As discussed in our recent note on the implications and investment opportunities of AI for the energy sector (see From Gridlock to Grid Smart: AI’s Role in Electricity Efficiency, April 4, 2024), AI has the potential to severely disrupt most of the themes that are dear to sustainability-minded investors. This week, we continue on the AI theme by focusing on other environmental subsectors – waste and water management – to identify the winners of the AI revolution in these two subsectors.
In order to find potential winners of the AI revolution, we looked for firms that have been discussing AI-related concepts in transcripts and disclosures as well as firms that have been aggressively investing through AI-related acquisitions, R&D budget, and Capex spending.
The most AI-vocal sector in transcripts and company disclosures is, to no one’s surprise, Technology. However, as the internet revolution in the early 2000s showed us, the next wave of AI-related return on investment might come from the application in other sectors of the economy, not the hardware itself. From that perspective, the industrials sector is punching above its class when it comes to the share of AI mentions relative to its weight in the global equity indices.
More specifically, AI mentions in transcripts and disclosures of companies that are part of our proprietary SMS Environmental Solutions universe (400 global companies) have been rapidly increasing over the past five years, demonstrating an interest in AI solutions for the environmental sector. Let’s dive a little deeper in a few of these sub-sectors.
How AI Is Sorting Out Waste Management
In a previous note (see Trash Talk: Turning Waste into Money Makers, February 15, 2024), we discussed how the waste management sector, both municipal and industrial, has been struggling to increase its recycling rate for technical and economic reasons. The reality is that traditional sorting methods rely heavily on manual labor, leading to challenges such as human error and higher processing costs. There are a few different ways AI and machine learning can support waste management companies.
For instance, AI can help waste management companies analyze data about the place and time of waste generation to optimize pick-up schedules. But the most promising and impactful application is likely to be in optimizing waste flow monitoring, where autonomous robots use visual sensors and advanced operational skills to sort and classify waste. Optimization can easily increase material recovery rates by up to 40%, therefore lowering labor and disposal costs by 40-60%.
All major waste management companies have started to mention their interest in AI-powered sorting robots, but two companies also have the investments and partnerships to back it up.
Republic Services (RSG US) has quietly started to integrate an AI-enabled operating system for material recycling facilities (MRFs) in some of its new plants. The RecycleOS technology is provided by EverestLabs, a privately owned start-up. It allowed them to reopen their “Next-Gen” recycling facility in Plano, Texas (after a fire destroyed the original one) with half the labor thanks to automation and optical sorters such as those from EverestLabs.
The second waste management company making serious investments in the AI space is Tomra Systems (TOM NO). The Norwegian reverse vending machine maker recently officialized an important investment in PolyPerception, a start-up offering AI-based waste flow monitoring, to help build more efficient collection and waste sortation infrastructure. This news was welcomed by market participants, bumping Tomra’s stock price by 30% on the day of the announcement (this coincided with earnings, but the acquisition was mentioned as a positive by many analysts).
AI's Dive into Water Management
Water management, from both quality and quantity angles, is already impacting businesses’ operations and citizens’ lives every day. As per a recently released report by the WWF, the looming water crisis endangers water resources and freshwater ecosystems valued at approximately $58trn annually (a whopping 60% of global GDP!). Even if this number is grossly exaggerated, the reality is that since 1970, the world has lost one third of its remaining wetlands, while freshwater wildlife populations have, on average, dropped by 83%.
Without being the silver bullet, AI applications have significant potential to revolutionize water purification and treatment systems. For instance, AI-powered sensors can continuously monitor water quality parameters like pH, turbidity, and contaminants in real time, enabling early detection of pollution events. AI can also optimize water treatment processes by analyzing complex data to determine the most efficient and cost-effective methods, reducing chemical and energy usage. Another important application is for leak detection, where AI can analyze distribution network data to detect and address leaks promptly, thereby conserving water resources.
Once again, two companies stand out as early leaders in the water management sector. Both have invested heavily and early in AI-related IP and, ceteris paribus, should greatly benefit from these investments in the future.
First, we have Veolia Environnement SA (VIE FP), active in water management (65% of their revenue) but also in waste management (35% of their revenue), which makes it even more relevant in the context of this note. Veolia claims to be the first company to actively use AI across all its divisions via its digital solutions “Hubgrade” which enables smart monitoring of the production and consumption of their clients’ water, energy, and waste. From an operational perspective, efficiency gains increased to €350m annually thanks to AI. With over 10,000 connected sites worldwide and 500 data scientists dedicated to its digital solutions, Veolia is definitely putting its money where its mouth is when it comes to AI.
The second candidate in the water management subsector is Ecolab Inc (ECL US). The water and sanitation firm was also an early mover with regards to digitalisation of assets. Their platform, called “ECOLAB3D”, gathers over 90 billion data points annually from over 100,000 globally connected assets like cooling towers, wastewater plants, and data centers. Their Digital Technology department has over 1,000 employees and is clearly one of their growing verticals. Another related area of growth for Ecolab is their water-based cooling solution for data centers, which has been growing at over 30% year over year. Ecolab therefore has two rather than one potential growth opportunity from the AI revolution.
The North American Electric Reliability Corporation (NERC) released their latest Long-Term Reliability Assessment in December 2023 which showed an increased risk of shortfalls in electricity under both extreme and normal conditions. Within a single year, the projected five-year annual growth rate for electricity consumption nearly doubled from 0.5% to 0.9%! This projection could be even higher as some utilities have revised their numbers upward since the release of the NERC forecast. There are several drivers behind this surge in demand – the reshoring of manufacturing, adoption of EVs and fossil fuel free heating systems – but perhaps the most noteworthy is AI.
The process of training new AI models is very energy intensive. AI dedicated servers can use 4 to 14 times the electricity of a normal rack of servers and AI queries (like ChatGPT) use 3-30 times more energy than a typical google search. Furthermore, generative AI not only needs to be trained on large datasets, but also creates loads of new data, thereby increasing the need for data centers. One study estimates the share of U.S. electricity demand for data centers will triple from 2.5% in 2022 to 7.5% by 2030. Clearly, a broader adoption of AI will be a substantial source of electricity demand.
This increased demand creates a challenging scenario for electricity grids. Already, the integration of renewables into the system has proved challenging. While it may seem faster to develop more traditional power plants to increase electricity production, this counters decarbonization goals. A better idea would be to improve the efficiency of existing grids instead. Ironically there is a role for AI in grid optimization, meaning AI can be leveraged to address the very problem it is helping to create.
The Ultimate Power Move: Efficiency
We see two areas where AI can serve to improve efficiency challenges. Firstly, continuous optimization of electricity consumption in industry and buildings. AI can help to unlock efficiency improvements in the design of individual electronic devices, reducing their electricity consumption. AI can also be applied to assess the energy management of larger industrial processes or commercial buildings. These make up 64.4% of U.S. electricity demand; with the rest of demand mostly coming from residential heating and cooling needs (see chart). Use of advanced data analytics in these areas can identify opportunities for energy savings, regulate power consumption and dynamically adjust needs in accordance with changes in electricity demand. This contributes to reduced energy waste and lower operational costs.
