ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login

ISSUES OF DEVELOPING AND BEHAVIRAL STATISTICAL MODELS OF CREDIT RISK

Journal: International scientific journal "Internauka." Series: "Economic Sciences" (Vol.1, No. 33)

Publication Date:

Authors : ; ;

Page : 40-46

Keywords : statistical model; application scoring; behavioral scoring; credit portfolio; default; default probability; Bayesian analysis; odds ratio;

Source : Downloadexternal Find it from : Google Scholarexternal

Abstract

In the article deals theoretical approaches to the construction of scoring models, depending on the data type and purpose of the models. Credit score has been shown to play a stronger role in large banking organizations, thanks to the requirements of the Basel Capital Accord (Basel II). The metrics development process must be shared between information technology (IT), data processing and operational staff. It also leads to a reassessment of methodologies and strategies for indicators based on the Basel II recommendations. The application of monitoring using scoring models requires analysis of bank lending performance in the context of solving its business problems. The concept of building a risk profile is an analysis of indicators that represent the main information categories. A practical example of developing a behavioral scoring model, based on the availability of quality data, is demonstrated. It is proved that the consideration of indicators is an instrument of managerial decision making. Tables should be seen as a tool for better decision-making and should be understood and monitored. Development of indicators should not complicate the model as it should be sufficiently clear for decision making or diagnostics. It is proved that the statistical models of credit scoring of retail borrowers are a modern tool for monitoring and managing bank limits. Indicators used for lending should be statistically sound, empirically obtained and capable of separating creditworthy from non-creditworthy applicants at a statistically significant rate.

Last modified: 2021-03-18 22:44:41