ASMF Seminar: Julien Tomas (SCOR)
'Prospective mortality modelling by cause of death: a study on risk diversification according to the dependency structure'
Event details from: ASMF Seminar: Julien Tomas (SCOR)
- Date
- 19 April 2024
- Time
- 12:30 – 13:30
Abstract
Insurance companies are required to produce "what-if" type of scenarios by both regulators and internal stakeholders. By using stress test scenarios, regulators seek to validate internal models and to identify probable crisis situations that would threaten the viability of the company. Senior management is also particularly interested in probable scenario analyses which permit assessing business resilience to shocks and evaluating portfolio diversification impact. Building "what-if" type of scenarios for mortality and longevity risks translates into a need to create hypothetical scenarios on one or more causes of death. In this study, we follow a cause of death mortality modelling within the competing risks framework introduced by Li and Lu (2019) and show how assumptions on cause dependency structure affect future mortality projections. We observe how not considering the dependence between causes when modelling scenarios could yield hypothetical gains. On the other hand, working within the competing risks framework results in more moderate gains from diversification between mortality and longevity lines of business.
