An Evaluation of the Fed-Treasury Credit Programs — Jeremy Stein (May 11, 2020)
Summary
Slides from a Princeton Bendheim Center for Finance Webinar on Covid-19, dated May 11, 2020, with introductory remarks by Markus Brunnermeier followed by Jeremy Stein of Harvard University on An Evaluation of the Fed-Treasury Credit Programs, based on joint work with Sam Hanson, Adi Sunderam, and Eric Zwick. The slides list the Main Street Facilities at $600 billion and the Corporate Credit Facilities at $750 billion, and say existing programs miss 26% of employment at public firms, 19% of sales and 14% of profits. The proposed design principles include preserving optionality as a Venture Capitalist of Last Resort, making aid widely available, and providing aid with less senior claims.
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
- An Evaluation Of The Fed Treasury Credit Programs
- Case
- An Evaluation Of The Fed Treasury Credit Programs
Cite this document
“An Evaluation of the Fed-Treasury Credit Programs — Jeremy Stein (May 11, 2020).”
Original at economics.princeton.edu. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/an-evaluation-of-the-fed-treasury-credit-programs/
, SHA-256 d6e51cf6c54feb66909f9d6e60736dc1b6a7f7618928ed7e409885091f62a099.Full text
Intro: MARKUS BRUNNERMEIER Twitter: @MarkusEconomist
Markus’ intro
Previous/future webinars
Nellie Liang Evaluating early Fed programs
Daron Acemoglu On the benefits of targeted policies
Paul Krugman The audacity of hope
Speakers
The 3 crises
Health crisis
Health contagion
Gaining time
Shut-down
Economy crisis
Supply (chains), demand (hoarding)
Lockdown
Financial crisis
Liquidity, solvency contagion
5/11/2020 3
2008 2020
Pre-crisis Build-up of imbalances Well balanced
- Run-up of credit - US gov. debt expansion
- Thinly capitalized - Corporate debt
(shadow) banks
Trigger Re-evaluation: real estate Corporate cash flow crash
- Regional correlation
Amplification HH & banks’ Corporate sectors
balance sheets balance sheets
Fin- sector Shadow banks FinTechs for mortgages
(part of banks) Banks still for SMEs
Structured CDOs CLOs
finance
Policy Stimulus Survival
objective
Speed Fast Extremely fast 4
Central Bank: Lending vs. Buying
Liquidity programs: lend against collateral
Downside risk: only if haircut was not large enough
No upside risk
1. Funding liquidity provision
2. Market liquidity provision: improve secondary markets
functioning
- funding to existing market makers
- “Market maker of last resort” (Buiter & Sibert 2007)
Credit program: buy risky asset
Should government sector take on credit risk? (“I Theory” yes)
Upside and downside risk (diversify) SPV
Which part of government?
1. Central bank Asset FED
2. Development banks (co-invests in loans) Pool
3. Fiscal authority – take on junior/equity tranch (diversify) Treasury
5/11/2020 5
The I Theory of Money
Brunnermeier-Sannikov (2018)
Risk free rate
Spreads
Term spread Yield curve management
Credit spread Corporate Bonds
contain
Expected loss +
Risk premium = price of risk * (exogenous risk + endogenous risk)
amplifications, spirals, run, …
non-linearities
Government steps in as intermediary, redistributes risk and
creates money reduces deflationary pressures 7
I Theory: Which assets to buy?
Consumption demand management vs. Risk perspective
Bottleneck approach
Financial sector repair transmission channel
Household sector
Corporate sector
Asset Purchase Programs/QE
Which assets? Gov. bond, mortgages vs. corporate bonds
Sequencing
Interaction with DMO (at Treasuries)
Buy via auction or like FX intervention
Make most use to stabilize prices given limited ammunition ($4 tr.)
