Functional Analysis of Insurance Loss Reserves

Risk management is a fascinating field. The models are interesting and, more importantly, actuaries are some of my favorite people. They have a certain nerdy panache that I find both endearing and relatable. This project topic was suggested to me by the actuarial program director as part of a Bayesian methods class during my master's. The purpose is to compare how long it takes for accident claims to be fully paid across different insurance lines. And being totally objective, the model I built rocks. Bayesian B-splines with geoemtrically decaying variance and a smoothness penalty? That is sure to get the old heart racing. Fusing actuarial science with functional data analysis? Oh, the panache!