Abstract
The cornerstone of retail banking risk management is the estimation of the expected losses when granting a loan to the borrower. The expected losses are determined by three parameters. The first is the probability of default (PD) of the borrower. The methods of PD estimation were studied in detail by previous authors, and the most common method is credit scorecard development. The second parameter is exposure at default (EAD). Except for revolving loans, it is known in advance, it is the current balance (principal amount plus accrued interests) of the loan. Finally, there is a third parameter that defines the expected losses. This is the so-called loss given default (LGD) which is in effect the share of EAD, which is irretrievably lost in the event of default. This paper discusses several econometric techniques which allow one to obtain estimates of the LGD parameter.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
Seminar on time-to-event analysis is available at http://www.ats.ucla.edu/stat/stata/seminars/stata_survival/.
- 2.
The similar seminar but within SAS framework can be found at http://www.ats.ucla.edu/stat/sas/seminars/sas_survival/.
References
Kaplan, E.L.: Nonparametric estimation from incomplete observations. J. Am. Stat. Assoc. 53, 457–481 (1958)
Weibull, W.: Fatigue Testing and Analysis of Results, p. 250. Pergamon Press, New York (1961)
Cox, D.R.: Regression models and life-tables. J. R. Stat. Soc. 34(2), 187–220 (1972)
Ratnikova, T. A., Furmanov, K. K. Factual unemployment duration determinants in Russia (in Russian) In: Proceedings of \(<<\)Statistical Analysis Implementation in Economics and Quality Estimation\(>>\), pp. 202–206, NRU-HSE – Moscow (2010)
Thomas, L.C., Edelman, D.B., Crook, J.N.: Credit Scoring and Its Applications, p. 250. SIAM, Pliladelphia (2002)
Silvia, F., Cribari-Neto, F.: Beta regression for modelling rates and proportions. J. Appl. Stat. 31(7), 799–815 (2004)
Ganter, B., Kuznetsov, S.O.: Hypotheses and version spaces. In: Ganter, Bernhard, de Moor, Aldo, Lex, Wilfried (eds.) ICCS 2003. LNCS, vol. 2746. Springer, Heidelberg (2003)
Stepanova, M., Thomas, L.: Survival analysis methods for personal loan data. J. Oper. Res. 50(2), 277–289 (2002)
Witzany, J., Rychnovsky, M., Charamza, P.: Survival analysis in LGD modeling. Eur. Financ. Account. J. 7(1), 6–27 (2012)
Chava, S., Stefanescu, C., Turnbull, S.: Modeling the loss distribution. J. Manage. Sci. 57(7), 1267–1287 (2011)
Acknowledgments
Author would like to express his gratitude to Ivan Medvedev, Head of Retail Risks at RN Bank (former RCI Banque representative office) for being a guide in the world of banking risk management.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Appendix
Appendix
Survival function within groups by education
Survival function within groups by sex
Survival function within groups by loan amount
Rights and permissions
Copyright information
© 2014 Springer International Publishing Switzerland
About this paper
Cite this paper
Masyutin, A. (2014). Alternative Ways for Loss-Given-Default Estimation in Retail Banking. In: Ignatov, D., Khachay, M., Panchenko, A., Konstantinova, N., Yavorsky, R. (eds) Analysis of Images, Social Networks and Texts. AIST 2014. Communications in Computer and Information Science, vol 436. Springer, Cham. https://doi.org/10.1007/978-3-319-12580-0_15
Download citation
DOI: https://doi.org/10.1007/978-3-319-12580-0_15
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-12579-4
Online ISBN: 978-3-319-12580-0
eBook Packages: Computer ScienceComputer Science (R0)