讲座主题：Surrogate Model Assisted Nested Simulation for Large Variable Annuity Portfolios
X. Sheldon Lin
X. Sheldon Lin, ASA, ACIA, is a Professor of Actuarial Science at the University of Toronto and serves as an Editor for Insurance: Mathematics and Economics. His recent research is on data-driven nonlinear regression modelling for insurance rate-making and reserving, and risk management of large variable annuity and other insurance portfolios. The research aims to develop new and implementable methodology and technologies for insurance.
Variable annuities (VAs) are equity-linked deferred annuities with embedded investment guarantees. Nowadays, many insurance companies are managing large VA portfolios that have hundreds of thousands of policies. Since most of the VA contributions are invested in the equity market, the insurance companies are exposed to significant market risks. Risk managing the liability of their VA portfolios has become the central task. In practice, the stochastic-on-stochastic or nested simulation is commonly used for VA portfolio valuation and risk management. However, the path-dependency of the most embedded guarantees and the non-homogeneity of the VA policies make the nested simulation algorithms extremely complex and time-consuming to run.
In this talk, we use the idea of surrogate modeling to design efficient nested simulation algorithms so that the important quantities for risk management such as the predictive liability distribution, hedging strategies and profit and loss analysis for a VA portfolio can be obtained/performed within a very reasonable timeframe. Specifically, we propose some population sampling techniques to select a small set of policies from the portfolio to represent the entire portfolio. Further, a spline regression model combined with scenario clustering is employed to reduce the numbers of outer-loops and inner-loops in the proposed nested simulation. A large synthetic VA portfolio that mimics VA portfolios in the real world is used to illustrate the efficiency and accuracy of the proposed algorithms.
This is joint work with my former PhD student Shuai (Alex) Yang.
报告地点：腾讯会议（会议ID：466 909 828）
邀 请 人：池义春
版权：2021欧洲杯足球竞猜官方平台 本网站新闻 独家专稿,未经许可,严禁转载!地址：中国 北京 昌平区沙河高教园 邮编 102206