Please rotate your device to landscape mode to view the charts.

Role of substantive and rhetorical signals in the market reaction to announcements on AI adoption: a configurational study

Journal: European Journal of Information Systems

Date: 2024

Author: Rohit Nishant, Tuan (Kellan) Nguyen, Thompson S. H. Teo, Pei-Fang Hsu

Abstract:
How do shareholders respond to technologies hyped in general discourse, e.g., artificial intelligence (AI), if a common understanding is lacking and the technologies are still evolving? Do they respond primarily to substantive signals in technology announcements, such as AI capabilities, or do rhetorical signals also play a significant role? Adopting signalling theory as a theoretical lens, we conceptualise announcements of AI capabilities as substantive signals and linguistic elements in the announcements pertaining to organisational time horizon and risk-reward considerations as rhetorical signals. Departing from the typical focus on bijective relationships, we consider holistic, complex configurations of interdependent factors using the qualitative comparative analysis (QCA) methodology. Notably, announcements pertaining to AI capabilities are not necessarily associated with positive market reactions; in fact, when all three types of AI are included in announcements without explicit consideration of risks, shareholders react negatively. We find that shareholder response is based on joint evaluation of substantive and rhetorical signals, and that these signals interact in a complex way to produce positive and negative market reactions. These findings motivate several propositions for market reactions to IT announcements, providing implications for both theory and practice.

Link: Google Scholar

Background and Context

Study Focus

This research examines how stock markets react to corporate announcements about artificial intelligence (AI) adoption, analyzing whether substantive signals (AI capabilities) or rhetorical signals (linguistic elements) drive investor responses.

Theoretical Framework

The study employs signaling theory to understand how different types of information in AI announcements influence market perceptions of a firm's potential and value.

Methodology

Researchers analyzed 409 AI announcements from 238 publicly traded US firms between 2015-2019 using qualitative comparative analysis (QCA) to identify configurations associated with positive and negative market reactions.

Three Types of AI Capabilities Have Different Market Impacts

Automation AI Automates repetitive tasks Mature technology Never core for positive returns Insight AI Analyzes data for insights Moderate maturity Most common in positive returns Engagement AI Interacts with users Newest technology Mixed market reaction Market Perception of AI Maturity More familiar More novel Different AI types signal different maturity levels impacting how investors respond to announcements
  • Automation AI is viewed as more mature technology but is never a core driver of positive market reactions.
  • Insight AI appears in most configurations leading to positive market returns, striking a balance between novelty and understanding.
  • Engagement AI is the newest type with a mixed market reception, requiring specific rhetorical signals to gain positive reactions.

Combining AI Types Without Risk Considerations Leads to Negative Returns

AI Capability Combinations and Market Reactions Positive Market Reaction Negative Market Reaction Maximum of Two AI Types + Risk Consideration + Appropriate Rhetorical Signals All Three AI Types + No Risk Consideration = Market Skepticism Investors view combining all AI types without risk assessment as overly ambitious and unrealistic
  • Configurations with a maximum of two AI capabilities coupled with appropriate risk considerations often received positive market reactions.
  • Announcements featuring all three AI types without explicit risk considerations led to negative market reactions.
  • This finding suggests investors are skeptical of firms attempting to implement all AI types simultaneously.

Different AI Types Require Different Rhetorical Signals for Positive Reception

  • Different AI capabilities require specific rhetorical framings to generate positive market reactions.
  • Automation AI needs future and risk focus to overcome skepticism about its incremental value over traditional automation.
  • Insight AI benefits from emphasis on present-future-reward or generally low rhetorical signals.

Market Reaction Varies Based on AI Capability Combinations

  • Insight AI appears in the highest proportion of configurations leading to positive market reactions.
  • The presence of all three AI capabilities together is consistently associated with negative market reactions.
  • Automation AI is more frequently associated with negative market reactions than positive ones.

Rhetorical Signals Play Significant Role in Market Response to AI Announcements

  • Future-focused rhetorical framing is dominant in configurations leading to positive market responses.
  • Reward-focused language without risk considerations often leads to negative market reactions.
  • This suggests investors view AI as a futuristic technology requiring balanced presentation of both opportunities and challenges.

How Substantive and Rhetorical Signals Combine to Influence Market Reactions

Key Winning and Losing Configurations Insight AI + Present/Future/Reward focus Proposition 2 Engagement AI + Reward focus Proposition 3 All 3 AI Types + Low Risk focus Proposition 4c Insight+Engagement AI + Risk focus without Present focus Market reactions depend on complex configurations of AI capabilities and rhetorical signals
  • Specific configurations of AI capabilities and rhetorical signals lead to distinctly positive or negative market reactions.
  • Insight AI with present, future, and reward focus creates a compelling narrative that resonates with investors.
  • Attempting to implement all three AI capabilities without addressing risks indicates poor strategic planning to investors.

Contribution and Implications

  • The study extends signaling theory by showing how substantive and rhetorical signals interact to influence market perceptions.
  • Companies should carefully consider which AI types to emphasize and how to frame announcements to optimize market reception.
  • Managers should avoid announcing all three AI types simultaneously and should clearly address risks in AI announcements.
  • Different AI types require different rhetorical framing strategies to achieve positive market reactions.
  • Investors appear to reward focused AI strategies rather than broad, potentially overambitious implementation plans.

Data Sources

  • Visualization 1 is based on the authors' conceptualization of AI capabilities in section 2.3 and findings in section 4.
  • Visualization 2 is based on the finding in section 4.4c that combining all three AI types leads to negative reactions.
  • Visualization 3 uses data from Table 7 and Figure 4, showing proportions of rhetorical signals in configurations.
  • Visualization 4 is derived from Table 6 and Figure 4a showing AI capability distribution in positive/negative configurations.
  • Visualization 5 uses data from Figure 4b showing the proportion of rhetorical signals in positive vs. negative reactions.
  • Visualization 6 represents the key propositions (2, 3, and 4c) developed from the QCA analysis in section 4.