Seminar by Gee Lee

Understanding insurance risk retention using loss models

In this talk, I will explain my joint work with Dr. Haiyan Liu and Dr. Peng Shi, where we are investigating the use of loss models to understand an insurance company’s risk retention decision. I will start by explaining the hierarchical claims model (loss model) that we have developed for individual policyholders. Models like these have been used extensively to solve the rate-making problem as well as the reserving problem in the actuarial literature. In our recent collaborative work, we have been trying to explore the use of loss models for solving the risk retention problem empirically. I will provide an outline of our approach, and discuss the managerial implications of such approaches to an insurance company. In addition, I will briefly talk about the multivariate risk retention problem, where the goal is to figure out the retention amounts for a bundle of insurance policies. I will illustrate how we can solve a thousand-dimensional retention problem using numerical methods, and also explain where I think there is opportunity for future work.