Ph.D. Theses
Ph.D. Theses
The theses listed below are sorted by Ph.D. and, within each Ph.D., by year of defense.
Title: Corporate credit risk modelling
| Programme: | Finance |
| Student: | João Dias Fernandes |
| Supervisor(s): | Miguel Ferreira |
| Abstract: | Corporate credit risk modeling for privately-held firms is limited, although these firms represent a large fraction of the corporate sector worldwide. This study is an empirical application of credit scoring and rating techniques to a unique dataset on private firms bank loans of a European bank. It is divided in two chapters. The first chapter is concerned with modeling the probability of default. Several alternative scoring methodologies are presented, validated and compared. These methodologies include a multiple industry model, and a weighted sample model. Furthermore, two distinct strategies for grouping the individual scores into rating classes with PDs are developed, the first uses cluster algorithms and the second maps internal ratings to an external rating scale. Finally, the regulatory capital requirements under the New Basel Capital Accord are calculated for a simulated portfolio, and compared to the capital requirements under the current regulation. On the second chapter, we model long-term Loss-Given-Default on loan, guarantee and customer characteristics using a random, 7-year sample. Two alternative modeling strategies are tested, taking in consideration the highly non-normal shape of the recovery rate distribution, and a fractional dependent variable. The first strategy is based on Beta transformation of the dependent variable, while the second is based on Generalizes Linear Models. The methodology can be used for long-term LGD prediction of a corporate bank loan portfolio and to comply with the New Basel Capital Accord Advanced Internal Ratings Based approach requirements. |
| Placement: | Banco Montepio Geral, Risk Officer |
| Defense Year: | 2007 |