Research

Publications

Distortions Caused by Lending Fee Retention, (with Travis L. Johnson)

Forthcoming, Management Science. See SoFi's Guide to Share Lending discussing our paper 

The Term Structure of Short Selling Costs

Review of Finance 27 (2023), 2125–2161

Under Revision

Bank Information Production Over the Business Cycle, (with Cooper Howes)

Revise and Resubmit, Review of Economic Studies 

~CICF, ETH Macro-finance Conference, FDIC Bank Research Conference, NFA, Fixed Income and Financial Institutions Conference, IBEFA, MFA, Procyclicality Symposium, BSE Summer Forum, EFA

The information banks have about borrowers drives their lending decisions and macroeconomic outcomes, but this information is inherently difficult to analyze because it is private. We construct a novel measure of bank information quality from confidential regulatory data that include banks’ private risk assessments for US corporate loans. We show that our measure of information quality improves as local economic conditions deteriorate, particularly for newly originated loans and loans with greater information sensitivity. Our results provide empirical support for theories of countercyclical information production in credit markets. 

Adverse Selection in Corporate Loan Markets, (formerly titled "Bank Loan Markups and Adverse Selection, with Mehdi Beyhaghi and Cesare Fracassi)

~SFS Cavalcade (Recording of Talk), FDIC/JFSR Bank Research Conference (Recording of Talk), MFA, IBEFA, CICF, Community Banking Research Conference (Recording of Talk), Corporate Finance Conference at WashU, FTG Bridging Theoretical and Empirical Research in Finance, AFA, Utah Winter Finance Conference (Recording of Talk) 

John W. Ryan Award for Most Significant Contribution to Community Banking Research, 2022 Community Banking Research Conference

Revise and Resubmit, Journal of Finance

Traditional models of competition predict a positive relationship between market concentration and prices. However, in loan markets, adverse selection can reverse this relationship as riskier borrowers become more likely to receive funding. Using supervisory data, we show that interest rates, banks' private risk assessments and collateralization are higher in markets with more banks. We also create a measure of markup that is orthogonal to borrower risk and show that markups are higher in markets with more banks and after repeated borrowing relationships. We provide causal support for the adverse selection channel by using a shock to large banks' lending costs.

What's in a Debt? Rating Agency Methodologies and Firms’ Financing and Investment Decisions, (with Cesare Fracassi)

Revise and Resubmit, Review of Corporate Finance Studies

~EFA, SFS Cavalcade, UNC/Duke Corporate Finance Conference, WFA 

In July 2013, Moody’s unexpectedly increased the amount of equity credit that speculative-grade firms receive for preferred stock from 50% to 100%. Firms affected by the rule change were suddenly considered less levered by Moody’s even though their balance sheets did not change. These firms responded by issuing debt, targeting a leverage ratio as defined by Moody’s, and growing their assets. The rule change transferred value from bond to equity holders, and led to an increase in preferred stock issuance. How rating agencies assess risk thus has a significant causal impact on firms’ financing, investment, and security design decisions. 

Working Papers

Reputational Algorithm Aversion

~Columbia & RFS AI in Finance Conference, AEA Annual Meeting

People are often reluctant to incorporate information produced by algorithms into their decisions, a phenomenon called ``algorithm aversion''. This paper shows how algorithm aversion arises when the choice to follow an algorithm conveys information about a human's ability. I develop a model in which workers make forecasts of a random outcome based on their own private information and an algorithm's signal. Low-skill workers receive worse information than the algorithm and hence should always follow the algorithm's signal, while high-skill workers receive better information than the algorithm and should sometimes override it. However, due to reputational concerns, low-skill workers inefficiently override the algorithm to increase the likelihood they are perceived as high-skill. The model provides a fully rational microfoundation for algorithm aversion that aligns with the broad concern that AI systems will displace many types of workers. 