Negative Interest rate: ZLB vs. Reversal Interest Rate
5/11/2020 8
Inverse policy prescription
with Arvind Krishnamurthy
Dos-and-don'ts are reversed
Usual recession:
Stimulus focused
interest rates to stimulate spending
and investment
COVID recession:
Survival focused
5/10/2020 9
Inverse policy prescription
Chapter 11 is good solution for large firms, but
SME need a pause
Usual recession:
Avoid evergreening is a problem b/c it crowds out credit to new
firms/start-ups (Japan …)
COVID recession:
Promote evergreening
offer banks cheap central bank refinancing
for rolled-over loans (special program x% below policy rate)
stabilize existing businesses
5/10/2020 10
Poll 01:
1. Did the Fed benefit from its crisis experience in 2007-09
1. Yes
2. No
2. Paradox of Prudence (I Theory)
1. Assuming risk and redistributing it across all nominal claim
holders can reduce overall risk
2. Risk is a zero sum game (whack a mole game)
3. Should Fed take on (credit) risk and distributed among
many nominal claim holders?
1. Yes
2. No, it risk its independence
5/11/2020 13
End of MARKUS’ INTRODUCTORY REMARKS
Now Please ask questions in Q&A box
An Evaluation of the Fed-Treasury
Credit Programs
(based on joint work with Sam Hanson, Adi Sunderam, and Eric Zwick)
Princeton Bendheim Center for Finance Webinar on Covid-19
Jeremy Stein
Harvard University
May 11, 2020
Why Support Businesses in the First Place?
• COVID-19 shock is a massive shock
→ Inward shifts in both aggregate supply and demand
→ Large loss of output, staggering unemployment
• Overarching goals of economic policy:
• Smooth consumption
• Allocate losses fairly and efficiently
• Contain forces that can amplify the initial economic shock
• Broad consensus that there should be a large expansion of social
insurance programs for households
• Unemployment insurance
• SBA Paycheck Protection Program: $659 billion
• Up to $10 million loan, converts to cash grant if 75% spent on payroll
• Rationale for providing direct support to firms (capital) is
somewhat less obvious and policy design more controversial
Classic Lender-of-Last Resort Logic
• Bagehot’s rule: lend freely to solvent firms, against good collateral,
at a penalty rate.
• I.e., lend to firms that are illiquid but fundamentally solvent.
• Underlying theory of the case: think of Diamond-Dybvig (1983)
• Lending by central bank eliminates the bad run equilibrium. In good no-run
equilibrium, everybody is solvent, can pay back their loans.
• In hindsight (and with a good bit of luck) this LOLR approach is a
decent superficial characterization of 2008-09.
• TARP funds were almost entirely repaid, Fed didn’t lose a nickel.
• Looks ex post to have been in significant part a liquidity crisis.
• Not to downplay importance of solvency-driven interventions, e.g. stress tests.
Mnuchin as Bagehot?
This Is Not 2008-09!
• Magnitude of the fundamental shock is much larger: this is not
primarily a liquidity event for most firms.
• Uncertainty about firm solvency is first order:
• We do not know how long the public health emergency will last.
• We do not know what the post-COVID world will look like (e.g., how
much business travel there will be 3 years from now).
• And economic interpretation of solvency is less clear-cut
• Normally, near-zero revenues for 12 months and inability to service debt
are strong signals that a business is not economically viable.
• Less obviously the case here.
• Is a dental practice that has minimal revenues over the next year non-
viable? What should become of the assets?
High-Level Policy Goals: Micro and Macro
• Micro-efficiency: prevent destruction of socially valuable business
capital: both physical and organizational.
• Goal is not to shield debt/equity investors from losses
• But technology for allocating losses—the bankruptcy system—is
imperfect and risks excessive destruction of productive capital
• And bankruptcy system becomes less efficient when capacity is strained
(Iverson, 2018)
• Even if many firms need to ultimately be restructured, benefits to
“flattening the curve”
• Macro: contain amplification mechanisms
• Financial accelerators: firm, household and bank balance sheets
• Fire sales in credit markets: implications for credit spreads on new loans
• Aggregate demand externalities and Keynesian multipliers
• Congestion externalities in bankruptcies courts
Core Design Principles
• Preserve optionality: “Venture Capitalist of Last Resort”
• In environment of high uncertainty, control government’s exposure
not with ex ante credit standards, but with staged-financing approach
• Provide enough aid for firms to stay alive for a few months, then
reassess as more is learned.
