Reproduce “Needles in a haystack: using forensic network science to uncover insider trading” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Gian Jaeger, Wang Ngai Yeung, Renaud Lambiotte
A network-based screening method was developed to identify groups of corporate insiders exhibiting statistically improbable coordinated trading patterns from 2.9 million SEC Form 4 filings. This approach, validated through two distinct null models, successfully detected known insider trading cases and high-risk clusters within corporate networks.
Key insights
- The empirically derived network of coordinated insiders exhibits connectivity vastly exceeding expectations from both null models (Z-scores > 1,000 and 170-180), confirming that observed coordination is not random or due to regulatory constraints or common market responses.
- The method shifts focus from atomistic individual anomalies to relational misconduct, providing group-level evidence that supports prosecution by highlighting statistically improbable coordinated trading among insiders.
- Anomalous egonets and central nodes identified through network analysis reveal specific individuals and groups whose trading patterns align with regulatory notions of illicit coordination, even in the absence of labeled training data.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.