Skip to main content
SpringerLink
Log in

Exploring Population Drift on Consumer Credit Behavioral Scoring

  • Conference paper
  • First Online:
Operational Research in Business and Economics

Part of the book series: Springer Proceedings in Business and Economics ((SPBE))

Abstract

Behavioral credit scoring models are a specific kind of credit scoring models, where time-evolving data about delinquency pattern, outstanding amounts, and account activity, is used. These data have a dynamic nature as they evolve over time in accordance with the economic environment. On the other hand, scoring models are usually static, implicitly assuming that the relationship between the performance characteristics and the subsequent performance of a customer will be the same under the current situation as it was when the information on which the scorecard was built was collected, no matter what economic changes have occurred in that period. In this study we investigate how this assumption affects the predictive power of behavioral scoring models, using a large data set from Greece, where consumer credit has been heavily affected by the economic crisis that hit the country since 2009.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Log in via an institution
Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 129.00
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 169.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info
Hardcover Book
USD 169.99
Price excludes VAT (USA)
  • Durable hardcover edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Notes

  1. 1.

    Just to name a few constraints the European Consumer Credit Directive 2008/48/EC (http://ec.europa.eu/consumers/financial_services/consumer_credit_directive/index_en.htm) stipulates among other that an applicant has the right to be comprehensively informed about the reasons of a rejection; The Basel II Accord (http://www.bis.org/publ/bcbsca.htm) imposes specific requirements for risk evaluation that have to be accredited.

  2. 2.

    We shall note here that there is an expanding research in credit scoring –and especially behavioral scoring- to support decisions in areas such as marketing, through the use of propensity scores (Bijak 2011; Thomas 2003; Thomas et al. 2005): there are response models (will the consumer respond to marketing offers), usage models (will the consumer use a credit line) and attrition models (will a customer continue with the lender). A recent trend is also profit scoring, that is the use of scorecards to maximize profit (Andreeva et al. 2007; Crook et al. 2007; Finlay 2010). Additional areas in which scoring models find application are collection scoring (dividing insolvent customers into groups, separating those who require decisive actions from those who don’t need to be attended to immediately) and fraud detection (ranking applicants according to the relative likelihood that their application may be fraudulent) (Phua et al. 2010).

  3. 3.

    A customer may be an existing one who has already a credit history or a new one applying for first time. This distinction is important in the context of behavioral score.

  4. 4.

    Multi-class credit risk classification has not being extensively studied or applied in practice (see (Chen 2012; Hsieh et al. 2010; Tang and Qiu 2012) for an example of multi-class SVM for credit scoring).

  5. 5.

    We shall note here that these data reflect the traditional and well-established academic and financial industry perspective. However, there is an increasing fad (eg FICO score with use of alternative data, see Andriotis (2015) and similar approaches from Equifax and TransUnion) for using alternative data (mainly from sources such as utilities & telecommunication bill payments, but also from data from social networks, rentals etc). The aim is to outreach those with very thin data or serious delinquencies, sometimes mentioned as “credit invisibles” [see http://www.perc.net (Turner et al. 2015; 2009)].

  6. 6.

    http://www.bis.org/publ/bcbsca.htm

  7. 7.

    Concept drift is a wider phenomenon than population drift; it can refer eg to situations where the classes of the classification problem change over time (Kelly and Hand 1999).

  8. 8.

    IMF: Compilation Guide on Financial Soundness Indicators, https://www.imf.org/external/np/sta/fsi/eng/2004/guide/index.htm

References

  • Allen L, Saunders A (2002) A survey of cyclical effects in credit risk measurement models (SSRN Scholarly Paper No. ID 315561). Social Science Research Network, Rochester, NY

    Google Scholar 

  • Anderson R (2007) The credit scoring toolkit: theory and practice for retail credit risk management and decision automation. Oxford University Press, Oxford

    Google Scholar 

  • Andreeva G, Ansell J, Crook J (2007) Modelling profitability using survival combination scores. Eur J Oper Res 183:1537–1549. doi:10.1016/j.ejor.2006.10.064

