.
Limits on executive
compensation,
dividends, and
distributions of capital.
KfW Entrepreneur Guarantee of loan for Firms operating in Guarantee of 90 percent
Loan investment or operating Germany in existence of loans for SMEs and
costs. for more than 5 years. 80 percent of loans for
(Loan can’t be used to Firms not in troubled larger firms. Maximum
payoff other loans or to condition at the end of loan size of €1 billion
invest in other firms.) 2019. (with other limits based
on annual turnover,
(SMEs are firms with wage costs, total debt,
fewer than 250 and financing needs).
employees and turnover Term of up to 10 years.
of less than €50 million
31
OR assets of less than Rate is based on risk.
€43 million.) Current range is from
1.03 to 7.43 percent.
Limits on executive
compensation,
dividends, and
distributions of capital.
KfW Syndicate Loans Participation in loan for Firms operating in KfW provides up to 80
investment and Germany with 50 or percent of the loan,
operating costs. more employees. with a limit of 50
Firm can’t be in percent of the total debt
default, and the bank of the firm. Minimum
must state that the firm size of €25 million,
can repay the loan. with limits on size.
Maturity of up to 6
years.
Rate and other terms
are the same as for the
private lenders.
Limits on executive
compensation,
dividends, and
distributions of capital.
Economic Stabilization Debt guarantees for up Troubled German firms Total guarantees of up
Fund (WSF) to 60 months that are two of: to €400 billion.
Assets over €43
million, revenues over
€50 million, and 250 or
more employees; and
have experienced
financial hardship as a
result of COVID-19.
Economic Stabilization Purchases of capital Troubled German firms Total investments of up
Fund (WSF) instruments that are two of: to €100 billion.
Assets over €43
million, revenues over
€50 million, and 250 or
more employees; and
have experienced
financial hardship as a
result of COVID-19.
There must be an
important interest of the
federal government in
the stabilization of the
company.
ECB Pandemic Low-cost funding Euro area banks Full allotment
Emergency Longer- operations providing
term Refinancing liquidity at a rate 25
Operations (PELTROs) bps below the rate on
32
Main Refinancing
Operations (currently
-25 bps).
ECB Targeted Long- Low-cost funding Euro area banks Rate on TLTRO III
term Refinancing operations can be as
Operations (TLTROs) low as -100 bps
through March 2021.
Banks will receive this
low rate if their loans
do not decline over the
year.
Note: The German government and the state governments also have programs to provide support to firms.
Source: KfW (2020); Juenemnn, Wirtz, and Leitmann (2020); Lane (2020).
33
The mission of the Hutchins Center on
Fiscal and Monetary Policy is to improve
the quality and efficacy of fiscal and
monetary policies and public
understanding of them.
Questions about the research? Email communications@brookings.edu.
Be sure to include the title of this paper in your inquiry.
© 2020 The Brookings Institution | 1775 Massachusetts Ave., NW, Washington, DC 20036 | 202.797.6000
34
File and source
- File
- designing-the-main-street-lending-program-challenges-and-options.pdf
- Size
- 585,739 bytes
- SHA-256
- 67863aa5d92a860b4f99715b65d50271ea6526722727c3642efacd4a322d266e
- Original
- www.brookings.edu