by: Professor Sheldon Lin, Department of Statistical Sciences, University of Toronto
Sheldon Lin is a Professor of Actuarial Science in the Department of Statistical Sciences at the University of Toronto. His research interests include insurance loss modeling and valuation of equity-linked insurance products.
Faculty Host: Professor Gord Willmot, Department of Statistics and Actuarial Science, University of Waterloo
Variable annuities (VA) are equity-linked insurance contracts that have rapidly grown in popularity around the world in recent years. Research up to date on variable annuities largely focuses on the valuation and hedging of guarantees embedded in a single VA contract. However, methods developed for individual VA contracts cannot directly be extended to a large VA portfolio due to the high inhomogeneity of the portfolio. Insurance companies currently use nested simulation to value and design hedging strategies for VA portfolios but the implementation of nested simulation for a very large VA portfolio has been a real challenge. The computation in nested simulation is highly intensive and often prohibitive.
In this talk, I will present a novel method that combines a clustering technique with a functional data analysis technique to address this issue. We create a highly non-homogeneous synthetic VA portfolio of 100,000 contracts and use it to estimate the dollar Deltas of the portfolio at each time step of outer loop scenarios under the nested simulation framework over a period of 30 years. Our test results show that the proposed method performs well in terms of accuracy, speed and efficiency.
200 University Avenue West
Waterloo, ON N2L 3G1