Today’s note looks at four leading nature markets – or environmental markets – i.e. markets where the underlying asset or commodity has positive environmental benefits or attributes. We look at recent advances, setbacks, leaders and the implications for the investability of environmental markets.
COPped out
Neither Conference of Parties (COP) meetings of the past few weeks changes where the four environmental markets (water, compliance and voluntary carbon, biodiversity) are heading. Below is a summary of the main developments in each market:
Compliance carbon markets: COP31 closes on November 20 and policymakers have not yet produced a final text. So far, the draft proposes a global effort to lift electricity to 35% of final energy by 2035, which supports demand for grids, storage and electrified end uses, while the fossil-fuel language is still contested and may be dropped. Interestingly, the fossil-fuel roadmap is being built outside the formal process by a voluntary coalition of 57 countries. This in itself tells a lot about the usefulness of the COP process). The case for grids and storage therefore rests on economics and national policy, not on a COP mandate.
Voluntary carbon markets: A coalition of governments has published a playbook to build demand for high-integrity credits. That should support quality credits but not a broad rebound in voluntary prices, because no COP decision sets a price.
Biodiversity: COP17 on Biodiversity closed on October 31 and changed disclosure and process with little if any implication for markets. It produced nearly USD 2bn of new funding – against USD 7.3 trillion of nature-negative flows.
Water : Water has no COP of its own, and its nearest catalyst is the UN Water Conference on December 8 to 10. The link to the climate COPs runs through adaptation finance, where the COP31 draft proposes that the Belém pledge to triple it by 2035 should reach at least USD 120bn, against estimated needs of about USD 305bn a year.
1. Energy Transition: Clean Power Outruns Demand
Clean power covered all of the growth in global electricity demand in 2025. Ember’s Global Electricity Review 2026 reports that solar rose by a record 636 TWh, that wind and solar together met 99% of the increase in demand, and that fossil generation fell 0.2%, the first decline since 2020. Renewables (33.8%) also overtook coal (33.0%) for the first time in the modern power system. Carbon Brief’s summary of the review adds that the global electric vehicle fleet now displaces 1.8 million barrels of oil a day.
Capital continues to move towards energy transition complex – but its mix is changing. The IEA’s World Energy Investment 2026 puts energy investment at USD 3.4 trillion in 2026, with about USD 2 trillion going to grids, power supply and electrification and USD 940bn going to China alone. Renewables investment has flattened after several years of rapid growth, partly because equipment got cheaper, so the growth in spending came from grids, storage and gas. Gas plant orders reached 130 GW in 2025, driven by US data centres. Energy security now ranks alongside cost in many import-dependent economies, which has also revived interest in coal in some of them.
Leadership sits with a handful of countries. China supplied 336 TWh of solar growth, and India cut fossil generation by 3.3% as record additions of solar and wind outran demand. For investors, generation is becoming a commodity, so the margin goes to the owners of the scarce pieces around it, such as interconnection, transformers, storage and dispatchable capacity. Equity markets have drawn a similar line. In the year to November 17, the grid infrastructure ETF (GRID-US) returned 5.1% while the broad clean energy ETF (ICLN-US) lost 9.6%, and both trailed the 12.2% return of the S&P 500.
2. Carbon Markets: Still on the Way Up
Carbon pricing has spread further than most investors assume. The World Bank’s State and Trends of Carbon Pricing 2026 counts 87 pricing instruments covering 29% of global emissions, with India and Viet Nam the newest to join, and finds that carbon pricing raised more than USD 107bn for public budgets in 2025. The EU adopted its 90% target for 2040 but delayed ETS2 (European Trading Scheme) to 2028, and in July the Commission proposed an Industrial Decarbonisation Bank and a route for permanent removals inside the ETS. California extended Cap-and-Invest to 2045, and CARB’s May update lets the emissions ceiling fall 11% a year this decade.
Compliance prices have firmed up while voluntary prices have slipped. The latest California-Québec joint auction, held on August 19, settled at USD 32.48 (CAD 45.11), up from USD 28.81 in May and USD 4.54 above the USD 27.94 floor. All 49.0 million current-vintage allowances sold for the fifth consecutive time, at a cover ratio of 1.31. Compliance entities took 77.8% of the allowances against 81.3% in May, which points to more participants coming from the financial sector. It was also the first auction since CARB finalised its May update, and California, Washington and Québec signed a linkage agreement in the interval. Carbon ETFs have recovered from spring lows but still trail equities: in the year to November 17, the California allowance ETF (KCCA-US) returned 7.9% and the global carbon ETF (KRBN-US) 3.4%, against 12.2% for the S&P 500.
In the voluntary market, MSCI reports that the average price index slipped to USD 3.5 per tonne in 2025 from USD 4.3, even though credits rated BBB and above rose by more than 20% to USD 6.8. The market for renewable energy credits shrank by a quarter, and the market as a whole has held near USD 1.4bn for four years. Dependence on a single buyer is a further weakness, as Microsoft’s April pause in removal purchases showed.
3. Biodiversity: All Talk, and Some Reporting
A recent UN report suggests that 22 of 23 targets biodiversity targets were missed by national entities, Furthermore, the USD 20bn international finance pledge for 2025 was missed, and 84% of countries missed the subsidy-identification deadline. Still, the agreed TNFD framework now counts 830 adopters, and Ceres notes that more than 680 investors now integrate nature into their processes.
The gap between ambition and allocation of capital is clearest if one looks at financial flows towards nature. UNEP’s State of Finance for Nature 2026 shows USD 7.3 trillion of nature-negative flows in 2023 – i.e. an economic activity or financial flow that contributes to the degradation of nature, rather than its protection or restoration – against USD 220bn for nature-based solutions, of which only USD 23bn was private. That works out to roughly thirty dollars for activities that degrade nature for every dollar that protects it.
4. Water: Infrastructure and Treatment, Not a Price
Water remains stressed in many regions of the world. UNU-INWEH reports that about 70% of major aquifers are declining and that drought costs USD 307bn a year. Investment of about USD 380bn a year compares with a need of more than USD 1 trillion, and UN DESA and Goldman Sachs Asset Management put the public share of that at 85% to 90%. In Canada, the gap is local as well as global. IISD cites a C$270bn municipal infrastructure deficit, which shows that water is constrained as much by governance and municipal capacity as by capital.
