Year: 2025
Type: article
Abstract: Credit card line assignment, a crucial component of the underwriting process, has traditionally relied on segmentation and profitability-based approaches. This article presents a novel machine-learning framework that enhances the precision and efficiency of credit line determinations. The proposed methodology employs clustering algorithms to identify distinct customer segments based on payment behaviors, spending patterns, and credit utilization trends. By leveraging premium bureau variables and engineered trend metrics, the framework develops sophisticated behavioral profiles that serve as the foundation for optimized line assignments. The optimization process incorporates both risk minimization and profit maximization objectives while adhering to regulatory constraints and portfolio management guidelines. This article demonstrates that the machine learning approach offers superior granularity in customer segmentation and improved adaptability to evolving market conditions compared to traditional methods. While the framework presents certain challenges regarding data quality requirements and model interpretability, the observed improvements in portfolio performance and operational efficiency suggest that machine learning-based line assignment represents a significant advancement in credit card underwriting practices.
Author: Pavan Rupanguntla
Institution: Bank of America
Language: English
Subfield: Marketing
Domain: Social Sciences
Open Access status: diamond
OpenAlex ID: W4407694440