The second more direct approach is integration of AI into the electricity grid. This can be achieved using smart grid technologies that allow for real-time monitoring and control of grid operations. An example of grid optimization would be to use AI algorithms to dynamically adjust the distribution of electricity. So in times of low demand, surplus energy is directed to power-hungry processes or energy storage, and when demand peaks, the grid adapts in real-time to reduce strain and prioritize energy efficient consumption. Virtual power plants (VPPs) will be an important element to this. These systems aggregate small scale energy sources such as solar panels, wind turbines and energy storage enabling coordinated management of those resources to optimize energy supply. The U.S. Department of Energy has estimated that tripling the scale of virtual power plants (VPPs) could lower America’s grid costs by $10 billion a year.
Deploying smart meters and IoT devices embedded with AI into grid infrastructure is paramount to deploying these technologies. It facilitates the collection of real-time data which can then be processed by AI algorithms to optimize grid performance. This will be especially important as the share of intermittent energy sources like wind and solar increases.
Investment in digital grid infrastructure has grown significantly in recent years (see chart below). Countries are taking advantage of technological advances to modernize their systems. However, this level of investment will need to accelerate to keep pace with the newly anticipated surge in electricity demand.
Making Gains in Power Savings
In our note Finding Real Returns in Artificial Intelligence (April 27, 2023), we explored how AI is helping to improve industrial resource efficiency across impactful sectors, including energy. We expect continued sales growth for established players in industrial automation and IoT connectivity such as ABB, Honeywell and Siemens as their solutions support enhanced energy efficiency in several key areas: building energy management, industrial processes, and grid infrastructure.
Similarly, information technology companies like Oracle, Fujitsu and IBM are poised to benefit from the expansion of AI. Any innovations by these companies that improve the energy efficiency of IT infrastructure will help reduce strain on the grid and lower electricity costs for customers as prices are likely to rise given greater demand. In addition to higher sales associated with more data centers and servers, these technology companies will see increased revenues from products with direct applications in electricity grid optimization.
- Fujitsu is already leveraging AI to improve energy efficiency for communications networks.
- IBM’s expanding consulting services aim to support clients with digital transformation and AI deployment and include services specific to energy utilities.
- Oracle was identified as a leader for advanced electricity distribution management systems and has a variety of data collection and analytics tools for both energy and water utilities.
Companies with energy management as their main business segment offer the greatest impact with regards to sustainability. They are key in helping to lower emissions and integrate renewables in our electricity infrastructure. Investors should pay close attention to how these energy management companies are trying to leverage AI to optimize the grid.
There are two small cap names of interest: Itron and Landis+Gyr Group, with market caps of $4 billion USD and $2 billion USD respectively. Landis+Gyr is exclusively dedicated to smart grid. Itron earns 67% of revenues from smart grid products but also sells smart metering devices for water management, another growing sector. General Electric is also a leader in smart grid technology, earning nearly $3.5 billion in revenues from that segment. The former conglomerate completed the spin-off of its energy segment into GE Vernova earlier this week, this is a great opportunity for sustainability investors looking for exposure to innovative energy solutions.
Eaton and Schneider Electric are two other large companies with a strong focus on energy efficiency. These more specialized companies cover all bases, improving industrial companies’ consumption and the efficiency of data centers while also supporting utilities with the integration of digital technologies aimed at optimizing the grid. Both of these companies have slightly higher valuations (Fwd PE of 25.1 for Schneider and 30.5 for Eaton) than the information technology and industrial automation companies mentioned above (average fwd PE of 21.9 for ABB, Honeywell and Siemens and 19.7 for Fujitsu, IBM and Oracle), but this may be justified given the outlook for electricity demand.
R&D spend will be a critical factor in these companies’ ability to keep pace with AI developments. There is a notable increase in recent years relative to historical levels. One can also pay attention to the mentions of AI in earnings calls and business strategies. Though we recognize that some companies may be mentioning AI to take advantage of the current hype, it can serve as an indicator of how much they are leveraging AI capabilities. Investors just need to be diligent in their research to filter out any fluff. In Eaton’s documents, “AI” jumped from 6 mentions in 2022 to 182 in 2023 and is already at 48 for 2024. For Schneider, this year’s mentions of AI are already 5% higher than last year’s total; it is afterall one of the five key focus areas in their latest strategy.
Similar to other technology stocks, many of these companies are currently trading at elevated levels. Investors can watch for potential pullbacks to add exposure. Unlike some other AI-hyped names, we expect digitization of the grid to have staying power as society comes to terms with the significant increase in electricity demand.
The opening months of 2024 have brought four macro developments worth monitoring that will impact investment outcomes for sustainability portfolios for the remainder of the year. In this installment of our quarterly update on Macro Meets Sustainability, we address the following:
- Fed interest rate expectations: what comes next and what does it mean for interest-rate sensitive plays like renewable energy?
- Implications on sustainability-focused investment strategies should Trump win the U.S. elections in November;
- Electrification and the copper rally; and
- China dumping and what is next for the EU.
A Neutralized, Neutral Fed
Whenever big gaps form between the Fed’s expectations and market participants’ expectations of interest rate policy, there is opportunity (and risk) for investors. That is NOT the current environment. Recall that as recently as January, the bond market was predicting a less than 5% chance that the Fed policy rate would still be at 525 basis points at this point in the year, fully expecting that the Fed would cut interest rates up to six times in 2024. Expectations have now fallen to a rather benign three rate cuts, beginning in June or later, which is incidentally, what the FOMC also expects of themselves (as per their March summary of projections). Although we wouldn’t be surprised to see even fewer rate hikes this year (due to sticky inflation), a Fed on hold with market participants that agree makes for a reasonably docile rate backdrop.
This should be viewed as especially good news for the “pain trade” in the renewable energy sector. In this sector especially, monetary policy had become incredibly restrictive. As we wrote in a note last week (“A Second Wind: What Peak Interest Rates Mean for Renewables”), simply removing the rate headwind improves the prospects of many renewable energy projects, since most of the operational costs associated with running utility-scale wind and solar projects are derived from financing costs (rates). To be sure, without interest rates coming down, the operating costs of running renewable projects do not improve. But with credit spreads in this sector suggestive of a low risk of default, we believe that “bottom fishing” in renewable energy production – though not equipment manufacturers due to the risks related to China as described below – is timely given the incredible long-term prospects.
A Trump World?
Although there are still seven months before the next U.S. president will be decided, the enormous gap between the two parties’ platforms suggests that investors of all stripes should consider how their portfolio will stand up to a change of guard.
President Biden recently released his proposed 2025 budget: the trajectory is one of very tepid deficit reduction, albeit a hefty increase in corporate tax rates is included. In contrast, Trump has indicated that if elected he would push to extend for another decade all the individual tax cuts that are set to expire at the end of this year. On the face of it, a Trump government appears likely be more “inflationary” than a Biden government.
More notably, the “flavor” of fiscal policy in the U.S. will certainly change under Trump. In 2022, President Biden’s government signed into law the Inflation Reduction Act (IRA) as well as the Chips and Science Act. Both acts represent an Industrial Policy approach to fiscal expansion: they target specific industries that are in the national interest. We expect that industrial policy will increasingly become the norm across the globe (please see our note from July 21, 2021, “Macro Goes ESG”). But the problem is that Trump does not share the Biden administration’s view of which issues are in the national interest – ignorance about climate change being the most glaring one.
That said, Biden and Trump do agree that re-shoring manufacturing and protecting American jobs are priorities. In this sense, many “green industries” may not suffer much under a Trump government. After all, many climate-focused companies truly are helping to make America great again.
Our preliminary work suggests that from a sector perspective, outside of the fossil fuel industry and a possible knee-jerk election-day reaction, a changing of the guard may not be overly deleterious for sustainability portfolios; we intend to delve further into this topic in an upcoming research note.
