Suyeol Yun
Research Subjects
Suyeol Yun has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Alejandro Lopez-Lira
University of Florida, Gainesville, FL, USA
1 shared publication - Chaewoon Kim
LinqAlpha, New York, USA
1 shared publication - Chanyeol Choi
LinqAlpha, New York, USA
1 shared publication - Hojun Choi
LinqAlpha, New York, USA
1 shared publication - Jaeseon Ha
LinqAlpha, New York, USA
1 shared publication - Jihoon Kwon
LinqAlpha, New York, USA
1 shared publication
Latest Publications
- Open Access
- 14 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 632-637
FinAgentBench: A Benchmark Dataset for Agentic Retrieval in Financial Question Answering
- 14 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 632-637
Open AccessFinAgentBench: A Benchmark Dataset for Agentic Retrieval in Financial Question Answering
- Open Access
- 1 January 2025
- Published by Elsevier in SSRN Electronic Journal
Decoding Congressional Stock Trades: An Industry and Committee-Focused Analysis with Graph Neural Network and Large Language Model
In this study on congressional stock trading, I refine the estimation of excess returns by focusing on the congressperson-ticker-timing level, departing from the traditional congress-aggregate level, calendar-based yearly approach methods which are less specific. This enhanced estimation provides a more detailed and relevant analysis for individual trades. Additionally, I advance the understanding of influential factors beyond traditional methods that relied on manual feature design and regression on tabular data. By employing a Graph Neural Network (GNN), adept at machine learning in graph-structured data, I effectively train the model to learn how to aggregate information channeled through the network, identifying industry sectors and committee assignments as pivotal factors in congressional stock trading. To address the interpretability challenges of the GNN, I initially use GNNExplainer, and subsequently introduce a Large Language Model (LLM) for predictive analysis and theorizing. Although the LLM does not yet equal the GNN in predictive power, it provides crucial interpretability, enabling deeper insights into the motivations and behaviors in congressional stock trading. This comprehensive approach emphasizes the importance of specific industry and committee assignments in these transactions and underscores the necessity for more stringent scrutiny and regulatory frameworks to ensure ethical financial practices.
- 1 January 2025
- Published by Elsevier in SSRN Electronic Journal
Open AccessDecoding Congressional Stock Trades: An Industry and Committee-Focused Analysis with Graph Neural Network and Large Language Model
In this study on congressional stock trading, I refine the estimation of excess returns by focusing on the congressperson-ticker-timing level, departing from the traditional congress-aggregate level, calendar-based yearly approach methods which are less specific. This enhanced estimation provides a more detailed and relevant analysis for individual trades. Additionally, I advance the understanding of influential factors beyond traditional methods that relied on manual feature design and regression on tabular data. By employing a Graph Neural Network (GNN), adept at machine learning in graph-structured data, I effectively train the model to learn how to aggregate information channeled through the network, identifying industry sectors and committee assignments as pivotal factors in congressional stock trading. To address the interpretability challenges of the GNN, I initially use GNNExplainer, and subsequently introduce a Large Language Model (LLM) for predictive analysis and theorizing. Although the LLM does not yet equal the GNN in predictive power, it provides crucial interpretability, enabling deeper insights into the motivations and behaviors in congressional stock trading. This comprehensive approach emphasizes the importance of specific industry and committee assignments in these transactions and underscores the necessity for more stringent scrutiny and regulatory frameworks to ensure ethical financial practices.