Swiss Re NatCat Modelling Engine
Cutting-edge natural catastrophe models for better risk selection, pricing and accumulation control
The natural catastrophe (NatCat) protection gap has continued its upward trend over the past decade, reaching USD 181 billion in 2024. The accumulation of human and physical assets in vulnerable areas as well as the increase in climate-related risks amplifies the issues many insurers face already today. In this context, access to next-in-class NatCat models becomes critical for business success, ultimately improving the quality of risk selection, pricing accuracy and long-term profitability.
NCME combines our hazard models with key parameters from the vulnerability and exposure modules for single- or multi-location policies. Taking into account asset distribution and specific insurance conditions, our proprietary models yield accurate prices in a time-efficient manner.
Further Information
Hazard: Where, how often, and with what intensity do events occur?
While historical event catalogues may be a starting point to quantify a natural hazard, they are often too short, limited and biased to be deemed representative. Therefore, a large number of simulated/hypothetical events are generated by varying specific parameters (e.g., geographical location, intensity, etc.) based on climatological, physical or statistical insights and assumptions. The resulting probabilistic event catalogue contains hundreds of thousands of events, thus resembling statistical quantities of the observed history and including much more unlikely or extreme events.
Vulnerability: What is the degree of damage expected for a given hazard?
When natural catastrophes strike, the extent of damage can vary significantly between two risks in locations which experience the same hazard. Key parameters to take into account in this module vary by peril, but may include: building code, construction, quality/age, protection measure and occupancy type. The knowledge basis for modelling vulnerability is a mix of authentic claims information, damage observations, expert opinions, and engineering judgement. This results in a large suite of specific vulnerability curves.
Exposure: Where are the insured objects located and what are their values?
The input to all loss models is exposure information, which is preferably as detailed and comprehensive as possible. However, granular exposure data is not always available, and models need to be able to adequately represent diverse qualities and resolutions. The ability to handle various levels of data quantity and quality is an important feature of probabilistic NatCat models. To do so, the model needs to make assumptions on the aspects not covered in detail in the exposure data.
Insurance conditions: What proportion of the loss is insured, and how?
This module calculates the insurer's net loss resulting from a given ground up loss. Insurance conditions such as deductibles, limits or exclusions are important characteristics that vary widely between insurance products. This module supports all possible insurance policy conditions. It samples uncertainty distributions around the losses simulated and applies the insurance conditions to each sample individually. This way, the effect of any combination of these conditions can be defined and adequately modelled on individual risks or groups of risks.
