In this era of inescapable climate dangers, from wildfires to floods and extreme heat, businesses need to know how vulnerable they are to global warming.
To meet this demand, new firms have emerged offering to analyze “financial climate risk.” The idea is to show where climate-related hazards might hurt a company’s bottom line, or a fund’s returns, often using exact coordinates to pinpoint which assets are exposed. Sometimes, these companies use the climate models that academics have spent years developing to understand how we are heating our planet.
The scientists who create these models and work closely with them say they aren’t designed to provide such specific information. Moreover, the lack of transparency in the nascent climate risk analysis business means it’s often impossible for even the developers of these models to know whether they’re being used correctly.
This dismay from experts is the topic of the paper I co-wrote — with academics from climate science and accounting — that was published in Nature Climate Change earlier this week.
The paper focuses on a situation that is, in a way, a triumph for climate modellers. The corporate world has developed great faith in their models, which have performed wonderfully at predicting how annual average global temperatures would change over decades in response to burning fossil fuels and other human activities.
Executives assessing “physical climate risk,” however, don’t really care about climate dynamics out to 2100. They want to know whether the assets they own might be affected in the next few years. Are they exposed to wildfires? Droughts? Hailstorms? And where, exactly?
The reason climate models can’t give them an easy answer is hard to describe in a 5,000-word peer reviewed paper by six authors, let alone an 1,000-word column. But it comes down, at least in part, to this: weather events are relatively small, local phenomena.
Extreme events — the ones we care about most — are by nature rare and thus harder to get a good sample size from historical data. Another challenge is accounting for the many intricate features of the globe: landmasses, ice sheets and mountains, all of which affect local weather patterns but make the models – some of which take months on a supercomputer to produce a single simulation – more cumbersome.
Then there’s “climate variability.” These are natural climate patterns that are longer than weather events, making it harder to detect the “fingerprint” of global warming. These variations are one reason that climate model projections are very robust for the year 2070 or 2100, but not so much for 2025 or 2030 – which is the outer limit for many investors and businesses.
On climate scales, hurricanes are tiny, ephemeral things even though they wreak considerable humanitarian and financial devastation. This is why, even though physics tells us that warmer air and oceans will likely make them worse, United Nations reports will use phrases like “medium confidence” when discussing the possibility of increasingly intense storms.
Most companies that offer climate risk analytics aren’t claiming that they’ll solve all your hazard problems. But it’s a hot market right now. Companies are scrambling to report their physical climate risk, in part because initiatives like the Taskforce on Climate-related Financial Disclosures set expectations around that metric. Regulators are also demanding more information. The Bank of England has alluded to wanting a “high degree of geographic granularity” when it stress tests financial institutions later this year, while U.S. Federal Reserve officials have talked about the need for geospatial locations of assets in relation to climate risk.
On climate scales, hurricanes are tiny even though they can wreak humanitarian and financial devastation.
Source: NOAA
The allure of combining climate science with other technologies like big data, artificial intelligence and satellite imaging is compelling. Venture capitalists are already backing several startups that do so.
For scientists who spent their careers trying to develop more robust projections, it’s frustrating to see their efforts misconstrued and even misused at times. Instead of “climate risk analysts,” our paper proposes “climate translators.” Think of them as something similar to professional weather forecasters who can translate numerical weather predictions into useful information for non-experts.
Some existing climate risk analysts say that’s exactly what they’re doing, but client expectations can be unrealistic.
Karl Mallon started Climate Risk Pty Ltd in Australia in 2006 and is now the chief technology office at physical risk specialists XDI. I met him seven years ago, when the phrase “climate change” was under attack from some of the country’s state and federal government agencies. Today, his services are more in demand than ever.
One of the biggest challenges, he says, is hurried clients who are unable to differentiate between the very different levels of analytic sophistication. “We’ve been offered project fees that don’t even cover the cost of firing up the cloud servers, but someone did the job. And that becomes their disclosure,” says Mallon.
Top-level analysis costs money, and the companies that can do it the best aren’t going to work for free.
“The private sector isn’t there to produce public goods,” says Bob Kopp, a professor at Rutgers University and a co-director of the Climate Impact Lab — part of a research collaboration with Rhodium Group. (Rhodium, an independent research firm, has used approaches developed by the Lab to provide climate risk analytics to partners including BlackRock Inc.) “But everyone in our society is impacted by climate change, and often times the poor more so than the rich.”
Instead, Kopp proposes that the U.S. use a model based on land grant-funded universities that would support climate translation at a local level, similar to the way many U.S. state universities collaborated with the agriculture and manufacturing industries in the early 20th century.
What might look like the growing pains of a niche financial analysis industry, or a rarefied debate over complicated models, reflects the devastating failure of governments around the world who have not taken the climate crisis anywhere near as seriously as warranted. That means not just cutting emissions rapidly, but helping prepare people for changes in the climate, even at the current levels of warming. We should already have legions of climate translators to augment the public good that is climate models; in their absence it was inevitable that businesses would try to take matters into their own hands.
Kate Mackenzie writes the Stranded Assets column for Bloomberg Green. She advises organizations working to limit climate change to the Paris Agreement goals. Follow her on Twitter: @kmac. This column does not necessarily reflect the opinion of Bloomberg LP and its owners.
Like getting the Green Daily newsletter? Subscribe to Bloomberg.com for unlimited access to breaking news on climate and energy, data-driven reporting and graphics, Bloomberg Green magazine and more.