On the 13-14^th September, OASIS Loss Modelling Framework hosted ‘The Good, the Bad and the Ugly’, a London conference that brought the catastrophe community together. The event consisted of insightful presentations, featuring keynote speakers from the industry. In the run-up to the event, we published a blog outlining ‘The Good, the Bad and the Ugly’ of catastrophe models. We’ve pounced upon the theme again, bringing you ‘A Fistful of Cat Models’, a debrief of the insightful event. This blog aims to summarise the main topics that were presented through the course of the event.

The Conference:

The Conference was spread out over 2 days and was presented by a diverse group of experts in their respective disciplines. Ambiental attended and exhibited at the event, showcasing our flood modelling and catastrophe modelling expertise. The first day focussed more specifically on earthquakes and hurricanes, which was rather timely in the context of Hurricane Florence, which at the time was eminently approaching the west coast of the USA.  The second day had more of an emphasis on flooding, and for that reason, this blog will focus on day 2. The main topics included: ‘Understanding Cat Risk in a Warming Climate’; ‘What Lessons can be Drawn From Recent Events and the Implications for Cat Modelling’; ‘What Key Challenges the Industry Faces in Realising’; and ‘Understanding Flood Risk and the Future of Cat Modelling’.

Below is an overview of some of the presentations and their key takeaways.

The main conclusions that can be drawn from the “Understanding Cat Risk Better in a Warming Climate” workshop was to be aware that the past is not prologue and catastrophic norms of the future are in the tail norms of today. This is crucial to understand when modelling flood risk, which is, by its very nature, a peril that is challenging for the insurance industry to predict. Incorporating climate change into catastrophe models remains one of the key challenges the industry faces. Ambiental, being at the forefront of flood risk management, has developed the capacity with the associated workflows to predict flooding accurately. We have additionally produced the FloodFutures® flood map, which incorporates the most up to date climate change scenarios for Great Britain.

Among the many conclusions to be drawn from the presentations included in the workshop “Recent Cat Events – Lessons Learnt and Implications for Cat Modelling”, was that loss adjusters’ expenses were a key unmodelled risk and that non-structural elements are perhaps more expensive than structural elements, and this was not included in Vulnerability Curves, which help define the loss calculators within the financial component of the catastrophe model. Another argument made the case that depending on the insurance company, claims may not be filed leading to an underrepresentation of risk, ultimately undermining the reliability of claims data. A final argument put forth from within this workshop was that in order to better represent the risk of peril in the context of climate change, it could be beneficial to have continuous data (rainfall data, tremors data, river gauge data etc..) inputted to the catastrophe model. This is because, in a warming climate, historical records cannot be used to predict the tail distributions of the perils.

The final two-panel discussions were the “Key Challenges Faced in Realising and Understanding Flood Risk” and “The Future of Cat Modelling”. The main conclusions from these two high calibre discussions boil down to three main factors; it is imperative that catastrophe models seek to portray natural perils in the context of climate change, the drive should be focused on acquiring better exposure data as without this, the financial component will not represent the risk appropriately, and finally, transparency is key to success. Transparency relies on each stakeholder being transparent in communicating the limitations, the specifications and the constraints of each model, but it is vital in order to mitigate the risk of inadvertently modifying the models and cause unnecessary modifications, that can ultimately result in less reliable results.

Ambiental endorses the open and transparent philosophy where model parameters, limitations and methodologies are communicated to the user to ensure the unknowns are known. Ambiental’s FloodCat Australia model is available through the transparent and open platforms of Aon Benfield ELEMENTS and Simplitium ModEx.

Better ways of forecasting events and dynamically modelling their impacts is also of growing interest to the insurance industry. Ambiental has been supporting the Malaysian Government through the development of FloodWatch; a software platform designed to provide high value decision support to tackle environmental challenges. It enables an early warning forecasting system for sensitive and flood prone areas several days in advance. This helps the government manage and mitigate the impact of flooding in the highly flood-prone country.

To discuss Ambiental’s latest work in catastrophe modelling or how we can assist with your flood risk strategy, please email us on info@ambientalrisk.com. Alternatively, call us on + 44 (0) 203 857 8545 .