How is climate change influencing natural perils and weather extremes, and what should reinsurance companies do to respond?
Reinsurance companies may feel they are relatively insulated from the immediate effects of climate change on their business, given that most property catastrophe policies are renewed on an annual basis. However, with signs that we are already moving off the historical baseline when it comes to natural perils, there is evidence to suggest that underwriters should already be selectively factoring the influence of climate change into their day-to-day decision-making.
Most climate scientists agree that some of the extreme weather anticipated by the United Nations Intergovernmental Panel on Climate Change (IPCC) in 2013 is already here and can be linked to climate change in real time via the burgeoning field of extreme weather attribution. “It’s a new area of science that has grown up in the last 10 to 15 years,” explains Dr. Robert Muir-Wood, chief research officer at RMS. “Scientists run two climate models for the whole globe, both of them starting in 1950. One keeps the atmospheric chemistry static since then, while the other reflects the actual increase in greenhouse gases. By simulating thousands of years of these alternative worlds, we can find the difference in the probability of a particular weather extreme.”
“Underwriters should be factoring the influence of climate change into their day-to-day decision-making”
For instance, climate scientists have run their models in an effort to determine how much the intensity of the precipitation that caused such devastating flooding during last year’s Hurricane Harvey can be attributed to anthropogenic climate change. Research conducted by scientists at the World Weather Attribution (WWA) project has found that the record rainfall produced by Harvey was at least three times more likely to be due to the influence of global warming.
This suggests, for certain perils and geographies, reinsurers need to be considering the implications of an increased potential for certain climate extremes in their underwriting. “If we can’t rely on the long-term baseline, how and where do we modify our perspective?” asks Muir-Wood. “We need to attempt to answer this question peril by peril, region by region and by return period. You cannot generalize and say that all perils are getting worse everywhere, because they’re not. In some countries and perils there is evidence that the changes are already material, and then in many other areas the jury is out and it’s not clear.”
While the last IPCC Assessment Report (AR5) was published in 2014 (the next is due in 2021), there is some consensus on how climate change is beginning to influence natural perils and climate extremes. Many regional climates naturally have large variations at interannual and even interdecadal timescales, which makes observation of climate change, and validation of predictions, more difficult.
“There is always going to be uncertainty when it comes to climate change,” emphasizes Swenja Surminski, head of adaptation research at the Grantham Research Institute on Climate Change and the Environment, part of the London School of Economics and Political Science (LSE). “But when you look at the scientific evidence, it’s very clear what’s happening to temperature, how the average temperature is increasing, and the impact that this can have on fundamental things, including extreme events.”
According to the World Economic Forum’s Global Risks Report in 2018, “Too little has been done to mitigate climate change and … our own analysis shows that the likelihood of missing the Paris Agreement target of limiting global warming to two degrees Celsius or below is greater than the likelihood of achieving it.”
The report cites extreme weather events and natural disasters as the top two “most likely” risks to happen in the next 10 years and the second- and third-highest risks (in the same order) to have the “biggest impact” over the next decade, after weapons of mass destruction. The failure of climate change mitigation and adaptation is also ranked in the top five for both likelihood and impact. It notes that 2017 was among the three hottest years on record and the hottest ever without an El Niño.
It is clear that climate change is already exacerbating climate extremes, says Surminski, causing dry regions to become drier and hot regions to become hotter. “By now, based on our scientific understanding and also thanks to modeling, we get a much better picture of what our current exposure is and how that might be changing over the next 10, 20, even 50 to 100 years,” she says.
“There is also an expectation we will have more freak events, when suddenly the weather produces really unexpected, very unusual phenomena,” she continues. “That’s not just climate change. It’s also tied into El Niño and other weather phenomena occurring, so it’s a complex mix. But right now, we’re in a much better position to understand what’s going on and to appreciate that climate change is having an impact.”
For insurance and reinsurance underwriters, the challenge is to understand the extent to which we have already deviated from the historical record and to manage and price for that appropriately. It is not an easy task given the inherent variability in existing weather patterns, according to Andy Bord, CEO of Flood Re, the U.K.’s flood risk pool, which has a panel of international reinsurers.
“The existing models are calibrated against data that already includes at least some of the impact of climate change,” he says. “Some model vendors have also recently produced models that aim to assess the impact of climate change on the future level of flood risk in the U.K. We know at least one larger reinsurer has undertaken their own climate change impact analyses.
“We view improving the understanding of the potential variability of weather given today’s climate as being the immediate challenge for the insurance industry, given the relatively short-term view of markets,” he adds.
The need for underwriters to appreciate the extent to which we may have already moved off the historical baseline is compounded by the conflicting evidence on how climate change is influencing different perils. And by the counterinfluence or confluence, in many cases, of naturally occurring climate patterns, such as El Niño and the Atlantic Multidecadal Oscillation (AMO).