    Article  Google Scholar 

  • Andriotis A (2015) FICO announces new credit score based on alternative data. Wall Street J. http://www.wsj.com/articles/fico-announces-new-credit-score-based-on-alternative-data-1427989748

  • Baesens B, Gestel TV, Stepanova M, Van den Poel D, Vanthienen J (2005) Neural network survival analysis for personal loan data. J Oper Res Soc 56:1089–1098. doi:10.1057/palgrave.jors.2601990

    Article  Google Scholar 

  • Baesens B, Van Gestel T, Viaene S, Stepanova M, Suykens J, Vanthienen J (2003) Benchmarking state-of-the-art classification algorithms for credit scoring. J Oper Res Soc 54:627–635

    Article  Google Scholar 

  • Bellotti T, Crook J (2014) Retail credit stress testing using a discrete hazard model with macroeconomic factors. J Oper Res Soc 65:340–350. doi:10.1057/jors.2013.91

    Article  Google Scholar 

  • Bellotti T, Crook J (2008a) Credit scoring with macroeconomic variables using survival analysis. J Oper Res Soc 60:1699–1707

    Article  Google Scholar 

  • Bellotti T, Crook J (2008b) Support vector machines for credit scoring and discovery of significant features. Expert Syst Appl 36:3302–3308

    Article  Google Scholar 

  • Besanko D, Thakor AV (1987) Competitive equilibrium in the credit market under asymmetric information. J Econ Theory 42:167–182. doi:10.1016/0022-0531(87)90108-6

    Article  Google Scholar 

  • Bijak K (2011) Kalman filtering as a performance monitoring technique for a propensity scorecard. J Oper Res Soc 62:29–37. doi:10.1057/jors.2009.183

    Article  Google Scholar 

  • Bijak K, Thomas LC (2012) Does segmentation always improve model performance in credit scoring? Expert Syst Appl 39:2433–2442. doi:10.1016/j.eswa.2011.08.093

    Article  Google Scholar 

  • Breeden J, Thomas L, McDonald J III (2007) Stress testing retail load portfolios with dual-time dynamics. J Risk Model Valid 2:1–19

    Google Scholar 

  • Chang S-Y, Yeh T-Y (2012) An artificial immune classifier for credit scoring analysis. Appl Soft Comput 12:611–618. doi:10.1016/j.asoc.2011.11.002

    Article  Google Scholar 

  • Chen Y (2012) Research on multi-classification of credit rating of small and medium-sized enterprises in growth enterprises board based on fuzzy ordinal regression support vector machine. Int J Econ Finance 4:248–252. doi:10.5539/ijef.v4n3p248

    Google Scholar 

  • Clémençon S, Vayatis N (2010) Overlaying classifiers: a practical approach to optimal scoring. Constr Approx 32:619–648. doi:10.1007/s00365-010-9084-9

    Article  Google Scholar 

  • Crook J, Banasik J (2005) Explaining aggregate consumer delinquency behaviour over time. Working Paper Series-University of Edinburgh Management School 5.

    Google Scholar 

  • Crook J, Bellotti T (2010) Time varying and dynamic models for default risk in consumer loans. J R Stat Soc A 173:283–305. doi:10.1111/j.1467-985X.2009.00617.x

    Article  Google Scholar 

  • Crook JN, Edelman DB, Thomas LC (2007) Recent developments in consumer credit risk assessment. Eur J Oper Res 183:1447–1465

    Article  Google Scholar 

  • De Jongh P, De Jongh E, Pienaar M, Gordon-Grant H, Oberholzer M, Santana L (2015) The impact of pre-selected variance inflation factor thresholds on the stability and predictive power of logistic regression models in credit scoring. ORiON 31:17. doi:10.5784/31-1-162

    Article  Google Scholar 

  • Durand D (1941) Credit-rating formulae. In: Durand D (Ed) Risk elements in consumer installment financing. NBER, pp. 83–91.

