Hongyan Liang

Using the 2004–2011 Kauffman Firm Survey, we examine how entrepreneurial firms balance supplier financing (trade credit) and bank financing (revolving credit lines issued to the business, excluding owners’ personal loans). We distinguish trade credit used for transactions from trade credit used for financing by leveraging Dun & Bradstreet’s Paydex measure of payment timeliness. We document that banks’ lending to young firms appears weakly tied to borrowers’ financial ratios, whereas suppliers extend financing primarily to customers with strong sales and exhibit notable tolerance for late payments. Although start-ups are highly credit constrained and frequently pay suppliers late, they rely more on bank revolving credit than on supplier lending. Paradoxically, many firms that borrow from suppliers hold sufficient unused bank credit; we attribute this to the interest burden that accrues on unpaid bank balances, which is absent or less explicit for overdue trade credit. We find little evidence that trade credit availability enhances subsequent access to bank credit, or that bank and supplier financing substitute for one another—even during the 2007–2009 credit crunch. Our results highlight the distinct roles and screening criteria of banks and suppliers in financing young firms and the resulting coexistence of these two debt channels.

This study examines whether fintech lenders outperform traditional banks and non-fintech non-banks in risk assessment for U.S. 30-year fixed-rate conforming mortgages. Analyzing Fannie Mae and Freddie Mac loans from Q1 2012 to Q1 2020 using ROC/AUC and risk-pricing regressions, we find fintech lenders have lower predictive accuracy and pricing misalignment, charging higher rates to borrowers who remain current and lower rates to those who default or prepay. These results indicate that conforming mortgage regulations and rapid loan sales to government-sponsored enterprises (GSEs) diminish fintech firms’ incentives for enhanced borrower screening, thus reducing their risk assessment effectiveness.

Macroeconomic variables are crucial inputs used in the credit loss forecasting (LF) models and the use of macro effects is also mandated by regulators for stress testing purposes which allow banks to project the potential credit loss under different hypothetical macroeconomic scenarios. The COVID-19 pandemic has caused an unprecedented level of volatility in the macroeconomic variables, leading to new challenges to use macroeconomic variables in the LF and the current expected credit loss (CECL) modeling framework. Especially, the historical observed strong relationship between the macroeconomic variables and credit loss rate seems to disappear. This study examines the dynamic relationship between the charge-off rates on loans of all U.S. commercial banks and the macroeconomic factor from 1985 to 2020. This paper seeks to explore the correlation and predictive power of key macroeconomic indicators on the charge-off rates for three different portfolios – consumer, commercial and industry, and real estate loans. This study will shed light on which macroeconomic factors have high importance in the credit loss forecasting modeling. The results presented in this paper can also be served as a guideline for the macroeconomic variables selection process in the credit LF/CECL models development or model overlay design.

This paper assesses whether fintech mortgage lenders align pricing with borrower risk using conforming 30-year mortgages (2012–2020). We estimate default probabilities using machine learning (logit, random forest, gradient boosting, LightGBM, XGBoost), finding that non-fintech lenders achieve the highest predictive accuracy (AUC = 0.860), followed closely by banks (0.857), with fintech lenders trailing (0.852). In pricing analysis, banks adjust the origination rates most sharply with borrower risk (7.20 basis points per percentage-point increase in default probability) compared to fintech (4.18 bp) and non-fintech lenders (5.43 bp). Fintechs underprice 32% of high-risk loans, highlighting limited incentive alignment under GSE securitization structures. Expanding the allowable alternative data and modest risk-retention policies could enhance fintechs’ analytical effectiveness in mortgage markets.

This study examines racial and ethnic disparities in mortgage-lending outcomes across different lender types—large banks, fintech lenders, non-bank lenders, small banks, and credit unions—using Home Mortgage Disclosure Act (HMDA) data from 2018 to 2023. By analyzing approval rates, rate spreads, and origination charges, we evaluate how borrower outcomes vary by race and ethnicity, controlling for loan characteristics, borrower attributes, and regional factors. Our findings reveal that Black and Hispanic borrowers consistently face less favorable terms than White borrowers, with disparities differing by lender type. Large banks, operating under stringent regulatory oversight, demonstrate relatively equitable pricing but impose higher loan denial rates on minorities. Credit unions, despite offering the lowest rate spreads overall, penalize minority borrowers more severely in pricing than other lender types. Fintech lenders, while charging higher-rate spreads and fees, show smaller credit access disparities for minority borrowers. Non-bank and small banks display mixed results, with inconsistencies in their treatment of minorities across pricing and credit access. These results highlight that neither technological innovations nor alternative lending models alone suffice to eliminate systemic inequities. Achieving equitable mortgage lending requires enhanced regulatory oversight, greater transparency in algorithmic decision-making, and stricter enforcement of fair lending practices.

