As the costs from devastating wildfires in different parts of the world have mounted in recent years, the issue of risk and exposure for re/insurers has rapidly climbed the agenda.
A panel of experts has told Intelligent Insurer that a lot more needs to be done to truly tackle a peril that one described as “a key canary in the coal mine” in terms of how climate change is already manifesting for the re/insurance industry.
In the panel discussion, “Grappling with uncertainty: wildfire risk and how the industry can better manage it”, broadcast on Intelligent Insurer’s Re/insurance Lounge platform, it became clear that a “rapid evolution” in wildfire re/insurance is needed as the impact of climate change on the peril was described as “dramatic”.
To put the scale of the issue in context, panellist Tom Larsen, principal, insurance and spatial solutions at CoreLogic, says that in the case of the US it is useful to look at the past 20 years. He says that US federal government figures show that fire losses per capita have been declining over the last 20 years when wildfires are excluded.
“When we include the wildfires we’re seeing an increase. What that means to fire insurance is that the regular losses, the uncorrelated losses, eg, from kitchen fires, are declining and the volatility portion is increasing.
“Through that context, we see the recent reinsurance attachments for wildfires that were unprecedented in California more prescient, an omen, of what we’re going to see in the future.”
Larsen says there have been significant wildfire losses in the past 10 years.
“In 2020 over 10,000 homes in the US have been burned,” he explains. “The losses from 2020 in the US are not likely to exceed what we saw in 2017 and 2018, but they are likely to end up in the $5 billion to $10 billion range. having already peaked at $20 billion in previous years of wildfire seasons.”
This risk is also on re/insurers’ radars for another reason, says Simon Young, senior director at Willis Towers Watson’s (WTW) Climate & Resilience Hub.
“Wildfire risk has been a key canary in the coal mine in terms of climate change. The growing impact on the insurance industry and the insurers on wildfire risk is dramatic, it’s more dramatic than any other feature of climate change if you look at the raw numbers.
“There are other factors related to wildfire risk but in particular climate change is a big driver if you look at activity in the western US, Australia, and southern Europe,” Young says.
In the US tackling the peril is complex, explains panellist Jason Schupp, founder and managing member at independent think tank Centers for Better Insurance.
He points out that in the US there are standard fire policy statutes in a significant number of jurisdictions including California. “As the name suggests these are built around the peril of fire, which is core and fundamental to what insurance does.
“The fuel loads are enormous, so the fires are much more severe—absent any climate change stressor.”
Tom Larsen, CoreLogic
“They’re very different from other catastrophic risks such as hurricane, earthquake, terrorism and even pandemic, where you can plausibly think about the underlying product and suggest the possibility of sub-limits or exclusions or other types of tools to manage the risk. It’s just not possible with fire.”
Schupp doesn’t think it would be possible to sell a homeowners policy that doesn’t cover fire.
“I see it as a particularly challenging area because the tools that are there and available to the industry—as this has turned from a core bread-and-butter risk into a highly correlated peak peril—are quite limited,” he says.
“That’s going to push us, probably more quickly than we have seen with the other exposures, into thinking about how to bring a larger balance sheet to the table, most notably how to bring the government balance sheet here.”
Regarding insured losses from wildfire, says WTW’s Young, they have doubled each decade but as of 2009 the total was still below $10 billion globally.
“On the other hand, in this decade we’ve had two years which have been over $10 billion in a single year.”
The whole scale of insured losses has changed, he says.
“If we go back to the fundamentals of wildfire activity we see the impacts, particularly of climate change, in drying and changing climate conditions, in reducing snowpack, for example, in the Sierra Nevada for the western US, plus all sorts of other conditions, which all seem to be moving in a similar direction.
“The fundamentals of wildfire as a hazard are certainly changing very quickly as well as the more traditional changes in risk that we’ve been seeing over a longer period. Residential areas have been moving closer into the wildlife urban interface (WUI)—becoming more urbanised, effectively—so more properties would have been at risk anyway, even if the hazard had been at the same place. We have compounding issues,” Young explains.