Listed water stocks have not consistently beaten global equities over ten years. It’s true that Goldman Sachs Asset Management reports outperformance against the MSCI ACWI IMI over the ten years to mid-2025 on a broad definition of water providers, and sector capex is rising. However, on investable water indices, the lead of the global funds has gone. To November 17, 2026, the S&P Global Water and Nasdaq Global Water funds returned 10.2% and 9.3% a year in US dollars against 12.2% for the MSCI ACWI proxy, having led it until April 2025 and May 2024 respectively, while the two US water funds returned 12.3% and 12.6% a year, slightly ahead of it.
Except for a few occurrences in the U.S. and Australia (mostly water rights in the Ag sector), water remains a non-tradable commodity. For their part, costs associated to delivering water, i.e. municipal infrastructure, have lagged overall inflation through the 2022 shock but have outpaced it since mid-2024. Over the five years to October 2026, US consumer prices for water, sewer and trash collection rose 27.7% against 22.0% for all items, and EU water supply prices rose 31.9% against 26.0%. That supports the revenue case for regulated utilities, although bills that outrun household inflation also raise the risk of affordability caps. Overall, it seems that anything related to water theme has been gaining in value these past few years.
There is an old market cliché that investors like divided government because politicians cannot do very much. The logic is straightforward: when neither party can pass its agenda unilaterally, the odds of sweeping changes to taxes, regulation and spending fall, and companies can plan around a narrower range of outcomes. The 2026 midterms provide an unusually revealing test of that idea. Democrats won control of both chambers of Congress while President Trump remains in the White House, leaving neither side with the votes required to impose a broad policy reset. The likely result is not stability in the conventional sense, but a narrower range of federal outcomes, more aggressive congressional oversight, and a larger role for states in determining how energy, infrastructure, and healthcare policy evolves over the next two years.
The initial market response was consistent with this interpretation, although it was far from uniform. We track four SMS policy baskets, labeled B1 to B4 in the chart below: Transition Utilities (B1), Grid & AI Enablers (B2), the Solar Value Chain (B3) and Medicaid & ACA-Exposed (B4). The SMS Solar Value Chain and Medicaid & ACA-Exposed baskets rallied sharply around the election, Transition Utilities also advanced, and the Grid & AI Enablers basket sold off. The split looks like a repricing of political downside where it had been concentrated, alongside a reassessment of areas where expectations, valuations and public resistance had already climbed.
The return of the policy floor
For much of the past two years, investors in clean energy and healthcare faced more than ordinary regulatory uncertainty. Unified Republican control created the possibility that executive actions could be reinforced through budget reconciliation, appropriations and durable statutory changes. Democratic control of Congress does not make policy supportive, but it reduces the probability that another large legislative package will deepen the retrenchment already enacted under the Trump administration.
That difference matters for valuation because a sector does not need a new subsidy or entitlement expansion to rerate. When expectations are depressed, a lower probability of the worst outcome can be sufficient. The immediate rallies in solar and healthcare therefore look less like forecasts of a new progressive policy cycle than recognition that the distribution of possible outcomes has narrowed. Investors are pricing a floor, not yet a catalyst.
Basket 1: Transition Utilities
The Transition Utilities basket captures companies positioned to earn regulated or contracted returns from the buildout of cleaner generation, transmission, distribution, and grid resilience. The basket is less dependent on a new federal subsidy package than the solar manufacturers and installers, but it is highly sensitive to state regulation, allowed returns, permitting, and the allocation of infrastructure costs.
States may matter more than Congress
The federal result will dominate headlines, but the state-level shift may have more direct investment consequences. Five governorships moved from Republican to Democratic control, while Democratic gains in attorney-general and regulatory offices change who shapes implementation, enforcement and utility planning. Energy policy in the United States is already highly federalized: states influence utility rate cases, transmission siting, integrated resource plans, renewable procurement, net metering and building standards.
This creates an important asymmetry for Transition Utilities. The basket does not require Washington to pass another national climate package. The companies need permission to deploy capital, recover investments through rates and earn returns on transmission, generation and resilience spending. A supportive governor or public utility commission can therefore influence earnings visibility even when Congress remains deadlocked.
There is, however, a less comfortable side to the state-policy argument. If voters have become more sensitive to electricity affordability, utilities are not automatically beneficiaries simply because regulators shift costs away from data centers or favor clean investment. The same commissions that approve grid spending also determine how quickly utilities recover that spending and what returns shareholders are allowed to earn. A campaign against electricity bills can therefore become a campaign against the pace of capital spending, cost recovery or allowed returns. The investment question is not simply whether infrastructure is necessary, but whether utilities can demonstrate that it lowers system costs and allocate those costs in a way regulators will continue to accept.
Basket 2: Grid & AI Enablers
The Grid & AI Enablers basket represents the physical layer behind electrification and digital growth. These companies benefit from the need for substations, transformers, switchgear, transmission lines, cooling systems and specialized construction. Their exposure is broader than clean energy, which gives the basket a stronger secular demand base but also links it to data-center capital spending and the politics of electricity affordability.
Why Grid & AI Enablers sold off
The decline in Grid & AI Enablers is the most useful counterweight to a simple blue-wave narrative. The basket had already been one of the strongest trades of 2026, supported by data-center construction, manufacturing investment and the expectation of a long grid-capital cycle.
Campaign politics supports part of the sell-off. Many Democrats ran against data centers: campaigns spent more than $380 million on television ads mentioning them, and Democrats outspent Republicans on the topic by more than three to one. Of the 34 governors elected in races AP called, 32 back stronger regulation and 12 back bans or moratoria. Tougher permitting, ratepayer-protection rules and local moratoria are now more likely, and they land on the demand that Grid & AI Enablers had been valued on.
The size of the move may still overstate the damage to this basket. Developers facing multi-year interconnection queues have been turning to behind-the-meter generation, and one tracker counts roughly 90 GW of announced behind-the-meter capacity, more than a quarter of planned US data-center capacity. Projects that supply their own power sidestep much of the ratepayer-cost argument, but they still need switchgear, transformers and electrical construction. Rules that make data centers pay for their own grid upgrades could push more projects on-site instead of cancelling them
In short, the election results do not eliminate the infrastructure opportunity for Eaton [ETN-US], Quanta Services [PWR-US] or EMCOR Group [EME-US], but it can slow the conversion of a powerful long-term theme into near-term revenue and make elevated valuations more sensitive to disappointment.
Basket 3: Solar Value Chain
The Solar Value Chain basket is the most direct policy-duration trade in the group. It spans domestic manufacturing, utility-scale equipment and residential deployment, so the companies respond differently to tariffs, tax credits, interest rates and state-level rules. What unites the basket is that the probability of further policy deterioration can matter almost as much as the probability of new support.
