AI Adoption and Corporate Investment Efficiency: Evidence from Chinese A-Share Listed Firms

Authors

  • Haoran Yu School of Economics, Guizhou University, Guiyang, 550025, China

Keywords:

Artificial Intelligence, Investment Efficiency, Over-Investment, Richardson Model

Abstract

This paper examines the association between AI adoption and corporate investment efficiency using panel data of Chinese A-share listed firms from 2003 to 2024. Drawing on the Richardson investment-efficiency framework, AI adoption is measured by the frequency of AI-related keywords in the MD&A section of annual reports, while investment inefficiency is measured by the absolute value of residuals from the expected-investment model. In the preferred specification with industry and year fixed effects and the full set of control variables, AI adoption is negatively associated with corporate investment inefficiency, with a coefficient of −0.0016 that is statistically significant at the 5% level. Given that the standard deviation of AI adoption is 1.108, a one-standard-deviation increase in AI adoption is associated with a 0.0018 decrease in investment inefficiency, equivalent to approximately 1.5% of its sample mean. The negative association is concentrated in the over-investment subsample, whereas the coefficient in the under-investment subsample is statistically insignificant. This asymmetric pattern is consistent with an agency-based governance interpretation. Heterogeneity analysis further shows that the negative association is descriptively more pronounced among non-state-owned enterprises, high-tech firms, and firms in the Eastern region. These patterns remain qualitatively similar across a series of robustness checks. The paper provides additional empirical evidence on the relationship between AI adoption and corporate capital allocation.

Downloads

Published

2026-11-21

How to Cite

AI Adoption and Corporate Investment Efficiency: Evidence from Chinese A-Share Listed Firms. (2026). CPS Digital Library - Series of Conferences, 42-52. https://seriesofconference.com/index.php/SCJ/article/view/566