Εκτίμηση του κινδύνου σε ένα ασφαλιστικό χαρτοφυλάκιο εξαρτημένων κινδύνων μέσω της θεωρίας των συνδέσμων
Risk estimation of an insurance portfolio of dependent risks using copulas

View/Open
Keywords
Διαχείριση κινδύνου ; Εξάρτηση ; Συνάρτηση σύζευξης ; Συναρτήσεις σύζευξης ; Σύζευξη ; Copula ; Θεώρημα Sklar ; Μέτρα κινδύνου ; Μέτρο κινδύνου ; Αρχιμήδεια οικογένεια σύζευξης ; Αρχιμήδεια σύζευξη ; Σχέσεις εξάρτησης ; Αξία σε κίνδυνο ; Ασφαλιστικό χαρτοφυλάκιο ; Value at RiskAbstract
Risk management is a daily challenge for insurance companies and financial institutions. Α
substantial step in risk analysis is the understanding of the dependence structure between one
dimensional results. The current dissertation focuses on Copulas, a tool for understanding the
relationship between random variables. In the first chapter, we present risk measures which are
widely used in insurance and finance along with their properties and methods for their
assessment. In the second chapter we describe Sklar’s theorem which is a fundamental result of
the theory of Copulas. We focus on bivariate Copulas and their properties and at the end of the
chapter we present the class of Archimedean Copulas. In the final 3rd chapter we study a non-
life insurance portfolio of Motor third part liability losses. We numerically investigate the class
of optimally fitted copula on our bivariate data through R statistical package. Finally, based on
the most appropriate copula, we calculate portfolio’s conditional Value at Risk.