    Google Scholar 

  • Elton EJ, Gruber MJ, Blake CR (1996) Survivor bias and mutual fund performance. Rev Financ Stud 9:1097–1120

    Article  Google Scholar 

  • Fawcett T (2006) An introduction to ROC analysis. Pattern Recogn Lett 27:861–874

    Article  Google Scholar 

  • Finlay S (2010) Credit scoring for profitability objectives. Eur J Oper Res 202:528–537. doi:10.1016/j.ejor.2009.05.025

    Article  Google Scholar 

  • Gama J, Žliobait\.e I, Bifet A, Pechenizkiy M, Bouchachia A (2014) A survey on concept drift adaptation. ACM Comput Surv 46, 44:1–44:37. doi:10.1145/2523813

  • Gao J, Fan W, Han J, Yu PS (2007) A general framework for mining concept-drifting data streams with skewed distributions. In: Proceedings of SDM’07

    Google Scholar 

  • Giambona F (2012) A discrete-time hazard model for loans: some evidence from Italian banking system. Am J Appl Sci 9:1337–1346

    Article  Google Scholar 

  • Guyon I, Elisseeff A (2003) An introduction to variable and feature selection. J Mach Learn Res 3:1157–1182

    Google Scholar 

  • Hajek P, Michalak K (2013) Feature selection in corporate credit rating prediction. Knowl-Based Syst 51:72–84. doi:10.1016/j.knosys.2013.07.008

    Article  Google Scholar 

  • Hand DJ (2009) Measuring classifier performance: a coherent alternative to the area under the ROC curve. Mach Learn 77:103–123

    Article  Google Scholar 

  • Hand DJ (2006) Classifier technology and the illusion of progress. Statist Sci 21:1–14. doi:10.1214/088342306000000060

    Article  Google Scholar 

  • Hand DJ, Anagnostopoulos C (2013) When is the area under the receiver operating characteristic curve an appropriate measure of classifier performance? Pattern Recogn Lett 34:492–495. doi:10.1016/j.patrec.2012.12.004

    Article  Google Scholar 

  • Hand DJ, Henley WE (1997) Statistical classification methods in consumer credit scoring: a review. J R Statist Soc A 160:523–541

    Article  Google Scholar 

  • Harrell F (2015) Regression modeling strategies: with applications to linear models, logistic and ordinal regression, and survival analysis. Springer, Heidelberg

    Book  Google Scholar 

  • Hofer V, Krempl G (2013) Drift mining in data: a framework for addressing drift in classification. Comput Stat Data Anal 57:377–391. doi:10.1016/j.csda.2012.07.007

    Article  Google Scholar 

  • Holte RC (1993) Very simple classification rules perform well on most commonly used datasets. Mach Learn 11:63–90. doi:10.1023/A:1022631118932

    Article  Google Scholar 

  • Hsieh H-I, Lee T-P, Lee T-S (2010) Data mining in building behavioral scoring models. In: 2010 international conference on computational intelligence and software engineering (CiSE). IEEE, pp 1–4

    Google Scholar 

  • Im J, Apley DW, Qi C, Shan X (2012) A time-dependent proportional hazards survival model for credit risk analysis. J Oper Res Soc 63:306–321. doi:10.1057/jors.2011.34

    Article  Google Scholar 

  • Ingolfsson S, Elvarsson BT (2010) Cyclical adjustment of point-in-time PD. J Oper Res Soc 61:374–380. doi:10.1057/jors.2009.136

    Article  Google Scholar 

  • Kelly MG, Hand DJ (1999) Credit scoring with uncertain class definitions. IMA J Manag Math 10:331–345. doi:10.1093/imaman/10.4.331

    Article  Google Scholar 

  • Kelly, M.G., Hand, D.J., Adams, N.M., 1999. The impact of changing populations on classifier performance. In: Proceedings of the 5th ACM SIGKDD international conference on knowledge discovery and data mining, pp 367–371

