Work detail

Macroeconomic Determinants of Probability of Default and Loss Given Default for Household

Author: Mgr. Pavel Dvořák
Year: 2009 - summer
Leaders: doc. PhDr. Ing. Ing. Petr Jakubík Ph.D. Ph.D.
Consultants:
Work type: Economic Theory
Masters
Language: English
Pages: 79
Awards and prizes:
Link:
Abstract: This thesis is focused on the estimation of expected loss for the consumer credit card portfolio. For the expected loss modelling we have developed in this thesis a model that satisfies conditions defined by New Basel accord (Basel II). The final model is therefore compound of the key credit risk components as they are identified by Basel II document and s separate model is built for each of them. These credit risk components are probability of default (PD), exposure at default (EAD) and loss given default (LGD). For the model development were used credit scoring techniques, especially classification tree method represented by CHAID algorithm. The empirical model then gives us an opportunity to test several hypotheses regarding the influence of macroeconomic indicators and demographics on expected loss, or correlation between PD and LGD.

Partners

Deloitte

Sponsors

CRIF
McKinsey
Patria Finance