Case Study: Swiss Re halves prototype development time for natural-disaster loss modeling with MathWorks MATLAB

Swiss Re Calculates Potential Loss from Natural Disasters

Swiss Re engaged MathWorks to tackle a tight, eight‑month deadline to build a prototype that could accurately estimate losses from natural disasters using decades of historical data and large simulated datasets. The Natural Perils Unit needed to combine multiple in‑house models (originally written in different languages), run fast Monte Carlo simulations—up to millions of artificial storm tracks—and handle geographic projections; Excel was inadequate and rewriting in C++/Java would have taken too long. MathWorks products used included MATLAB and Mapping Toolbox.

Using MathWorks tools, Swiss Re imported and unified its legacy models and historical data in MATLAB, used Monte Carlo techniques to generate and validate artificial storm tracks, and leveraged Mapping Toolbox for map projections and terrain handling; the prototype was then productionized as a Java application for underwriters. MathWorks enabled Swiss Re to address complex analyses, reduce development time by 50%, increase function speed, and deploy the solution to more than 50 underwriters worldwide.

Swiss Re

Gerry Lemcke

Deputy Head of Catastrophe Perils