Developing Aggregate Loss Models For Obscure Insurance Exposures

University of Minnesota M.S. thesis. August 2020. Major: Mathematics. Advisor: Fadil Santosa. 1 computer file (PDF); xi, 53 pages.In this paper we will examine how to apply parametric modeling with some simulation techniques to generate potential frequency and severity of claims distributions. The paper outlines the general theory of selecting appropriate statistical distributions for claims data and in maximizing model accuracy or fit. Two specific exposures, critical customer loss and supply chain disruption loss, are used as illustrations. From these distributions, an aggregate loss model is then proposed. The paper will discuss: 1. Possible ways to develop claims data from available market information for obscure insurance risks with application to specific examples. 2. Outline of the theory of parametric estimation. 3. Applying primarily parametric techniques to optimize the ‘fit’ of the estimated distributions for severity and frequency of claims distributions. 4. Development of possible aggregate loss models for the two specific exposures. 5. Proposed areas for further study