STOP-LOSS BOUNDS ON FUNCTIONS OF POSSIBLY DEPENDENT RISKS IN THE PRESENCE OF PARTIAL INFORMATION ON THEIR MARGINALS
(2004)
Files
dp0408.pdf
Open Access - Adobe PDF
- 191.67 KB
Details
- Authors
- Abstract
- A general strategy is proposed for deriving upper and lower stop-loss bounds on directionally convex, increasing functions of possibly dependent, non-negative random variables when partial information is available about their joint copula or their marginal distributions. Illustrations are presented which involve looking for a stop-loss bound on a sum of insurance risks whose means, variances or range are finite and known. The key to these developments, which generalize recent findings of Genest, Marceau and Mesfioui (2002), is a result due to Müller and Scarsini (2001) on the stochastic comparison of random vectors having a common copula.
- Affiliations
Citations
Denuit, M., Genest, C., & Mesfioui, M. (2004). STOP-LOSS BOUNDS ON FUNCTIONS OF POSSIBLY DEPENDENT RISKS IN THE PRESENCE OF PARTIAL INFORMATION ON THEIR MARGINALS (STAT Discussion Paper 0408). https://hdl.handle.net/2078.5/33569