Advanced Modeling, Visualization, and Data Mining Techniques for a New Risk Landscape

The risk landscape that confronts financial institutions in the 21st century presents an unprecedented departure from past experience. Traditional mathematical tools based on linear statistical mathematics are failing actuarial science by not being able to deliver credible analyses in an environment characterized by issues such as multiple correlations, extreme events and cascading risks. Complexity science, and area of scientific investigation that has been largely limited to powerful laboratory environments until recent decades, offers new tools and methodologies with which to address this new environment. InsuranceWorld© is an adaptive agent-based simulator designed through application of complexity science to allow an individual company to see how markets will evolve under various catastrophe scenarios. This paper describes this tool and its scientific and mathematical foundations.