From regulatory data to quantitative investment signals in equity markets

This dissertation presents five empirical studies that leverage regulatory disclosures filed with the U.S. Securities and Exchange Commission (SEC) as primary information sources to identify structural patterns, anomalies, and actionable signals for equity portfolio strategies. First, a large-scale exploratory analysis of the EDGAR database covers over 15 million disclosures from more than 800,000 entities between 1994 and 2024, revealing persistent declines in public company filings, rising private capital activity, pronounced seasonal trends in insider trading and disclosure volume, and filing surges preceding major market events. Notably, just 2% of form types account for 80% of annual submissions, highlighting concentration in regulatory reporting. Building on these findings, the second study documents a strong seasonal cycle in regulatory disclosure behavior, with winter months showing significantly higher activity across various filing types. This seasonality aligns with the “Halloween effect” in equity markets, where stock returns tend to outperform from November to April. The analysis suggests that patterns in disclosure timing contribute to this well-known market anomaly and are consistent across both U.S. and European markets. The third study investigates the effectiveness of replicating hedge fund strategies using holdings reported in quarterly SEC Form 13F filings. Analyzing over 150,000 portfolios between 2013 and 2023, it finds that cloned portfolios based on top-quartile hedge fund filings outperform the S&P 500 by 24.3% annually on a risk-adjusted basis, closely mirroring original fund returns when rebalanced at disclosure dates. The fourth study focuses on market reactions to Item 4.02 disclosures in Form 8-K, which report non-reliance on previously issued financial statements. Examining over 8,000 such filings from 2004 to 2023, it finds these disclosures are associated with significant negative abnormal returns, especially when related to revenue recognition errors or when the disclosure lacks clarity about the impact or magnitude of the misstatement. Finally, the fifth study develops an equity portfolio construction strategy based on insider trading activity disclosed in SEC Form 4 filings. Using Monte Carlo simulations on 1.8 million insider transactions, the strategy identifies high-conviction insider purchases that deliver a 5-day cumulative abnormal return of 3.4%. The top-performing backtested portfolio consistently outperforms the S&P 500 in both in- and out-of-sample periods, achieving strong risk-adjusted returns. Together, these studies demonstrate the value of regulatory disclosure data for understanding capital market behavior and for developing robust quantitative investment strategies.