Multimodal Readability, Linguistic Complexity, and Governance Signals in Corporate Financial Disclosures: An Integrated Theoretical and Empirical Analysis

Abstract

Corporate financial disclosures represent one of the most critical communication channels between firms and their diverse stakeholders. Over the past several decades, research has increasingly demonstrated that financial reporting is not merely a neutral transmission of numerical facts but a complex linguistic, strategic, and governance-driven process. This study develops a comprehensive, theory-driven examination of readability, linguistic complexity, multimodal sentiment, and corporate governance mechanisms in financial disclosures, with particular emphasis on annual reports, integrated reports, earnings calls, and regulatory filings. Drawing strictly on established literature in accounting, finance, communication, and corporate governance, this article synthesizes insights from textual readability research, agency theory, voluntary disclosure theory, and emerging multimodal deep learning approaches to sentiment analysis.

The study advances the argument that readability and linguistic clarity are not only outcomes of managerial communication skills but also reflections of underlying governance quality, agency costs, managerial incentives, and investor sophistication. It further argues that the evolution from purely textual disclosures to multimodal communication environments—where tone, sentiment, and narrative complexity interact across documents and spoken disclosures—fundamentally reshapes how investors interpret firm performance and risk. By conceptually integrating deep learning–based multimodal sentiment analysis with traditional readability measures, the article provides a unified framework explaining how language, governance, and disclosure channels jointly influence market judgments, analyst behavior, and firm valuation.

Methodologically, the study adopts a qualitative, theory-anchored synthesis approach, explaining in detail how prior empirical strategies measure readability, sentiment, and disclosure quality, and how these methods conceptually align with governance and performance outcomes. The findings synthesized from prior studies consistently suggest that more readable and linguistically accessible disclosures are associated with lower agency costs, stronger governance mechanisms, more persistent earnings, and improved investor decision-making, particularly among less sophisticated investors. Conversely, excessive complexity, obfuscation, or strategic ambiguity often signals managerial opportunism, weak monitoring, or heightened information asymmetry.

The discussion elaborates theoretical implications for agency theory, signaling theory, and behavioral finance, while also addressing counterarguments regarding managerial discretion and proprietary costs. Limitations of existing research are critically examined, including contextual dependencies, cross-country institutional differences, and the evolving nature of multimodal disclosures. The article concludes by outlining a forward-looking research agenda that emphasizes integrated governance–language models, cross-linguistic disclosure environments, and ethically grounded applications of artificial intelligence in financial communication analysis.