The past two decades have seen below-normal European windstorm activity, for instance, and evidence builds that the unprecedented reduction in Arctic sea ice during the autumn months is the main cause, according to Dr. Stephen Cusack, director of model development at RMS. “In turn, the sea ice declines have been driven both by the ‘polar amplification’ aspect of anthropogenic climate change and the positive phase of the AMO over the past two decades, though their relative roles are uncertain.
“We view improving the understanding of the potential variability of weather given today’s climate as being the immediate challenge for the insurance industry, given the relatively short-term view of markets”
“The (re)insurance market right now is saying, ‘Your model has higher losses than our recent experience.’ And what we are saying is that the recent lull is not well understood, and we are unsure how long it will last. Though for pricing future risk, the question is when, and not if, the rebound in European windstorm activity happens. Regarding anthropogenic climate change, other mechanisms will strengthen and counter the currently dominant ‘polar amplification’ process. Also, the AMO goes into positive and negative phases,” he continues. “It’s been positive for the last 20 to 25 years and that’s likely to change within the next decade or so.”
And while European windstorm activity has been somewhat muted by the AMO, the same cannot be said for North Atlantic hurricane activity. Hurricanes Harvey, Irma and Maria (HIM) caused an estimated US$92 billion in insured losses, making 2017 the second costliest North Atlantic hurricane season, according to Swiss Re Sigma. “The North Atlantic seems to remain in an active phase of hurricane activity, irrespective of climate change influences that may come on top of it,” the study states.
While individual storms are never caused by one factor alone, stressed the Sigma study, “Some of the characteristics observed in HIM are those predicted to occur more frequently in a warmer world.” In particular, it notes the high level of rainfall over Houston and hurricane intensification. While storm surge was only a marginal contributor to the losses from Hurricane Harvey, Swiss Re anticipates the probability of extreme storm surge damage in the northeastern U.S. due to higher seas will almost double in the next 40 years.
“From a hurricane perspective, we can talk about the frequency of hurricanes in a given year related to the long-term average, but what’s important from the climate change point of view is that the frequency and the intensity on both sides of the distribution are increasing,” says Dr. Pete Dailey, vice president at RMS. “This means there’s more likelihood of quiet years and more likelihood of very active years, so you’re moving away from the mean, which is another way of thinking about moving away from the baseline.
“So, we need to make sure that we are modeling the tail of the distribution really well, and that we’re capturing the really wet years — the years where there’s a higher frequency of torrential rain in association with events that we model.”
Over the long term, the industry likely will be increasingly insuring the impact of anthropogenic climate change. One question is whether we will see “no-go” areas in the future, where the risk is simply too high for insurance and reinsurance companies to take on. As Robert Muir-Wood of RMS explains, there is often a tension between the need for (re)insurers to charge an accurate price for the risk and the political pressure to ensure cover remains available and affordable.
He cites the community at Queen’s Cove in Grand Bahama, where homes were unable to secure insurance given the repeated storm surge flood losses they have sustained over the years from a number of hurricanes. Unable to maintain a mortgage without insurance, properties were left to fall into disrepair. “Natural selection came up with a solution,” says Muir-Wood, whereby some homeowners elevated buildings on concrete stilts thereby making them once again insurable.
“In high-income, flood-prone countries, such as Holland, there has been sustained investment in excellent flood defenses,” he says. “The challenge in developing countries is there may not be the money or the political will to build adequate flood walls. In a coastal city like Jakarta, Indonesia, where the land is sinking as a result of pumping out the groundwater, it’s a huge challenge.
“It’s not black and white as to when it becomes untenable to live somewhere. People will find a way of responding to increased incidence of flooding. They may simply move their life up a level, as already happens in Venice, but insurability will be a key factor and accommodating the changes in flood hazard is going to be a shared challenge in coastal areas everywhere.”
Political pressure to maintain affordable catastrophe insurance was a major driver of the U.S. residual market, with state-backed Fair Access to Insurance Requirements (FAIR) plans providing basic property insurance for homes that are highly exposed to natural catastrophes. Examples include the California Earthquake Association, Texas Windstorm Insurance Association and Florida Citizens Property Insurance Corporation (and state reinsurer, the FHCF).
However, the financial woes experienced by FEMA’s National Flood Insurance Program (NFIP), currently the principal provider of residential flood insurance in the U.S., demonstrates the difficulties such programs face in terms of being sustainable over the long term.
With the U.K.’s Flood Re scheme, investment in disaster mitigation is a big part of the solution, explains CEO Andy Bord. However, even then he acknowledges that “for some homes at the very greatest risk of flooding, the necessary investment needed to reduce risks and costs would simply be uneconomic.”