Dr. Copper
Copper - the "metal of electrification" — is vital to all energy transition plans. Demand for the metal is expected to double between now and 2035, but the supply side is constrained on several fronts (underinvestment over the past decade, environmental backlash and bans, cost overruns, lower ore grades, etc).
The logical outcome of this demand/supply imbalance is a rise in copper prices. The breakout in copper above 4$/lb in March is just a small sign of things to come. However, before investors pile into the long-term copper trade, it’s worthwhile reviewing recent dynamics (after all, copper as part of the industrial metal complex, is subject to significant boom/bust cycles):
- Most of the demand for copper over the past year has been due to China’s massive ramp up of renewable energy production. Solar capacity increased by nearly 150% last year, from an ever-larger base. It is reasonable to expect that demand for copper from this sector will slow in the year ahead. After all, some of the largest solar producers in China, such as Longi, have admitted that supply glut has now formed in China and will force a significant slowdown in production.
- The more conventional source of demand for copper, China’s residential sector, is still in the doldrums and is unlikely to be revived anytime soon. Indeed, the Chinese government has shown little impetus to use previous tactics to re-ignite the construction sector.
- Part of the March rise in copper prices is due to a concerted effort among Chinese copper smelters to refine less copper at several loss-making plants. The implication of this move is that global copper inventories have begun building again and are no longer precariously close to all-time lows.
The bottom line is that we remain structural bulls on copper due to the enormous demands from electrification, however a more sustained rally and/or “copper mania” may only take place once some of China’s power glut is re-absorbed. Note that copper is a highly recyclable commodity; sustainability investors should first consider companies with strong potential in the recycling space. However, given the enormous amounts of copper required, high-quality miners with superior ESG standards can also be considered. Since most of the rest of the industrial metal complex does not share copper’s bullish dynamics, we suggest investors focus on miners that are producing exclusively (or nearly exclusively) on copper. Freeport-McMoRan (FCX US) and Ero Copper (ERO TSE) are two companies that fit this bill.
China Dumping: No Lift Off For Europe
As mentioned above, China currently appears stuck with overcapacity in solar panel production, but also a whole host of other industries from steel to electric vehicles. Indeed, China has (re)begun flooding its export markets with deeply price discounted goods to help support its domestic manufacturing sector.
But China’s goods dumping is perhaps nowhere more contentious than in the “green goods trade”. After all, cheap accessible green products should be a godsend to global decarbonisation. Unfortunately, both U.S. and EU efforts to address climate change have been wrapped around another goal: to reshore manufacturing capacity and create jobs. Thus, while China solar panel production (as well as wind power equipment, batteries and EVs) may be good for the global energy transition, they are problematic for competing nations’ goals.
For now, it appears that the EU is unlikely to impose new anti-dumping tariffs on Chinese solar panels or other renewable energy equipment, opting instead to find more ways to subsidize local production. We reiterate our view that the odds of other regions catching up to China are very low (see our note March 8, 2023 “China’s Solar Power and Deglobalization”). Investors keen to gain exposure to renewables should avoid solar equipment manufacturers and focus on renewable energy operators.
Today’s note looks at pricing discrepancies in equity vs. credit markets within our Energy Transition theme. While both ESG and sustainable investing had a hard time since 2022, within the Energy Transition theme, our renewable energy subtheme has suffered the most. Indeed, the combination of higher interest rates, ESG cultural wars and backlash against renewables in some jurisdictions (Texas, Alberta) have all contributed to the derating of transition-focused strategies. As a result, many companies in the space are now trading close to their 2017 values – while both the NASDAQ and the S&P 500 have doubled and continue to make new records.
On the face of it, many highly levered stocks in the energy transition space are facing costs associated with higher interest rates. This has affected profit margins and free-cash flow yields – which have deteriorated in the past three years as total debt to EBITDA ratios have doubled.
Conversely, the credit market points towards a different story. First, credit spreads over U.S. Treasuries for the global alternative energy sector – renewables and energy efficiency – have come back to their pre-COVID levels, a time when energy transition was one of the hottest investment themes.
Energy Transition: The Tide Will Turn Soon
In addition to the disconnect between equity and credit markets, we think that it is time for equity investors to move back into the renewable energy theme, essentially for two economic reasons:
- Global Demand for (clean) Energy: Can’t Stop Won’t Stop
Electricity demand is set to boom globally over the next few years. Increased demand coming from AI computing capacity, the growth in the number of data centers (see our note “Catching The Next Wave: Positioning For 2nd Derivative AI Plays in Sustainable Themes”, February 29), increasing demand for heat pumps, additional electrification of the grid, electrification of the automotive fleet, the reshoring and re-industrialisation western economies are indicative trends of a stronger demand for electric power – both in Europe and in the U.S.Furthermore, some studies suggest that in the U.S. alone, the additional electrical production capacity that needs to be built is equivalent to bringing all of California’s current production online within the next five years. In this context, electricity prices will remain elevated and so will demand for renewable energy suppliers to build the required. This should support utilities and equipment makers in the renewable space.
Source: World Bank via IEA
- The Pipeline for Green Projects is Full
Furthermore, the sheer amount of renewable energy projects set to come online globally will support renewable equipment makers. While several renewable projects were cancelled in the U.S. in recent months due to growing costs and soaring interest rates (see our note: “Finding a Floor in Renewables”, October 26, 2023), it still makes sense for these projects to come off the shelf now that interest rates have stabilized and will potentially decline later this year.
After all, the U.S. Energy Information Administration (EIA) forecasts that renewable energy production will soon match the capacity of natural gas – yet another indication that companies operating in renewable energy production, distribution and equipment making are set to benefit from a renewed second tailwind.

Finding Nuggets Among the Wreckage
To find gems within the renewable energy space, one should first look at companies that have lost their downside momentum. Interestingly, companies like Innergex and NextEra (producers and developers) have recently bounced off their recent 3-month low.
What’s more, the credit market is also suggesting that the default risk of these companies is relatively low. That said, one also sees that most credit spreads against their benchmarks remain substantially higher than the MSCI Global Green Bond Alternative Energy Index (which itself was of 72 bps as of March 20) – suggesting that the companies in our sample have also somewhat derated with credit investors compared to the overall sector.
On September 10, 2020, around the first time when an ounce of gold traded at 2,000, we wrote a note: “Finding a place for gold in an ESG portfolio” whereby we identified sustainable companies involved in gold recycling and mining equipment. While the price of gold has traded sideways for the last three years, with the recent gold breakout to new all-time highs, we are revisiting the topic.
Gold is a scarce commodity. A cube of gold
with an edge measuring only 22 m would contain all the gold that has ever been mined! Other than its distinctive color, gold has attractive chemical properties: it is inert (does not corrode or tarnish), highly malleable and ductile (it can be stretched very thin), has high conductivity (electricity and heat) and is nontoxic (compatible with humans in medical devices or dental work).
What Is Gold Used For?
Due to the above properties, gold has been used as a medium of exchange and store of value (currency), in art (jewelry, decoration, craftmanship) and in technology (telecom, electronics, aerospace, medical devices, dentistry).
Despite having other useful applications, the main sources of demand for gold are jewelry and finance.
The image below, from the World Gold Council, shows the interaction between annual supply and demand for gold over the last 10 years. Gold supply has come mainly from new mining (75%) with the remainder sourced from recycling (25%). It is worth noting that the average annual amount devoted to functional use in technology (8%: 329 tonnes) is lower than the supply from recycling (25%: 1,172 tonnes). In other words, the quantity of gold needed for functional applications could be entirely derived from recycling.

