Detecting Information Channels in Congressional Trading via Temporal Graph Learning

TL;DR

This paper presents a novel temporal graph learning framework to detect potential information channels in congressional trading, revealing complex temporal dependencies linked to trading outperformance.

Contribution

It introduces a multimodal dynamic graph approach and a walk-forward validation method to identify congressional trading signals with temporal dependencies.

Findings

TGN effectively captures temporal dependencies in congressional trading data.

The framework identifies trades with significant outperformance over the S&P 500.

The approach reduces look-ahead bias in dynamic edge classification.

Abstract

Congressional stock trading has raised concerns about potential information asymmetries and conflicts of interest in financial markets. We introduce a temporal graph network (TGN) framework to identify information channels through which members of Congress may possess advantageous knowledge when trading company stocks. We construct a multimodal dynamic graph integrating diverse publicly available datasets, including congressional stock transactions, lobbying relationships, campaign finance contributions, and geographical connections between legislators and corporations. Our approach formulates the detection problem as a dynamic edge classification task, where we identify trades that exhibit statistically significant outperformance relative to the S&P 500 across long time horizons. To handle the temporal nature of these relationships, we develop a two-step walk-forward validation architecture that respects information availability constraints and prevents look-ahead bias. We evaluate several labeling strategies based on risk-adjusted returns and demonstrate that the TGN successfully captures complex temporal dependencies between congressional-corporate interactions and subsequent trading performance.