Chen Li

Research

Working Papers

Artificial Intelligence in Recruitment: Examine the Effects on Hiring and Firm Outcomes

(Job Market Paper)

This paper examines the impact of AI-recruitment tools on human capital acquisition and firm outcomes. I construct a novel dataset by hand-collecting firm-level AI adoption events and linking them to employer–employee matched data administered by the U.S. Census Bureau. Using a staggered difference-in-differences design, I find that the adoption of AI recruitment tools is associated with significant improvements in the quality of newly hired workers, measured by estimated worker fixed effects, as well as increases in gender and racial diversity. Turnover rates among new hires decline after adoption, suggesting better candidate matching and stronger retention. I also find that higher-quality, instead of more diverse workers, are associated with increases in firm revenue and productivity.


Thy Bust, My Boom: Micro Evidence on Small Firms’ Tech Evolution after Dot Com Bubble Burst

with John (Jianqiu) Bai and Wenting Ma

This study investigates the impact of mass tech layoffs on non-tech firms. Using micro-level data from the U.S. Census, we find that non-tech firms in regions affected by tech layoffs experienced significant employment growth, particularly among small firms with fewer than 50 employees. This employment growth drives long-term gains in revenue and productivity for a subset of small firms that successfully hire displaced high-skill workers and navigate the challenges of adopting new technologies. These results highlight a crucial, yet often overlooked, externality: disruptions in the tech sector labor market can act as a catalyst for technology advancement and growth in traditionally less dynamic sectors.


Local Income Uncertainty and Peer-to-Peer (P2P) Lending

with Jue Wang

This paper investigates the effect of local income uncertainty on borrowing and lending behavior in peer-to-peer (P2P) credit markets. Using loan-level data from Prosper Marketplace LLC, we examine the relationship between state-level income uncertainty—distinguishing between employed and unemployed households—and consumer loan volumes. We find that rising income uncertainty narrows the loan volume gap between traditional intermediaries and P2P platforms. We further explore the role of the COVID-19 pandemic and the CARES Act in shaping borrower and lender responses. Our results show that the CARES Act, by reducing income uncertainty, increased borrowing from traditional institutions relative to P2P platforms. During the pandemic, lenders exhibited heightened caution toward unemployed applicants, while P2P credit remained an important financing channel for employed borrowers.

Published Papers

Sector Option Correlation Premiums and Predictable Changes in Implied Volatility

(with Apoorva Koticha and Joseph M. Marks)

We examine options listed on sector ETFs that constitute the S&P 500 and find evidence of predictability in implied volatilities associated with abnormally high or low implied correlations. We show that sector implied volatilities evolve to maintain stable relations between sector correlation premiums and the correlation premium on the S&P 500, allowing the calculation of a sector-specific, idiosyncratic correlation premium. The sector-specific correlation premium is a more reliable signal of future changes in sector implied volatility relative to simple level measures of the volatility or correlation premiums due to its focus on correlation rather than volatility, and its adjustment for aggregate levels. Moreover, we find that one-day reversals in sector implied volatilities are related only to reversals in the sector-specific correlation premium, and that information extracted from stock implied volatilities has little or no predictive ability for sector implied volatility. The predictable variation in sector implied volatilities associated with the sector-specific component of the correlation premium forms the basis for profitable trading signals that dominate strategies based directly on sector volatility premiums.