Copula-Based Multivariate Models with Applications to Risk Management and Insurance
27 Pages Posted: 12 Sep 2005
Date Written: September 2005
The purpose of this paper consists in analysing the relevance of dependence concepts in finance, insurance and risk management, exploring how these concepts can be implemented in a statistical model via copula functions and pointing out some difficulties related to this methodology. In particular, we first review the statistical models currently used in the actuarial and financial fields when dealing with loss data; then we show, by means of two risk management applications, that copula-based models are very flexible but sometimes difficult to set up and to estimate; finally we study, by means of a simulation experiment, the properties of the maximum likelihood estimators of the Gaussian and Gumbel copula.
Keywords: Copula, Tail dependence, Loss model, Portfolio model
JEL Classification: C15, C51, C63
Suggested Citation: Suggested Citation