Climate change: reduced time and location data veils model opportunities

Existing nat cat models are too limited in time frame and geography to prepare global society and re/insurers for a dramatically different climate in the future, a panel of experts has warned.

The four experts on the Intelligent Insurer panel argued that such models, and the decision-makers who rely on them, must broaden their geographical scope and work with predictions that go beyond the typical five years that C-suite executives and politicians currently work with.

“In my old job at RMS, we focused on models that covered where most of the insurance premium was but we need to be thinking about it a bit more globally,” said Meghan Purdy, senior product manager at Jupiter Intelligence. “The places where we are considering these climate risks need to expand in the future.”

Purdy was speaking as part of an Intelligent Insurer panel titled “Climate change: adapting models to better understand the systemic threat” ( part 1 of the session report detailed a discussion on the need to fundamentally overhaul the way nat cat models are designed and built).

Purdy’s fellow panel members were Bhaskar Chattaraj, partner, head of modelling and R&D at TigerRisk; Sébastien Piguet, co-founder and head of underwriting at Descartes Underwriting; and Daniel Stander, special advisor at the United Nations Development Programme (UNDP); with Claire Churchard, deputy editor of Intelligent Insurer, as moderator.

Purdy said that the narrower model view was a problem right now.

“We have the SEC and other regulators asking people to model not just where their own assets are but also their supply chains and distribution networks,” she said.

This means going to places in the world that have a lower insurance penetration and therefore fewer catastrophe models.

“Expanding to a more global reach is certainly important because, from a climate justice perspective, it tends to be the developing countries who bear the brunt of climate change. If we can produce metrics to help them understand what they’re going to be facing then they can do more to adapt to it in the future,” she added.

Stander agreed with Purdy, saying: “The vast majority of the models that exist in the industry today are focused on a minority of regions such as North America, Western Europe and Japan. There’s very little that is modelled at all, let alone well, in Africa, Central or South America, or South East Asia, so there’s a lot of room for the models to improve.”

Buy some time
Data limitations in the models are not just about geography. A perceived lack of interest in predictions that look beyond four or five years is leading to missed opportunities.

Chattaraj explained the issue faced by analytics firms. “You go to an insurance company and present the things that can affect their portfolio and explain what will happen. You say: ‘In the short term we cannot say anything very clearly but in the medium to long term we can give you some pretty robust ways of doing things’,” he said.

“The problem is the C-level of insurance companies often don’t care about beyond five years. So all these big companies have a committee, they’ll hire three people and create big reports, but when it comes to action, it’s bobcus [a load of baloney],” he said.

This seeming lack of interest in taking action on the longer-term data has also been seen by Purdy.

“There’s the idea that prices change every year, especially for traditional insurance underwriting, so it can be changed as it happens,” she said. “But I think that they’re missing out on an opportunity. It’s pretty myopic.”

The panel touched on the issue of the current models being “undercooked” and Purdy added that underwriters may be making decisions based on a poor understanding of present-day risk. “They’re missing out on chances to build stronger partnerships with their insureds and be that trusted provider going forward,” she said.

“They’re missing out on understanding that their company may need to start moving away from these different markets and, for example, how quickly they need to flee Florida to avoid sea level rise. I don’t think they have a good understanding of that.”

Increasing the amount of data fed into nat cat models brings its own challenges. Chattaraj was positive about the opportunity presented by such “immense amounts” of data from radar and geophysical data, and data from synthetic aperture radar (SAR) and the internet of things.

“The global sub-climate models produce petabytes of data (one petabyte = 1,000 terabytes), but to go through that data is not humanly possible, so this is where artificial intelligence (AI) and machine learning (ML) come in,” he explained.

“We have to use AI and ML. Where they will come in very handy is to give the data some structure and see if it’s useful. Then the human comes in to make more sense out of it. But using AI and ML is almost a necessity.

“If you want to work with this much data, which tells you about climate a few years in the future, you have to use these techniques.”

Silent and invisible risk

The discussion flagged up heat as a chronic risk that doesn’t yet get the attention it deserves, even with high profile events such as the heat wave/heat dome over North America and Canada in June and July 2021, and the serious heat wave in India and Pakistan in March and April this year.

“Heat might not be an event that triggers a reinsurance payout but it does have problems on P&C and life insurance,” said Purdy. “So we could see heat deaths, workers’ comp claims, business interruption on high heat days or aeroplanes that can’t fly at full capacity above certain temperature thresholds. I think we’re going to see a lot more on heat that we need to understand.”

The Atlantic Council think tank has done good work looking at the economic impacts of heat, especially on productivity, Stander said.

“They’re predicting that by 2050 there will be half a trillion dollars of productivity lost per year in North America alone. It’s a huge amount. I don’t know how well that’s understood but more than that I don’t know how well that is insured today. So for me it’s also a conversation about new insurance products.”

Some new products are already available for this type of event, with Descartes, for example, offering heatwave cover. Piguet said: “Many growers and agri businesses are interested in buying cover that can include heat waves.”

The wider take-up of this kind of product is hindered by a lack of understanding of heat risk, with the result that companies often underestimate the risk, he said.

“When we come up with a price which takes into account the impact of climate change and rising comparators, it’s very hard to sell because client companies are still massively underestimating the risk.”

He highlighted a case in the past six months where he was working with a large agri business in Northern America. “We had to come up with a price that was twice what they were expecting to pay and that the broker was expecting us to charge.

“There was an intense discussion with the broker because they were claiming that there was no clear trend in temperatures in Northern America, whereas it’s straightforward.

“We still have many people in the industry who are massively underestimating the trends, which is making our life more complicated, and as a risk it’s harder to sell innovative covers.”

Piguet said that this lack of understanding was “not serving” societies in developing or developed countries because to be able to adapt you first have to realise how bad the situation is.

He pointed to the limitations of reports such as those by the Intergovernmental Panel on Climate Change (IPCC) to reach wider audiences and build understanding. “When you look at IPCC reports and their conclusions, the conclusions are not widely shared by people around the globe. It’s a small report which is published on a regular basis.

“There is a lot of work just to share the conclusions of those reports among all countries all over the world and to discuss these conclusions with decision-makers from business and the developmental side.”

Stander agreed with Piguet that leadership and advocacy were key to a wider understanding of these issues. He said that reports such as the IPCC’s and data need to influence policy makers and the business community.

“The community needs to fully imbibe those findings and change regulatory, legal, and policy frameworks. But equally we need to change business practices to embrace what the science is telling us.

“You’re right, there’s a lot more work to be done on the advocacy and I’d like the insurance industry to partner with civil society and with governments to make sure we keep banging that drum so that behaviours are eventually changed,” he concluded.