Solar: a repricing of survival, not a return to the IRA
The post-election solar rally should not be mistaken for a restoration of the Biden-era policy architecture. President Trump retains control of the executive branch and the veto, while many of the incentives weakened under unified Republican government cannot be readily restored. Trade policy, federal permitting and agency implementation also remain important sources of risk.
What changed is that the clean-energy industry now has additional institutional defenses. Democratic committee chairs can conduct oversight, slow appropriations language, scrutinize agency implementation and use must-pass legislation to seek targeted concessions. State governments can also sustain renewable portfolio standards, storage programs, distributed-solar rules and procurement mandates. For a basket that remained deeply negative for the year even after the November rally, reducing the probability of another policy shock can have a disproportionate effect.
Basket 4: Medicaid & ACA-Exposed Healthcare
The healthcare basket combines managed-care organizations, hospital operators and a healthcare property company. The common thread is exposure to insured patient volumes and public reimbursement, but the transmission mechanism differs: Centene and Molina are directly affected by enrollment and rate setting, while hospitals are affected through payer mix, uncompensated care and utilization.
Healthcare: gridlock can still protect earnings
The healthcare basket may be the clearest expression of the policy-floor thesis. Democratic control of Congress does not automatically repeal the Medicaid financing changes and ACA provisions enacted under President Trump. The White House can veto a broad reversal, and Senate rules remain an obstacle to standalone legislation. Providers and managed-care companies should therefore not assume that the pre-2025 policy environment will return.
Nevertheless, the election changes the negotiating environment around additional cuts and implementation. Healthcare affordability was a major voting concern, and Democrats now control the committees, hearings and legislative calendar in Congress. Partial mitigation through budget negotiations, implementation delays, targeted ACA support or relief for hospitals is more plausible than it was before the election. Those changes can matter materially at the company level because managed-care enrollment, reimbursement and hospital uncompensated care respond to policy at the margin.
The next two years are about floors, not catalysts
The four baskets point to a differentiated conclusion. Transition Utilities gain from a lower federal downside tail and potentially friendlier state regulation. Solar receives the largest policy rerating because the sector had absorbed the greatest policy damage, although the rally does not by itself establish a new bull market. Medicaid- and ACA-exposed companies benefit from a reduced probability of further retrenchment and a greater chance of targeted mitigation. Grid & AI Enablers retain the strongest secular demand case, but they also face the clearest risk that electricity affordability and local opposition complicate the investment cycle.
The central question after an election is usually what the winners will do. In this case, investors may get more insight by asking what each side can no longer do alone. Democrats cannot rebuild the old policy regime while Trump controls the White House, but Trump can no longer rely on a compliant Congress to deepen it. For sectors carrying significant policy scars, fragmented government can function as downside protection. The 2026 blue wave may matter less because it puts Democrats in control of Congress than because it fragments control of policy across Washington, state capitals and regulatory bodies.
Most professionals in the energy sector, whether in renewables or conventional sources, know Jevons paradox: when efficiency lowers the cost of using a resource, total consumption tends to rise rather than fall. The pattern is visible today in two places. Cheaper solar and wind have added to global supply without displacing existing sources, and AI and data centers have pushed electricity demand growth to its fastest pace in decades, even as the efficiency of computing improves. But a second paradox sits alongside it: at a time when the politics of climate policy have rarely been more contested, the practical ingredients for meaningful progress are finally almost uniformly available, because the cost, technology and security arguments now point the same way in most countries.
It is easy to be a skeptic. Global energy use since the early 1900s shows new sources layered on top of old ones, with none yet displaced (Chart 1). For all the talk of a transition, the aggregate data show little evidence of one.
Indeed, looking at total global energy use from primary sources (i.e. before turned into electricity), fossil fuels and traditional biomass that are high in carbon emissions still account for about 87% of our energy use (see the chart below). There is no universally accepted threshold of a low-carbon transition, but leveraging the climate scenarios of organizations like the IEA, a net-zero-aligned transition generally means fossil fuels fall to about 20% to 25% of primary energy, while non-fossil sources (renewables, nuclear, hydro) provide roughly 75 to 80%. If we extrapolate the current pace of transition, it will take at least another 100 years to get anywhere close to a low-carbon energy system!
But it’s unlikely that the energy transition will follow the trajectory of the past because many of the pre-conditions that took decades to arrive are now in place.
In today’s note, we give four reasons why the ability to remove carbon from our energy supply has never been higher.
Reason #1: The Economics of Solar and Wind Are Winning
In the spring of 2025 (A Transition Tipping Point, May 29, 2025), we argued that many renewable technologies are entering the exciting phase of the S-curve, i.e. when growth in usage becomes exponential. The uptake of new technologies usually occurs when the cost is sufficiently competitive to make switching from the incumbent economically enticing.
This is now becoming the case for renewables: The International Renewable Energy Agency (IRENA) found that more than 90% of utility-scale renewable capacity commissioned in 2025 delivered power at a lower cost than the cheapest new fossil fuel alternative in its market when measured by the levelized cost of electricity (LCOE), i.e. the average cost of producing a megawatt hour (MWh) over a plant’s life, before the cost of integrating it into the grid.
The gap appears widest where capital is cheap and fuel is imported. For example, onshore wind in China averaged US$27/MWh, against gas-fired LCOE of US$50 to 60/MWh where gas is cheap. To be sure, this is the current extreme example of renewable costs that are favorable relative to fossil fuels; in a recent analysis, Thunder Said Energy shows that LCOE by region varies remarkably. In the U.S., gas and coal remain competitive with solar and wind, whereas in Europe, China and Australia, renewables are already the clear winner. The key takeaway is that in most parts of the world, economics now favor the technology that backs an aggressive energy transition.
Reason #2: Energy security aligns with renewables
For most of the world, energy security favors renewables because the vulnerability they replace is a recurring one: a fossil fuel-importing country must buy, ship and receive fuel every year, so its supply, price and routes are exposed continuously. In contrast, a solar farm, wind turbine or geothermal plant becomes a long-life domestic asset once it is built and is far less dependent on global trading systems. Since 2022, the need for domestic energy security has been proven: Russia’s invasion of Ukraine, China’s limits on critical materials in response to U.S. tariffs, and U.S. involvement in Venezuela have forced most nations to reconsider their reliance on foreign energy.
