In this video, Lisa Schilling from the Society of Actuaries Research Institute presents the challenges of estimating mortality improvement rates, emphasizing their importance in actuarial modeling for life insurance and pensions. She explains that limited data increases uncertainty, and the reliability of estimates depends on the number of observed deaths, the time frame, and the chosen methodology. The SOA has developed an Excel-based simulation tool to help users assess and compare the reliability of their mortality improvement estimates, demonstrating that using more years of data and regression methods can significantly reduce uncertainty. Lisa also highlights practical considerations such as demographic shifts and outlier years when analyzing multi-year mortality data.
Contributors: Lisa Schilling, FSA, EA, FCA, MAAA; Jon Forster, ASA, MAAA