Survival Mixture Model for Credit Risk Analysis

Mo Leo S. F. and Yau Kelvin K. W.
Additional contact information
Mo Leo S. F.: City University of Hong Kong
Yau Kelvin K. W.: City University of Hong Kong

Asia-Pacific Journal of Risk and Insurance, 2010, vol. 4, issue 2, pages 1-20

Abstract: The survival mixture model, which is an extension of the ordinary survival model that allows the existence of a fraction of the borrowers to be risk-free, is applied to credit risk analysis. In a regression setting, the effect of borrowers' characteristics on both the risk-free probability and default risk can be assessed simultaneously. Using the C statistic as a measure of accuracy, the survival mixture model shows improved power to discriminate between ‘good' and ‘bad' customers, when compared with other commonly used statistical models for credit risk analysis. A simulation study is conducted to assess the performance of the proposed numerical estimation method. The survival mixture model not only concentrates on the time-to-default of the borrowers, it also predicts the probability of being risk-free. It provides additional information about the borrowers' default risk in relation to their characteristics, which assists the lending institutions to better manage credit risk.

Date: 2010
References: Add references at CitEc
Citations Track citations by RSS feed

Downloads: (external link)
http://www.degruyter.com/view/j/apjri.2010.4.2/apj ... .1061.xml?format=INT (text/html)
For access to full text, subscription to the journal or payment for the individual article is required.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: http://EconPapers.repec.org/RePEc:bpj:apjrin:v:4:y:2010:i:2:n:5

Ordering information: This journal article can be ordered from
http://www.degruyter.com/view/j/apjri

Access Statistics for this article

More articles in Asia-Pacific Journal of Risk and Insurance from De Gruyter
Series data maintained by Peter Golla ().