Type: article
Abstract: Artificial intelligence (AI) has become a central driver of firm productivity and innovation, yet little is known about how earnings pressure influences firms' AI investment decisions. We develop a model that yields two competing predictions. Under a cash-financing motive, AI is an inefficient target for cost relief, and any reduction would not depend on exploration intensity. Under a tolerance motive, exploration-loaded AI would be disproportionately affected because its value is harder to verify and its interim signals more closely resemble failure. Using a regression discontinuity design comparing firms just missing the analyst earnings per share (EPS) forecast with those just meeting it, we find that narrowly missing the target reduces AI investment over three years. The reduction concentrates among firms with high AI exploration share and is negligible among those with low exploration share. These findings show that earnings pressure selectively erodes tolerance for failure, the organizational precondition on which exploratory AI activity depends.
Authors: Zhao, Junyong, Liu, Jialu
Institution: Shanghai Jiao Tong University
Language: English
Subfield: Finance
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7222989402