Improving credit scoring by differentiating default behaviour
Improving credit scoring by differentiating default behaviour
We present a methodology for improving credit scoring models by distinguishing two forms of rational behaviour of loan defaulters. It is common knowledge among practitioners that there are two types of defaulters, those who do not pay because of cash flow problems (‘Can’t Pay’), and those that do not pay because of lack of willingness to pay (‘Won’t Pay’). This work proposes to differentiate them using a game theory model that describes their behaviour. This separation of behaviours is represented by a set of constraints that form part of a semi-supervised constrained clustering algorithm, constructing a new target variable summarizing relevant future information. Within this approach the results of several supervised models are benchmarked, in which the models deliver the probability of belonging to one of these three new classes (good payers, ‘Can’t Pays’, and ‘Won’t Pays’). The process improves classification accuracy significantly, and delivers strong insights regarding the behaviour of defaulters
credit scoring, statistics, domain-knowledge, constrained clustering, banking, game theory
771-781
Bravo, Cristian
b22c4145-644e-40ee-85d8-431c59c3c71b
Thomas, Lyn C.
a3ce3068-328b-4bce-889f-965b0b9d2362
Weber, Richard
da9918d6-bc84-4c98-8ffe-2aaf7b58cf1b
May 2015
Bravo, Cristian
b22c4145-644e-40ee-85d8-431c59c3c71b
Thomas, Lyn C.
a3ce3068-328b-4bce-889f-965b0b9d2362
Weber, Richard
da9918d6-bc84-4c98-8ffe-2aaf7b58cf1b
Bravo, Cristian, Thomas, Lyn C. and Weber, Richard
(2015)
Improving credit scoring by differentiating default behaviour.
Journal of the Operational Research Society, 66, .
(doi:10.1057/jors.2014.50).
Abstract
We present a methodology for improving credit scoring models by distinguishing two forms of rational behaviour of loan defaulters. It is common knowledge among practitioners that there are two types of defaulters, those who do not pay because of cash flow problems (‘Can’t Pay’), and those that do not pay because of lack of willingness to pay (‘Won’t Pay’). This work proposes to differentiate them using a game theory model that describes their behaviour. This separation of behaviours is represented by a set of constraints that form part of a semi-supervised constrained clustering algorithm, constructing a new target variable summarizing relevant future information. Within this approach the results of several supervised models are benchmarked, in which the models deliver the probability of belonging to one of these three new classes (good payers, ‘Can’t Pays’, and ‘Won’t Pays’). The process improves classification accuracy significantly, and delivers strong insights regarding the behaviour of defaulters
Text
ONLINE Improving Credit Scoring by Differentiating Defaulter Behavior.pdf
- Other
More information
Accepted/In Press date: 25 March 2014
e-pub ahead of print date: May 2015
Published date: May 2015
Keywords:
credit scoring, statistics, domain-knowledge, constrained clustering, banking, game theory
Organisations:
Centre of Excellence in Decision, Analytics & Risk Research
Identifiers
Local EPrints ID: 376179
URI: http://eprints.soton.ac.uk/id/eprint/376179
ISSN: 0160-5682
PURE UUID: 842c7f54-7938-4982-83ce-4530138c14b3
Catalogue record
Date deposited: 24 Apr 2015 13:17
Last modified: 15 Mar 2024 03:33
Export record
Altmetrics
Contributors
Author:
Lyn C. Thomas
Author:
Richard Weber
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics