Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/de-gruyter/an-empirical-comparison-of-some-experimental-designs-for-the-valuation-RstqW45lIk
References (6)
-
Loeppky (2009)
Choosing the sample size of a computer experiment : A practical guideTechnometrics, 51
-
McKay (2008)
A census of small latin hypercubes SIAM J DiscreteMath, 22
-
Cathcart (2015)
Calculating variable annuity liability greeks using Monte Carlo simulation ASTINBull, 45
-
Khuri (2006)
Design issues for generalized linear models : a reviewStat Sci, 21
-
Crombecq (2011)
Efficient space - filling and non - collapsing sequential design strategies for simulation - based modeling EurOper Res, 214
-
Myers (1999)
Response surface methodology current status and future directions QualTech, 31
- Publisher
- de Gruyter
- Copyright
- © 2016 Guojun Gan and Emiliano A. Valdez
- ISSN
- 2300-2298
- eISSN
- 2300-2298
- DOI
- 10.1515/demo-2016-0022
- Publisher site
- See Article on Publisher Site
Abstract
AbstractVariable annuities contain complex guarantees, whose fair market value cannot be calculated inclosed form. To value the guarantees, insurance companies rely heavily on Monte Carlo simulation, whichis extremely computationally demanding for large portfolios of variable annuity policies. Metamodeling approacheshave been proposed to address these computational issues. An important step of metamodelingapproaches is the experimental design that selects a small number of representative variable annuity policiesfor building metamodels. In this paper, we compare empirically several multivariate experimental designmethods for the GB2 regression model, which has been recently discovered to be an attractive modelto estimate the fair market value of variable annuity guarantees. Among the experimental design methodsexamined, we found that the data clustering method and the conditional Latin hypercube sampling methodproduce the most accurate results.
Journal
Dependence Modeling – de Gruyter
Published: Dec 14, 2016