KIT | KIT-Bibliothek | Impressum | Datenschutz

A Vertical Mixture Cure Model for Credit Risk Analysis

Wycinka, Ewa; Jurkiewicz, Tomasz


Credit risk assessment is one of the most important tasks of banks and other financial institutions. There are three main reasons of credit termination: maturity, early repayment and default. Credits that mature can be considered as not susceptible to early termination, whereas early repayments can be treated as competing risk to default. Most credits end on time (mature) or are repaid early, default happens only for a few percentage of credits. Modelling probability of default requires taking into account the probability of early repayment and maturity. We propose the use of a vertical mixture cure model with a cured fraction to analyse the probability of default. Empirical research was conducted on the sample of 5,000 consumer credit accounts of a Polish financial institution. Credits were observed 24 months since origination. The vertical mixture cure model was estimated with characteristics of borrowers as predictors. The discrimination ability of the model through 24 months of the credit life span was compared with a mixture model that has been earlier proposed in the literature.

Verlagsausgabe §
DOI: 10.5445/KSP/1000085951/21
Veröffentlicht am 22.12.2021
Cover der Publikation
Zugehörige Institution(en) am KIT Institut für Wirtschaftsinformatik und Marketing (IISM)
Publikationstyp Zeitschriftenaufsatz
Publikationsjahr 2018
Sprache Englisch
Identifikator ISSN: 2363-9881
KITopen-ID: 1000141452
Erschienen in Archives of Data Science, Series A
Band 4
Heft 1
Seiten P21, 15 S. online
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page