Traders, data scientists and vendors explain how they are incorporating climate risk into their portfolios.
Hurricane season in the Atlantic Ocean typically lasts from June to November, with the brunt usually hitting in the months of August and September. In recorded history, there have been a total of 33 Category 5 hurricanes, which are storms that have sustained wind speeds of over 157 miles per hour, or 136 knots. Between 1924 and 2002, there were 22 Category 5 hurricanes recorded in the Atlantic. From 2003 through 2017 there were 11, with three coming in the last two years. In addition, since 2001, there have been 24 Category 4 hurricanes (storms with wind speeds of 130 to 156 mph)—compared to a total of 47 over the previous 50 years. Whatever your political beliefs are relating to climate change, it’s a proven fact that catastrophic storms have become bigger, more frequent, and more costly in the 21st century.
Led by Hurricanes Harvey, Irma and Maria, the 2017 hurricane season was the most costly in the history of the US, with damages exceeding more than $200 billion, besting 2005—which included the single most costly storm in history, Hurricane Katrina—which, all told, cost $159 billion. And these calculations do not take into account the lives lost as a result of these storms.
Founded in 2012 after Hurricane Sandy crippled islands in the Caribbean and parts of New Jersey and New York, Four Twenty Seven—based in Berkeley, Calif., and taking its name from the state’s goal of lowering its greenhouse gas emissions to 427 million metric tons of carbon dioxide by 2020—is one of a growing number of data vendors looking to provide more concrete metrics around the effects of climate change and, specifically, on physical climate risk. Four Twenty Seven’s flagship product, Equity Risk Score, takes in an ensemble of climate models that are available via open-source tools and various other scientific agencies to produce forward-looking predictions for data points such as average precipitation, extreme precipitation or heat, sea-level rise, and hurricane, typhoon and tornado activity.
From there, it has more than one million facilities mapped around the world and provides an exposure score for a facility exposed to climate risk, based on its location and industry sector. A portfolio manager or trader would then use that information to determine if they have outsized risk, whether it makes sense to trim positions that are disadvantageously positioned to climate-change events going forward, or add positions where they feel like there is relatively less risk either in total space or relative to their benchmark, says Frank Freitas, chief development officer at Four Twenty Seven.
“The market is maturing to the point where it makes sense to focus a lot of energy on it. You’ve got Article 173 in France that mandates reporting of climate risk, you’ve got the Financial Stability Board’s task force for climate-related financial disclosure that’s very prescriptive about telling people that they need to be aware of, and reporting on, their climate risk exposure from both a transition risk perspective and a physical risk perspective,” Freitas says. “So there’s a lot of demand for this information because regulators are telling people that they need to prepare to be able to discuss meaningfully what their exposure is, but also because as climate change intensifies, and as you go from one [Hurricane] Sandy every 50 years to three Hurricanes in the Gulf [of Mexico] that cost over $1 billion in damages in a year, people are beginning to realize that they really need to account for this.”
The climate change data space is growing. In addition to FourTwentySeven and other specialists like TruValue Labs and RepRisk, data giants Bloomberg, Thomson Reuters and MSCI are also active in this environment, as well as Sustainalytics, which is 40-percent owned by Morningstar, and Trucost, which is part of S&P Dow Jones Indices.
The industry is expanding rapidly because investors in Europe, Canada, Australia, New Zealand, Japan, and the US are starting to ask more and more questions about climate risk and how it’s being incorporated into the investment process, says Rob Furdak, chief investment officer of Man Numeric, the quantitative equity management arm of Man Group. “Over the last three to four years, things have been ramping up in terms of investor interest on all ESG metrics, but particularly climate change,” he says. “If it’s on the top of asset owners’ minds, it’s on the top of investment managers’ minds.”
Kathryn McDonald, head of sustainable investing at Axa Investment Management’s Axa Rosenberg Equities unit, sees the same interest, and says that while ESG data has been around for decades, the climate risk space is growing in importance.
“As investors, and especially as quants, we’re on the hook to really be forward-looking and thinking about how might the future look different than in the past,” she says. “And we see climate change—and its ancillary impact on production, on human capital, on even big things like migration or economic participation—as being much more important going forward than it has been in the past.”
Climate change data falls under the “E” sector of environmental, social and governance (ESG) data. Over the past 15 years or so, investors have increasingly incorporated ESG information into their decision-making processes as numerous scientific reports have shown that firms with superior performance on material sustainability issues outperform firms with inferior performance on these issues.
