1. UKM Learning and Research Repository
2. PHD Thesis / Tesis PHD
3. Faculty of Science and Technology / Fakulti Sains dan Teknologi

Title:	Modeling of corporate credit risks using actuarial techniques
Authors:	Jaber Jamil Jamal Jamil (P80683)
Supervisor:	Noriszura Ismail, Prof. Dr.
Keywords:	Financial institutions Corporate credit risks Universiti Kebangsaan Malaysia -- Dissertations Dissertations, Academic -- Malaysia
Issue Date:	27-Jan-2021
Description:	In credit risk management, the Basel committee provides three approaches to calculate the required capital for financial institutions; standardized approach, Internal Ratings-Based approach and Advanced IRB approach. The IRB approach is preferable than the standard approach for estimating probability of default (PD..) and loss given default (LGD) because it provides more accurate estimates and lower capital charges. The PD and LGD are the elements of credit risk portfolio required to estimate the capital specified under the Basel committee. This thesis develops a framework for the estimation of PD and LGD using survival analysis. The advantage of using survival model is that it deals with the analysis of lifetime data and thus, is capable of analysing censored data which are often found in the credit portfolio of a financial institution. In this thesis, the data for corporate credit portfolio is obtained from a bank in Jordan. Several parametric and non-parametric models are suggested to estimate PD, whilst the beta distribution (without time) and gamma distribution (with time) are suggested to estimate LGD. Several financial variables are then incorporated in the LGD using beta and gamma regression models. The LGD is then forecasted using ARIMA and ARIMA-Wavelet transform. Finally, the logistic regression (LR) and ordinal logistic regression (OLR..) are applied to analyse the role of corporate governance in improving the creditworthiness of financial institutions. This thesis focuses on five variables that can represent the key features of good governance practices for the Jordanian firms, namely board independent directors, role duality, board expertise, board stock and board size. The results show that the Gompertz distribution is the best parametric model for estimating PD, and the Nelson-Aalen estimator is the best non-parametric model for estimating PD. The results also show that gamma regression model (with time) is the best model for incorporating the financial variables in the LGD, and ARIMA-WT is a suitable model for forecasting LGD. The empirical results also show that both variables of board stock and board expertise are moderately significant to the creditworthiness in the LR model, whilst the variables of board independent directors and role duality are quite significant in the OLR model. These empirical results reveal that the considered variables of corporate governance provide a significant impact on the creditworthiness of the Jordanian firms.,Ph.D
Pages:	226
Call Number:	HG181.J333 2020 tesis
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Science and Technology / Fakulti Sains dan Teknologi

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.