What Are the Factors Determining the Price of Gold?
Let’s look at the supply and demand picture:
Supply: there is not a lot of gold left to mine, according to the World Gold Council, approximately 80% of identified gold reserves have already been mined leaving just 20% of tougher to access reserves, where the concentration of gold is lower. Meanwhile, recycling could bring a lot of existing gold back towards more useful endeavours. The chart below shows the main cost differences between mining and recycling. In mining, costs are associated with exploration, geology, and mining methods. Recycling involves expenses related to collection, sorting, and processing. There are other economic factors that determine the marginal cost of gold like interest rates, currency and inflation. Even with the price of gold at its historical high, gold supply is expected to experience only modest growth over the next decade given declining ore grade, higher extraction costs and stricter environmental regulations.
Demand: Gold experiences demand from various sectors, with jewelry, industry, financial products, and central banks are the primary sources.
- Jewelry: Gold holds cultural significance and serves as a symbol of wealth and fashion.
- Industry: In electronics and renewable energy, gold is prized for its conductivity and corrosion resistance. Medical devices like pacemakers, implants and electrodes utilize gold for its biocompatibility and inertness. Aerospace applications such as satellites benefit from gold's ability to reflect infrared radiation and protect from heat. Gold nanoparticles are used in chemical processes, while the telecom and auto industries also rely on gold for various applications.
- Financial Products: Gold is often sought after as an investment safe haven during times of economic uncertainty and also serves as a hedge against inflation. It is also considered as a form of currency, functioning as a medium of exchange, store of value, and unit of fungibility. Exchange-traded funds (ETFs), futures contracts, certificates, structured products, and gold bars offer investors ways to access the gold market.
- Central Banks: Significant amounts of gold are held by central banks as part of their reserves. They view it as a means of portfolio diversification that provides stability amidst economic fluctuations. Gold acts as a store of value, a reference point for currency valuation, and a hedge against inflation and currency devaluation.
While new supply of gold is expected to be limited, demand could increase at a faster pace from greater jewelry demand and increased industrial use. Financial product demand is a wild card; the advent of Bitcoin ETFs has taken financial flows that would normally have gone to gold (even though the gold price increased in 2023, physical gold ETF saw outflows). The most promising source of demand lies in central bank reserves that are reacting to the confiscation of Russian assets. The chart below highlights that for central banks surveyed, they expect the portion of gold in their reserves to increase (63% expect it will increase compared to 9% that believe it will decrease from the current 15%).
How to Gain Exposure to Gold?
Since the price of gold broke out to all-time highs, should a sustainable investor gain exposure through gold miners, gold royalties, gold recyclers or some form of gold bullion ownership?
Gold Miners:
The practices of gold miners have improved significantly over the years, especially for the larger firms, but many issues still exist. From an environmental standpoint, deforestation, water and air pollution, soil degradation and energy needs are all major concerns. From a social standpoint, human rights violations, health and safety, and other issues such as worker displacement, labor exploitation, and cultural disruption are worrisome. Finally, in terms of governance, challenges include corruption, lack of transparency, weak regulations, and the abuse of land and indigenous rights.
For all of these reasons, it is difficult to find long investment candidates in the mining sector from a sustainability perspective. We find it much easier to find companies to potentially short sell. Gold miners are levered to the price of gold because they have fixed costs and benefit disproportionately from higher prices. Moreover, the value of their reserves increases with the price of gold. As a result, their beta to the underlying global gold price is generally much higher than one.
Gold Royalties:
The role of royalty companies is to provide upfront financing to support initial mining costs in exchange for a percentage of the revenue generated from the sale of gold. They are exposed to the same issues as miners: instead of being the operator, they finance gold mining activities. Royalties have a lower gold beta than miners because they have lower fixed costs.
Gold Recyclers:
Considering the environmental and social damage from gold extraction, recycling offers a more sustainable option. Gold can be recycled without losing its purity, it is much less energy intensive to recycle and recycling avoids the bulk of the environmental and social risks highlighted above. Recyclers have an even lower beta to gold than royalties (less than 1).
Gold Bullion:
The sustainability argument for gold bullion is that it has already been mined. This is a weak argument in our opinion because gold mining is hard on the environment and society but some investment products, like Royal Mint’s Responsibly Sourced Physical Gold ETC, try to improve the sustainability of gold bullion by having a high percentage of recycled gold. Holding gold bullion or investment products linked to gold is not as damaging as miners or royalties, but to us, recyclers are still the most sustainable.
Which Companies Caught Our Attention?
The table below highlights gold miners with weak ESG scores versus recyclers with strong ESG scores, using CSRHub data. Barrick, B2B, Anglo and Kirkland Lake stand out as strong potential gold miner short candidates, while Sims, Umicore, Are Holdings and Johnson Matthey are potential long recyclers.
Next, from a financial perspective, the table below highlights a similarity in revenue growth over the past five years for both miners and recyclers. Valuation remains attractive for both groups, though it is worth noting that miners currently exhibit more favorable profit margins than recyclers. However, as highlighted above, 80% of identified reserves have already been mined. Given that the remaining 20% will be harder to access and yield lower concentrations of gold, miners will need to inject significant capital to explore and increase production. This will inevitably be reflected in higher capital expenditures and interest costs, thereby impacting the overall profitability of mining companies.
How Has the Gold Complex Behaved Recently?
The chart below shows an index of the four miners compared to an index of the four recyclers and the price of gold. Even if the price of gold has increased steadily since the beginning of the decade, gold miners and recyclers have struggled.
How to Position?
The resilience of the price of gold to much higher interest rates, falling inflation and reduced financial market volatility has been surprising. As we discussed earlier, physical gold ETFs saw outflows in 2023 but the main reason for gold’s resilience lies in increased central bank demand. As we have shown, most central banks expect to continue to accumulate gold in the next 5 years. Recyclers that have not yet followed the price of bullion higher should be the investment of choice for sustainable investors looking for gold exposure.
The share of consumers’ budgets dedicated to spending on food has made new multi-decade highs; in the U.S., households have not spent this much of their income on food since the 1980s. Is the long era of cheap and plentiful food behind us? We think so.
Although cyclical forces such as a tight labor market are also contributing to higher food prices, much of the increase in food costs are due to the headwinds to agricultural productivity from climate change, biodiversity loss and other environmental and social factors at a time when the global population is still growing. Indeed, these headwinds are affecting a multitude of food supply chains. Work at Ceres reveals that within each of the eight major food commodity supply chains, there exists at least one environmental or social factor considered a ‘high’ risk. This includes factors that pose threats to food production itself, as well as instances where food production contributes to environmental and social harm.
These risks are not simply theoretical and/or something to consider in the future. Take cocoa for example; West Africa produces most of the world’s cocoa, but two years of hot and dry weather have severely impacted crops; both current yields and the viability of farming in the future are at risk. U.S. cocoa futures reached a nearly fifty-year high in February!
From an environmental perspective, no food industry is more problematic than beef. Risks are rated as “high” or “medium” across all major environmental and social factors according to Ceres.
If the status quo prevails, it is likely to lead to levels of price inflation that can not be fully borne by the consumer. Put simply, industry profit margins are at risk from the environmental and social factors in the table above. In today’s note, we analyze the food supply chain of beef and cocoa. We explore how investors can both protect their capital and improve environmental and social outcomes.