Relationship Finance in Bond Markets? Evidence from Corporate Call Policy, (with Paul Beaumont  and David Schumacher)

~Aarhus Workshop on Strategic Interaction in Corporate Finance, NFA 

The FinReg Blog Summary of Paper 

Can relationship lending be sustained in public financial markets? We use firms' call decisions as a laboratory to study this question. After forcing existing bondholders to sell their bonds back at below market prices, i.e., a fixed-price call, existing bondholders are far less likely to participate in the firm’s subsequent bond issuances. The effects are strongest for the firms most valuable bondholders, leading to a deterioration in firms' bondholder bases. In turn, firms delay calling their bonds when they have more valuable investors in their bondholder base. Finally, firms' borrowing costs are affected by the reputation they develop from their historical call decisions. Our results reveal the importance of relationship lending in public markets and show how call policy can serve as a mechanism to build and sustain bondholder relationships.

The Information Advantage of Banks: Evidence From Their Private Credit Assessments, (with Mehdi Beyhaghi and Cooper Howes)

~Conference in Financial Economics Research by Eagle Labs, UT Finance PhD Alumni Conference, IBEFA Summer Meeting, Finance Forum, Stress Testing Research Conference, MFA, Midwest Macro, Society for Economic Dynamics, EFA, NFA, Conference on Markets and Intermediaries, Summer Workshop on Money, Banking, Payments, and Finance, FDIC Bank Research Conference

 In classic theories of financial intermediation, banks mitigate information frictions by monitoring and producing information about borrowers. However, it is difficult to test these theories without being able to observe banks' private information. In this paper, we use supervisory data containing banks' private assessments of their loans' expected losses. We show that changes in expected losses predict firms' future stock returns, bond returns and earnings surprises. The predictability is concentrated among small firms and growth firms, and only occurs when banks become more pessimistic. Using within-firm variation in borrowing across banks, we identify sources of private information for banks and show that this information affects banks' credit allocation decisions. Our findings show that banks' information production and monitoring create an information advantage over financial markets, even among publicly traded firms.

The Impact of Beliefs on Credit Markets: Evidence from Rating Agencies, (with Chen Wang)

~CICF, Boulder Summer Conference on Consumer Financial Decision Making, NFA, AFA, MFA

We analyze the impact of rating agencies’ beliefs on credit markets. We measure their beliefs as the difference between their forecasts of aggregate credit spreads and the consensus. When rating agencies become more optimistic, they issue higher ratings even though their forecasts do not predict future credit spreads. This optimism leads to lower initial bond yields and subsequent negative excess returns. Firms respond by increasing their leverage and investment. Finally, rating agencies become more optimistic as their head economists’ property values increase. Our analysis shows how subjective beliefs drive aggregate financing and investment through mispricing in credit markets.

Information Externalities in Opaque Credit Markets, (with Mahyar Sefidgaran)

~OxFIT, BSE Summer Forum,  Finance Forum, Midwest Theory, Vienna Festival of Finance Theory, Cambridge Corporate Theory Symposium

In many opaque markets plagued by asymmetric information, e.g., interbank and OTC markets, firms borrow from many lenders at once and individual contracts are not observable to other lenders. We identify a novel information externality in a model based on this type of setting. Due to adverse selection, lenders use their private information to adjust the size of loans rather than the prices they offer to borrowers. Each lender’s individual rationing decision creates an information externality that raises both lender profits and the efficiency of trade. This information externality occurs even though information is not shared and lenders compete with each other. The model provides a microfoundation for adverse selection-based peer monitoring in opaque credit markets and has implications for their optimal structure.  

Margin Lending and Information Production

~OFR PhD Symposium on Financial Stability, Paris December Meeting, AFBC, MFA

Many types of financial institutions borrow using margin loans. I propose a new explanation for the widespread use of these contracts in the financial system. In my model, margin loans prevent lenders from producing too much information about borrowers' assets at origination, by forcing borrowers to liquidate prior to maturity following a negative shock. However, while margin loans deter information production in the primary market, they induce it in the secondary market, depressing asset prices. The analysis can rationalize policies that curb margin lending in normal times and support it in periods of market stress.

Media Coverage:

"Distortions Caused by Lending Fee Retention": Reuters, Canadian Investment Review