The Model Validator™ allows an analyst to validate the core assumptions underlying a property-casualty loss model. Specifically, it enables users to empirically validate many complex actuarial modeling concepts in a theoretical environment using pseudo-random loss data. Examples: How sensitive are the results to changes in the data collection threshold? What is the impact of a severity cap? Is it practical to independently fit frequency and severity distributions when the data are heterogeneous and contain outliers? Are some frequency or severity distributions universally better or worse than others? How reliable are the standard goodness-of-fit tests and how useful is graphical analysis? How can one combine data from different sources (e.g., empirical loss data and expert opinion) in a practical and theoretically valid manner? How does severity fitting under the Truncated (MLE method compare to joint frequency and severity fitting under the ALEC fitting method?

Products and Services Brochure