This year’s conflict in the Middle East reinforced the point and China is the clearest test of the argument so far. The country imports about 70% of its oil, roughly half of it from the Gulf, yet it has been the least disrupted of the large importers during the Strait of Hormuz crisis. After the strait closed in March, China’s crude imports fell by about a third. About 60% of the unreplaced shortfall was met from storage, but the remainder was absorbed by a real fall in the oil burned. That fall was manageable because of electrification, one of several buffers: electric vehicles accounted for more than 60% of domestic monthly vehicle sales in the first half of 2026, and China added more wind capacity in that period than in any full calendar year before 2025. The crisis has also raised demand for Chinese equipment, with solar exports reaching a new records in March 2026.
Countries will try to increase energy security according to the resources they hold; China is relatively poor in fossil fuels and rich in renewable energy inputs, so it’s easy to understand the ramp up in renewables in this context. The U.S. has relatively abundant domestic fossil resources, so there is far less urgency from a security perspective to transition to renewables. As we wrote earlier this year (please see Macro Meets Sustainability: Outlook for Q2 and Beyond, March 26, 2026), other countries less endowed will necessarily pursue diversification both among energy trading partners and energy sources; this will favor renewables relative to the status quo.
Reason #3: Storage Is Closing the Intermittency Gap
Falling battery costs have turned solar’s main physical weakness – that it produces only when the sun shines – into an economic problem that can be priced and solved. BloombergNEF reported that lithium-ion battery prices fell to a record US$108 per kilowatt hour (kWh) in 2025, 93% below 2010 levels, and that stationary storage (batteries installed at fixed sites such as solar farms and substations to store electricity for later use on the grid) fell 45% in a single year. The drop reflects intense competition in China, where annual stationary cell production capacity is roughly double global demand. In other words, there is a strong likelihood that China can flood the market with cheap batteries that will make the intermittency problem a thing of the past.
In fact, the effect is already visible on grids: Ember estimates that solar plus batteries met more than a quarter of California’s evening peak demand on the average day in the first half of 2026, up from 6.8% in 2023, while Texas batteries supplied a record 20% of evening demand in March.
The investment conclusion is that storage turns the intermittency problem from a question of physics into a question of cost. The cost is falling quickly, making renewables + battery storage competitive with gas in a growing number of markets.
Reason #4: Policy Can Solve The Remaining Hurdle
Policy still shapes the pace of the energy transition, but the problem policy needs to solve is no longer the prohibitive costs of renewables, but rather the administrative process. The IEA reports that more than 2,500 GW of renewable, large-load and storage projects are stalled in grid connection queues worldwide, that grid investment lags far behind generation, and that congestion-related curtailment is already rising. As we previously argued (Power Hungry: The Real Cost of a Digital Future, October 23, 2025), large power projects can take five to ten years to permit and build, against about two years for a data center. Europe is on a similar path, with the European Commission identifying slow permitting as a binding constraint and aiming to adopt the European Grids Package by the end of the year.
The NOAA, the key U.S. weather forecaster, gives an 83% chance that El Niño reaches historic strength in October to December, exceeding any event since 1950, and likely persisting into spring 2027, with effects on crops and supply chains running into 2028. Weaker trade winds will let warm water spread east across the Pacific and shift rainfall: drought in Southern Africa, Southeast Asia and Australia, heavy rain on the Peruvian coast, and wind shear that suppresses Atlantic hurricanes.
El Niño is a natural cycle, but climate change raises the cost of each event by amplifying heat, evaporation and rainfall extremes, a pattern that underpins our view that weather is now a lasting driver of food price inflation. This year’s event is already disrupting industries such as fisheries, while for others the shocks will play out later. The table below illustrates when various industries are likely to be affected: we offer three trade ideas that leverage a strong El Niño effect.
Trade #1: Marine Feed: The Visible Shock
One of the first major disruptions from this year’s El Niño to occur is in marine feed. The term marine feed refers to the ingredients derived from marine organisms used in animal feed, particularly for aquaculture (e.g. fishmeal, fish oil and algae). It is produced mainly from small pelagic fish like anchovy, sardine, and herring.
On October 16th, Peru cancelled its second anchovy season, after the first season was stopped in June with only about a quarter of its quota caught. Peru supplies about 20% of global fishmeal and fish oil, and IFFO (The Marine Ingredients Organization) projects 2026 fishmeal supply at its lowest since 2016. According to the IMF, fishmeal prices increased 89% in the past year. Oceanographic conditions were the culprit behind the closure: warm waters caused by El Niño have changed anchovy distribution and forced schools into deeper water.
Fishmeal and fish oil are globally traded, so the shortfall raises feed costs for fish farmers everywhere, but not equally. In the case of Mowi (MOWI-NO), the company manufactures its own feed for its own seafood farms, so Mowi retains the feed margin and has procurement scale and forward cover, whereas Lerøy (LSG-NO) and most other producers buy feed from external manufacturers with no retained margin. Share prices among salmon farmers are highly correlated, but El Niño opens a window for Mowi to outperform in 2027, which we would capture with a pair trade, long Mowi against short Lerøy, or simply by overweighting Mowi.
- Mowi and its competitors both attribute rising feed costs to weak marine fisheries, and Lerøy names El Niño explicitly, expecting the effect in the fourth quarter of 2026 and mostly in 2027. Its third-quarter harvest fell 7% year on year.
- Mowi’s feed is mainly covered through the end of the growth season, and its October 19 third-quarter update beat estimates a record harvest. That result reflects salmon prices and volumes before feed inflation arrived, so it shows operational strength, not yet the feed advantage.
- Mowi ranks first in the FAIRR Protein Producer Index (57) and the World Benchmarking Alliance’s Seafood Stewardship Index also rates Mowi a leader.
Trade #2: Copper in El Niño’s path?
El Niño matters for copper because supply sits in its path. Countries like Chile, Peru, Zambia and Indonesia are known to experience strong El Niño effects and together mine about 42% of the world’s copper: Chile about a quarter, Peru about an eighth, Zambia about 4% and Indonesia about 3%.
The mechanism and timing differ by region. In South America (Chile and Peru), excess rainfall, flooding, landslides, and port closures directly halt or slow mine operations and logistics. Meanwhile in southern Africa the rains fail, the Zambezi basin dries and the hydropower grid feeding the Copperbelt has less water. Similarly, Indonesia is facing severe drought due to El Niño, which perhaps surprisingly, is actually helping Freeport-McMoRan’s repair its Grasberg mine after massive mud slides last year.