What’s The Beef About Beef
The U.S. and Brazil are the largest producers of beef, accounting for one-third of global production. The main difference is that beef raised in the U.S. is primarily nourished on grains (corn, soybeans, and alfalfa), while beef produced in Brazil is mostly grassfed. In the U.S., the beef value chain has five major links:
- Animal Feed Suppliers: Producers of corn, soybean and alfalfa. Three companies produce nearly 40 million metric tons of feed annually.
- Cow-calf (ranching) Operations: calves are birthed, raised, and weaned on site. This stage of production is extremely fragmented, with over 700,000 farm operations across the U.S.
- Cattle feeding (feedlots): Cattle spend the last six months being fed a high-energy diet. Until recently, JBS, Tyson and Cargill owned about 95% of the feedlot capacity. However, since around 2020 most of the U.S. feedlot capacity is owned by private market firms.
- Meat packing and processing: There are only four companies that slaughter, cut and pack about 85% of all U.S. cattle.
- Food Retailers & Restaurants: About two-thirds of U.S. beef supply goes into food service—restaurants and cafeterias—while one-third is sold in supermarkets.
For each stage of the beef value chain, we use public company reports and government data to estimate profit margins. The table below summarizes each stage of production and relevant characteristics.
There are two key takeaways from the table above:
- With the exception of feed suppliers, each stage of production tends to deliver healthy profit margins, albeit it is the cow-calf operations (ranchers) that have the most variability (uncertainty) in profits.
- At least two stages in the value chain can be considered extremely concentrated. Four companies – JBS, Tyson, Cargill and NFC – control virtually all of the meatpacking in the U.S. and these companies tend to have the highest margins in the value chain.
Cracking the Cocoa Conundrum
Cocoa production is heavily concentrated in Ivory Coast and Ghana, which collectively account for nearly 60% of the world's production. The cocoa supply chain involves four distinct stages:
- Production: Cocoa is cultivated in plantations, then collected by collector members of cooperatives and processed at fermentation and drying centers to produce cocoa beans. Millions of farmers participate in this highly fragmented part of the chain.
- Processing, transport, trade and distribution: Beans are refined and ground to produce cocoa butter, cocoa liquor, and cocoa powder. Seven traders handle 62% of cocoa export volumes, making it a concentrated market.
- Manufacturing: Cocoa products are used by food manufacturers to produce chocolate confectionery and by industries such as cosmetics and pharmaceuticals. Mars, Hershey, Cargill, and Blommer—some of the world’s biggest chocolate manufacturers and processors—are headquartered in the U.S.
- Retail and consumption: Chocolate products are sold through food retailers and consumer outlets (grocery stores, wholesalers, and online).
As with beef, in the table below, we use public company reports and government data to estimate profit margins and we summarize each stage of production of cocoa and provide relevant characteristics.
Again, there are two key takeaways from the table above:
- While farmers receive a disproportionately small share of profits, manufacturers and retailers command a significant portion.
- Market concentration is exceptionally high in processing, trade, and manufacturing, with only a few major players dominating these sectors, while small-scale farmers bear the brunt of the risks at the production level. This concentration means that processing, trade, and manufacturing entities have the ability to diversify some of the risks they face, contrasting sharply with the vulnerability of small-scale farmers.
Investment Implications
So far, our analysis has shown that in the two foodstuffs we analyzed, there is at least one stage in the value chain that is completely dominated by a limited number of companies. In the cocoa chain, Mondelez, Olam, Cargill, Barry Callebaut and Nestle dominate. For beef, JBS, Tyson, and Cargill dominate meat processing, while ADM, Cargill and Bunge have the stronghold on animal feed production.
The table below revisits the initial eight foodstuffs analyzed by CERES and shows that there are a handful of companies involved in ALMOST EVERY food value chain.

In other words, there are select public companies that are likely to participate in virtually every meal of the typical western diet!
At the early stages in the value chain, ADM, Cargill, and Bunge are involved in all eight foodstuffs. At the transformation stage, Nestle, Mondelez and Kellogg are each involved in at least 50% of the foodstuffs.
From a consumers’ perspective, the realization that a mere handful of companies have such dominance in what we eat can be unsettling. However, from an investor’s perspective, particularly one that is concerned with sustainability, it is clear that influencing the operations of just a handful of companies can potentially have enormous impacts throughout our food system. In other words, food is an industry whereby aggressive investor engagement, when done at key links in the value chain, can be effective by concentrating efforts on just a short list of companies.
Engagement at the Agricultural Production Stage: ADM, Bunge and Cargill
ADM, Bunge and Cargill together are the main agribusiness companies that buy, trade, process and ship agricultural commodities such as corn, soybean, palm, wheat, etc. ADM and Bunge are both listed companies while Cargill is private. Another major difference is that ADM and Bunge have similar, elevated ESG scores across all material SASB categories (energy management, water management, product quality, employee safety, supply management, materials sourcing) while Cargill is severely lagging (in 2019, it was named the worst company in the world by Mighty Earth!)
That said, all three companies have a multitude of controversies; ADM scores a D- according to Refinitiv Eikon and Bunge is not much better with a C-. Controversies range from animal testing, workplace conditions, child labor, deforestation, shareholder rights, anticompetition and more.
Engagement at the Processing & Transformation Stage: Nestlé, Olam, Mondelez and JBS
Nestlé: Nestlé demonstrates comprehensive disclosures and proactive approaches to water management, energy efficiency, and product safety, and is considered better than its peers on these topics. However, they lack disclosures on corrective audits in supply chains, GMO products, and ingredient sourcing in regions with high water stress. Furthermore, Nestlé has faced a series of recent controversies (scoring a D according to Refinitiv Eikon), including lawsuits over anti-competition practices (infant formula), allegations of illegal water bottling (California), labor disputes and working condition controversies. They have long been accused of child labor in their cocoa supply chain and following stakeholders’ pressure, they announced a US$1.4 billion investment plan in 2022 to address the issue, even though there are limitations in their ability to effectively tackle this pervasive issue.
Olam Group Ltd: Olam provides operational disclosures that are slightly behind their peers, although the good news is this company is relatively free of controversies. However, they acknowledge that child labor remains a persistent issue in cocoa production. Olam relies on a vast network of suppliers spanning over 60 countries, with a direct and indirect supply base predominantly farmed by small-scale farmers in emerging markets. While the company is aware of child labor, their efforts remain paltry: they provide training on the topic to 240,000 farmers which represents a mere 5% of all farmers they work with.
Mondelez International: Mondelez stands out not only for its operational performance but also for its product portfolio, which primarily consists of highly-processed and sugary snacks. Unlike other companies mentioned, Mondelez specializes in indulgent treats that often contain high levels of sugar, unhealthy fats, and artificial additives. This focus contributes to potential health risks such as weight gain, obesity, diabetes, and heart disease associated with excessive consumption of these products. In addition to these concerns, Mondelez, like its peers, faces regular controversies, scoring a C- according to Refinitiv Eikon, regarding product recalls, data breaches, advertisement norms violations, and child labor in its supply chain.
JBS: JBS has been long-considered a major laggard on most ESG-related issues, with particularly low scores related to working conditions, human rights and its supply chain. But the company’s controversies go even beyond these issues and include environmental scandals, child labor issues and anti-competitive behaviour (Refinitiv Eikon ratings is a D). All of these areas are ripe for investor engagement.
