Objectifs

Modern actuarial models with strong business relevance such as price and loss reserve optimisation, customer segmentation, dynamic monitoring, risk and customer management require deep understanding of the underlying dependence of multidimensional risks.

The principal goal of this course is to provide a solid background of key building blocks for actuarial modeling of dependent risks with the main emphasis on copulas, multivariate dependence measures, risk measures, risk aggregation, models for multivariate extremes and measures of extremal dependences. The R-Project will provide an essential experimental platform for testing various models. Additional facultative materials include several past projects.

Contenus

1. Modelling dependent risks
2. Copulas and transforms
3. Dependence measures
4. Dependence models for large risks using mulivaiate Extrem Value Theory
5. Risk measures & aggregation
6. Projects on price optimisation, portfolio cleaning, customer future value, modelling of large losses, SST (facultative reading)

Références

1. Denuit, M., Dhaene, J., Goovaerts, M., and Kass, R. (2006) Actuarial Theory for Dependent Risks: Measures, Orders and Models. Wiley
2. Asmussen, S., and Steffensen, M. (2020) Risk and Insurance. Springer
3. Reiss, R-D., and Thomas, M. (2007) Statistical Analysis of Extreme Values: From Insurance, Finance, Hydrology and Other Fields. 3rd Edt, Birkhäuser
4. Embrechts, P., Klüpelberg, C., and Mikosch, T. (1997) Modeling extremal events for finance and insurance. Springer

Evaluation

1ère tentative

Examen:

Ecrit 3h00 heures

Documentation:

Non autorisée

Calculatrice:

Non autorisée

Evaluation:

Course Grade = Grade Final Exam

Rattrapage

Examen:

Ecrit 3h00 heures

Documentation:

Non autorisée

Calculatrice:

Non autorisée

Evaluation:

Course Grade = Grade Final Exam