The El Niño effect for 2026 in Chile has already largely occurred and total output losses have been relatively subdued. Companies like Antofagasta (ANTO-GB) and Lundin Mining (LUN-CA) were forced to halt production in July after heavy snow and rain and power outages. Both companies slightly cut 2026 production guidance and the Chilean Copper Commission, cut its national 2026 forecast to 5.27 million tonnes, a 2.6% decline from 2025.
In Africa, the impact on copper mining is still ahead of us. In the African Copperbelt, significant production depends on Kariba lake hydropower, which is filled by rain in the upper Zambezi basin and drawn down through the dry season. A severe drought would mean the impact on copper would be felt starting in March, 2027 should lack of hydroelectric power affect mine production. Many companies do have contingency plans in place. For example, First Quantum (FM-CA) has the ability to import much of its power needs from neighbouring companies, albeit at higher prices which add to mining cash costs.
In sum, El Niño is a short-term supply risk for copper. The Chilean disruption has largely played out, while Zambian hydropower risks from March 2027 are still ahead. Investors who want to reduce geopolitical and carbon risk can own copper recyclers rather than miners like Aurubis (NDA-DE), please see Dig, Squeeze and Recycle: Rewiring the Copper Market July 17, 2025. Alternatively, investors can increase exposure to miners that have limited or no operations in Africa such as Freeport McMoRan (FCX-US), Lundin Mining (LUN-CA); they will benefit from a global rise in copper prices while avoiding production challenges related to the remaining Niño weather impacts.
Trade #3: A Palm Oil Squeeze
Palm oil has one of the strongest documented links to El Niño, and its biology means the damage arrives late. Indonesia and Malaysia grow roughly 80% of the world’s palm oil, a crop that provides about a third of global vegetable oil. Palm is a perennial that cannot be replanted in response to a drought: water stress cuts the share of fruit-bearing flowers about two years before harvest and causes flowers to abort in the final months, so a dry spell depresses output across two seasons rather than one.
August rainfall across the palm belt was 53% below normal, and 2027 output estimates are down 4.0% in Malaysia and 3.0% in Indonesia. This is coming at a time when Indonesia’s new B50 biodiesel mandate adds a further claim on supply, with an estimated 35% of Indonesian output about to go to fuel. We therefore expect tightening of palm oil supplies, though it is hard to trade directly because parties that bear the loss are mostly outside our investable universe: smallholders with no price hedge, standalone mills squeezed on fruit costs, and low-income households in importing countries such as India, Pakistan and Bangladesh. Nonetheless, the El Niño effect on palm oil (and other ag products) supports our house view that climate change and extreme weather is a structural driver of sustained food price inflation, and we have written extensively about the resulting margin squeeze for food manufacturers (please see Weather, Not War: Why Higher Food Inflation is Here to Stay October 15, 2026). Palm is a smaller input than cocoa, sugar and dairy for companies like Mondelez (MDLZ-US) and Hershey (HSY-US), but it is hard to replace in confectionery and bakery fats and it adds to an ingredient basket where costs are already rising. Both companies also face deforestation and sourcing scrutiny, a separate regulatory and reputational exposure. We maintain our negative stance.
If any investor has the scale and resources to build adaptation exposure deliberately, it is Norges Bank Investment Management (NBIM). The fund manages about NOK 22,700bn, holds 7,077 companies in 54 countries and owns an average of 1.2% of each. Its returns therefore rise and fall with the global economy and climate rather than with any single company. Like many peers, NBIM has raised climate issues with companies since 2006, and its mandate has treated climate change as an investment risk since 2022. It is also a useful case study because it publishes its holdings, allowing us to measure what it owns rather than estimate its exposure. NBIM’s 2030 Climate Action Plan shows how a leading owner approaches adaptation. One priority is a stronger focus on nature, physical climate risk, adaptation and resilience. The fund stress-tests its equity portfolio against extreme climate scenarios and conducts physical-risk analysis of real assets. It has also added government bonds to its analysis and created a climate focus list that includes physically vulnerable companies while bringing adaptation more explicitly into engagement meetings. Even so, many of these actions remain geared toward climate risk management and mitigation, with limited detail on capital allocated directly to adaptation. Many institutional climate strategies have the same limitation because they were built around decarbonisation targets and stewardship rather than adaptation.
In past notes, ‘Climate Change and Strategic Asset Allocation‘ (August 24, 2022), we argued that climate belongs in long-term capital market assumptions, with inflation, especially food inflation, the main channel and a case for more inflation-protected and natural-resource assets. Moreover, our ‘Weatherproofing Portfolios: Resilience as the New Alpha‘ (September 11, 2025) and ‘Adaptation Investing: Survival of the Funded‘ (May 7, 2026) notes showed private adaptation finance at only about 0.5% of climate finance, kept low because adaptation pays in avoided losses, not cash flows. Today’s note looks at how to take climate adaptation into SAA a step further.
How we measure exposure: the SMS Climate Adaptation Index
To measure exposure to climate adaptation, we use the SMS Climate Adaptation Index, a rules-based thematic index built in four steps. We first define five drivers that frame the theme: water and flood resilience, urban and built environment resilience, grid and energy resilience, agriculture adaptation, and climate risk intelligence and insurance. We then screen a developed-market all-cap universe for companies whose reported revenues align with products and services tied to those drivers, using revenue data classified into granular business activities rather than broad sector codes. Our research team adds a bottom-up review where disclosure is thin, and the final portfolio applies controversy and other responsible-investment screens, then weights companies by free-float market cap with a 5% cap per name. The index holds 72 companies, 71 of which are priced: 15 in water and flood, 18 in urban and built environment, 14 in grid and energy, 12 in agriculture and 13 in climate risk intelligence and insurance. The index measures what a company sells, not how it describes itself, which gives peers a transparent yardstick.
NBIM’s 1.3% of NAV is low only against a deliberate allocation, not against the market. For its part, the SMS index covers about 1.6% of the free-float value of the MSCI ACWI IMI, and NBIM holds it at 1.8% of its equities, at or slightly above market weight and more than twice its unlisted renewable infrastructure (0.5% of NAV). The gap is therefore one of intent: a benchmark-hugging fund owns the theme in proportion to its small market size, so only an explicit mandate can raise it. Looking at recent performance, the SMS Climate Adaptation Index has tracked the MSCI World (URTH) closely in recent months.