For instance, a 2015 report by the University of Hamburg, in conjunction with Deutsche Asset and Wealth Management, looked at about 2,200 individual studies to examine the entirety of research directed toward the field of ESG and financial performance. The report found that roughly 90 percent of studies find a “nonnegative ESG-CFP (corporate financial performance) relation,” and that “the large majority of studies report positive findings.” In total, 47.9 percent of vote-counted studies and 62.6 percent of meta-analyses studies found a positive relation between ESG and CFP, while only 6.9 percent and 8 percent, respectively, found negative results, with the remainder falling between neutral and mixed results.
Still, most of the outperformance gains have been found in the “governance” arena, followed by “social,” with “environmental” traditionally pulling up the rear. ESG is, after all, a wide-ranging field, and the incorporation of physical risk factors stemming from climate change is still in its early days.
Craig Davies, head of climate resilience investments for the European Bank of Reconstruction and Development (EBRD), blames a lack of standardization for the slow rollout in this space. In May, the EBRD, along with the Task Force on Climate-related Financial Disclosure (TCFD)—a committee set up by the Group of 20 (G20) nations—released a set of recommendations for companies reporting on physical climate risks and opportunities.
Davies says that what became “very obvious” as EBRD put together the 50-plus-page report was that carbon transition, as a topic, is much better understood by the markets because a whole system of metrics exists that can be used to quantify and assess those risks, and the same system and metrics can be used to quantify the improvements that are being made—or not made—to manage those risks, such as for greenhouse gas accounting.
“They are very widely disbursed across the market, they’re very well understood by the market, and everyone is speaking the same language when it comes to measuring greenhouse gas emissions or reductions in greenhouse gas emissions. So whether you’re a bank, a business, or a regulator, you’re all talking about the same thing and we have a comparable and widely-understood system of methodologies and metrics behind carbon transition risk,” he says. “We don’t have anything like that for physical climate risk. As we see it, that’s a barrier on market action. That’s one of the reasons we’re not seeing the same kind of widespread market action on understanding and tackling physical climate risks as compared to carbon transition risk.”
One surprise, Davies says, is that the market doesn’t quite understand that while there are clear and present risks related to climate change, it also presents opportunities. For example, he notes that farmers in the Chiltern Hills, about 35 miles west of central London, are starting to think about growing new crops and, specifically, stepping up cultivation of Chardonnay and Pinot Noir grapes in the assumption that within 50 years, the region will have a similar climate to that of central France today—ideal for making Champagne.
“Climate change is overall a bad thing, but we as businesses need to think about what the opportunities are so we can be more mobile than our competitors are; so that we can we be more resilient,” Davies says. “I was quite surprised at how difficult it was for the firms and financial institutions that we engage with to understand how physical climate can present opportunities.”
As a result of increasing numbers of players in the climate change space as more vendors look to capitalize on investor interest, firms oftentimes have to do their best to blend this information to obtain a useful signal.
Furdak says Man Numeric uses data from Sustainalytics, MSCI and Trucost, among others, and the quants take that messy data, synthesize it, clean it, then automate the incorporation of that data into its system to see where the sources are consistent and agree, where they disagree, and how they disagree. Then the quants try to incorporate that blended look into the quantitative equity manager’s portfolio construction process.
So, for example, Man Numeric has a client with specific investment criteria that the amount of carbon emitted by companies in its portfolio must not exceed a certain level. Based on its holding in a stock, what percentage of the company it owns, and how many tons of carbon a company emits in total, the quants can imply how many tons of carbon the client’s overall portfolio emits based on those third-party data sources.
“So, if we have two stocks that we like equally based on our models and one is a big carbon emitter and one isn’t, we tend to favor the one that emits less carbon,” he says. “You can apply a similar approach in terms of gallons of water used, or similar metrics.”
The other way to incorporate climate data is to create more generic constraints by coming up with some combined measure of environmental scores from different data providers and then—say, if you’re managing a long-only fund on a benchmark-weighted basis—try to beat the benchmark score by 10 or 20 percent.
Furdak says one of the bigger challenges still facing the space is that there are two segments of data to manage—slow-moving, long-term data, such as that produced by MSCI, Sustainalytics and Trucost, where the frequency of updates is quarterly or annually and the data is fairly static over time; and then shorter-term data, such as that of RepRisk, Four Twenty Seven or TruValue Labs, which provide much shorter-term information with more-frequent updates.
“For us, it’s about trying to strike the right balance between the stability and constancy of the longer-term data and trying to marry it with a shorter-term signal that may be more dynamic to a specific event, such as a Facebook data breach or a big chemical spill, which will take a while to filter into Sustainalytics, where a TruValue or other shorter-term signal will pick it up the next day. But one also shouldn’t want to be trading on every little story and rumor, because that will incur a lot of turnover,” he says. “You need to be able to strike the right balance between being reactive to what’s going on in the world, but not so reactive that you’re spinning your portfolio based on things that aren’t relevant.”