    Google Scholar 

  • Kennedy K, Mac Namee B, Delany SJ, O’Sullivan M, Watson N (2013) A window of opportunity: assessing behavioural scoring. Expert Syst Appl 40:1372–1380. doi:10.1016/j.eswa.2012.08.052

    Article  Google Scholar 

  • Kraft H, Kroisandt G, Müller M (2002) Assessing the discriminatory power of credit scores. Discussion papers. Interdisciplinary Research Project 373: quantification and simulation of economic processes

    Google Scholar 

  • Liu Y, Schumann M (2005) Data mining feature selection for credit scoring models. J Oper Res Soc 56:1099–1108. doi:10.1057/palgrave.jors.2601976

    Article  Google Scholar 

  • Thomas L (2007) Measuring the discrimination quality of suites of scorecards : ROCs Ginis, bounds and segmentation. Presented at the Credit Scoring and Credit Control X, Edinburgh

    Google Scholar 

  • Thomas L, Jung KM (2013) When to rebuild or when to recalibrate scorecards. Presented at the Credit Scoring and Credit Control XIII, Edinburgh

    Google Scholar 

  • Maldonado S, Weber R, Basak J (2011) Simultaneous feature selection and classification using kernel-penalized support vector machines. Inf Sci 181:115–128. doi:10.1016/j.ins.2010.08.047

    Article  Google Scholar 

  • Marrs GR, Hickey RJ, Black MM (2010) The impact of latency on online classification learning with concept drift. In: Bi Y, Williams M-A (eds) Knowledge science, engineering and management. Lecture notes in computer science. Springer, Berlin, Heidelberg, pp 459–469

    Google Scholar 

  • Mays E (2005) Handbook of credit scoring. Publishers Group Uk, [S.l.]

    Google Scholar 

  • Medema L, Koning RH, Lensink R (2009) A practical approach to validating a PD model. J Bank Financ 33:701–708

    Article  Google Scholar 

  • Turner M, Varghese R, Walker P (2015) Research consensus confirms benefits of alternative data (Technical Report). PERC

    Google Scholar 

  • Niklis D, Doumpos M, Zopounidis C (2014) Combining market and accounting-based models for credit scoring using a classification scheme based on support vector machines. Appl Math Comput 234:69–81. doi:10.1016/j.amc.2014.02.028

    Google Scholar 

  • O’brien RM (2007) A caution regarding rules of thumb for variance inflation factors. Qual Quant 41:673–690

    Article  Google Scholar 

  • Overstreet GA, Bradley EL, Kemp RS (1992) The flat-maximum effect and generic linear scoring models: a test. IMA J Manag Math 4:97–109. doi:10.1093/imaman/4.1.97

    Article  Google Scholar 

  • Pavlidis NG, Tasoulis DK, Adams NM, Hand DJ (2012) Adaptive consumer credit classification. J Oper Res Soc 63:1645–1654. doi:10.1057/jors.2012.15

    Article  Google Scholar 

  • Phua C, Lee V, Smith K, Gayler R (2010) A comprehensive survey of data mining-based fraud detection research. arXiv preprint arXiv:1009.6119.

    Google Scholar 

  • Rezac M, Rezac F (2011) How to measure the quality of credit scoring models. Czech J Econ Finance (Finance a uver) 61:486–507

    Google Scholar 

  • Rona-Tas A, Hiss S (2008) Consumer and corporate credit ratings and the subprime crisis in the US with some lessons for Germany. SCHUFA, Wiesbaden

    Google Scholar 

  • Saha A, Siddiqi N (2011) Survival analysis workflow: assessing the impact of macro-economic shocks on credit portfolios and predicting the time of default

    Google Scholar 

  • Sarlija N, Bensic M, Zekic-Susac M (2009) Comparison procedure of predicting the time to default in behavioural scoring. Expert Syst Appl 36:8778–8788. doi:10.1016/j.eswa.2008.11.042

