Print
E-MailJie (Jack) He is a Professor of Finance at the University of Georgia. This post is based on a recent paper by Professor He, Sean Cao, Associate Professor of Accounting at the University of Maryland, Itay Goldstein, Professor of Finance at the University of Pennsylvania, and Yabo Zhao, Assistant Professor of Finance at the Chinese University of Hong Kong, Shenzhen.
The emergence of new technologies confronts firms with difficult decisions on resource allocation. Contemplating investments in these emerging technologies, firm managers have to assess the opportunities in a constantly changing environment without past records to learn from and with limited models of the costs and benefits involved. In recent years, we have seen two such emerging technologies – artificial intelligence (AI) and green (i.e., climate/environment-related) technologies – rising to huge prominence and increasingly capturing the attention of firm managers. How do they decide whether and how to invest in these technologies? What sources of information do they rely on for such decisions? Having better answers to these questions is crucial not only for understanding firms’ investment behaviors but also for understanding modern societies’ technological transformation more broadly. After all, the collective decisions of different firms determine the extent of emerging technology investment and the scope of economic transformation that could follow.
In this paper, we look at a prominent source of information – the stock market – and investigate how it has been used in the process of emerging technology investment. Financial markets have been found to be a powerful source of information: They aggregate the opinions of a diverse body of investors, are forward-looking, and respond rapidly to announcements and economic developments. A growing strand of literature has argued and shown that informational feedback from the financial markets can help guide the decision-making of corporate managers (see, e.g., Bond, Edmans, and Goldstein (2012) and Goldstein (2023)). Building on these insights, we examine how firms use information from the stock market when deciding on their investments in AI and green technologies.
To identify the information from the market, we measure the stock market’s reaction to firms’ announcements of upcoming AI and green investments extracted from earnings conference calls and 8-K filings. We then study how firms adjust their investments in these emerging technologies in response to this market reaction. To measure firms’ AI (green) investments, we use a new proxy: the number of AI (green) job postings. These job postings provide a real-time indicator of a firm’s motive to invest in and its commitment of resource allocation to these technological areas.
We find that changes in firm-level AI and green investments from one year before to one year after a corresponding announcement are positively correlated with the cumulative abnormal return over a short window (e.g., 5 days) surrounding the disclosure event. This result is consistent with the idea that firm managers adjust upward (downward) their emerging technology investments when the stock market reacts favorably (unfavorably) to related discussions in their major corporate disclosures.
Our results point to the presence and importance of active managerial learning from the market feedback, as opposed to just a passive reflection of information in both market reactions and firm actions (e.g., the unobservable quality/prospects of the proposed AI or green technology investments): Nearly half of our sample disclosures involve negative market reactions, and the documented investment-price association is stronger in such cases. After exploring the possible reasons behind these negative feedback cases and conducting further analyses, we conclude that active managerial learning plays an essential role in explaining firms’ emerging technology investment response to market reaction. The positive investment-price association is also unlikely to be driven by other non-learning-based explanations, such as confounding effects from non-AI/green-related components of the corporate disclosures, financial constraints, or pre-disclosure trends in such investments.
Our analysis reveals several other nuanced dynamics regarding how firms learn from the market.
- Market Participants’ Domain Expertise and Managerial Uncertainty: The investment-price association is more pronounced when market participants possess more domain knowledge in emerging technologies and when managers are more uncertain about the proposed investments.
- Emerging vs. Conventional Technologies: We do not find similar feedback effects for conventional-technology investments, highlighting the unique uncertainty about emerging technologies.
- Differential “Peer Learning Effect” by Technology Types: Interestingly, firms learn from the market feedback of their competitors regarding green investments, but not AI. This indicates that green investments may be more affected by industry factors, while AI investments are viewed as firm-specific.
- Long-Term Payoff of Learning: Firms following the market feedback on emerging technology investments experience better long-run operating and stock performance.
In all, our paper constructs a comprehensive database of AI/green investment-related corporate disclosures and documents the trend and extent of such feedback seeking behavior by firm managers. Our findings suggest that actively seeking and utilizing market feedback can help mitigate managers’ ex-ante concerns and improve their ex-post investment decisions when they venture into uncertain areas such as emerging technologies. Our analyses also shed new light on what kind of information managers actually learn from the market. Establishing that managers exhibit a pattern of learning when it comes to emerging technologies, such as AI and green technologies, but not for traditional technologies, alongside other patterns of differential learning (e.g., managers learn from peers’ green feedback but not AI feedback) highlights what kind of feedback managers seek from the market. Moreover, the setting we study here – observing the short-window market reaction to a firm’s specific announcements on AI/green investment plans and then tracking how its corresponding investments evolve in response – enables us to achieve a degree of precision in identifying managers’ active learning from the market, which is generally difficult to obtain in this literature.
Full paper is available for download here.
