442. Statistical Properties and Power Analysis of Divergence Measures for Credit Risk Model Monitoring]{Statistical Properties and Power Analysis of Divergence Measures for Cr
Invited abstract in session Business Management in Dynamic Emerging Markets, stream Selected Aspects of International Finance and OR.
Authors (first author is the speaker)
| 1. | Abdullah Karasan
|
| Department of Computer Science and Electrical Engineering, UMBC |
Abstract
Monitoring distributional shifts is vital in volatile financial sectors. While PSI is standard, Jensen-Shannon Divergence (JSD) and Kullback-Leibler Divergence (KLD) offer unique benefits, such as symmetry and Bayesian utility. Extending Yurdakul and Naranjo (2020), we derive chi-square benchmarks for JSD and KLD and test them on credit default models. Results show JSD provides superior Type I error control, minimizing false positives, whereas PSI and KLD offer higher power in small samples. This allows practitioners to balance false alarms against detection sensitivity.
Keywords
- Data Science
- Finance
- Performance Measurement
Status: accepted
Back to the list of papers