Can Generative AI Disrupt Post-Earnings Announcement Drift (PEAD)?

One of the most persistent market anomalies is the post-earnings announcement drift (PEAD) — the tendency of stock prices to keep moving in the direction of an earnings surprise well after the news is public. But could the rise of generative artificial intelligence (AI), with its ability to parse and summarize information instantly, change that?

PEAD contradicts the semi-strong form of the efficient market hypothesis, which suggests prices immediately reflect all publicly available information. Investors have long debated whether PEAD signals genuine inefficiency or simply reflects delays in information processing.

Traditionally, PEAD has been attributed to factors like limited investor attention, behavioral biases, and informational asymmetry. Academic research has documented its persistence across markets and timeframe. Bernard and Thomas (1989), for instance, found that stocks continued to drift in the direction of earnings surprises for up to 60 days.

More recently, technological advances in data processing and distribution have raised the question of whether such anomalies may disappear—or at least narrow. One of the most disruptive developments is generative AI, such as ChatGPT. Could these tools reshape how investors interpret earnings and act on new information?

Can Generative AI Eliminate — or Evolve — PEAD?

As generative AI models — specifically large language models (LLMs) like ChatGPT — redefine how quickly and broadly financial data is processed, they significantly enhance investors’ ability to analyze and interpret textual information. These tools can rapidly summarize earnings reports, assess sentiment, interpret nuanced managerial commentary, and generate concise, actionable insights — potentially reducing the informational lag that underpins PEAD.

By substantially reducing the time and cognitive load required to parse complex financial disclosures, generative AI theoretically diminishes the informational lag that has historically contributed to PEAD.

Several academic studies provide indirect support for this potential. For instance, Tetlock et al. (2008) and Loughran and McDonald (2011) demonstrated that sentiment extracted from corporate disclosures could predict stock returns, suggesting that timely and accurate text analysis can enhance investor decision-making. As generative AI further automates and refines sentiment analysis and information summarization, both institutional and retail investors gain unprecedented access to sophisticated analytical tools previously limited to expert analysts.

Moreover, retail investor participation in markets has surged in recent years, driven by digital platforms and social media. Generative AI’s ease of use and broad accessibility could further empower these less-sophisticated investors by reducing informational disadvantages relative to institutional players. As retail investors become better informed and react more swiftly to earnings announcements, market reactions might accelerate, potentially compressing the timeframe over which PEAD has historically unfolded.

Why Information Asymmetry Matters

PEAD is often linked closely to informational asymmetry — the uneven distribution of financial information among market participants. Prior research highlights that firms with lower analyst coverage or higher volatility tend to exhibit stronger drift due to higher uncertainty and slower dissemination of information (Foster, Olsen, and Shevlin, 1984; Collins and Hribar, 2000). By significantly enhancing the speed and quality of information processing, generative AI tools could systematically reduce such asymmetries.

Consider how quickly AI-driven tools can disseminate nuanced information from earnings calls compared to traditional human-driven analyses. The widespread adoption of these tools could equalize the informational playing field, ensuring more rapid and accurate market responses to new earnings data. This scenario aligns closely with Grossman and Stiglitz’s (1980) proposition, where improved information efficiency reduces arbitrage opportunities inherent in anomalies like PEAD.

Implications for Investment Professionals

As generative AI accelerates the interpretation and dissemination of financial information, its impact on market behavior could be profound. For investment professionals, this means traditional strategies that rely on delayed price reactions — such as those exploiting PEAD —  may lose their edge. Analysts and portfolio managers will need to recalibrate models and approaches to account for the faster flow of information and potentially compressed reaction windows.

However, the widespread use of AI may also introduce new inefficiencies. If many market participants act on similar AI-generated summaries or sentiment signals, this could lead to overreactions, volatility spikes, or herding behaviors, replacing one form of inefficiency with another.

Paradoxically, as AI tools become mainstream, the value of human judgment may increase. In situations involving ambiguity, qualitative nuance, or incomplete data, experienced professionals may be better equipped to interpret what the algorithms miss. Those who blend AI capabilities with human insight may gain a distinct competitive advantage.

Key Takeaways

• Old strategies may fade: PEAD-based trades may lose effectiveness as markets become more information-efficient.
• New inefficiencies may emerge: Uniform AI-driven responses could trigger short-term distortions.
• Human insight still matters: In nuanced or uncertain scenarios, expert judgment remains critical.

Future Directions

Looking ahead, researchers have a vital role to play. Longitudinal studies that compare market behavior before and after the adoption of AI-driven tools will be key to understanding the technology’s lasting impact. Additionally, exploring pre-announcement drift — where investors anticipate earnings news — may reveal whether generative AI improves forecasting or simply shifts inefficiencies earlier in the timeline.

While the long-term implications of generative AI remain uncertain, its ability to process and distribute information at scale is already transforming how markets react. Investment professionals must remain agile, continuously evolving their strategies to keep pace with a rapidly changing informational landscape.