A non‐default rate regression model for credit scoring

Abstract

In this paper, we propose a new non‐default rate survival model. Our approach enables different underlying activation mechanisms which lead to the event of interest. The number of competing causes, which may be responsible for the occurrence of the event of interest, is assumed to follow a geometric distribution, while the time to event is assumed to follow an inverse Weibull distribution. An advantage of our approach is to accommodate all activation mechanisms based on order statistics. We explore the use of maximum likelihood estimation procedure. Simulation studies are performed and experimental results are illustrated based on a real Brazilian bank personal loan portfolio data. Copyright © 2015 John Wiley & Sons, Ltd.

Generalized cure rate model for infectious diseases with possible co-infections

“…Some researchers [ 24 – 26 ] have applied the cure rate model to evaluate loan performance and determine loan recovery. Additionally, the model has been applied to determine the percentage of convicts that would return to jail [ 27 ].…”

A zero-inflated non default rate regression model for credit scoring data

“…Consequently, as already mentioned, it can be said that bankers expect it to be rarely recorded within loan portfolios. This has been addressed in different papers, such as Abad et al [1], Banasik et al [4], Barriga et al [5], Bellotti & Crook [7], Leow & Crook [20], Louzada et al [24], Stepanova & Thomas [37] and Tong et al [38]. The reason for the widespread use of survival analysis in credit risk rather than other modeling techniques, besides monitoring the loan portfolio credit risk over time, is that it can accommodate censored data, which is not supported.…”

“…From these cases based on clinical studies, models that accommodate cure fractions of the data events, known as cure rate models, were introduced into the literature. In Barriga et al [5], the authors used a different terminology in order to clarify its use in a credit scoring setting. They denoted cure by non-default, leading to what they called by non-default rate models.…”

“…Furthermore, there is clear evidence that most of the loans are granted to people who actually want to pay it back and will, therefore there are customers who can be considered immune to default. Hence, the methodology to be introduced in this paper is intended for credit risk analysis and generalizes the usual cure rate model due to Berkson & Gage [8], by taking into account a non-zero probability of failure at time zero by STD borrowers, together with the (usual) non-zero probability of surviving up to any time t. As in Barriga et al [5], we maintain the notation of the non-default rate to make reference to the cured rate, leading to what we call the zero-inflated non-default rate model.…”

The Zero-Inflated Cure Rate Regression Model: Applications to Fraud Detection in Bank Loan Portfolios

“…Such data must be addressed to make a holistic risk management of the loan portfolio, that is, dealing with fraud prevention, delinquency control and ensure the customer loyalty growth.For using survival analysis techniques, we must consider the modelling outcome of interest be the survival time after loan concession, also mentioned as customer or loan survival time, which is represented by time to occurrence of the event of default. This has been done in different papers, such as [1,2,3,22]. The reason for the increased use of survival analysis in credit risk over other modelling techniques, besides allowing monitor over time the credit risk of the loan portfolio, is that it can accommodate censored data, which are not supported, for example, in credit scoring techniques based purely on good and bad client classification, see for instance [11,13,21].Notwithstanding, survival analysis deals with non-negative and censored data, however, generally without excess, or even, the presence of zeros.…”