Research
Research
This paper examines how AI exposure shapes productivity and creativity in innovation. Using an instrumental-variable approach, I find that inventors with higher AI exposure generate patents that are more explorative, less textually distinctive, and span a broader set of technological domains, while patent productivity remains unchanged. Heterogeneity analyses reveal a clear specialization between AI developers and AI users: AI developers contribute primarily through fewer but more influential and distinctive inventions within a more focused set of technologies, while AI users translate AI exposure into broader scaling, exploration, and diffusion across domains. Additionally, firms with higher AI exposure tend to generate more customized innovation outcomes. These findings suggest that AI creates value through a division of innovative labor, where deep AI-building capabilities and broad deployment of AI tools jointly shape innovation across R&D.
Presented at: Goizueta Doctoral Research Conference (2025)
This paper studies how acquirers with AI-equipped teams perform on target selection, deal efficiency, synergy realization, and innovation. AI-adopting acquirers expand their search radius on target candidates to less similar industries and more distant locations. Post AI-adoption acquirers reduce headcounts in M&A teams yet initiate more deals, and reduce the duration to deal closure. These acquirers enjoy higher synergies by product differentiation and preempting competitive threats. Finally, these acquirers improve post-merger innovation performance, as evidenced by filed patents.
(With Tetyana Balyuk)
We study how e-commerce affects credit provided to small businesses in the U.S. Theoretical predictions are ambiguous as e-commerce can affect credit supply and demand in opposite directions. We find that small businesses have higher sales, larger orders, and more stable revenue after e-commerce adoption, suggesting that online sales are more efficient than offline trade. Lower cash flow volatility should reduce credit demand, because small businesses often borrow to finance liquidity shortfalls. Yet, we find that businesses obtain more credit after e-commerce adoption, even after controlling for higher efficiency. We confirm these findings using the staggered entry of Uber Eats in the food industry and other strategies. Our results are consistent with e-commerce relieving credit constraints both directly by increasing expected payoffs to lenders due to higher sales and indirectly by lowering lending costs due to hard data generated in e-commerce. We also find redistributional effects of e-commerce on credit across industries in local credit markets.