The neoclassical economic theories predict that increased competition can lead to lower prices; however, this prediction has yet to find empirical support in the consumer lending market. We find that, given the same ex post level of credit and prepayment risks, large banks charge the highest interest rates on GSE mortgages, fintech firms charge the lowest interest rates, and the interest rates charged by small banks and non-fintech non-banks fall in the middle with marginal differences between them. With the entrance of non-banks, the GSE mortgage interest rates decrease by roughly 7 basis points from 2010 to 2019.

The authors apply machine learning techniques to Loss Given Default estimation, identifying key variables in LGD prediction and evaluating the performance of

Purpose: In theory, the impact of debt liquidity risk (DLR) on the firm's future growth is ambiguous. This study aims to examine the empirical relationship between the DLR and firms' growth rate using annual data for USA companies from 1976 to 2020.Design/methodology/approach: Given the longitudinal nature of the data, the author uses OLS (ordinary least squares) regression methodology with fixed effects to control for unobserved characteristics that might affect the dependent variable. Instrument variable regression is also used to address the potential endogeneity problem.Findings: The results show that firms having higher DLR, as proxied by more short-term debt, experience lower growth rate. An increase in firms' short-term debt decreases the firms' future growth rate as evidenced by lower assets, revenue and employee growth rate. Moreover, the authors' results show that small firms or firms with more investment opportunities grow fast if the firms take higher DLR. Finally, cyclical firms with higher DLR exhibit lower growth rate during the credit tighten period. The authors' results hold for both the pre-zero lower bound (ZLB) era and ZLB period.Originality/value: To the authors' best knowledge, this is one of the earliest studies to carefully examine the effects of DLR on firms' growth rate. While prior research finds that firms with higher growth potential, measured by market-to-book (MTB) ratio, use more short-term debt, the authors' research directly addresses whether DLR affects firms' future growth rate. The authors’ findings also help explain why firms with high growth potential use more short-term debt.

Purpose: The unemployment rate (UR) is the leading macroeconomic indicator used in the credit card loss forecasting. COVID-19 pandemic has caused an unprecedented level of volatility in the labor market variables, leading to new challenges to use UR in the credit risk modeling framework. This paper examines the dynamic relationship between the credit card charge-off rate and the unemployment rate over time.Design/methodology/approach: This study uses quarterly observations of charge-off rates on credit card loans of all commercial banks from Q1 1990 to Q4 2020. Univariate, multivariable, machine learning, and regime-switching time series modeling are employed in this research.Findings: The authors decompose UR into two components – temporary and permanent UR. The authors find the spike in UR during COVID-19 is mainly attributed to the surge in temporary layoffs. More importantly, the authors find that the credit card charge-off rate is primarily driven by permanent UR while temporary UR has little predictive power. During recessions, permanent UR seems to be a stronger indicator than total UR. This research highlights the importance of using permanent UR for credit risk modeling.Originality/value: The findings in the research can be applied to the credit card loss forecasting and CECL reserve models. In addition, this research also has implications for banks, macroeconomic data vendors, regulators, and policymakers.

The purpose of this paper is to investigate the relationship between investors' divergence of opinions and the asset prices of foreign stocks and also examine the effect of home market country-level factors on the influence of divergency of opinions on stock price. The authors employ panel data estimation with fixed effects to examine the host market response in divergent opinions to the earnings announcements. The paper uses the American Depositary Receipts (ADRs) of 42 countries from 1985 to 2011. The authors find a negative relationship between differences of opinions and excess quarterly earnings announcement returns, and investors do process information asymmetrically based on good and bad earnings shocks. In addition, the authors find the negative relationship between divergent opinions and excess earnings announcement returns in ADRs is more pronounced in countries with short-sales restrictions, while other home-market country-level factors – the enforcement of insider trading law, legal origin, investor protection and rating on accounting standard – do not influence the relationship between investors' divergency of opinion and stock returns. This paper is among the first to bring asymmetric effects on convergence in Miller framework and enhance the understanding of price convergence documented in Miller (1977). In addition, this study incorporates home-market country-level factors in explaining the relationship between investors' divergency of opinions and stock returns.