Carriers Take Cautious Climate Change Approach

It has been said that the only constant is change, and that certainly is the case with earth’s climate. The rancorous debate about whether human activity is abetting this change is largely immaterial to an insurer, which, as an assessor of risk, needs to account for the vagaries of climate such as floods, wild fires and hurricanes, regardless of the underlying cause.

While the political heat around the issue in the United States may mean that American insurers are less likely to trumpet their efforts to account for a changing climate, little such hesitance exists in Europe, where insurers such as Zurich Financial Services Group, Munich Re Group and Allianz Group have well-publicized climate change initiatives underway. “Munich Re and its experts have been drawing attention to man-made climate change and its effects since 1973,” says Torsten Jeworrek, a member of the company’s board. “In the long term, global warming will lead to a further increase in weather-related catastrophes, the financial impact of which will have to be borne by insurers and the public.”

This concern over climate change also is evident among underwriters. In its fourth annual underwriting survey, published in February, Lloyd’s of London found that 57% of underwriters believe more needs to be done to prepare for the impact of climate change.

THE BROAD APPROACH

Lindene Patton, senior VP and counsel for Zurich North America, says the company seeks to deal with climate change in a comprehensive manner, and is entrenching the tools and mechanisms needed to address climate change throughout Zurich’s operations. The most obvious manifestation of this is the Internal Climate Office, which Zurich established and housed within the office of its global chief underwriter. “The concept is to make issues associated with climate change fully embedded in our underwriting infrastructure so people are clear that we have a market-oriented approach,” Patton says. “Our intention is to make this part of our business rather than deal with it as a separate activity.”

In order to address issues such as tropical storms arising out climate change, one tool Zurich employs across the enterprise is monitoring. “In the natural catastrophe area, we have historically used sophisticated monitoring techniques to model all of our exposures,” Patton says, adding the company uses the technology not just for underwriting, but also for claims management.

For example, monitoring enables the company to precisely and safely preposition resources (everything from adjusters to generators) once the track of a storm has been determined. “We’re using this technology not just to improve pricing, but also performance.”

In addition to new monitoring techniques, the other primary area where Zurich employs technology is in catastrophe modeling. Like most in the industry, the company uses models compiled with the help of outside risk modeling firms.

“We take those models and augment them with our proprietary information,” Patton says. “Our expectations are that as we, and scientists, learn more about climate change, these models will get more sophisticated.”

TERRA INCOGNITA

The onus of improving catastrophe models falls largely on the shoulders of people such as Dr. Peter Dailey, director of atmospheric science for Boston-based AIR Worldwide Corp., which develops risk models for insurance companies. Dailey says AIR develops its catastrophe models based largely on historical data. What vexes him and other climatologists is that climate change is an unknown, and there are no records to refer back to when compiling new models. “Our models really depend on the historical record in order to formulate the model and understand the risk,” he says. “Without the data, it’s virtually impossible to start on the problem.”

Dailey is also quick to delineate between forecasts and the long-term models, which are often cited when discussing the ultimate implications of a warming planet 50 or 100 years from now. “Those are all simulations—models that project climates in 100 years and simulate a world that hasn’t been historically observed. This adds a lot of uncertainty to the output,” he says. “Whereas, a weather forecast model can be calibrated on history.”

One area where an accurate and long data set is available is the record of sea surface temperatures (SSTs). While it is commonly accepted that warmer oceans will beget more tropical storms, this also is regarded as something of an oversimplification. Dailey notes that a unique combination of atmospheric and temperature conditions are necessary for hurricanes to be created and, consequently, the storms are relatively rare and susceptible to decay. Yet, for now, SSTs are the metric of choice. “The main driver for tropical storm activity is the warmth of the ocean,” he says. “There are many other factors, but the most highly correlated is SST.”

However, SSTs may not be best the metric for measuring tropical storm creation for much longer. In the last decade, due to advances in satellite imaging, scientists have been able to ascertain ocean heat content, which is a measurement of how deep SSTs persist, and is considered a more precise indicator for storm creation. However, climatologists need a longer record of ocean heat content data before they can put it to practical use. “Ten or 20 years of ocean heat content data is not sufficient to develop a full climatology, so we have to revert to SSTs,” Dailey says.

It is this insufficient data, coupled with the tendency of models to accumulate errors as they age, that led AIR to eschew long-range forecasts in favor conditioned climatology, Dailey says. Instead of attempting to predict specific outcomes in the future, conditioned climate projections give insurers guidance based on what similar climate conditions, such as a cold SST season or a warm SST season, have meant in the past. “What we’ve determined in trying to forecast the next five years is that there is not sufficient skill in these models in forecasting SSTs beyond a year. That’s why we’ve focused our research on conditioned climate,” he says. “The bottom line is that, even in a five-to-10-year time frame, climate model SST predictions are not all that useful. They are helpful as a sort of guidance, but not as helpful in prediction.”

NEW THEORIES, NEW CHALLENGES

It is not only new technologies and metrics altering the understanding of climate, but new theories too. Dailey mentions a study published in the Dec.13, 2007 issue of the journal Nature that is causing some to reconsider the link between rising SSTs and an increased instance of landfall from Atlantic hurricanes. The paper, by Gabriel Vecchi of the Geophysical Fluid Dynamics Laboratory of the National Oceanic and Atmospheric Administration (NOAA) and Brian Soden of the Rosenstiel School for Marine and Atmospheric Science at the University of Miami, posits that the effects of increasing SSTs on tropical storm creation are mitigated by other atmospheric conditions, such as increased wind sheer high in the atmosphere.

