Credit Risk - methods of scorecards development in R

nieobowiązkowe

The course gives both theoretical knowledge and practical skills to model a credit scorecard. During the course all necessary steps to develop a scorecard would be discussed and presented. Starting from data preparation (handling a missing data and outliers, derived variables preparation, data sampling), going through model estimation (i.e. logistic regression) and model quality assessment (discriminatory power, stability) and ending on optimal cut-off choice.

During the course examples would be presented in R.

A detailed course plan:

1. Statistical inference basics

2. Modelling sample definition

3. Risk factors specificity

• Application data

• Behavioral data

4. Data preparation

• GB flag

• Discretization and different methods of data preparation

• Preliminary variable selection

5. Probability of Default prediction

• Logistic regression and other methods

6. Method of scorecard building and transformation to Masterscale

7. Scorecard Quality assessment

• Functional form selection

• Goodness-of-fit tests

• Discriminatory power

• Stability analysis

• Dimensions of Quality assessment

8. Optimal cut-off point choice

Banasik, J., & Crook, J. (2004). Does reject inference really improve the performance of application scoring models? Journal of Banking & Finance, vol 28, pp. 857-874.

Banasik, J., & Crook, J. (2007). Reject inference, augmentation, and sample selection. European Journal of Operational Research, 183 (2007) pp. 1582–1594.

Hosmer, D. W., Lemeshow, S., & Sturdivant, R. X. (2013). Applied logistic regression. Hoboken, NJ: Wiley.

King, G., & Zeng, L. (2003). Logistic Regression in Rare Events Data. Journal of Statistical Software, 8(2).

Kleinbaum, D. G., Klein, M., & Pryor, E. R. (2010). Logistic regression: a self-learning text. New York: Springer.

Löffler, G., & Posch, P. N. (2013). Credit risk modeling using Excel and VBA. Chichester: John Wiley & Sons.

Siddiqi, N. (2006). Credit risk scorecards developing and implementing intelligent credit scoring. Hoboken (N.J.): Wiley.

Thomas, L. C., Edelman, D. B., & Crook, J. N. (2002). Credit scoring and its applications. Philadelphia: Society for Industrial and Applied Mathematics.

The students will learn how to perform a whole scorecard development project (from modeler perspective). Starting with data preparation (handling a missing data and outliers, derived variables preparation, data sampling), through model estimation (i.e. logistic regression) and model quality assessment (discriminatory power, stability) to optimal cut-off choice.

KW01, KW02, KW03, KU01, KU02, KU03, KK01, KK02, KK03

All students will be obliged to:

• be present at the classes (according to common University of Warsaw rules),

• prepare a project (code + paper) in which they will present a comparison of different scorecards quality