Not All Information Hits the Market at Once
Financial markets are normally described as hyper-efficient systems, capable of instantly absorbing all available information. But reality tells a different story. Investors (especially institutional ones) operate under severe attention constraints, juggling vast portfolios and reacting to constant streams of news. In this crowded informational environment, not everything gets processed equally or immediately. The complexity of corporate narratives, specifically, is an overlooked driver of market dynamics.
The hidden timing of financial storytelling
In their recent paper “Investor distraction and multi-dimensional financial narrative”, published in the Review of Accounting Studies, Miles Gietzmann (Department of Accounting, Bocconi University), Francesco Grossetti (Department of Accounting and Bocconi Institute for Data Science and Analytics, Bocconi University), and Craig M. Lewis (Vanderbilt University) explore how investors process the language used in corporate disclosures.
Turning their attention to the Management Discussion and Analysis (MD&A) sections of 10-K filings (annual reports that publicly traded companies in the United States must submit to the Securities and Exchange Commission), the authors move beyond traditional but one-dimensional measures like tone and introduce a broader concept: multi-dimensional narrative.
Inside the method: how AI reads corporate language
To move beyond traditional measures of tone, the authors adopt a sophisticated methodology that blends human judgment with machine learning. They begin by collecting over 12,000 sentences from MD&A sections and having them manually classified by hundreds of respondents across multiple dimensions—such as optimism, specificity, and directness. These annotated data are then used to train Naïve Bayes classifiers, a statistical learning technique capable of recognizing linguistic patterns at scale. Building on this foundation, the researchers construct the Aggregate Attribute Index (AAI), a composite measure that captures the overall richness and structure of corporate narratives. Crucially, the empirical design isolates the effect of language by focusing on firms that release financial numbers before the full 10-K, allowing the study to attribute subsequent market reactions specifically to narrative content rather than to new quantitative information. This combination of human insight, algorithmic analysis, and clever identification strategy enables a much more precise understanding of how investors process complex disclosures.
When complexity meets distraction
The central insight of the paper is both intuitive and counterintuitive: investors do react to multi-dimensional narratives—but not right away.
“Narrative complexity does not trigger short-term return responses but significantly affects stock prices over longer horizons”
In the short term, especially when investors are distracted by other events in their portfolios, complex disclosures are only partially processed. Over time, however, as attention returns, these narratives are gradually incorporated into stock prices.
This creates a delayed pricing effect, challenging the established view of instant market efficiency.
The mechanics of delayed market reactions
The study shows that investors initially rely on simple cues like overall sentiment but struggle to immediately decode richer, multi-dimensional narratives.
“Disclosures convey multi-dimensional linguistic signals… that require cognitive effort to interpret and are incorporated into prices more gradually”
This means that complexity is not ignored—it is simply processed in stages.
The result is a staggered reaction pattern:
- Immediate but incomplete adjustment
- Gradual refinement as attention increases
Why traditional metrics fall short
Standard tools—like readability scores or dictionary-based sentiment—fail to capture the real informational content of narratives. The AAI, by contrast, reveals that what truly matters is the interaction of multiple linguistic dimensions, not just tone. This suggests that markets are sensitive to deeper layers of communication, layers that require time and cognitive effort to unpack.
The broader implication is therefore that markets are not just shaped by information, but by attention dynamics.
When investors are distracted, complex information is underweighted, price adjustments are delayed and opportunities for predictability emerge.
For firms, this raises strategic questions about how to design disclosures. For investors, it opens the door to identifying mispriced assets driven by temporary inattention.
In an era of information overload, the real bottleneck is not data but attention. This study shows that corporate narratives, especially complex ones, play a crucial role in shaping market outcomes.