    Article  Google Scholar 

  • Shi J, Zhang S, Qiu L (2013) Credit scoring by feature-weighted support vector machines. J Zhejiang Univ Sci C 14:197–204

    Article  Google Scholar 

  • Shmueli G (2010) To explain or to predict? Stat Sci 25:289–310. doi:10.1214/10-STS330

    Article  Google Scholar 

  • Shtatland ES, Cain E, Barton MB (2001) The perils of stepwise logistic regression and how to escape them using information criteria and the output delivery system. In: 26th Annual SAS users group international conference, Long Beach, California

    Google Scholar 

  • Siami M, Gholamian MR, Basiri J (2013) An application of locally linear model tree algorithm with combination of feature selection in credit scoring. Int J Syst Sci 1–10. doi:10.1080/00207721.2013.767395

  • Siarka P (2012) Quality measures of scoring models. J Risk Manag Financ Instit 5:432–446

    Google Scholar 

  • Siddiqi N (2005) Credit risk scorecards: developing and implementing intelligent credit scoring (Wiley and SAS Business Series), 1st ed. Wiley

    Google Scholar 

  • Song X, Ding Y, Huang J, Ge Y (2010) Feature selection for support vector machine in financial crisis prediction: a case study in China. Expert Syst 27:299–310. doi:10.1111/j.1468-0394.2010.00546.x

    Article  Google Scholar 

  • Sousa MR, Gama J, Brandão E (2013) Introducing time-changing economics into credit scoring (No. 513), FEP Working Papers. Universidade do Porto, Faculdade de Economia do Porto

    Google Scholar 

  • Sousa MR, Gama J, Gonçalves MJS (2014) A two-stage model for dealing with temporal degradation of credit scoring. arXiv:1406.7775 [q-fin]

    Google Scholar 

  • Lessmann S, Thomas LC, Seow H-V, Baesens B (2013). Benchmarking state-of-the-art classification algorithms for credit scoring: A ten-year update. Presented at the Credit Scoring and Credit Control XIII

    Google Scholar 

  • Stepanova M, Thomas L (2002) Survival analysis methods for personal loan data. Oper Res:277–289

    Google Scholar 

  • Stiglitz JE, Weiss A (1981) Credit rationing in markets with imperfect information. Am Econ Rev 71:393–410

    Google Scholar 

  • Tang B, Qiu SB (2012) Multi-class support vector machine for credit scoring. Appl Mech Mater 235. doi:10.4028/www.scientific.net/AMM.235.419

  • Thomas LC (2003) Consumer credit modelling: context and current issues, Workshop Paper Presented on the Banff Credit Risk conference

    Google Scholar 

  • Thomas LC, Edelman DB, Crook JN (2002) Credit scoring & its applications (monographs on mathematical modeling and computation), 1st ed. Society for Industrial and Applied Mathematics

    Google Scholar 

  • Thomas LC, Ho J, Scherer WT (2001) Time will tell: behavioural scoring and the dynamics of consumer credit assessment. IMA J Manag Math 12:89

    Article  Google Scholar 

  • Thomas LC, Malik M (2010) Comparison of credit risk models for portfolios of retail loans based on behavioural scores. In: Rausch D, Scheule H (Eds) Model risk in financial crises. Risk Books, pp 209–232.

    Google Scholar 

  • Thomas LC, Oliver RW, Hand DJ (2005) A survey of the issues in consumer credit modelling research. J Oper Res Soc 56:1006–1015. doi:10.1057/palgrave.jors.2602018

    Article  Google Scholar 

  • Tsai C-F (2009) Feature selection in bankruptcy prediction. Knowl-Based Syst 22:120–127. doi:10.1016/j.knosys.2008.08.002

    Article  Google Scholar 

  • Tsai C-F, Chen M-L (2010) Credit rating by hybrid machine learning techniques. Appl Soft Comput 10:374–380. doi:10.1016/j.asoc.2009.08.003

