“You can’t manage what you don’t measure”

Weatherman-turned-risk professional stresses the importance of data in managing climate risk

“You can’t manage what you don’t measure”

Risk Management News

By Gabriel Olano

Climate change remains as one of the top risks all organizations need to contend with due to its wide-reaching and financially costly effects. Scientists have long collected weather data and noticed long-term shifts, which indeed confirmed that the climate is changing and that humans have a huge influence in that.

Stephen Bennett (pictured), a former meteorologist and now chief product officer of climate risk advisory Demex, believes that data holds the key in making an organization more climate-resilient.

“First of all, you can’t manage what you don’t measure,” Bennett told Corporate Risk and Insurance. “Data and analytics drive our ability to measure the changing climate, which then informs management decisions. Data and analytics empower accounting for weather-related emergencies, crisis management, and operational planning. If organizations don’t have those plans in place now, they should take steps to plan as soon as possible.”

The number of extreme weather events in recent years has proved that being “weather-ready” is of utmost importance to organizational leaders, he added.

Bennett cited the example of extremely hot summer days in the New England region of the United States. While temperatures have been strongly increasing in a city like Boston, year-to-year variability has been strongly decreasing. Since 2009, the hottest summer days in the city have almost always been above 35 degrees Celsius.

“Reliably warmer temperatures year-after-year tells infrastructure managers they can expect strong return-on-investment for capital investment in infrastructure projects that mitigate the effects of heat on transportation and public health,” he said.

Meanwhile, in Chicago, its snowiest winter on record and one of the least snowy both occurred within the last 10 years. According to Bennett, the huge variability from year to year can cause dramatic cost fluctuations for property managers, who pay to maintain passable sidewalks and parking lots.

“These companies spend very little in years with only little snow but they pay a lot more during snowy winters,” he said. “The reverse is true from the companies that service those sidewalks and parking lots. They make big money in snowy winters and can have extreme losses in a winter like 2011 when it rarely snowed at all.”

Highly technological solutions that allow organizations to better assess weather patterns now exist, such as hyper-local weather data, aided by blockchain technology.

“The same technology that empowers hyper-local weather forecasts also leads to hyper local insurance applications,” Bennett said. “Suddenly an underwriter can assess the weather history directly at a property address as opposed to just the nearby airport or city-center. This innovation is driving a whole new scale for risk management.”

Meanwhile, he singled out blockchain as a vital prerequisite for climate-linked risk management, because it provides the foundation for “smart contracts” that provide total visibility to all elements of the transaction.

“Take, for example, a large property management firm paying for snow and ice removal at hundreds of different properties,” Bennett said. “They must develop a budget for these costs but how do they know how to set that budget months before winter arrives? Year after year, they have no idea how much it's going to snow. Risk management involves assessing and modeling the risk at each of the properties independently. Then, combine them together to create a risk transfer policy, either through insurance or derivatives. The policy provides for cost control. The policy kicks-in a payment for a very snowy winter and client doesn’t suffer cost overruns.”

Blockchain, which provides smart contract audit tracking, is important in letting every involved party, such as the property manager, the insurer/broker, and the plow contractor, can know and verify what happens after each storm, Bennett explained. 

“Data vendors provide the real-time and historic snowfall data and those tallies go onto the blockchain,” he added. “It’s visible to anybody who wants to see a storm-by-storm accumulation of snow. They also immediately see the changes and corrections that come from data vendors. Weather data is often subject to revision and correction over time. These revisions are very common but can lead to disagreement between parties to the risk transaction. The audit function in the smart contract establishes a fundamental source of secure information demonstrating how the contract is progressing through the risk period and which party is driving any changes. Transparency inspires trust.”

As a former weatherman, Bennett appreciates the impact technology has had on the field of weather forecasting. Nowadays, he said, a five-day weather forecast is as accurate as a one-day forecast was in 1980. Reasonably accurate and useful forecasts can now extend up to 15 days forward, as well.

Technology has also improved the predictions for many hazardous weather conditions such as hurricanes, flash floods and tornadoes.

“Much of this advancement is due to computing processor power doubling every 1.5 years,” Bennett said. “Today’s processor is roughly four times more powerful than one from three years ago and 16 times more powerful than one from six years ago.

“For example, in 2010, my company (EarthRisk Technologies) was developing a long-range weather forecasting model. We suddenly realized that it would take us a full day to run one of the necessary calculations using the single server that we owned. We needed thousands of these calculations. It was going to take us over three years to run the model! We couldn’t simply buy more computers. Thankfully Amazon Web Services (AWS) was just coming onto the scene.  We pivoted our entire approach to technology and built our model on AWS. It ran in a matter of days instead of years.” 

According to Bennett, there are many cloud computing solutions available today that can make weather models more accurate and faster to create. These same solutions are being used by Demex to bring together financial models and weather models to create tools such as the Demex Climate Indicator and provide what Bennett calls “high-resolution risk management.”     

Keep up with the latest news and events

Join our mailing list, it’s free!