Our hypothesis at the end of 2023 that the artificial intelligence hype had become extreme with respect to related stock valuations has not panned out so far in 2024. Not only did Nvidia’s stock price soar on last week’s earnings announcement, but the company’s CEO Jensen Huang has further fuelled excitement related to generative AI with his optimistic guidance, “[AI] Demand is surging worldwide across companies and nations… Auto, health, robotics, financial services… multibillion-dollar industries are embracing our generative AI”.
Although the valuations of many “AI companies” make our eyes water (NVIDIA is trading at 66 times earnings), there is certainly much to be excited about from a productivity perspective. After all, as we highlighted in our April 2023 note, “Finding Real Returns in Artificial Intelligence” ChatGPT displays an IQ of above 120, placing it among the top 10% of humans. This is quite the progress, given that less than a decade ago, researchers ranked Google’s AI IQ at 47, just shy of the average IQ (55) of a human 6-year-old. That said, for AI to make its way into all spheres of the economy like Internet did in the late 1990s, a massive economy-wide CAPEX effort will be needed – and as of now, that has yet to happen.
And what about from a sustainability perspective? There has been healthy debate (although perhaps not enough action) surrounding ethical aspects of AI. Questions have been raised about the energy intensity of AI tools, about data biases that are leading to discriminatory outcomes, about potential job losses, etc. On the flip side, AI promises to address challenges across industries, including resource optimization and medical breakthroughs. In short, artificial intelligence’s influence on our future is complex and heterogeneous. For investors looking to put capital to work beyond the semiconductor companies that have supplied AI with the chips to flourish, we note the key Sustainable Market Strategies’ themes where leveraging AI should produce superior real-world outcomes alongside superior returns.
Productivity with Purpose
With labor shortages persisting in many developed countries and an ever-increasing need to make better use of physical resources, the first major allure of artificial intelligence is its potential to increase productivity, either by replacing less efficient technologies (and/or labor) or via optimisation of resources.
Software companies were one of the “early movers” with respect to excitement in the application of AI tools. Indeed, four software companies with strong ESG credentials that have been widely lauded for their AI integration – Salesforce, ServiceNow, Adobe and Workday – have seen their stock prices jump between 70- 80% in the past year: a less stellar performance than Nvidia perhaps, but nonetheless extraordinary relative to the S&P500.
That said, investors would be justified if concerned that last year’s hype will not be repeated in 2024: ServiceNow and Workday both reported quarterly earnings in February and the stock market reaction was negative. Both companies disappointed on revenue growth. Investors seemed particularly underwhelmed by Workday’s announcement of the acquisition of an “AI talent solution company”. What is more, Workday has recently come under pressure for facilitating discriminatory biases with its existing AI tools. At valuations still sky high (the lowest P/E of these companies is Adobe at 47 times earnings!), it makes sense to look elsewhere for productive and sustainable applications of AI.
Beyond software, one area that we find exciting as an AI sustainability opportunity is in industrial automation and digital design. In our note last year, we screened for companies that were best positioned to leverage AI advances for purposeful applications, i.e. supporting decarbonization, optimizing physical resource use and improving safety in industries like energy and food production.
The companies we identified at the time were already making good use of artificial intelligence tools and this has continued, with several companies like Schneider Electric and Rockwell Automation announcing new partnerships with Microsoft to leverage their OpenAI tools.
- Rockwell Automation (ROK) provides automation and software solutions to the manufacturing sector, primarily in the U.S. Like many parts of the U.S. manufacturing sector, the company struggled over the past year with an inventory bulge and supply chain constraints, which have been reflected in the company’s muted stock price performance. With the U.S. manufacturing sector showing signs of revival (e.g. the ISM manufacturing survey has bottomed), companies like Rockwell offer an interesting entry point.
- Schneider Electric SE is an automation and energy management company, with about three-quarters of revenue generated from energy management and one-quarter from industrial automation. This makes the company particularly well-positioned to leverage AI advances for purposeful applications as well as to capitalize on the structural buildout of electricity grids (see last section below).
Leave No Person Behind
At Sustainable Market Strategies, investing to reduce social inequalities centers around three key ideas: improving health outcomes through greater access to care and medical breakthroughs, improving education and increasing labor fairness. Of these three sub-themes, we believe education is the area where artificial intelligence is already playing out in stock prices and where sustainability investors can leverage near-term trends.
Artificial intelligence apps, such as ChatGPT, are increasingly part of an educational toolbox just like calculators, and simple web searches. Artificial intelligence employs adaptive learning, intelligent content curation, and smart monitoring and reporting to teach skills: dozens of learning apps powered by AI have sprung up in recent months. One stand-out is Duolingo. The company’s main focus is on language learning, which is in high demand due to ever-increasing international worker flows and remote work. In a note last June “Education Stocks: Will ChatGPT Decide Who Passes or Fails”, we contended that Duolingo would be an AI winner. Since then, the company has added personalized language lessons fuelled by AI-powered feedback and has reached all-time high engagement rates, as measured by the activity of daily and monthly users. The stock price is up over 40% since our June note and Q4 earnings that were reported earlier this week confirm that the company is able attract paid subscribers to their products.
As for health care, this is the sector that holds the most promise for positive disruption from AI. Already, there are several research studies showing that AI can perform as well as or better than humans at some health care tasks. Studies show, for example, that patients often prefer responses to medical questions from chatbots over physicians, giving the chatbot higher ratings on both quality and empathy!
But the best medical use cases of AI involve settings where large data sets (especially images) that can be applied to well-defined problems are available. Diagnosing diseases, by reviewing X-rays for example, is a very effective use of AI. Companies like Quest Diagnostics, which offers access to diagnostic testing services for cancer, cardiovascular disease, neurological disorders, already leverage AI in their diagnostic tools, most notably for dementia.
A major risk, however, is that current medical AI tools tend to amplify existing discriminative biases in health care outcomes: since AI-based health care technology depends on the data that it is fed, a lack of representative data risks leading to biased models that ultimately produce incorrect health assessments especially according to race. Moreover, although the health care sector is heavily regulated, no regulations directly target the use of AI in medical devices. This is likely to change and will significantly alter the risk/reward profile of health care companies actively investing in AI tools.
Finally, in terms of labor fairness, there remain many challenges to ensure that AI tools do not accentuate existing cognitive biases. In addition, at a macro level, there are good odds that artificial intelligence will impact income inequality in ways similar to previous technology booms: if AI is levered mainly to increase automation, then a widening in the gap between capital and labor and between the professional class and other workers is likely.
A recent Pew survey in the U.S. is illustrative: the survey showed that workers with a university education are twice as likely as those without to be exposed to AI in their jobs. These same workers were far more likely to say that AI will help them more than hurt them in their jobs. On the flipside, retail trade, construction and accommodation and food services are all industries where workers are unlikely to see much benefit from AI tools and are also at risk of being replaced by technology. These are industries that tend to be lower wage jobs with low levels of unionization or other worker protections. In other words, there is some evidence to suggest that gains from artificial intelligence will accrue in already high-paying sectors and largely bypass and/or harm employment in lower paying industries. It is challenging to find AI focused investment ideas related to alleviating income/labor fairness, particularly since, as described above, some of the best company software tools (Salesforce, Workday) allegedly create increased discriminatory biases.
Achieving The Energy Transition
Artificial intelligence is a voracious consumer of energy. An oft-quoted study published in late 2023 projects that based on current and future sales of Nvidia microchips specifically for AI, the amount of electricity required to power the world’s data centers could jump by 50% by 2027, thanks to AI alone. According to the study, that’s about the power consumption of the Netherlands, and would require the entire output of about 15 average-size nuclear power plants.
