The objective of this thesis work is to present and develop a statistical economic model for estimating and forecasting the probability of default, similar to what is also used by Banks in the context of Stress Test exercises. Such model will be based on the Credit Portfolio View framework theorized by Wilson (1997), which consists in using the principal macroeconomic factors as regressors to provide predictions of future values of default rates for a specific economic sector. This approach will be implemented in a case study to the Italian corporate sector, and the model selection process as well as the results of the estimation methods used for the regression will be discussed. Finally, it will be observed whether there is a worsening of the expected default rates under stress applied to each of the selected explanatory variables.
Credit Portfolio View: probability of default modelling through macroeconomic factors
TURCHET, MARTINA
2022/2023
Abstract
The objective of this thesis work is to present and develop a statistical economic model for estimating and forecasting the probability of default, similar to what is also used by Banks in the context of Stress Test exercises. Such model will be based on the Credit Portfolio View framework theorized by Wilson (1997), which consists in using the principal macroeconomic factors as regressors to provide predictions of future values of default rates for a specific economic sector. This approach will be implemented in a case study to the Italian corporate sector, and the model selection process as well as the results of the estimation methods used for the regression will be discussed. Finally, it will be observed whether there is a worsening of the expected default rates under stress applied to each of the selected explanatory variables.File | Dimensione | Formato | |
---|---|---|---|
Turchet_Martina.pdf
accesso riservato
Dimensione
11.31 MB
Formato
Adobe PDF
|
11.31 MB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/48564