Aon Benfield, Lloyd’s and Intersys on new trends for Catastrophe modelling

Catastrophe modelling – at one time a simple, single-event spreadsheet in an insurer’s toolbox – is now a multi-tiered, multi-event modelled approach to risk management. New versions of catastrophe models are designed to take into account almost every aspect of potential natural and man-made disasters, alongside a company’s own risk portfolio, to provide a much broader picture of what is at stake. 

Catastrophe models also have the ability to move and change with an ongoing event, which makes them an indispensable tool for risk managers.

The future for catastrophe modelling looks bright, with the incorporation of new information from scientific models of hazard and engineering perspectives on damage. However, experts are also warning businesses not to rely exclusively on catastrophe modelling as a tool for determining a business’s potential catastrophe risk. 

Following a number of extreme natural catastrophes in the past year – super typhoon Haiyan, polar vortex winter storms in the US and some of the worst flooding on record in Canada, Australia, India, China, Indonesia, southern Africa and Argentina – experts are concerned about the current scope of models. They believe they do not sufficiently address the unpredictable nature of disasters to provide a single solution for business.

Aon Benfield head of impact forecasting Adam Podlaha says the trend for catastrophe modelling is heading in the right direction by including more scientific analysis, as well as alternative model providers such as academics and proprietary firms. 

“These are basically taking into account the things we did not know with the models already available,” he says. “An example is [modelling for] earthquakes in New Zealand, which takes into account the same modelling as the Japanese earthquake.” 

He points to a similar trend with tsunami modelling. “There is a big shift towards a multi-model approach. There are the big companies that are modelling, but there are also lots of new modelling companies coming out as well.”

One such firm, Intersys, is looking at ways to combine catastrophe modelling with a company’s own risk assessment data to create a more holistic approach to catastrophe modelling. 

Intersys director of risk management Cath Geyman says: “We are looking to overlay flood and earthquake zones on the data that our clients work up. Our clients are assessing their own organisations to understand where their critical points are and plot them on a map in order to work out their true exposure. 

“What they really then want to do is take the flood and earthquake maps produced by traditional catastrophe modelling firms and combine the two data sets in order to access their risk.”

She adds: “At the moment, there is a lot of focus in the modelling world on really understanding each hazard in a lot of detail and being able to model and predict it in a lot of detail. Perhaps what’s missing is the company’s view on that – each organisation understanding its own complexity, every critical part of it and what the impact would be on it financially should one of these natural catastrophes hit.”

Developing economies

Lloyd’s head of exposure management and reinsurance Trevor Maynard believes that the future of catastrophe modelling – as well as the likely insurance options – is linked closely to developing economies. 

“As economies develop around the world and populations become wealthier, there will be an increasing demand for insurance,” he says. “These regions will become more material and therefore there will be an increasing need for more granular modelling in those regions.”

But experts in developing nations believe that despite a high rate of natural catastrophes in these regions, the future of catastrophe modelling is still uncertain and there is plenty of groundwork still to cover. 

Lockton Wattana Insurance Brokers (Thailand) executive chairman Wattana Wongvisesnopakun says: “The use of catastrophe modelling in Thailand is still very small compared with the international market. It is not as advanced here, so there is not much use for modelling for the industry here just yet.” 

Recent extreme natural disasters, such as super typhoon Haiyan, which killed more than 7,000, are testing the scope and coping mechanisms of current catastrophe models. Rapidly changing conditions, in combination with the extreme nature of catastrophes, mean risk managers are often faced with using out-of-date information. 

“Generally, hazards do not change that rapidly, although our understanding of them can,” says Maynard. “The Japanese earthquake in 2011 showed us that mag 9 quakes can occur in that region – when models did not expect that. Inevitably, there will always be new information that surprises us. However, current models have a large spread of events already, which has led to greater financial strength. 

“It should be noted, for example, that the Japanese earthquake event in financial terms was much less than some other larger events contained in the model.”

Maynard points out that research by Lloyd’s and catastrophe modelling firms shows that models incorporate changing conditions implicitly but not explicitly. 

“For example, the approximate 20cm of sea level rise at the Battery in New York since the 1950s increased [Cyclone] Sandy’s storm surge loss by 30%. Climate change is therefore clearly having an effect on damages,” he says.

Lack of flexibility

Podlaha believes it is the unpredictable nature of catastrophes and the innate lack of manoeuvrability of catastrophe models that leave businesses most exposed. 

“Most of the current models come as they are,” he says. “They are designed to model extreme weather. They are optimised for big events and the models are officially designed to yield information on big events. 

“But from time to time, things happen that the modelling company doesn’t take into consideration. The result is big, it is bad, and [the company] knew it was always there, but because it did not happen previously, it was not considered properly. 

“In general, the models can cope, but incrementally. You do the basic risk assessment. Following every new event, we learn something new.”

A lack of accurate predictability is a driving factor behind the extreme exposure that most
businesses face in catastrophic situations. Experts say this unpredictability will continue to underpin catastrophe models for the foreseeable future. 

“It is important to realise that catastrophe models are not forecasts. They do not predict what events will occur in the coming year. They show the range of possible events that might occur and their likely costs as insurers deal in probabilities,”
says Maynard.

“Current models already build in features such as recent higher sea surface temperatures into their projections. The intention, specified by insurance regulation, is to predict current levels of risk and if these are different to the past then adjustments should be (and are) made,” Maynard adds.

Oasis framework

The need for more predictable information for risk managers has led to the development of the Oasis Loss Modelling Framework, owned by 21 insurers, reinsurers and brokers, including Lloyd’s. The initiative aims to provide a framework for independent catastrophe modelling. 

Oasis project director Dickie Whitaker says: “Risk managers from multinational companies will be able to take advantage of catastrophe modelling techniques in a way that has not been practical with large models built for the insurance industry. 

“Corporate risk managers seek better information that helps to understand the risks associated with possible catastrophic events. In many cases, this information is not easy to access in a form that is easily useable and some of the avenues available can be very expensive.”

The insurance options available to businesses are still costly, say risk managers, but increased uptake of government pools is making this more viable. 

“The insurance industry pools risks from around the world to help communities and businesses to offset the randomness of extreme events,” says Maynard. “Governments are increasingly seeing the relevance of insurance protection. 

“For example, the CCRIF [Caribbean Catastrophe Risk Insurance Facility] is a parametric product that provides fast liquidity to Caribbean governments in the event of earthquakes, excess rainfall or hurricanes.”  

Risk managers need to be more aware of the potential scale of natural hazards facing them and how these might change. FERMA president and DLA Piper chief risk officer Julia Graham believes catastrophe modelling is merely one string to the bow of risk managers and needs to be part of the wider scope of enterprise risk management.

“Risk managers need to work with insurance managers and business continuity managers as a team. That’s what enterprise risk management is all about,” she says. “It is also about risk managers operating at the most senior level in their organisation. The risk manager needs to be there when businesses are thinking of developing their plans or their businesses in places that perhaps are not totally advisable.

“I think that’s getting the risk manager at the right place at the boardroom table to be acting as an adviser in some of these areas, not after the decision has already been made.”

Veolia Water risk and business continuity manager Lenny-Baptiste Conil sums up the complex situation in simple terms. “Predicting is impossible,” he says, “so planning, training people and keeping up-to-date procedures is the way to go.”