Many companies are working on smaller, less power-hungry AI models that, but even if AI-related energy is half of the current estimate, it will require substantial increases in power production. Most problematic is that the energy “needs” of AI are arriving at a time when renewable energy supply is still falling short of overall energy demand: without the complicating factor of AI, the energy transition is already a massive challenge.
To be sure, artificial intelligence can be used to make the grid smarter and optimize energy usage. But it’s likely that over the next five years, demand from the production of data across industries will far outpace energy savings from optimization within the energy sector.
Crucially, it is the demands on distribution and transmission of electricity (i.e. the grid) that is particularly challenged by artificial intelligence demands. Transporting electricity from where it is produced to where data centres are desired requires a buildout of the electricity grid.
This problem is particularly acute in North America, where grid planners have recently nearly doubled their 5-year load growth forecast; in the U.S. electricity demand is now expected to increase by 4.7% over the next five years. By way of comparison, over the past decade, grid planners have been forecasting a mere 0.5% annual growth rate.
We see this as a major bottleneck for artificial intelligence usage and a major opportunity for companies involved in the buildout of energy transmission. Schneider Electric, Eaton Corporation, ABB, Nexans, Prysmian should become household names for investors that want exposure to the energy transition.
There is a significant perception gap between the actual operations of municipal waste management companies and the public’s perception of their environmental impact. It’s understandable that the public might view these companies as environmental saviors, given their green-themed logos and websites.
Currently, municipal waste management companies generally derive less than 10-15% of their revenue from recycling activities, with the majority (80-90%) coming from the collection and disposal of trash at landfill sites. Industrial waste management companies, however, tend to have a higher share of revenue from recycling activities, particularly in the metals and batteries sectors.
In this week’s note, we will focus on the municipal waste management companies and see if sustainability-minded investors should consider them in their portfolios.
From an investment outlook perspective, the municipal waste management sector is quite attractive. Annually, our planet generates about 2.01 billion tons of municipal solid waste (MSW), a figure expected to reach 3.40 billion in about 30 years (a 70% increase). According to the World Bank, daily per capita waste in developed markets is expected to increase by 19%, while in emerging and frontier markets, it could rise by 40% or more.
The good news from a sustainability perspective is that the MSW business model is evolving in the right direction. A few years ago, these companies were even less involved in recycling activities, as the norm was to ship most recycling to China and other Asian countries.
As discussed in our note Somebody's Trash is Someone Else's Treasure, July 11, 2019, this situation started to change in 2018 after China stopped accepting plastic waste and other recycling materials from the rest of the world due to environmental concerns. The sudden ban – called Operation National Sword – drastically changed how plastic recycling had been handled in developing countries. Up until then, China was the final destination for half of the world’s plastics destined to be recycled. To avoid becoming large plastic dumpsites, Indonesia, the Philippines, Thailand, Malaysia and Vietnam introduced regulations to limit, if not outright ban, the import of contaminated recycled materials – a definition that de facto excludes most of what was being shipped at the time. To avoid a race to the bottom, the EU even banned the export of waste plastic to non-OECD countries. MSW companies in North America, Europe and Japan had to step up and fill the gap, which they are still in the process of doing.
Another positive change in the MSW industry is the revenue potential from renewable natural gas (RNG) and landfill gas-to-electricity (LFGTE) projects due to methane capture at landfills. Thanks to incentive programs like the Inflation Reduction Act (IRA) and the Renewable Fuel Standard (RFS), MSW companies can generate more revenue from environmentally-friendly activities.
The main incentive in the US is the opportunity for landfill operators to generate Renewable Identification Number (RIN) credits that can be sold to regulated entities under the RFS scheme, such as refiners. The different types of RINs (D3, D4, D5, D6) correspond to different categories of renewable fuels.
The different RIN credits have vastly different environmental credentials, and this can be reflected in their value.
Fortunately for municipal waste management companies involved in RNG production at landfills, they qualify for the highest integrity credits (D3), which carry a very lucrative price these days.
Some MSW companies have started to notice. Waste Management (WM US), for instance, will invest up to $2.9B in its recycling and renewable energy platforms through 2026. In particular, WM expects to commission five new renewable natural gas facilities by the end of the year, with 15 more in the coming years. These renewable energy investments are expected to generate over $510M in incremental operating EBITDA per year by 2026, which should surpass the company's recycling revenues.
Waste Connections (WCN CN) is also investing significantly in RNG and expects about $200M in incremental operating EBITDA per year by 2026 to come from their landfill methane capture projects, which is proportional to WM's numbers.
Republic Services (RSG US), on the other hand, seems to have chosen to focus on its advanced polymers recycling facilities, which should rapidly become a profit center as they will charge other MSW companies to process their plastics. The first Polymer Centre opened in Las Vegas in 2023, with a handful more planned in the next 5 years across the US. Contrary to WM and Waste Connections, Republic does not seem to believe in the RNG market, as they have given the operationalization of these projects to BP, only retaining a minority stake in the business.
Finally, while being slightly behind in terms of RNG projects, GFL Environmental (GFL CN) still expects about $175M in incremental operating EBITDA by 2026. In addition, another $50M to $75M is expected to come from recycling facilities in Canada as a result of the Extended Producer Responsibility (EPR) regulations coming into effect in various provinces. As GFL derives almost 40% of its revenues from Canada (versus 13% for Waste Connections and 3% for WM), they are probably better protected than the others in the event of a change of tack at the White House this November – an event that would more than likely affect the subsidies currently available through the IRA and RFS programs.
Despite the differences in strategies and markets of these four companies, the stock performance and valuation charts show a certain homogeneity in the MSW sector. For us, picking winners will be a question of total sustainability efforts and future-proofing of the business model in case of a change in direction in Washington come this November.
In terms of total sustainability efforts, WM stands out as the company with the highest share of revenues directly linked to recycling and renewable energy projects (approximately 20% by 2026). Also, GFL Environmental, with its larger presence in Canada, could be an interesting political hedge in case some of the juicy subsidies don't survive past the next White House Administration.
News last week that the U.S. was now putting restraints on new liquified natural gas shipments to countries with whom it does not have a free trade agreement may mark the “peak gas” moment in the world economy. This comes at a time when environmental concerns are growing regarding the role of natural gas as a “clean” source of energy. After all, total global CO2 emissions from natural gas are approaching those of oil and to a lesser extent, coal.

Furthermore, numerous studies suggest that methane leaks coming from LNG operations and other natural gas facilities are at least as potent as coal with regards to global warming. Recently published research suggests that assuming a 0.2% loss of methane across natural gas and LNG facilities, this would put the LNG/natural gas industry on par with coal and its effects on climate change. What’s more, other researchers found that methane leaks are often largely underreported by the Oil & Gas sector in countries such as Canada, adding to the overall climate warming estimates of the industry. Finally, a recently published article in Nature suggests that the extraction and use of fossil fuels account for about 15–22% of total methane emissions and those same methane emissions have been underestimated in the range of 20–60%. Given that natural gas is mostly made up of methane, these numbers are especially alarming considering that global natural gas production has boomed since 2000, in fact doubling in volume, and growing much faster pace than oil.
Despite the greenhouse gas potency coming of gas-related emissions, some countries are still accelerating their dependence on the fuel. For example, Germany has decided not to restart some of their decommissioned nuclear facilities despite the energy crisis resulting from Russia’s invasion of Ukraine. As a surprise to many - and despite the critique from several experts and environmental groups - the country recently announced that it would launch the construction of several new LNG terminals instead.