• Must be willing to lose money: there are few dead-safe loans to be
made. Need to pay to preserve option value.
• Make aid widely available
• Do not impose excessively stringent ex ante credit standards
• Bagehot “solvency and good collateral” criteria not appropriate here
• Provide aid with less senior claims (i.e., more like preferred plus
warrants than senior debt)
• Preserves firm balance sheets and reduces future cashflow problems
• Mitigate debt overhang that would otherwise slow recovery
Existing Programs: Overview
• Corporate Credit Facilities: $750 billion
• Investment grade firms (+ recent fallen angels)
• Few other restrictions
• Main Street Facilities: $600 billion
• Firms up to 15,000 employees or up to $5b in annual revenues
• Many restrictions: leverage, bank risk retention, executive
compensation, distributions to shareholders
• Facilities akin to special purpose banks: Fed lends, equity
investment from Treasury funds appropriated in CARES Act
Existing Programs: Public Firm Coverage
Program Eligibility N Employment Sales EBITDA
% of Total
Included: PMCCF 12.0% 57.9% 65.4% 69.6%
Included: Main Street 33.0% 15.8% 15.3% 16.2%
Included: SBA PPP 24.5% 0.3% 1.0% 0.9%
Excluded: Large, high-yield (or unrated) 4.7% 19.2% 13.0% 10.7%
Excluded: Mid-sized, levered, high-yield 25.8% 6.7% 5.2% 2.6%
Excluded: Investment grade with no debt 0.0% 0.1% 0.0% 0.0%
Total 100.0% 100.0% 100.0% 100.0%
• Existing programs miss 26% of employment at public firms
• 19% of sales and 14% of profits
• Significant employment at
• Large, high-yield firms (too big to qualify for Main Street)
• Mid-sized, high-yield (too levered to qualify for Main Street)
Existing Programs: Private Firm Coverage
Program Eligibility N Payroll Sales EBITDA
% of Total
Included: Main Street 2.4% 42.6% 45.9% 42.9%
Excluded: Big retail 0.6% 6.1% 13.6% 7.7%
Excluded: Too small 97.0% 51.2% 40.5% 49.4%
Total 100.0% 100.0% 100.0% 100.0%
Average debt per firm in the Too Small bucket: $81,212
Fed programs also miss vast majority of private firms
◦ Too small, not very profitable
◦ Some of these firms are potentially covered by PPP.
Why the Gaps?
• Overall program design suggests aversion to taking credit risk
• Bagehot dictum at work?
• No junk-rated firms allowed in Primary Market Corporate Credit
Facility (PMCCF)
• Leaves a large chunk of public-company employment uncovered.
• Suggested tweak: admit BB and B-rated firms. But take warrants to
improve government’s expected return, align incentives.
• Also: caps on exec comp, shareholder distributions to better screen.
• And more tightly control quantity that they can initially borrow.
Leverage limits in Main Street programs.
Main Street Programs in More Detail
Main Street Design Concerns
• Bank risk retention: designed to ensure government is making loans on
“commercial” terms—seen as positive-NPV by banks.
• Given all the externalities at play, this is not the right social criterion.
• Banks more likely to participate when new loan bails out an existing troubled
position—allocative distortion.
• And probably won’t participate otherwise, even if loan is socially positive-NPV
• Accelerated repayment schedules
• E.g. in new loan program, 1/3 of principal in each of years 2-4.
• Likely to create cashflow problems for firms that borrow.
• Somewhat more moderate in priority loan program, but at cost of giving
lenders a more senior claim—thereby exacerbating debt overhang problems.
