Στατιστικά μοντέλα μέτρησης πιστωτικών κινδύνων
Βαμβακάρης, Πέτρος Φ.
SubjectΔιαχείριση κινδύνου -- Οικονομετρικά μοντέλα ; Καταναλωτική πίστη ; Διαχείριση κινδύνου -- Στατιστικές μέθοδοι ; Πιστωτικός κίνδυνος
Credit risk constitutes the oldest type of risk that financial institutions have to face. Over the past years and especially after the first regulatory framework of 1988 set by the Basel Committee on Banking Supervision, risk models have become necessary for critical decision making in every financial institution. These models use statistical techniques, probability theory and stochastic processes. In the present thesis an integrated approach is presented on credit risk management techniques. For this reason the structure and the operational procedures of the financial system is outlined and the operating framework of the financial institutions is analyzed. Moreover a historical overview of the evolution of this area is provided along with the description of the most essential statistical models and estimation methods that have been suggested for the measurement of credit risk. Finally a brief introduction of basic concepts, definitions and properties of Markov chains is offered and then it is outlined how these can be used for the description of credit risk as well as for the estimation of the (credit) transition matrices.