    Article  Google Scholar 

  • Tsymbal A (2004) The problem of concept drift: definitions and related work. Computer Science Department, Trinity College Dublin

    Google Scholar 

  • Turner MA, Walker PD, Dusek K (2009) New to credit from alternative data. Information Policy Institute/Political and Economic Research Council

    Google Scholar 

  • Van Gestel T, Baesens B, Martens D (2010) From linear to non-linear kernel based classifiers for bankruptcy prediction. Neurocomput Int J 73:2955–2970. doi:10.1016/j.neucom.2010.07.002

    Article  Google Scholar 

  • Verikas A, Kalsyte Z, Bacauskiene M, Gelzinis A (2010) Hybrid and ensemble-based soft computing techniques in bankruptcy prediction: a survey. Soft Comput 14:995–1010. doi:10.1007/s00500-009-0490-5

    Article  Google Scholar 

  • Waad B, Ghazi BM, Mohamed L (2013) A three-stage feature selection using quadratic programming for credit scoring. Appl Artif Intell 27:721–742. doi:10.1080/08839514.2013.823327

    Article  Google Scholar 

  • Wang J, Hedar A-R, Wang S, Ma J (2012) Rough set and scatter search metaheuristic based feature selection for credit scoring. Expert Syst Appl 39:6123–6128. doi:10.1016/j.eswa.2011.11.011

    Article  Google Scholar 

  • Widmer G, Kubat M (1996) Learning in the presence of concept drift and hidden contexts. Mach Learn 23:69–101

    Google Scholar 

  • Yao P (2009) Fuzzy rough set and information entropy based feature selection for credit scoring. In: 6th international conference on fuzzy systems and knowledge discovery, 2009. FSKD’09. Presented at the 6th International Conference on Fuzzy Systems and Knowledge Discovery, 2009. FSKD’09, pp. 247–251. doi:10.1109/FSKD.2009.713

  • Yu L, Wang S, Lai KK (2008a) Credit risk assessment with a multistage neural network ensemble learning approach. Expert Syst Appl 34:1434–1444. doi:10.1016/j.eswa.2007.01.009

    Article  Google Scholar 

  • Yu L, Wang S, Lai KK, Zhou L (2008b) Bio-inspired credit risk analysis: computational intelligence with support vector machines. Springer, Heidelberg

    Book  Google Scholar 

  • Žliobaitė I (2009) Learning under concept drift: an overview (Technical Report). Faculty of Mathematics and Informatics, Vilnius University, Vilnius, Lithuania

    Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Financial Engineering Laboratory, Technical University of Crete, Chania, Greece

    Dimitris Nikolaidis, Michael Doumpos & Constantin Zopounidis

  2. Audencia Business School, Nantes, France

    Constantin Zopounidis

Authors
  1. Dimitris Nikolaidis
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Michael Doumpos
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Constantin Zopounidis
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Michael Doumpos .

Editor information

Editors and Affiliations

  1. Decision Support Systems Laboratory, Technical University of Crete, Chania, Greece

    Evangelos Grigoroudis

  2. Financial Engineering Laboratory, Technical University of Crete, Chania, Greece

    Michael Doumpos

Appendices

Appendix 1: Scoring Parameters

Table 3 Observation point (T0) definitions
Full size table
Table 4 Performance point (T1) definitions
Full size table

Appendix 2: Variables

Table 5 Behavioral scoring variables
Full size table

Rights and permissions

Reprints and permissions

Copyright information

© 2017 Springer International Publishing Switzerland

About this paper

Cite this paper

Nikolaidis, D., Doumpos, M., Zopounidis, C. (2017). Exploring Population Drift on Consumer Credit Behavioral Scoring. In: Grigoroudis, E., Doumpos, M. (eds) Operational Research in Business and Economics. Springer Proceedings in Business and Economics. Springer, Cham. https://doi.org/10.1007/978-3-319-33003-7_7

Download citation

Publish with us

Policies and ethics

Search

Navigation