Underwriting Loans in the Recovery Phase of COVID-19 - A Government-Private Small Business Recovery Loan Program

Sep 19, 2026

Designing a government-subsidized small business recovery loan program requires aligning capital constraints, pricing, and default risk with practical underwriting. The note argues that insufficient public capital should de-risk private investment rather than fund all borrowing, while high expected defaults make true-cost pricing burdensome for survival. It recommends government loss-capital backstops via tranches or waterfall structures, careful use of guarantees, and underwriting rules that remain simple, automatable, and focused on viable firms across different business sizes.

LanguageEnglish

CategoryLetters

FormatPDF

Pages4

Reading Time2 min

AccessFree to download

Section I: Structure of lending program based on economic “truths”

Not Enough Government Capital

Expected Default

Backstops

Government & Bank Skin-in-the-Game

Underwriting to Viable Businesses

Underwriting to Different Size Businesses

A government-private structure for small business loans in the Recovery Phase of COVID-19 By Adair Morse Soloman P. Lee Chair in Business Ethics, Associate Professor of Finance, and Faculty CoDirector of the Sustainable and Impact Finance Initiative, Haas School of Business, University of California, Berkeley Fellow, Berkeley Center for Law & Business, Law School, University of California, Berkeley Research Associate, NBER The point of this note is to lay out how I would design a small business recovery loan program, subsidized by the government at some level, in the Covid-19 recovery summer of 2020. Contents: Section I: Structure of lending program based on economic “truths” Section II: Underwriting / eligibility standards for a loan fund Section I: Structure of lending program based on economic “truths” 1) Not Enough Government Capital My first truth is there is not sufficient government capital to provide subsidized loans for all the demand. Now that we are starting to see openings from shelterin-place rules, governments should use public funds to de-risk other capital, not to provide loans directly. The capital will reach manifold more small businesses and provide more sustaining loan sizes. 2) Expected Default The second truth is that the expected default rate of a small business recovery loan today is way too high for loans to be priced at their true costs. Many small businesses simply may not survive. Thus, the “fair rate” of interest that a lender must charge is way above normal times. My guess is that the “fair rate” is in the order of 25%, if not a lot more. A twenty-five percent interest rate implies a very large debt payment burden on companies, such that the debt itself may tip the company into failure and bankruptcy. Therefore, a small business recovery loan program must be at least partially subsidized by government or philanthropy backstops. 3) Backstops Backstops in this context are financial contract mechanisms whereby governments or philanthropies put up “walls” to protect private investors from at least some of the defaults. They can set up these backstops as guarantees or, preferably, as tranches or waterfall cash flow structures. My third truth is that governments, especially local governments, must be careful about guarantees. If the virus re-emerges and we go into more shelterings, the default rates might be much higher ; locking a government in to untenable liabilities. Thus, my preferred way for governments to subsidize loan programs is to use their capital as “loss capital” in debt tranches or waterfall distributions to de-risk the loans. Importantly, such structures allow for leveraging-up government loss capital. For example, consider a simple waterfall mechanism, commonly used in private equity. All payments coming in from borrowers (in sum, not on a loan-by-loan basis) go to the private investors until they are repaid their investment. Governments and philanthropist get no loan repayments until the water falls to the next level down. Variants of this mechanism through tranching is easily designed. Using waterfalls or tranches is marginally harder to set up ex ante for loan programs, as it requires an ex ante structure to pool funds, rather than just alending platform to originate and service the loans. 4) Government & Bank Skin-in-the-Game The fourth truth is that governments and banks are depending on recovery for their revenues into the future, and thus they have a monetary incentive tied to loan programs. Banks hope the programs will help support existing loans to struggling small businesses. Governments hope the programs will restore the bases for sales, income, property, and other taxes that support the entire system of public goods. These implications mean not only are they going to be biased toward any programs that help their future revenues (i.e. , toward their customers and constituents), but also that they should be most willing to participate in t

FAQs

Why does the note argue that governments should not provide subsidized loans directly to all demand?

It states that government capital is not sufficient to cover all demand. Governments should use public funds to de-risk other capital so funding reaches more small businesses and supports larger loan sizes.

What is the central pricing concern raised about recovery loans?

The note claims the expected default rate is too high for loans to be priced at their true costs, implying a “fair” interest rate around 25% or higher. Such rates create heavy debt burdens that may push firms into failure, so programs need at least partial subsidization.

What backstop structures does the note prefer for government or philanthropy support?

It prefers using government loss capital in debt tranches or waterfall cash-flow structures rather than guarantees, aiming to protect private investors from some defaults while avoiding untenable government liabilities if defaults rise again.

Sep 19, 2026

Upload by supergirl