Description: Reserving and ratemaking calculations are vital to how insurance companies perform financially. Current-day practices for predicting future loss ratio or pure premiums for reserving and ratemaking rates are not based on low prediction errors—mean squared error (MSE) criterion. This report provides mathematically optimal weights that lead to minimum MSE prediction errors, and the formulas are validated using retro-testing with real-world datasets. In this report, two main theorems are discussed, each addressing prediction of future loss ratio or pure premium and improving upon current practices for how rates and reserves are calculated.

Description: This report identifies and makes recommendations regarding possible exposure measures for pricing and analyzing the risks in cyber insurance. Cyber insurance is an insurance product that is designed to provide protection against the financial consequences of a failure or compromise of an organization’s information system as a result of a cyber event.

Description: An analytical basis about national trends and state specific factors that may be driving the auto insurance loss costs.

Description: The Spatial Hazards Events and Losses Database for the United States (SHELDUS) provides county-level data for direct economic losses due to natural hazards from 1960 to the present.

Description: The transportation industry is in the early stages of profound changes, due to development of sophisticated driving safety and automation technologies.

Description: An interview with Weili Lu, the president of ARIA

Description: The aim of this project has three aspects: (1) providing a comprehensive literature review on telematics car driving data analytics; (2) developing various visual tools and a data cleaning procedure for telematics data; and (3) extracting risk factors from individual trips and improving classic classification ratemaking models with those risk factors.