“I’m not saying I agree or disagree (with Vecchi and Soden), but as scientists look into this problem more, they end up answering some questions, but raise many more,” Dailey says. “To make the assumption that storms will behave in the way they do today, even though the atmosphere is warming, is an assumption that requires more research.” Indeed, this complex competition between climate factors demands that both scientists and modelers need to avoid simple assumptions and approach the issue scientifically, Dailey says. What’s more, modelers have the added challenge of synthesizing the bountiful, rapidly changing and, often contested pieces of research emerging from government and academia into tools their clients can use. “Ultimately, it’s about how we use this complex information from science and translate it into a real application,” Dailey says. “That requires a thorough understanding of the data, statistics and science and persistence to learn more.”

For insurers, one problem this embarrassment of modeling riches presents is how to choose from the various models at their disposal. In instances where models vary significantly, finding a middle ground is especially difficult. Simply accepting the mean or average of differing models may not be indicative of the range of possibilities.

Likewise, Tom Hettinger, managing director of San Diego-based actuarial consultancy EMB America LLC, says the challenge for insurers concerning climate change has less to do with employing new models than reconsidering the parameters of their existing ones. “Explicitly, nobody has come out with a new type of cat modeling just because of global warming, but they are adjusting parameters for severity of loss,” he says. “If your parameters are not right, then your model is going to look like it’s wrong.”

FIRES AND FLOODS

The importance of the nexus between research science and industry is also evident when addressing issues on land, such as floods and wild fires. Dr. Howard Botts, a professor of geography at the University of Wisconsin at Whitewater and also VP and director of database development at First American Proxix Solutions. The Austin, Texas-based spatial solutions division provides location intelligence, personifies the interrelation between industry and academia. Botts says insurers are increasingly turning to location intelligence solutions to help get a better sense of their exposures, and to write policies that have a lower potential for loss.

One area in which location intelligence is at a premium is for insurers attempting to understand the potential range of wild fires. A factor greatly impacting this is the northward march of invasive species such as the bark beetle, which historically couldn’t survive cold winters, but is now flourishing in places such as Colorado—to disastrous effect. “The impact of climate change is evident in the changing patterns of bark beetle outbreaks in the western United States,” notes the U.S. Department of Agriculture in a November 2007 briefing paper. “Several outbreaks are occurring outside historic pest ranges and at intensities not previously observed.”

Botts adds that vegetation that’s under drought stress is more susceptible to infestations by the bark beetle. “It’s a worse case scenario because now you have tall, dead trees that, in a fire, burn rapidly, when the dead crown catches fire, the fire can spread quickly,” he says.

Dr. Tom Jeffery, also a professor in geography at Whitewater, and chief hazard scientist at First American, says advances in satellite technology, which allow areas of land to be imaged in different wavelengths such as the near-infrared, are paying big dividends, especially when assessing the health of vegetation and its propensity to catch fire.

“The big benefit working in this area now as opposed to 15, or even five years ago, is that we have so much imaging technology available,” Jeffery says. “We can actually evaluate the health of vegetation that has been affected by the bark beetle or disease long before the tree turns brown because internally there are changes that are reflected in different wavelengths.”

These satellite imaging advances coupled with improvement in 3-D visualization are giving insurers better elevation data, with hundreds of elevation points now plotted per acre. This is important in coastal area prone to flooding. “By combining the models that predict how big an envelope of water is going to be pushed onshore with granular levels of elevation data, you can build precise models that can identify the homes at risk for coastal surge,” Botts says, adding that in the post-Katrina world, underwriting and actuarial departments need a more refined understanding of their environmental risks. “Before, it was enough to understand what the risk was at a county or zip code level. Now companies are trying to understand it at a property-by-property basis.”

New Technologies Present ... Challenges and Opportunities

One way climate change impacts insurers is by altering the business models of the companies they cover. If climate change recasts the way we power and transport ourselves, insurers will have to adapt.”Our risk profiles are reflective of our customers,” says Lindene Patton senior VP and counsel for Zurich NA. “Our techniques and technologies will change as the underlying business changes.”

Nowhere is this more evident than in the energy sector where, as greener technologies emerge, insurers are forced to reconsider how they underwrite the business. From a risk management perspective, alternative energy technologies and local power sourcing present challenges. For example, the risk profile of a wind farm or solar thermal plant will differ greatly from a traditional, coal-based power plant. Likewise, underwriters need to know how the risk of fire at an ethanol plant differs from the risk at an oil refinery. Additionally, due to legislative pressures such as the Climate Security Act of 2007, which, if enacted, would prompt U.S. businesses to invest in cleaner technology and greater energy efficiency, insurers will have more opportunities to offer liability, E&O and D&O insurance coverage.

“There are very unique challenges on the property and liability side that present themselves whenever you have a new technology.” Patton says, noting in the energy sector, especially where the variables are changing rapidly, technology is crucial to properly assess risk. “There are unique risks presented by both the engineering technology and the information technology. We have to use IT and basic knowledge gathering to figure out what their exact capacity is in case a natural catastrophe or other event takes one of these alternative energy sources offline.”

The disruptive impact of new technologies is hardly limited to the energy sector. Another area where the move to go green is felt is in the construction business, where Zurich covers many green builders.

What’s more, insurers are finding many strategies that mitigate risk also have ancillary green benefits. One possible instance of this is the installation of global positioning systems in trucking fleets. By helping drivers find safer and more efficient routes, the monitoring via GPS can reduce both accidents and emissions.

Patton says the key for insurers contemplating their approach to climate change is linking different areas of knowledge. “We don’t see any single technology — whether its IT or engineering technology — being the single insurance solution,” Patton says. “Just like there’s no single solution to climate change, there’s no single solution for insurers to manage climate risk. That’s why we’re looking at a multi-disciplinary, completely embedded approach that highly dependent on technology.”

For reprint and licensing requests for this article, click here.
Security risk Policy adminstration Data security
MORE FROM DIGITAL INSURANCE