On top of this, he points to the issue of forest management in the western US and the effect that has had over the past few decades of increasing the amount fuel in the forests.
“You’ve got the tinder, which climate change has now set alight to, to draw a relevant analogy. But it’s real: the lack of traditional forest management approaches in the western US, particularly putting out fires rather than letting them burn out, has really come home to roost now.”
“It’s more dramatic than any other feature of climate change if you look at the raw numbers.”
Simon Young, Willis Towers Watson
Larsen agrees that climate change is a stressor for wildfire and that it is exacerbating a bad situation. “We must look at it again in the context of the trees and forests that take decades to grow,” he adds.
Focusing on the US, he says that in the past decade a bad year meant around 10 million acres burned.
“That seems extreme because in the decades prior to that maybe about 5 million acres a year were burned. But if we look further back, prior to World War 2, a normal year had 30 million acres of land burned.
“Those fires reduce the fuel loads, so we’re not having fires as severe. We’re in a situation now, after 50 years of successfully and very effectively extinguishing fires, where the fuel loads are enormous, so the fires are much more severe—absent any climate change stressor.
“Then you add in climate change, the extended drying season that we get from climate change, and you make it a lot worse.”
Larsen agrees with Young that in America the preference for living in the WUI areas has added to the exposure for re/insurers, adding “that statement is defended by the rapid growth in the population”.
“When you look at that signal alone, it’s driving about 40 percent of the increase in the losses because we’re putting all that value in these rural areas.”
There’s no one signal, says Larsen, although he’s clear that climate change certainly makes wildfires worse.
The panellists are in consensus that there is more than one reason behind worsening wildfires. Schupp points to three aggravating factors: climate change, encroachment into forest land by commercial and residential property, and public decisions that are made around forest management “specifically being aggressive about containing and extinguishing forest fires rather than having them clear out the mounting fuel”.
However, Schupp adds: “When we step back and ask who bears the cost of that, there is the challenge. There is only so much we can point at the homeowner, who is behind only one of those three causative facts, which is the desire to live in these areas that are at the WUI.
“The other two factors can’t be attributable to individual decisions to live in those areas. So it becomes challenging from a public policy standpoint to expect policyholders who maintain property in those locations to pay the full freight of that risk.
“That probably brings us to some sort of a public-private partnership (PPP) to spread the cost of larger policy decisions around forest management and ultimately the causes behind climate change.”
“Billions are being spent on fuel management and some of it is extremely effective.”
Overcoming a complex problem
Modelling is a key element for assessing risk although, Larsen says, it is trickier for wildfire.
“First let’s talk about the similarities of modelling wildfire to modelling other perils: they’re physics based, we understand how fires grow and spread and there are many different types of models out there and much research being done because these models are used today to anticipate the spread of existing fire so that we can fight it, and they’re very effective at it,” he explains.
These types of model can be used in conjunction with other important data layers, he says, such as fuel, topography, and the home locations, to be able to come up with a fire spread model and an ignition model that can help provide a rational understanding of what the risk is in an area.
“That’s the common part of it: we have the basic bones of a model so that we can understand how a fire is ignited and we can build a framework of a model that’s rational, similar to the earthquake models and hurricane models that are widely used in risk transfer and policy assessment.
“Where we are trying to improve the dynamic is by looking at the fuels and fuel management. Billions are being spent on fuel management and some of it is extremely effective and reducing our aggregate risk.
“Some of it is focused, in California, on protecting utilities and not so much on protecting society. So it’s quickly evolving. Fuel management is a moral hazard that is owned by federal authorities, states and a whole lot of other entities that are influencing the risk.
“These models are there but there’s a human element, a dynamic of it, that requires us to update the model as we change our defences and it’s much more influential than it is with the earthquake and hurricane models, which are more about physics, the geophysics of the earth.”
