Essays on digital financial reporting

As financial information becomes increasingly digital, understanding how market participants access, use, and are affected by digital financial reporting is crucial for academics, regulators, and investors alike. This dissertation presents three essays that explore different dimensions of digital financial reporting, with a particular focus on the U.S. Securities and Exchange Commission’s (SEC) EDGAR (Electronic Data Gathering, Analysis, and Retrieval) system. By leveraging novel datasets such as EDGAR server log files and outage data, this work provides new insights into how digital access to financial information shapes investor behavior, information acquisition patterns, and market outcomes. Chapter 2 offers a systematic literature review on the use of EDGAR log files in archival finance and accounting research. It describes the structure and content of the SEC’s server log data, documents how researchers have used this unique dataset to examine investor attention, and categorizes the literature into key themes. A topic modeling (LDA) analysis reveals distinct research clusters focused on retail investor behavior, automated information retrieval, and attention-driven trading. This chapter also highlights gaps in the literature and proposes new research questions that could be addressed using EDGAR logs, such as modeling investor sophistication, exploring seasonal attention shifts, or understanding the impact of machine-based queries on capital markets. Chapter 3, titled “503 Service Unavailable: When Acquiring Information Goes Wrong,” investigates the capital market consequences of temporary server outages on the EDGAR platform. Using an event-study methodology and a difference-in-differences design, this essay quantifies how access disruptions affect trading behavior and market liquidity. The results show that even short-term outages significantly reduce market liquidity and investor responsiveness, particularly among retail and short-horizon traders. Further analysis reveals heterogeneity in the impact depending on firm characteristics and the timing of earnings announcements. The study underscores the market’s dependence on stable, timely digital access to financial disclosures and provides evidence supporting investments in digital infrastructure resilience. Chapter 4 examines the relative usefulness of quarterly versus annual reporting through the lens of digital information acquisition. Using EDGAR log files to trace investor download activity, the study compares how users interact with 10-Q (quarterly) and 10-K (annual) filings. Results show that quarterly reports drive significantly more frequent and time-sensitive access, especially among investors with short-term horizons. Additionally, the findings suggest that machine-based queries increasingly dominate the retrieval of quarterly filings, indicating a rise in algorithmic trading based on interim data. The chapter offers important implications for the ongoing debate around reporting frequency, highlighting the trade-off between timeliness and comprehensiveness in financial disclosures. Collectively, the three essays contribute to the understanding of digital financial reporting by (1) mapping the existing research landscape, (2) providing causal evidence on the importance of uninterrupted information access, and (3) exploring how reporting periodicity aligns with evolving user behavior in digital environments. This dissertation advances the literature by demonstrating how digital trace data can serve as a powerful tool for measuring investor behavior and evaluating the efficiency of financial reporting systems. It also informs regulators on the critical role of digital infrastructure and the behavioral effects of reporting regimes in today’s fast-paced, data-driven capital markets.