Investors Gauge Future Climate Risks With Satellite Imaging

The world was watching end-of-days scenes: ­Firefighters in yellow jackets, blurry against a copper sky, battled to push back walls of flames. Veterinarians tended to badly burned koalas and kangaroos. Dazed survivors picked through the ruins of their torched houses.

Chris Kaminker was one of the many remote onlookers unnerved by the images of Australia’s most recent bushfires. In London, where he leads sustainable investment research and strategy at $65 billion Lombard Odier Investment Managers, Kaminker couldn’t escape the thought that he’d seen this apocalyptic vision before. Indeed he had—a 2008 report commissioned by the Australian government had predicted that by 2020 climate change would cause the country’s fire seasons to start earlier, end later, and be more intense.

“The physical reality leaping off of the pages of scientists’ reports that warned us decades ago was simply shocking to behold,” says Kaminker, a 37-year-old dual U.S.-French citizen. “What struck home was the realization that the scientists really did get this right, sadly.”

So why hadn’t financiers like Kaminker been better prepared for the extent of the damage to life and assets? Something must have been missing in the data. This is where Kaminker believed he could add some value.

An alum of Goldman Sachs Group Inc. and Société Générale SA, Kaminker ran sustainable finance research at SEB, the Swedish bank that structured the first green bond, before joining the asset management arm of Swiss private bank Lombard Odier last year. In a newly created role, he was responsible for developing ways to analyze how the warming planet—and increasing pressure to cut carbon emissions—will affect companies. The idea was, of course, to give the firm’s funds a unique investment advantage.

Kaminker made two big moves. First, he dedicated himself to building expertise in a field known as physical risk, which involves understanding and predicting the potential damage to assets and infrastructure from a changing climate and extreme events. The Central Banks and Supervisors Network for Greening the Financial System, an international group of institutions aiming to address global warming, said in June that as much as 25% of the world’s gross domestic product could be wiped out by losses from physical damage alone by 2100 if no further action is taken on climate change.

Second, he recruited Laura García Vélez from the World Wildlife Fund conservation group as a geospatial analyst, tasked with studying satellite images, geographic information system data, and historical cartographic records.

Vélez, 31, first learned about geospatial techniques while working for a utility in her hometown of Medellín, Colombia, where she assessed how suitable locations were for hydropower plants. Growing up in a country that’s both lush in rainforests and full of minerals and energy resources made Vélez see climate change as more than a scientific phenomenon or an investment strategy. “What tends to happen is that the places that have more biodiversity and are more important in terms of their ecosystems are also places which are very rich in natural resources. So it’s really difficult to just put in place those trade-offs if a country is developing and its economy depends substantially on commodities,” she says.

At WWF, Vélez was a key contributor to a joint report with Investec Asset Management (now called Ninety One) that explained how geospatial data and satellite imagery could detect warning signs for a sovereign debt portfolio by spotting potential environmental hazards and countries’ progress in managing them. In July, Ninety One, which manages £118 billion ($153 billion) in assets, launched a country risk index with WWF that builds on that research.

Vélez joined Kaminker in January, and they began studying the Australian fires. They were frustrated with the widely used climate models produced by the Coupled Model Intercomparison Project, or CMIP, which receives contributions from at least a thousand researchers and is cited in United Nations reports. They found the CMIP models failed to show that huge fires would ravage New South Wales, the southeastern state that’s home to Sydney, with such ferocity. They became convinced that geospatial analysis could do more than help calculate the damage from extreme events—it could help forecast where these calamities might occur.

It’s been almost 75 years since the first photo of Earth was taken from space. Insurers have used satellite data since the 1990s to model flood and hurricane risks; commodity traders and some hedge funds have used the images to track traffic at shopping malls and monitor usage of oil storage facilities.

The cost of launching a payload into space has fallen ­significantly—by a factor of 20 since the 2000s, according to one estimate—and now many more satellites are in orbit, pinging back an ever-larger volume of images. Virtually every building and tree on the planet is under daily surveillance. Computing breakthroughs, most notably developments in artificial intelligence, have made the tools to interpret and analyze the data much more readily available. And we’re just at the start of this new space race for financial data, says Rowan Douglas, head of Willis Towers Watson’s Climate and Resilience Hub. Douglas predicts that within five years spatial techniques will be deeply embedded in financial and risk analysis across the industry.

But there are limitations. The cost of purchasing data from satellite providers remains high, according to Kaminker. And, while satellite imagery of any point on the planet is available, there isn’t yet any corresponding or interlinked database showing who owns each piece of property. Investors may struggle to gain a full picture of a company’s impact on the environment or vice versa.

There are projects under way to remove some of these roadblocks. For instance, the U.K.-based Spatial Finance Initiative is working to make geospatial capabilities widely available for financial decision-making. It plans to introduce a database of the location and ownership of global cement and steel facilities by the end of 2020, says Ben Caldecott, a founder of the initiative and founding director of Oxford’s sustainable finance program.

Kaminker and Vélez developed an alternative to the CMIP models by compiling data on local temperature, rainfall, wind speeds, and humidity from meteorological agencies, adding their own analysis of images from satellites. After doing backtest trials, Kaminker and Vélez determined that the images could have predicted, with some degree of certainty, the location and severity of the Australian fires a month before they occurred. So Lombard Odier Investment Managers began making greater use of satellite data and Earth observation, and physical risk analysis has become one of its key inputs in investment decisions. Kaminker declined to provide details on the specifics of its current portfolio, but he said physical risk and geospatial data were material factors in decisions related to California utility PG&E, U.S. timberland company Weyerhaeuser, Brazilian iron ore giant Vale, and Brazil’s JBS, the world’s biggest meat company.

Once seen as a peripheral topic for money managers, climate change has become a material element in investment decisions. BlackRock Inc.’s Chief Executive Officer Larry Fink said in his annual letter to CEOs in January that “climate change has become a defining factor in companies’ long-term prospects” and would bring about a “fundamental reshaping of finance.” For most fund managers this manifests as a focus on incorporating environmental, social, and governance factors in their investment decisions—buying and selling stocks based on how well they score on factors such as carbon emissions. Now firms such as Lombard Odier are seeking an edge with a view from space. Instead of waiting for a company’s annual sustainability report, they’re using satellite images to get a real-time picture of its emissions. In this way they’re measuring and managing risks, but also arming themselves with more accurate data to push companies and governments for change.

In March, Lombard Odier created a climate transition fund. The strategy, with more than $500 million under management, invests in companies that will profit from the move to a lower-­carbon world—either by creating solutions to curb greenhouse gases or making aggressive efforts to cut their own emissions. The fund also includes companies building infrastructure for a warmer planet or those that monitor physical and financial risks related to climate damage. Its biggest holdings include engine maker Cummins, grocer Kroger, sports apparel manufacturer Nike, and carmaker Volkswagen, according to a Lombard Odier spokesman. The fund returned 13% in the three months to mid-October.

Meanwhile, Kaminker and his 10-person team are working on a host of additional tools to analyze the exposure and resilience of companies to climate change, including one that assesses the extent to which more than 23,000 companies are aligned with the temperature goals of the Paris climate agreement.

Kaminker says 10 years ago he’d have never imagined he’d be doing the work he’s doing today. “I’m a finance guy, I am not a climate scientist,” he says. “This is about risks and returns. We’re not doing this for many other reasons. It’s primarily because we’re an investment institution, and these issues could be material to financial considerations.”