In the carbon transition space, firms are able to take in public sources of information to create their own scores.
Last year, UBS Asset Management launched the Life Climate Aware World Equity Fund, which attempts to push CO2 benchmarking a step further by accounting for the future risk of climate change over the next 10, 15, 20 or more years. For it, UBS created a “glide path probability score” that looks at the trajectory of CO2 emissions reductions over time and compares that to the International Energy Association estimates of how much energy is used, and how much CO2 needs to be reduced per industry and sector to be on a glide path of a two-degree carbon reduction scenario, says Chris Greenwald, head of sustainable and impact investing research at UBS Asset Management.
“That’s a kind of proprietary scoring with a future-oriented approach,” he says. “It takes a bit more work than just looking at the current CO2 footprint in a portfolio. So you have to be a bit more analytical to account for future climate change risk.”
The fund is built around greenhouse gas emissions, and considers Scope 1 emissions (emissions that occur from pollution sources such as stationary combustion, mobile combustion or process emissions) and Scope 2 emissions (indirect emissions that occur through the use of purchased electricity, steam, heat, or cooling, according to the Environmental Protection Agency), and, to a far lesser extent, Scope 3 emissions (all other indirect emissions).
“Intensity levels, as well as the absolute levels, are the most important factors in reweighing the portfolio,” Greenwald says. “But, again, [we are] looking at change over time compared to a standard benchmark of where the industry needs to get to in a two-degree carbon reduction scenario. There are—for certain industries, like oil and gas and utilities—other factors that we’ll look into, like the energy mix, the percentage of energy coming from renewable sources versus fossil fuel sources, the exposure to fossil fuel reserves for oil and gas companies. There are some more granular datasets that we also look at for very carbon-intensive industries where we have additional information that can be used in the scoring process.”
Francis Condon, a sustainable and impact investing research analyst at UBS Asset Management, says he expects even greater use of climate change data in the future, when coupled with another growing sector of alternative data, that of geospatial data. This will be particularly useful in the physical climate risk space.
“It’s now possible for data providers to replicate the location of company assets. And they can place that on top of a climate-related grid or a water-stress grid and those sorts of things. That’s one of the more interesting areas of development in the field that we’re seeing,” Condon says.
Greenwald concurs, adding that while no vendor has “cracked” this space yet, “Using alternate data sources from a variety of different areas to account for physical climate change risk is something that is an emerging area that we’re quite interested in. And the strategy is meant to incorporate those new datasets over time. So as new datasets become available, we can incorporate them into the rules and how we rebalance the portfolios.”
In addition to geospatial information, another major factor in the proliferation of ESG datasets is improvements in—and access to—natural language processing (NLP) tools and other forms of artificial intelligence and (to a lesser, but growing, degree) machine learning.
Mauricio Bustos, a data architect at Axa Rosenberg Equities, says that not only is the number of data providers in this space growing quickly—the quality of their offerings is also “improving quickly.” Having data delivered in usable formats makes it easier for quants to incorporate that information into the investment process.
“When we talk about alternative data and, specifically the NLP techniques used, those are techniques that really only matured in the last few years. The availability of those techniques to be useful in our process only came around in the last couple of years,” Bustos says. “And so, quants can have this access to be able to piece out ESG data from things like 10-Ks and 10-Qs—it really is a very recent phenomenon.”
Michael Lewis, head of ESG thematic research at DWS (previously known as Deutsche Asset Management), says that while sustainability has been top-of-mind for the better part of two decades, the climate arena has picked up pace in part because of the Montreal Carbon Pledge of 2014 and the Paris Agreement of 2016. The former is a commitment by more than 150 investors worldwide to annually measure and publicly disclose their portfolios’ carbon footprint. The latter is a global effort to combat climate change with the aim of keeping global temperature rise in the 21st century to below 2 degrees Celsius above pre-industrial levels, and to pursue efforts to limit the temperature increase even further to 1.5 degrees Celsius, according to the United Nations Framework Convention on Climate Change (UNFCCC). Though the US has backed out of the agreement, 178 other parties (including nation states and supranational entities, such as the European Union) worldwide have ratified the agreement.
While DWS has a multitude of data providers, such as such as MSCI, Sustainalytics and RepRisk to assess ESG risks, it uses Four Twenty Seven and Trucost to better understand physical risk and transition risk as a result of climate change.
