Catastrophe models have become an invaluable tool in managing large and unpredictable losses across the insurance industry— and beyond—since their introduction in the late 1980s to the point that today catastrophe model output has become a currency for understanding and quantifying risk in insurance and reinsurance transactions.

Risk modelling is a substantial cost for the insurance sector, which spends over $1 bn annually in licence fees for risk models and the expertise to operate and interpret them for clients. Paradoxically, however, one of the biggest systemic risks of cat modelling is an over reliance on a small group of cat model vendors.

Pressure from the re/insurance industry to create more efficiency through greater interoperability is increasing, and major model providers are responding positively to assisting with this goal.

Researchers can support insurers in developing global databases of standardised information, such as a worldwide buildings database using unique building identification, as well as with efforts to compare different models.

Key Research Questions

To drive development in this field, we have put together a series of questions that we believe can help shape the right level of research – and fundamentally deliver results. These are as follows:

• How can we objectively determine whether a model contains decision relevant information? Can methodologies from other disciplines be adapted or used directly?
• How can models be used to produce ‘robust’ decisions which do not alter materially in the face of parameter, data or model uncertainty, rather produce a ‘good practice’ outcome?
• Can we determine a framework to carry out a retrospective study on model performance; including past versions of models to see how they have evolved?
• Can we get some more data-points from the “tail” of climate model runs to sit alongside existing catastrophe model output as a “second set of eyes”? Key Questions for Researchers
• How can we boost interoperability of models across modelling platforms and extend these models to public sector use and for expansion of risk?
• Can we develop a qualitative tool for model risk governance – ‘models of models’ and embrace a systems-based thinking for development of the next generations of Cat models?
• Could a focus on ensuring geospatial accuracy of exposures limit the degree to which the model portrays false positives and will reflect a more accurate picture of risk geospatially?
• What is the business case for industry databases, and what are the obstacles to creating them when other sectors like aviation have gained so much?
• Is there a way of standardising a definition of what makes a high frequency, low severity event?
• Can we design practical tools for the use of models at both the qualitative strategy level – to structure decision making at board level – and the quantitative decision level- to support business decisions by senior management?

If you are an academic and believe you can help fulfil or progress the questions above, or if you are an insurer wishing to expand or share your own expertise on the topic, please get in contact with us to see how you can help.