The MBBEFD Distribution After two years, the MBBEFDLite package for R is finally mature! As a brand-new actuary, in my first job, I had the privilege of working with Dr. Stefan Bernegger at Swiss Reinsurance. Albeit I did not have much to do with him—I was a one-exam, no experience new actuary and he was Read the full article... The post MBBEFDLite has its mature release appeared first on Strange Attractors.

R-bloggers

MBBEFDLite, an R package implementing the Maxwell-Boltzmann Bose-Einstein Fermi-Dirac (MBBEFD) distribution used in reinsurance exposure rating, has reached version 1.0.0 on CRAN. The package was created as a lightweight, dependency-free alternative to the existing mbbefd package. The key update is an improved method-of-moments fitting algorithm, inspired by a 2026 paper by Dr. Stefan Bernegger, using nested 1D fits based on mean and coefficient of variation rather than raw moments. The author also clarifies that some samples simply don't converge under MoM fitting, and that maximum likelihood estimation is generally more robust when raw observations are available.