Young says Larsen is absolutely right that looking at the raw hazard is challenging because the conditions are changing quite frequently.
“If you think about earthquake and shaking intensity at the surface, it’s to do with the rock and soil type for example for a given magnitude of earthquake in a given location. That isn’t changing very quickly, whereas conditions in the forest, which are driving the spread of wildfires, are changing quickly.”
Young is particularly interested in documenting the factors that are important for wildfire using modelling from space. He points to the improvements in the resolution and the frequency of satellite images to capture information as a real positive.
“Forest canopy, for example, and soil moisture are two very important criteria for short-term wildfire prediction and management, and also from a modelling perspective.
“Those are the sorts of things we’ll see becoming more important and feeding better data, the big data from space.”
The community factor
Young says the other piece of the puzzle is community-level resilience.
“If you protect your own house from an earthquake, you retrofit it and tell your insurance company, they come and check that you’ve done it, you know that is going to have an influence on the damage from a given earthquake.
“For wildfire it’s not quite the same—there is a controlling factor in what everybody else is doing around you.
“There’s a sense that the issue of wildfire risk and exposure has crept up on the industry.”
Jason Shupp, Centers for Better Insurance
“Community-based information around the resilience of the community, rather than your own home, is more of a factor in reducing your home’s risk for wildfire than for other perils. It’s an important aspect that we have to do more work on; it’s very important to identify and recognise.”
With regard to how the industry has managed its response to this growing area of risk, Young says that “he doesn’t want to dump on the insurance industry, but it’s crept up very suddenly”.
He adds: “From an industry-wide strategy perspective, there’s a bit of being caught unawares, and the public aspect of this is going to be extremely important.
“We’re going to have to think about how we can put together PPPs to manage it because what has happened, and we’ve seen this over the summer in terms of wildfire liability covers, is the rates being paid for those who have to buy it are astonishing from a traditional cat insurance perspective.
“The re/insurers’ response has been, ‘well we’ll sell it at the cost we think we need to charge for it to those who can possibly afford it, or who have to buy it for other reasons’, be they regulatory or financing reasons.”
That is not a sustainable response, particularly at the homeowner level, Young says.
“We’re going to have to see rapid evolution. There’s history with this on other perils: the setting up of the Terrorism Risk Insurance Act of 2002 after 9/11; and the creation of the California Earthquake Authority after Northridge in 1994,” he recalls.
More work needed
Schupp agrees that there’s a sense that the issue of wildfire risk and exposure has crept up on the industry—and it has certainly crept up on regulators and policy makers as well, he says.
“As a result what we see in California and the homeowners space is an effort to lock existing insurers into place. This makes it difficult, if not impossible, to non-renew or cancel policies.
“It’s the ‘Hotel California’ approach to insurance: ‘you’re in, so we’re going to hold you there as long as we possibly can’,” he says.
This is a “fairly desperate measure”, warns Schupp, because eventually insureds will be able to decouple from those commitments and it will be hard to encourage them to stay, and hard for other carriers to encourage them to come in when they’re dealing with the spectre of not being able to unwind a risk in the way they thought they had the right to do.
“We are seeing some maybe entirely necessary, but fairly short-term, solutions being put in place, but we need to shift towards longer-term more robust solutions.”
Schupp mentions the California Wildfire Fund in the US, “but that is quite focused and narrow in terms of its scope and responds only to limited circumstances where a utility is at fault with respect to the causation of a wildfire”—and many wildfires are not caused by utilities.
“There’s still a lot of work to be done on how to get our arms around the wildfire risk, and who is going to bear the ultimate cost of this.”
With so much more that can be done, particularly within the framework of PPPs, mitigation awareness and action, and greater use of data to produce more accurate models, the panellists looked ahead to a more positive future for managing the risks of this unique peril.
For more on PPPs, mitigation and the future of wildfire risk, click here to watch the full discussion on Intelligent Insurer’s Re/insurance Lounge.
Image: Shutterstock / Toa55