“The new innovation, I guess, is this physical climate risk mapping,” Lewis says. “We spoke to a number of clients and people and they’ve been trying to look at this topic for a while. So we were very fortunate to find Four Twenty Seven, really, because a lot of investors were interested and keen to get this sort of information. Four Twenty Seven has filled a very big gap, and we’re building on that data to embed physical climate resilience into investment portfolios.”
Lewis says that about 18 months ago, an insurance client came to the group. While the client was an expert on underwriting for catastrophic events and pricing climate risk and physical extreme weather events, it wasn’t as adept at doing that on the asset side. DWS’ job was to bridge that disconnect between the liability side of the insurance company and its investments.
This extends to oil and gas companies, as well. There have been advances in attribution science for measuring carbon emissions, and Lewis says that in the future, offending companies may well pay fines similar to those faced by tobacco companies in the US in the 1990s.
For a client looking to build client resilience into its portfolio, DWS starts by looking at a particular region, just to narrow down the risk involved. “From here, we can then start to solve the entire picture globally,” Lewis says.
“So, for example, we started by looking at an Asia-ex-Japan listed equity index, with Four Twenty Seven and Trucost data providing a physical climate and transition risk score to each company, respectively. We are then able to map those Asian companies in a quadrangle, or a matrix, with the top right being very resilient companies to both transition and physical risks, and the bottom left being less so. In terms of developing a passive index, we assigned a higher weight to the physical climate risk score compared to transition risk because the former is happening right now, and we don’t have much control over it, while the latter is more about risk over time, over which we have perhaps greater control.”
What you find, he says, is that you can’t use transition risk as a proxy for physical climate risk, as these data points are uncorrelated. So next, you look at the universe and start attaching weights to more resilient companies and, in a sense, overweighting the portfolio so you’re weighting towards more physical climate-resilient companies, and underweighting less climate-resilient companies. From there, you conduct your back-testing. The aim is to link extreme weather events and supply-side disruptions to financial market performance. So, for example, if a Korean manufacturer has the majority of its hard-drive production in facilities in a floodplain in Thailand, they are at risk of disruption even if the company, itself, is headquartered in a less susceptible location.
“So what your strategy is trying to do is to then take into account those operational and supply-chain risks and to minimize your exposure to companies that are going to suffer big draw-down events in the event of extreme weather events and the transition to a low-carbon economy. Consequently, this strategy is not to profit from extreme weather events. Rather, this is more of a risk management tool to build resilience into a portfolio so that when extreme weather events strike, your portfolio is more resilient to those events when they happen,” Lewis says.
While it’s still “early days,” the asset manager has been onboarding Four Twenty Seven’s dataset into its ESG Engine to improve its portfolio construction for the Asia ex-Japan index, and will use it to launch a family of passive indexes “during the remainder of the year,” which would then lead to spin-off indexes optimized for regional climate risk for listed equities.
“So at the moment, we’ve gone from Asia to the listed equity in Asia, then we’re building regionals of Europe, the US and global. And then we would move into fixed income. But that’s some way off,” Lewis says. “This way, though, we will have for a company [through our Aladdin system] information where any portfolio manager would be able to see the ESG risks—and specifically, climate-related risks.”
While US President Donald Trump’s decision to back the US out of the Paris Agreement came as a major blow to environmentalists, the whims of a government are unlikely to dissuade institutional investors from wanting climate risk incorporated into the portfolio construction process.
First, it’s a manager’s fiduciary responsibility to make sure a portfolio is resilient against catastrophic events. There’s a growing consensus that it would simply be irresponsible not to factor in climate effects. Second, while the US might not be as keen to tackle the effects of climate change under the current administration, Europe and Asia are. In addition to the Paris Agreement and the TCFD’s voluntary framework for disclosing climate-related impacts and opportunities, the European Commission (EC) recently took its first concrete actions to push the financial sector “to throw its full weight behind the fight against climate change.”
And in Asia, Japan’s Government Pension Investment Fund (GPIF), the world’s largest pension fund with about $1.5 trillion in assets, has said that its asset managers “should give careful consideration to ESG issues” to raise its allocation of socially-responsible investments from 3 percent in 2017 to 10 percent of its stock holdings, though it didn’t give a deadline. And in Australia, the Responsible Investment Association Australasia says socially responsible investment in the continent is up 9 percent from 2015, pulling in $622 billion in assets under management.
“Climate change is just undeniable now—you just can’t ignore it,” Davies says. “The market is really taking this issue seriously. When you get enough market action, it… means that you don’t have to rely on the government to give you the signals and instructions anymore.”
Bryan Cross, who heads UBS Asset Management's QED group, joins to discuss alternative data and AI.Subscribe to Weekly Wrap emails