Against the SMS index, NBIM holds 62 of the 71 priced names, worth about NOK 290bn, or 1.3% of the fund, so the companies are in the portfolio and the exposure is incidental. Market-cap weights put about 40% of that exposure into four industrial giants and 46% into grid and energy, against 31% in the SMS index, while water and flood account for 4% against 10%. A peer with a global market-cap equity book should expect a similar tilt.
What peers stand to gain: a better-diversified equity book, not a storm shelter
Measured against a replica of NBIM’s equity mix, the SMS index adds breadth, inflation sensitivity and a small-cap tilt, while giving up low volatility and commodity exposure (see figure below). Those loadings come mostly from the companies rather than the weighting. For a pension plan with inflation-linked liabilities, allocating to an index such as the SMS Climate Adaptation Index could broaden and diversify the fund’s equity exposure. In 2022, when equities and bonds fell together, the index lost far less than global equities, although that year sits within a hindsight-biased sample.
The table shows what a move toward higher targets would have produced over ten years. The result hinges on how much of the index’s excess return persists, and the record is a flattering, hindsight-biased sample led by Quanta Services (PWR-US), which gained more than 2,000%. At 5% of assets, the added return ranges from 0.16 pp a year if none of the SMS Climate Adaptation Index’s excess return persists in the future to 0.27 pp at full persistence, or roughly USD 160m to 270m a year per USD 100bn before costs. The benefit shrinks toward zero if the excess return vanishes.
The climate adaptation index will not shelter a portfolio from a potential financial storm. Its downside beta is above 1, and it has no significant loading on agriculture, duration or commodities, the channels used in NBIM’s extreme-weather scenario. A larger allocation would therefore lose slightly more than equities in that shock. What it offers is diversification and a modest return improvement, paid for in tracking error, which for a fund on a tight risk budget is the binding constraint.
Copper hit an intraday record of US$14,875 per tonne on September 10th, as planned U.S. tariff increases prompted hoarding across the global supply chain. In real terms, prices remain close to a 25-year high. Although recent demand has benefited from data centres riding the AI boom, AI still accounts for less than 5% of global copper consumption. Prices could therefore climb further: U.S. growth remains strong – the Atlanta Fed’s GDPNow estimate stands at 3.7% year on year – and demand from EVs, construction and defence shows little sign of easing.
Furthermore, military rearmament has emerged as a genuine, additive driver of copper demand, layering on top of the electrification and AI-driven grid buildout already straining supply. Industry estimates put the current incremental demand from rising military spending at roughly 500,000 tonnes per year, about 1.5% of global consumption.
Global copper inventories have nevertheless risen recently, even as declining ore grades and production difficulties constrain mine supply and deepen bottlenecks. Pressure on prices are therefore likely to persist.
Meanwhile, global copper demand is forecast to grow at roughly twice the pace of supply over the next decade. That imbalance underscores the market’s strength, even with prices already elevated by historical standards.
Can copper be replaced?
Substitution is a genuine but limited risk. Aluminum is the only viable alternative, yet its conductivity is roughly 61% that of copper. It is also only one-third as dense, making it attractive where weight matters. That is why aluminum already dominates high-voltage transmission and is gaining share in EV wiring harnesses, particularly with copper trading at about 4.2 times its price.
Industry estimates put aluminum substitution at approximately 2% of global copper demand in 2026, rising to about 6% by 2030 and concentrated almost entirely in low-voltage automotive circuits. EV battery busbars, where current density and thermal reliability are critical, remain roughly 85% copper. Traction-motor windings have seen almost no substitution because bulkier aluminum windings would sacrifice the power density manufacturers are working to preserve.
More importantly, the largest incremental source of demand – AI-driven data centre and grid construction – is largely immune to substitution. Copper represents less than 0.5% of total data centre project costs, leaving hyperscale developers relatively indifferent to its price when specifying grid connections. Other alternatives, including silver, graphene and superconducting materials, are either too expensive at scale or commercially immature. Copper will remain difficult to displace for the foreseeable future.
Copper: the case for a bull market
First, both copper and copper miners have outperformed the S&P 500 so far this year.
Second, market participants have remained net long copper for the past two years, and current positioning points to an increasingly bullish stance. This suggests that asset managers and speculators are still invested in the market and expect prices to remain elevated for some time.
Against this backdrop, diversified miners have outperformed pure-play copper producers year to date even if enthusiasm around AI has cooled, leading by nearly ten percentage points. In our May 26, 2026, note, “From Transition Play to Arms Race: A Critical Materials Reset”, we argued for pure-play copper companies over diversified groups within the broader critical-materials universe.
That call has yet to pay off. Much of the AI enthusiasm was already embedded in leading copper miners, and investors are now questioning whether valuations across AI – and the associated data centre build-out – can be sustained, even for indirect beneficiaries such as the copper “complex”. Even so, copper stocks still look relatively inexpensive: their average price-to-earnings ratio remains below that of the S&P 500, which trades at 26 times earnings.
Copper recyclers such as Aurubis, however, retain solid fundamentals and trade at attractive valuations within the “Copper Kings” universe. They have lagged diversified miners, some of which have risen by 1.5 to 2 times over the past year. Higher copper prices also bring more scrap to market, giving recyclers easier access to feedstock than miners confronting steadily declining ore grades. We therefore continue to prefer recyclers to diversified copper miners.
Four years ago, we argued that nuclear was coming back (please see Back to the Future: A Nuclear Revival, 9 June 2022). The thesis was, and still is, that nuclear is a necessary source of firm, low-carbon electricity in a system trying to exit coal and gas. The investments we recommended were uranium miners in geopolitically safer jurisdictions, above all Canada (CCO-CA), and small modular reactor vendors such as NuScale Power SMR-US, GE (now GEV-US), Hitachi 6501-JP and Rolls Royce (RR-GB).
The mining leg has played out as expected. The SMR contribution to the electricity grid has not. Stock-market performance reflects that split: CCO-CA and CEG-US have rerated sharply over four years, while NuScale, the listed name most levered to an SMR buildout, has not. Without that buildout, nuclear’s share of global electricity has declined from about 10% in 2022 to under 9% today. Today’s note examines what has changed in the outlook and where we would (and would not) put money between now and 2030.
The scorecard
Nuclear remains the principal widely deployed, low-carbon source of firm power that is not constrained by river geography or weather. That was true in 2022, and it is true now. Our 2022 expectation was that by the second half of this decade, high fossil-fuel prices, regulation and national-security concerns would make new reactors financeable relative to other sources, and that SMRs would be near ready for operation.
