Abstract
The valuation of large variable annuity portfolios is an important enterprise risk management task but is computationally challenging due to the need for simulation. Existing methods in the literature only use simple experimental designs with significant room for improvement. This article identifies three major components in an efficient valuation framework. In addition, we propose optimal experimental designs and provides analytical insights for each component. Our numerical results show that our proposal achieves significantly higher accuracy than state-of-the-art alternatives without requiring any additional computational resource.
ACKNOWLEDGEMENT
We thank the anonymous reviewers for their careful reading of our manuscript and their many insightful comments and suggestions.
Discussions on this article can be submitted until January 1, 2021. The authors reserve the right to reply to any discussion. Please see the Instructions for Authors found online at http://www.tandfonline.com/uaaj for submission instructions.
Notes
1 The inverse distance weighting (Hejazi, Jackson, and Gan 2017) method has similar computational costs. However, we found that its prediction accuracy is very low, which is consistent with findings in Hejazi, Jackson, and Gan (2017). Therefore, we do not consider this method any further.