*Accident Fund Insurance Company*
**Proposed Project for the
MSU Industrial Math Students**

Large Loss Predictive Model

Traditional insurance pricing involves determining expected losses for different types of risk. While this focus provides adequate rate levels for types of business in the aggregate, it is likely that the variance is a more important characteristic of the distribution from the standpoint of individual risk selection.

Accident Fund is interested in engaging with the MSU Industrial Math Department to perform a study that provides the ability to determine which policies have a high potential for a loss above a certain threshold. This goal will be obtained by using multivariate techniques that fall under the umbrella of “predictive modeling.”

Throughout the modeling process, Accident Fund’s Actuarial department will provide the Industrial Math group with data and guidance to getting to the ultimate goal of a finalized model. Accident Fund will also have the expectation of periodic updates as to the status of the project.