MASCA: LLM based-Multi Agents System for Credit Assessment
Gautam Jajoo, Atharva Pandey, Pranjal A Chitale, Saksham Agarwal
Why It Matters
What makes this one worth your time
This work is relevant for AI researchers and practitioners interested in applying LLMs to financial applications, particularly in improving the fairness and accuracy of credit scoring systems.
MASCA leverages LLMs in a multi-agent framework to enhance credit assessment processes.
Summary
The paper introduces MASCA, a multi-agent system utilizing large language models (LLMs) to improve credit assessment by simulating real-world decision-making processes. It employs a layered architecture with specialized agents for sub-tasks and integrates contrastive learning for risk and reward evaluation. Theoretical insights are provided using signaling game theory, and a bias analysis in credit assessment is conducted. Experimental results suggest MASCA outperforms baseline methods.
Key contributions
- Development of a multi-agent system using LLMs for credit assessment.
- Theoretical insights into hierarchical multi-agent systems using signaling game theory.
- Bias analysis in credit assessment to address fairness concerns.
Notable insights
- Integration of contrastive learning for risk and reward assessment in credit evaluation.
- Application of signaling game theory to analyze hierarchical multi-agent systems.
Possible limitations
- Not stated in the abstract
Abstract
arXiv:2507.22758v2 Announce Type: replace Abstract: Recent advancements in financial problem-solving have leveraged LLMs and agent-based systems, with a primary focus on trading and financial modeling. However, credit assessment remains an underexplored challenge, traditionally dependent on rule-based methods and statistical models. In this paper, we introduce MASCA, an LLM-driven multi-agent system designed to enhance credit evaluation by mirroring real-world decision-making processes. The framework employs a layered architecture where specialized LLM-based agents collaboratively tackle sub-tasks. Additionally, we integrate contrastive learning for risk and reward assessment to optimize decision-making. We further present a signaling game theory perspective on hierarchical multi-agent systems, offering theoretical insights into their structure and interactions. Our paper also includes a detailed bias analysis in credit assessment, addressing fairness concerns. Experimental results demonstrate that MASCA outperforms baseline approaches, highlighting the effectiveness of hierarchical LLM-based multi-agent systems in financial applications, particularly in credit scoring.