Oil prices were not uniformly high. However, the other supports (policy, security of supply, and now AI-driven electricity demand) have largely been in place; the long-run case for nuclear is intact. What did not happen is the SMR buildout. The most closely watched projects in North America were cancelled or delayed:
- NuScale’s Carbon Free Power Project with UAMPS was cancelled in November 2023 after estimated project cost rose from about $5 billion to about $9 billion on inflation and higher rates.
- The first North American grid-scale SMR actually under construction, GE Vernova Hitachi’s BWRX-300 at Ontario Power Generation’s Darlington site, is now aimed at commercial operation around 2030, later than first discussed.
Factory fabrication and modular add-on capacity remain the theoretical attractions. In practice, most of the 80-plus SMR designs are still at conceptual or detailed design. Only two SMR plants are in operation: Russia’s Akademik Lomonosov (floating) and China’s HTR-PM, which entered commercial operation in December 2023. A handful more are under construction in Argentina, China and Russia. The technology is still largely untested at commercial scale in North America.
So what happened?
Global electricity demand rose on the order of 3% in 2025, faster than total energy demand, with AI and data centres a visible incremental driver. Hyperscalers are demanding power as fast as it can be added to the grid. Earlier this summer we wrote why natural gas plus renewables (solar and batteries) is the current go-to for new U.S. data-centre load (please see July 2026, Data Centers: After the Gas Rush). SMRs are a long-term firm-power option but can’t address the needs today.
The four largest U.S. hyperscalers (GOOGL-US, AMZN-US, META-US, MSFT-US) have signed nuclear offtake. According to a recent Carnegie report, hyperscalers generally prefer buying needed energy through power purchase agreements (PPAs), which has resulted in restarts, life extensions, and uprates of existing plants. Thus far, their interest has not translated into the financial backing that would be required to generate an “order book” to build a fleet of new reactors.
The Bulletin of the Atomic Scientists’ July 2026 accounting is blunter: nuclear has played no meaningful role in meeting the AI load already on the grid, and many SMR announcements are feasibility funding and options, not engineering, procurement and construction cheques. Amazon’s plan with Energy Northwest in Washington is a case in point. Amazon has funded the initial feasibility phase and secured the right to purchase power from the first project (four modules), with no obligation to take that power if it is too expensive. Delivery is “some time in the 2030s”; no construction permit has been filed.
Google’s partnership with privately held Kairos Power in Tennessee is likely to be the first of these advanced-reactor data-centre deals to produce power, aimed at 2030. The project broke ground in April 2026. It is a 50 MW demonstration plant selling into the TVA system, not the 500 MW fleet originally advertised.
SMRs are therefore unlikely to fill hyperscaler demand this decade. The nuclear equity story through 2030 is primarily restarts, uprates and life extensions of existing plants, plus the fuel those plants burn (large new reactors continue to be built in Asia, which supports uranium demand even if Western SMRs slip). With that backdrop, we update the listed opportunities.
Investing in Nuclear: Today To 2030
Despite a slower timeline for small modular reactors, nuclear will remain part of the energy mix. In the section below we walk through the listed names in the nuclear chain, from the mine to the electricity that comes out of the plant, and where we would (and would not) put money between now and 2030.
The physical chain is simple. Uranium starts as ore, is milled to yellowcake, converted to UF6 so U-235 can be concentrated, then returned to ceramic pellets in fuel assemblies. Most listed stocks sit at the two ends, the mine and the power1. Mining and milling steps matter but are hard to buy as public equities.
Mining and milling. Our preferred uranium miner since 2022 has been CCO-CA, because most of its uranium is sold under long-term contracts and because the mines are in Canada. KAP-GB is the larger competitor, but Kazakhstan is a harder jurisdiction for an ESG-constrained book.
Alongside these established companies, there are four developers with potential to supply more uranium in the future: NexGen (NXE-CA), Denison (DML-CA), Uranium Energy (UEC-US) and Paladin (PDN-AU). The first two are not yet producing uranium, while UEC is in early production in the U.S. and PDN is producing in Namibia and developing assets in Canada.
Conversion. Only a handful of plants can convert yellowcake to UF6. Cameco operates one of them, while the other large Western converter, France’s Orano, is unlisted.
Enrichment. About 95% of world capacity sits with private companies. Centrus Energy (LEU-US) is the listed US exception.
Fuel fabrication. Utilities do not switch suppliers lightly, so this is a stable business. Incumbents are mostly private or inside larger groups. Westinghouse (Cameco and Brookfield) makes fuel for the common US reactor type. GE Vernova (GEV-US) and Hitachi (6501-JP) make fuel for boiling-water reactors. BWX Technologies (BWXT-US) makes fuel for the U.S. Navy and some advanced designs.
Designing and building new reactors. This layer is paid when a plant is actually built. GE Vernova and Hitachi are building the BWRX-300 at Darlington, aimed at 2030. Mitsubishi (7011-JP) and Doosan Enerbility (034020-KR) are large-reactor hardware in Asia. BWX Technologies (BWXT-US) and Curtiss-Wright (CW-US) sell into plants that already run. NuScale (SMR-US), Oklo (OKLO-US) and Rolls Royce (RR-GB) are the main listed SMR expressions.
Construction contractors. Once a reactor is financed: Jacobs (J-US), Fluor (FLR-US), Atkins Realis (ATRL-CA), Aecon (ARE-CA) and Samsung (028260-KR). Bechtel, often the lead, is private.
Generating electricity. In North America, since new plants are not being built, the main activity is restarting previously decommissioned plants, improving output from current plants and licence extensions.
The specialist is Constellation Energy (CEG-US). Microsoft is paying for the restart of Three Mile Island Unit 1; Meta’s contract keeps Clinton in Illinois running. Vistra (VST-US) is in the same business and is also a gas generator. Talen Energy (TLN-US) has a PPA with Amazon for Pennsylvania data centres.
Outside the US, longer operation of existing fleets is the same idea: KEPCO (015760-KR), Kansai Electric (9503-JP), CEZ (CEZ-CZ Fortum (FORTUM-FI). CGN Power (1816-H) is the China large-fleet name, with the usual ESG-jurisdiction caveat.
Waste and decommissioning. Used fuel is stored, usually on site; reactors are eventually taken apart. There is no satisfactory listed way to own the waste step. The liability stays with the generator. Cloud companies that enter power purchase agreements do not take the long-term environmental or community risk, nor the nuclear-safety residual.
