Editor’s Note: This article first appeared in Insurance Day on June 10, 2019.
Climate change has been ongoing since the 1800s. As the first puffs of fossil-fueled exhaust began rising through the atmosphere, Svante Arrhenius, a Swedish physicist and chemist, was busy estimating the net impact on global temperature from a tripling of CO2. He didn’t have access to general circulation models or even an electronic calculator, but he still came up with a plausible estimate of a rise in global temperature of 8–9° C.
It is not clear if anyone took him seriously then or even showed any concern. But since that time, climate scientists with increasingly more sophisticated tools have been estimating not only climate change impacts on global temperature but also regional impacts on extreme weather, wildlife, natural habitats, and even the economy.
Despite the long history of scientific studies and concerning results, governments, urban planners, and others have been taking note of climate change only in the last two to three decades. And industries, including the insurance industry, have largely been showing greater concern only in the last several years. A likely few reasons for the delayed industrial response include a lack of confidence in the studies, the gradual nature of climate change, and no real evidence of extreme weather that could be attributed to climate change.
After Hurricane Andrew struck Florida in 1992, 11 insurance companies became insolvent, yet no one asked whether this storm was the result of climate change. But when hurricanes Katrina, Rita, and Wilma all occurred in 2005—on the heels of Charley, Francis, Ivan, and Jeanne in 2004—the industry did ponder whether climate change would bring more hurricane seasons like those. And after an unprecedented nearly 12-year lull in U.S. major hurricane landfall activity, the 2017 U.S. hurricane season had insurers pondering climate change again.
There had been at least one significant storm globally in a few of the intervening years that provoked the industry’s curiosity about the link to climate change: Sandy 2012 in the U.S. Northeast, Haiyan 2013 in the Philippines, and Patricia 2015 near Mexico. But during 2017, in just shy of one month’s time, hurricanes Harvey, Irma, and Maria generated nearly USD 270 billion in economic losses. The severity, relentlessness, concentration, and cost of those storms got the industry’s attention. In addition, 2017 and 2018 were record wildfire years. The economic losses from wildfires in 2017 reached USD 18 billion, tripling the previous record of USD 6 billion in 1991. And in 2018, that 1991 mark quadrupled to USD 24 billion.
Extreme Event Attribution
The growth in the number of weather disasters—inflation and increased exposure factors aside—has spawned a new branch of climate science called extreme event attribution. The science does not answer whether climate change caused a particular weather catastrophe, but it can quantify how climate change is affecting the return period of similar events. The year 2017, for example, demonstrated how precipitation events such as Hurricane Harvey will increase in frequency from a 2,000-year event in the late 1900s to a 100-year event by late this century (at present it is approximately a 300-year event) because of climate change.
The insurance industry is focused on what the next year will bring as well as the year after that and beyond, so the focus has evolved from a rather myopic one-year view to a more telescopic 10- to 15-year perspective.
Some reinsurance companies have even longer-term concerns. Many insurance companies, particularly large reinsurers, are now taking a holistic approach to climate change, addressing the physical, transition, and liability risks Mark Carney, chairman of the G20’s Financial Stability Board, described in a speech in 2015. Physical risks are the ones almost everyone thinks about: damage from extreme weather and climate. Liability risks are those that will stem from parties seeking compensation for damage or loss from climate change because of actions from parties they hold responsible (for example, the fossil fuel industry).
Transition risks will stem from uncertainties (e.g., in investment opportunities) in converting to a lower-carbon economy. Insurers are transitioning by not only divesting from but also not insuring the fossil fuel industry. Other actions the industry is taking include providing microinsurance to low-income entities at high risk from climate change, issuing catastrophe bonds for climate change–related perils, participating in various risk insurance facilities, and working with governments to enable their knowledge and expertise to manage risk to the fullest extent possible.
A report from the Geneva Association in early 2018 noted: “[The insurance industry] is neither the polluter nor the climate policy setter, but it plays a critical role in building socio-economic resilience and enabling economic development and entrepreneurial pathways for achieving climate change goals and targets.” The Industry is a critical part of the climate change solution, and the topic is clearly part of core business strategies.
Strength in Numbers
Regarding the physical risks, at the heart of the insurance industry concern is whether the tools (catastrophe models) that have been used reflect the current risk—namely, the influence of climate change. If companies have been making decisions using models that account for climate change, then they have less concern. If the models do not fully account for climate change but companies know the extent to which they do not, that helps them make informed decisions, too. What’s important is being able to refine model output selectively and knowledgeably. Toward that end, insurance companies have been engaging more with model providers to better understand what is and is not accounted for in models.
Traditionally, catastrophe models are developed with a long time-history of data that uses physically based algorithms to simulate events that have not yet happened, but could. One climate change–related concern is that there’s less relevant historical data for simulating what could happen. In fact, trends in the historical record, if not properly handled, can work against catastrophe model efficacy by diluting the current climate signal. It’s a fine line for a model developer to limit the historical data used for the model to better capture the mean current climate characteristics (e.g., frequency, intensity, etc.) but still use enough data to account for the possible variability.
Climate change has long been a challenging issue—with potential climatological, ecological, and economics responses, among others. But insurance needs to have a response, too. Therefore, insurers, policymakers, scientists, engineers, and, of course, catastrophe modelers (developers) must continue to work together in the coming decades to mitigate the ever-increasing risks from climate change.