Selected work

Actuarial / Risk

Cross-Line Tail Risk Modelling

Dependence between mortality and catastrophe losses

A copula-based study of tail dependence between life-insurance mortality losses and property-catastrophe claims, examining how dependence assumptions affect aggregate losses, diversification and economic-capital estimates.

LIFE × PROPERTY CATASTROPHEPipeline output · synthetic data
0LIFE LOSS RANK u → 1PROPERTY LOSS RANK v → 1Do bad yearsarrive together?■ Joint tail: both lines in their top 10%● One synthetic loss year after fitting marginals┈ Dotted row at v ≈ 0.35: no-catastrophe years
Probability-integral transforms of each line

Problem

Aggregate capital estimates depend on how losses across lines are assumed to move together in the tail, where data is sparse and dependence assumptions matter most.

What I built

Mortality losses and property-catastrophe claims are aggregated under different copula specifications, then compared through aggregate loss distributions, diversification benefit and economic-capital estimates.

Methods

  • Marginal loss modelling by line
  • Copula-based dependence specification
  • Joint loss simulation
  • Aggregate loss distribution and economic-capital estimation
  • Comparison of diversification benefit across dependence assumptions

Technology

  • Python

Links