Boyuan Li - Research

Job Market Paper:

AI Washing

Awards: Best PhD Presentation Award, AI in Finance Conference 2025

This paper investigates AI washing -- the exaggeration or misrepresentation of corporate investment in artificial intelligence (AI). Using large language models (LLMs), I develop novel measures of forward-looking AI investment plans from earnings call transcripts ("AI talk") and AI-related workforce expertise from employee resumes ("AI walk") for U.S. public firms from 2016 to 2024. I validate these measures by demonstrating that AI walk, but not AI talk, predicts subsequent AI patent quantity and quality. Within firms, past AI talk does not predict future AI walk, and AI washing incidents have surged since 2019, particularly among smaller, less capital-intensive manufacturing firms. Market rewards talk in the short run but discounts it in the long run, while walk earns large, persistent valuation gains only over longer horizons. Institutional investors are more discerning than the broad market, allocating capital toward high‐walk firms early on. Finally, firms with strong managerial incentives are more likely to raise talk without increasing walk, consistent with strategic hype. Overall, the results reveal a measurable and growing disconnect between AI rhetoric and real investment, reflecting the tension between short‐term market incentives and long‐term value creation.

Other Working Papers:

One Hundred and Thirty Years of Corporate Responsibility

with Joel Houston and Sehoon Kim

U.S. history has been punctuated by time-varying attitudes and shifting public discourse about the externalities and responsibilities of business. Applying natural language processing (NLP) techniques that account for context evolution in historical news text, we develop a monthly time-series index dating back to the late 19th century that measures public attention to environmental and social (E&S) issues related to business (ESIX). We explore the properties of ESIX and relate it to macroeconomic fluctuations, asset prices, and corporate decisions. Public attention to social issues around business arises during times of macroeconomic and social instability, whereas attention to environmental issues is heightened during times of relative prosperity. At the firm-level, positive exposure to such public attention is associated with lower future stock returns. Heightened E&S concerns reduce the level of corporate investments and weaken the link between corporate investments and Tobin's q in the short-run (i.e., 1-2 years out), but ultimately improve both the level and efficiency of corporate investments in the long-run (i.e., up to 10 years out). These findings indicate that markets are unable to fully price the long-term real effects of the demands for corporate responsibility.

The Emerging Greenium

with Baolian Wang and Jiawei Yu

We examine whether and to what extent investors are willing to forego financial returns in exchange for non-pecuniary benefits in the United States municipal bond market. We match municipal green bonds to otherwise almost identical non-green bonds from 2013 to 2022. Comparing 1,027 pairs of exact matches, we find green bonds are issued at a lower yield after 2018, with the average greenium being 2.3 basis points. The underwriter discount difference between green bonds and their matches was positive before 2018 and has become negative in recent years. In the post-2018 period but not before, the size of greenium is positively correlated with bond-level greenness and a proxy for the absence of greenwashing. These findings are associated with the sharp increase in investor awareness of Environmental, Social, and Governance (ESG) issues starting in 2019. We also find that the term structure of greenium is downward sloped. An international corporate bond sample reveals a similar post-2018 emergence and term structure of the greenium.

Job Market Paper:

AI Washing

Other Working Papers:

One Hundred and Thirty Years of Corporate Responsibility

The Emerging Greenium

Work in Progress:

Workforce Composition and Firm Performance: Education versus Skills

with Murillo Campello and Mingyin Zhu

Corporate Automation with Artificial Intelligence

Corporate ESG Reports and Institutional Investors