In this context – with coal almost universally banned from sustainability-minded investor portfolios and several investors now divested out of oil – what are investors to do with assets in gas? After all, for many years gas-related companies and utilities have claimed that they were “cleaner than coal”. This was a convenient way to keep consumers and governments happy, permitting a business-as-usual approach in supplying the economy with a stable source of base power – all this despite its relatively high carbon footprint. But that narrative is changing rapidly – at least in regions where energy security seems a lesser concern than in Germany.
An increasing number of cities around the world have taken steps to ban new natural gas hookups to combat climate change. In the U.S., Berkeley (California) became the first American city to ban natural gas hookups in new buildings in 2019. Other cities such as San Jose, Mountain View, Santa Rosa and Brisbane have also banned natural gas equipment in new buildings. Chicago is also considering a ban on natural gas hookups in new construction. In Canada, Montreal just announced it will no longer allow gas hookups in new buildings of up to three storeys high as of October 2024, and ban the fossil fuel as of April 2025 in larger new builds.
If the trend stays the same, gas-related assets could realistically become stranded within the next decade or so. Aside from issues pertaining to global warming, this is especially true if one considers the fact that the levelized cost of electricity production from gas is increasingly expensive compared to that of renewable sources – and even nuclear.
The Last Mile Is the Hardest One
For some investors, the path towards “getting out” of gas will prove a hard feat. We see two main issues around weaning portfolios off gas. First, if will be hard for investors in the Canadian stock market, with a 17 percent weight in Energy in the S&P TSX equity index, to find diversified investment strategies if we start excluding gas and pipelines from the universe. Looking at the fixed income market, as much as 20 percent of all fixed income securities in Canada were in the gas and pipeline sector as of 2023. In this context, decarbonizing the “last mile” of Canada-focused fixed-income portfolios could also mean considerably shrinking the investment universe and increasing active risk.
The second issue comes from the desire of many investors to gain exposure to infrastructure assets through infrastructure funds. These multi-sector infrastructure funds often include pipelines, LNG terminals, and local gas distribution networks. In fact, while the Oil & Gas sector represents only 3.8% of the S&P 500’s market capitalization, it represents about 15% (on average) in infrastructure funds. Looking at 98 funds on Bloomberg with exposure to Oil & Gas assets included in the infrastructure theme, we see that Oil & Gas assets represented about USD 9.9 billion of a total AUM of USD 48.6 billion as of late 2023. The table below underlines the funds with the heaviest exposure to the Oil & Gas sector, with some funds displaying a 20-40 percent exposure to Oil and Gas.
Interestingly, the equal-weighted index of the 98 infrastructure funds underperformed by some 30 percentage points against the MSCI ACWI index since 2019 (on a total return basis). While the infra sector should benefit from lower interest rates towards the latter part of this year, it is hard to see a long-term case for holding infrastructure assets with a high exposure to Oil & Gas in your portfolio for the long run.
In previous notes, we wrote about the comparative advantage of Mexico and India relative to China from a global manufacturing standpoint: “Made in Mexico” on July 6, 2023, and “Slumdog Millionaire Could Eat Wolf Warrior’s Lunch” on August 3, 2023. This week, we examine how and where the opportunities should materialize within Mexico and India on their own merits, rather than simply as replacements to China.
Our analytical framework consistently examines four drivers that shape the global economy: demography, international relations, technology and the regulatory backdrop. For the foreseeable future, climate change and social inequalities are also important factors that should not be ignored. Our analysis indicates that the next ten years will be chaotic globally because the large Baby Boomer generation is retiring, being replaced by the much smaller Gen Z, while populism and zero-sum calculus is steering international relations. Meanwhile, artificial intelligence and robotics have reached the implementation phase which could bring great productivity improvement at a time when regulatory oversight is lagging. Meanwhile, the earth is getting warmer and social inequalities are reaching levels that have historically resulted in a popular sense of resentment.
Considering this global context, we find that India and Mexico are better positioned than most countries. On demographics, they have a young and growing population without the baby boomer retirement issue. Contrary to most other countries, they have more residents in the 10-20 year-old cohort than in the 50-60 year-old age group. The chart below highlights the difference between Mexico and India’s similar profiles versus China. We use China, but Europe and Japan also have a lot more upcoming retirees than new workers.
On geopolitics, Mexico is located next to the largest consuming country in the world, with which it has a free trade agreement. India, by not being strategically aligned to a specific group of countries, is courted by both the United States and China. Geographical factors can influence economic opportunities and trade dynamics. Both India and Mexico have strategic geographical locations that facilitate trade as the map below highlights. Mexico's proximity to the United States makes it a key player in North American supply chains, while India's location provides access to emerging markets in Asia, Africa, and the Middle East.
On the negative, the map also spotlights the risk resulting from social inequalities in Mexico and India with GINI coefficients of 46.1 and 35, respectively, in 2019. These inequality levels are not as dire as those in Brazil and South Africa – as measured by the GINI coefficient - but they are still relatively high. Also, since both countries are close to the Equator, it puts them at greater risk from climate change.
For the last two drivers: technology and regulation, both represent opportunities and risks. India is producing a large number of engineers that can harness technology, but artificial intelligence and robotization will reduce the demand for workers. For regulations, we cannot affirm that Mexico and India are regulatory leaders with political interference and lots of red tape, but they have shown some progress over the years.
Structural Trends
Considering our perspective on the six drivers examined above, we are optimistic on the structural paths of India and Mexico. The picture below highlights the areas where investors should focus for sustainability impact: infrastructure and essential services. More consumers and greater household formation levels lead to more purchases of food, cell phones, housing-related items, technology and financial services. To support this larger and younger population, infrastructure needs will increase for education, water, telecom, electricity, hospitals and waste management, among others.
Both India and Mexico are vulnerable to various climate change impacts, and these effects can manifest across different sectors. Some potential risks for India and Mexico are highlighted below.
These climate risks compound the need for improved infrastructure and essential services like water, electricity, food and medical supplies.
Infrastructure
On infrastructure, we looked for sustainability leaders that will participate in the infrastructure expansion and respond to climate change issues. Larsen & Toubro, in India, is a diversified company involved in infrastructure and heavy industries. It builds roads, bridges, airports, energy transmission and production. Orbia, in Mexico, is similar as it is involved in a similar sector with a greater emphasis on agriculture. In the materials industry, Cemex in Mexico, is a leader in low-carbon cement and has a Science-Based Target Initiative (SBTi) objective consistent with 1.5 degrees of warming. For telecom and the electric grid, Bharti Airtel and Power Grid of India are strong candidates.
Essential Services
With that in mind, we searched for locally focused sustainability leaders well positioned to respond to increased internal demand. Consumer staple companies that fit the bill include Hindustan Unilever in India and Walmex in Mexico, while in the financial services industry: HDFC in India and Banorte in Mexico tend to be leaders. Infosys is a promising technology play.
The chart below displays their overall sustainability performance using their sustainability percentile score from CSRHub.
Have these drivers led to strong equity performance since the beginning of the decade? The chart below shows the performance for the companies we discussed regrouped according to their sector of activity. We are using a short history and small sample of companies but a wide dispersion of results is evident: Infrastructure in India has done amazingly well but the same cannot be said for Mexico. Essentials have behaved decently in both countries. The stronger performance of infrastructure in India coincides with a strong push by the government with the “Make in India” program. In Mexico, political interventions have contributed to the poor performance of the infrastructure sector.