Modelling risks in a shifting environment


Modelling risks in a shifting environment

Adam Podlaha, head of Aon's Impact Forecasting, Aon

Modelling complex risks ranging from hurricanes to cyber attacks, wildfires and floods, is no easy task.

Over the last few years as the re/insurance industry has faced a series of large claims from events which previously may not have even been on the radar, the market has had to reconsider its use of catastrophe models and how it can accurately predict and price risk in a changing landscape.

Adam Podlaha, head of Aon’s Impact Forecasting, the firm’s specialist modelling unit, knows the scale of the challenge facing the industry better than most.

As floods in Europe and wildfires across the US provide a timely reminder of the work yet to be done, Podlaha sat down in the Re/insurance Lounge, Intelligent Insurer’s online, on-demand platform for interviews and panel discussions with industry leaders, to talk about how Aon is helping its clients get to grips with the issue.

“Climate change and conditioning the models to take into account the effects of climate change has become almost a standard.” Adam Podlaha, Aon’s Impact Forecasting

Climate models

Heading a team of around 130 people operating across 80 countries, Podlaha has an impressive vista over the catastrophe modelling world.

One issue casting a large shadow over that view is climate change. As leaders across the industry move to tackle the changing threats that shifts in the underlying environment are causing, the modelling community is working ever harder to get to grips with the issue and how it is impacting previous assumptions.

Podlaha said the issue is at the top of the firm’s list when it comes to developing models that the industry can rely on in the years to come.

“Climate change and conditioning the models to take into account the effects of climate change has become almost a standard, and there is enough of a market of scientific materials either already in existence or being developed which can support it,” he said.

“From the model developers’ point of view, as well as the risk carrier and re/insurance company points of view, it’s always important to prioritise the amount of work and the scope, which you want to do for certain perils.

“If hurricanes are peril number one, then you need to spend a lot of time and effort to quantify the effect of climate change.

“For US hurricanes and tropical cyclones, we partner with Columbia University—with Adam Sobel and the team around him—to help us to quantify the effects of climate change on tropical cyclones.”

For more content like this visit the Reinsurance Lounge 

“With some very prominent recent events, more developers will be considering how they can create effective flood models.”

Flood and fire

A related issue which has caught the more immediate attention of the modelling industry is how to tackle the notoriously tricky issue of floods.

During the summer, large swathes of Europe were suddenly faced with widespread flooding from unprecedented concentrated rainfall, causing loss of life and devastation across the region and handing the re/insurance sector significant losses which fell outside the scope of previously modelled scenarios.

Podlaha says the events have forced the industry to confront the issue head-on as it seeks to learn from the devastation and apply the lessons to future catastrophes.

“Being the costliest peril type globally and with some very prominent recent events, more developers will be considering how they can create effective flood models. On the hazard side, the models are getting increasingly precise with more, and better elevation data.

“There are also improved modelling techniques on the spread of the flood, and creating the flood extent, as well as the use of available flood defence data,” he explained.

“With the recent losses, what it will give us—despite its being a terrible tragedy—is lots of valuable insight, especially in terms of claims information, which we can incorporate in the data or into the model.

“We are working currently on a German flood model, which we don’t have in our suite yet. The recent flood event in Germany will be one of the major events which will define the model. Events like this, plus those happening in the US and elsewhere in Europe, are pushing the flood model forward.”

Wildfires are another area where the company is working hard to adapt models and use more data which has become available from a devastating series of events over the last four years to improve risk analysis and management for clients.

The increased frequency has provided more information for Impact Forecasting to work with, and Podlaha noted that the complexities of the regions where wildfires hit, especially in California, mean that it will remain a key challenge for the industry.

“With wildfires it’s about two things: climate change, which is having an effect, and the exposure growing in places which are vulnerable, and that is also having quite a big effect. There is much more time spent in general in creating wildfire models,” he said.

“Factors such as terrain, wind, directional shapes, and the impact of smoke, will need to be considered in a more detailed way in future models. We are working on our California and US Western states models, where we are using a dynamic way of modelling the fire spread.

“It’s a very localised peril and therefore very complex to model.”


To view the full Re/insurance Lounge session click here


Adam Podlaha is head of Aon’s Impact Forecasting. He can be contacted at:

Aon, Aon Impact Forecasting, Risk Modelling, Catastrophe, Cyber Attacks, Insurance, Reinsurance, Adam Podlaha, North America, Global

Intelligent Insurer