Capstone Projects

Insurance Renewal Pricing Optimization

Program: Data Science Master's
Location: Wisconsin (onsite)
Student: Sam Hurley

Surplus lines insurance carriers generally use soft insurance market intelligence and leadership experience to guide renewal pricing strategies. There is an opportunity for insurance carriers to leverage existing data assets to inform renewal pricing practices and improve underwriting profitability. A review of insurance-related literature and interviews with insurance professionals identified variables underwriters use to make discretionary renewal pricing decisions. Historical renewal transaction data was collected, transformed, and analyzed to determine which of the variables were significant for guiding pricing decisions. A series of statistical hypothesis tests identified the variables that could be useful in predicting optimal renewal pricing for future renewals. Classification and time series models were trained using the significant variables to predict the likelihood of winning a renewal account at different pricing levels. These models formed the foundation for a recommendation engine designed to identify the renewal price that optimizes the expected value of a renewal quote. The recommendation engine designed in this project can be integrated into the underwriting process to help underwriters make better pricing decisions on future renewal accounts and improve profitability.