Catching the Forward Wave on ESG Themes

As investors wrestle the challenges of incorporating key environmental, social and governance themes into their portfolios, data developments via machine learning and network data offer unique insights that can help identify potential opportunities and mitigate risks. At CFM, a quantitative and systematic asset manager, big data crunching and advanced modeling techniques provide a compelling path to understand and implement a sustainable approach to investments.
“When it comes to ESG themes, we look at phenomenon that haven’t really manifested themselves much in the past, but that we expect could manifest more in the future,” said Pierre Lenders, managing director and head of sustainability at CFM. “We know our entire production and consumption systems are in the middle of a massive overhaul towards becoming more sustainable, but it is hard to predict the next segment that will be disrupted or the next topic that will become financially material.”
CFM’s quantitative approach to sustainability investing is akin to a surfer watching the waves, then moving quickly to position her board on the most promising wave, Lenders said. The surfer needs to position herself before that wave reaches the shore, just as ESG investors need to position themselves for what’s coming next. Aiming for green-based financial performance, positions are built only when sustainability factors gain traction in specific sectors, he said. “Just as you only surf towards the beach, we never position ‘against’ sustainability.”

Language signals
Out of the full toolbox of quantitative techniques at its disposal, to identify investable sustainability-related themes, CFM increasingly uses natural-language processing, or NLP, a form of machine learning that uses multiple data sources and millions of texts to assess key changes in the investment ecosystem. “We track multiple ESG concerns and green technologies and identify those that are increasingly mentioned by financial sources,” Lenders said.
While NLP is not new, its deployment in finance so far has mostly consisted of tracking sentiment on individual stocks to create price predictability signals, he noted. But it could also be used to signal when a material ESG topic becomes financially material. “In a way, when it comes to sustainability investing, narratives drive prices more than fundamental factors, so NLP is even more applicable.” CFM uses NLP to track “indications that investors are catching up with non-governmental organizations or generalist commentators, that a wave is forming.”
Being talked about is not enough though. Another condition for a particular topic to gain financially materiality is that corresponding metrics indicative of specific corporate performance need to exist and need to be increasingly used by investors. “Simplistic modeling based on observed sector averages and implying footprints in proportion of revenues may be sufficient for reporting but won’t cause intra-sector reallocations,” Lenders explained.

Consider emerging factors
CFM’s ESG integration philosophy, “sustainability-aware investing,” is resolutely forward looking and focused on emerging financial materiality, with limited usage of ethical screening.
On climate, some companies have been moving faster than others to reduce their direct greenhouse gas emissions, and back-testing has suggested this has been reflected in stock performance over the last decade, Lenders said. But the same cannot be said about Scope 3 emissions, those related to a company’s upstream suppliers and downstream clients. “Quite recently, we have seen an alpha tilt though,” he said. While it’s not yet statistically significant, it “could indicate growing concern by investors that, even when a company doesn’t emit much carbon itself, it could get penalized in the future if its suppliers pay a heavier cost for emitting, or if its customers move to greener products. Scope 3 data is, still, only partially reflective of company-specific product mixes and value chains; but the urge to integrate this risk into investment decisions may nevertheless start impacting the price discovery mechanism,” he said.
CFM is also closely watching companies in the food sector that may be “more capable and faster than others at moving away from animal proteins into alternative proteins, or at promoting regenerative agriculture,” Lenders said. “This is a very material theme, with huge societal, biodiversity and climate impacts, but investors are not yet equipped with enough bottom-up data, so it is not yet financially material”, he noted, adding that “hopefully, the Sustainability Accounting Standards Board may soon suggest new metrics for the food sector, notably to assess innovation in alternative proteins, but regulators and data providers still need catching up on deforestation, pollution and soils degradation,” all quite challenging areas for data collection.

A part of the whole
Once CFM creates sustainability-related predictors, they become additional ingredients in its overall quantitative approach to security selection and portfolio construction, as is the case in its $8 billion flagship fund. “On a stand-alone basis, sustainability predictors may create exposure to temporary underperformance risk; when you combine them with other predictors, more short-term oriented and orthogonal to the transition, you can aim for enhanced Sharpe ratio potential,” Lenders said.
Another differentiating factor relates to portfolio concentration: originally, in green or clean tech funds, there was only room for discretionary managers, picking up pure-play stocks from a small subset of the universe, mostly small caps. “The transition is now a universal concern though, so we believe there is room for systematic and highly-diversified formats, relying on the law of large numbers and using NLP and networks data to identify how thousands of firms connect to financially-material ESG topics,” he said. “Positive overall performance may statistically emerge from enough of those different signals working on different time scales when applied in a large-enough universe.”

Double materiality
In the U.S., ESG’s integration is mostly about mitigating reputational and financial risks and detecting investment opportunities, Lenders noted. It is consistent with the Securities and Exchange Commission’s approach towards promoting climate risk disclosure as useful for investors. With the recently passed Inflation Reduction Act, green electricity production and transportation will be substantially incentivized.
In Europe, regulators use “sticks as well, not only carrots,” Lenders said. “Carbon prices are three times higher, and sales of internal combustion engines will stop before 2035. European regulators have also extended ESG disclosure requirements well beyond carbon, and asset owners are contending with the concept of double materiality” — risks faced by companies and companies’ impacts on outside stakeholders. There is a legitimate debate around whether an alpha-driven effort is compatible with this broader definition of sustainability “It is only when externalities are perceived to embark on a path towards becoming internalized by substantial regulatory action that the two dimensions coincide,” Lenders opined.
On climate, there is global recognition that “massive investment to reduce our dependency on fossil fuels is unavoidable,” he said. “It’s going to be messy, but it’s going to happen. Therefore, our investment programs should try to surf those waves.”

Tackling data quality
Data challenges persist. “We don’t need more aggregated scores and interpretations, but fact-based granular ESG data, similar to ingredients of traditional financial statements such as EBITDA,” Lenders said, referring to earnings before interest, taxes, depreciation and amortization. “We need mandatory disclosure to encompass all financially-material information, annual emission data included.”
Meanwhile, independently established, often closer to real-time ESG data has become increasingly available, just as there are more sector datasets that provide intelligence on meaningful dimensions, Lenders said. For example, “you’re starting to see, combining satellite data with sample data collected from the ground and AI, not just whether forest surfaces are being reduced, but also how actual biomass and biodiversity are evolving.”
As asset owners continue to focus on ESG investing, Lenders advocates they should not only demand ESG implementation from their asset managers but also lobby governments on implementing changes in the real economy to support progress towards a greener economy. “In the scaling up phase, government intervention is indispensable”, he noted, citing the example of Germany and other European countries that heavily subsidized solar power, which helped spark industry growth, adoption and going through grid parity. “For hydrogen it’s a bit of the same,” he noted. “Businesspeople and investors are, on average, not selfless enough to change behaviors unless properly incentivized, so regulations focusing on disclosure alone would be the equivalent of setting webcams on the Titanic.” Ultimately, if asset owners want ESG integration to be more than a fad and net zero targets to be met, they should ensure that changes in the real economy continue to happen at a pace allowing sustainable investing to remain financially rational.” ■

^{Any description or information involving investment process or allocations is provided for illustration purposes only. There can be no assurance that these statements are or will prove to be accurate or complete in any way. All figures are unaudited. This article does not constitute an offer or solicitation to subscribe for any security or interest.}