• Tight credit standards
• Debt/EBITDA < 4.0x in new loan program.
Hardness of Debt Claims a Particular Worry
• Combination of fast repayment plus senior claims likely to put many
Main Street borrowers in distress when economy is still fragile.
• Who will manage the workouts? Does Treasury delegate to a third party?
• Given participation of banks, workouts may have to be on private-market
terms; if so, reduced scope for socially-desirable recontracting ex post.
• Senior bank lenders may be relatively liquidation-prone.
• A better option: finance with preferred claims
• Interest payments can be deferred without forcing default.
• More junior status lessens debt overhang, makes it easier to attract future
rounds of financing.
• Can add warrants to strengthen government’s overall position, align
incentives.
• Again, venture-capital analogy is helpful: what is the right way to provide
finance to firms in an environment of high uncertainty?
How To Think About Staged-Finance Aspect?
Cap loan size: Protect taxpayers by limiting the amount firm can
borrow, not by excluding bad firms.
Easy to estimate recurring fixed obligations (rent, interest
payments, lease payments) from 2019 tax return as
Obligations =Revenues − Cost of Goods − Wages
− Depreciation − Profits
Might allow firms to borrow up to ¼ of this amount each quarter.
Dynamically adjust program in as public health conditions change.
◦ If situation deteriorates, gradually reduce the amount firms can borrow
in subsequent tranches needed wave of bankruptcies can take place
in an orderly way.
In Sum
• This is not 2008-09, and we need a different approach to
government financial support of firms. Classic LOLR does not
address the problem at hand.
• Given enormous uncertainty, desirable program features include:
• Wide access: not too stringent on ex ante credit quality
• Staged financing: control exposure via quantity allocated, preserve
optionality
• Relatively junior claims and ability to defer interest.
• Other mechanisms to align incentives and screen borrowers: warrants,
caps on executive comp and shareholder distributions.
• Government as venture capitalist of last resort.
• Have to be prepared to take significant losses if things go south.
• Not LOLR policy, but rather a Fed-leverage-enhanced form of fiscal policy.
Slightly more formal sketch
t=0 t=1 t=2
G
G
Y frictionless
Y frictionless
BG
Y frictionless
B
Y frictionless
BB
Y frictionless
Frictionless fundamentals determined by health outcomes
◦ Could also imagine frictionless fundamentals affected by past
decisions (i.e., liquidate or not), but ignore here for simplicity.
Slightly more formal sketch
t=0 t=1 t=2
G
G
Y frictionless
Y frictionless
BG
Y frictionless
B
Y frictionless
BB
Y frictionless
Private market outcome: Fixed point in failures and actual output:
fail = f Private (Yactual )
Yactual = g (Y frictionless , fail )
Slightly more formal sketch
failures
Yactual = g (Y frictionless
G
, fail )
fail = f Private (Yactual )
Yactual
Private market outcome without government intervention.
Slightly more formal sketch
failures
Yactual = g (Y frictionless
G
, fail )
fail = f Private (Yactual )
fail = f Social (Yactual )
Yactual
Government intervention shifts failure curve down big effect if
fundamental resolves to good state (though govt may still have losses).
Private actors take Yactual as given when choosing fail; a benevolent
government internalizes effect of fail on Yactual
Slightly more formal sketch
failures Yactual = g (Y frictionless
BB
, fail )
fail = f Private (Yactual )
fail = f Social (Yactual )
Yactual
In contrast, not much effect if we have a bad health outcome.
Funds to survive 2 periods + possibility of very negative outcome
Govt reluctant to commit Keep firms alive for 1 period and reassess
File and source
- File
- an-evaluation-of-the-fed-treasury-credit-programs.pdf
- Size
- 2,251,032 bytes
- SHA-256
- d6e51cf6c54feb66909f9d6e60736dc1b6a7f7618928ed7e409885091f62a099
- Original
- economics.princeton.edu