Given the value chain and our view that SMRs will not be a material 2026–30 solution, the best listed exposure is at steps 1 (mining) and 7 (generating). Cameco is the only listed name that reaches from the mine through conversion and, via Westinghouse, into fuel and plant services. Centrus and BWXT are small ways to own the scarce middle. Constellation, and to a lesser extent Vistra, have customers for any extra megawatt-hour they can extract from plants that already exist.
As of mid-August, Canada was on track for its fourth-worst forest fire season. The U.S. is faring little better, with national acreage burnt running at roughly double the ten-year average early in the season and air quality alerts active across more than a dozen states. France, Spain, and Portugal have all experienced worse than average wildfire seasons in 2026 and the consensus is now to expect increasingly devastating forest fire seasons as climate change progresses. Indeed, data from Global Nature Watch confirms that forest fires are becoming more widespread and destructive around the globe; annual forest fires now burn more than twice as much tree cover as in the early 2000s.
Why This Particle is Different
Forest fire smoke is not a diluted version of urban air pollution, and treating the two as interchangeable understates the health burden. Toxicological and epidemiological evidence indicates wildfire particulate is more harmful per unit mass than PM2.5 from other sources (PM2.5 is fine particulate matter of 2.5 micrometres or smaller, able to penetrate deep into the lungs and bloodstream).
Recent studies, using Southern California data, found that a 10 µg/m³ increase in forest fire-specific PM2.5 was associated with a 1.3% to 10% rise in respiratory hospitalisations, compared with 0.67% to 1.3% for the same increment of non-forest fire PM2.5. The per-unit-mass effect is on the order of ten times larger. One implication is that mass-based Air Quality Indexes systematically understate the health consequence of a smoke episode.
This matters because forest fire pollution sits largely outside the regulatory architecture that created the last generation of air-quality capex. In the United States, under the Clean Air Act’s Exceptional Events Rule, state and local air quality authorities may petition the EPA to exclude data collected during forest fire smoke episodes. Forest fire-driven exceedances can be struck from the record, so that areas affected by smoke are not flagged as failing national ambient air quality standards, even when concentrations were, in fact, over the limit.
An important corollary is that, unlike stack and tailpipe pollution, ambient forest fire smoke does not benefit from the Porter Hypothesis. Strict environmental regulations can force innovation and the introduction of cleaner technologies. As we wrote previously (please see June 1, 2023 note “Clearing the Air”), pollutants such as sulphur dioxide and nitrogen oxides come out of identifiable stacks and tailpipes owned by identifiable regulated entities, so exceedances translate into permit conditions, enforcement actions, and compliance deadlines. That legal obligation is what converts an environmental problem into capital expenditure, and it is what built the addressable markets for flue-gas desulphurisation, selective catalytic reduction, and diesel aftertreatment, which we examined through companies such as Alfa Laval (ALFA-SE), Cummins (CMI-US), and Johnson Matthey (JMAT-GB).
For forest fire smoke, that mechanism is missing. There is no compliance-driven capital expenditure and no obvious “environmental winner” analogous to the scrubber cycle: ambient smoke has no permit holder, so the same physical exposure produces no mandated spending on abatement. Ignition is a different matter: where a utility, a railroad, or a landowner can be shown to have started the fire, liability and wildfire-mitigation rules do create a defendant and a capex cycle. That distinction is the difference between the filtration discussion below and the grid-hardening discussion that follows.
What Can Investors Do
Unlike industrial air pollution, which created compliance obligations, forced capital expenditure, and therefore investable demand for control equipment, wildfire smoke creates health problems at scale with little regulatory consequence. Governments tend to treat wildfire pollution as an act of nature to be excluded from the record. Investors looking for the equivalent of the next scrubber cycle will not find it here.
That does not mean wildfires are without investment implications. We identify three investable sectors affected by more frequent forest fires in Western countries. We do not address voluntary carbon credit markets; we will do so in a future note.
1. Reinsurance repricing
For most reinsurers, wildfire remains classified as a secondary, or non-peak, peril. The economics no longer match that label. According to Swiss Re, North American wildfire insured losses have been growing at more than 14% per year, among the faster-growing catastrophe loss lines.
The January 2025 California fires showed that wildfire can materially affect reinsurer profitability. Renaissance Re (RNR-US) is a pure-play P&C reinsurer; the stock sold off sharply in the first quarter of 2025 as the company reported a large underwriting loss.
To be sure, not all wildfire seasons produce such large stock price declines. Stock prices of reinsurers continued to climb through the 2026 season because fires in Canada, France, and Spain largely avoided dense urban areas unlike California in 2025.With RNR as the notable exception, reinsurers tend to have diversified businesses (see the table below).
Another risk to reinsurers is that they, and the primary companies they support, will be increasingly compelled to hold exposure they would rather shed. In California, the Department of Insurance’s Sustainable Insurance Strategy has begun conditioning rate approvals on commitments to write in wildfire-distressed areas.
The implication is that forest fire is migrating, in economic if not in formal labelling terms, from a secondary to a primary risk. Several large P&C reinsurers have outperformed the S&P 500 over the past five years. A repricing of forest fire risk does not mean the end of that bull market, but it does mean underwriting results and California take-up commitments now deserve more weight in the thesis than they did when wildfire losses were considered a rounding error.
2. Utility wildfire liability and grid hardening
Earlier this month, Los Angeles County fire officials confirmed that Southern California Edison (EIX-US), the primary electric utility for much of Southern California, ignited the 2025 Eaton fire. Electrical arcing on an idle, out-of-service transmission tower sent hot metal into dry vegetation, killing 19 people and destroying more than 9,000 homes and businesses. Over the past decade, Edison has paid out billions in claims related to equipment-sparked fires. Utilities are a relatively common ignition source in North America. In 2019, Pacific Gas and Electric filed for bankruptcy after its equipment ignited wildfires that killed more than 100 people and destroyed entire towns.
These cases illustrate the need for grid hardening: upgrading and reinforcing the electrical system against wildfire ignition and against other physical disruptions. Budgets for wildfire-related grid hardening in California alone are already in the billions, large enough to sustain contractor backlogs at firms such as Quanta Services (PWR-US) and MasTec (MTZ-US), whose businesses include undergrounding, vegetation management, and system reinforcement for utilities.
Both stocks have suffered major sell-offs this summer after a strong prior twelve months. That decline should not be read as the market rejecting the grid-hardening thesis. Trailing earnings multiples remain elevated relative to history, so a correction was overdue.
3. Filtration and indoor air quality
Because ambient smoke has no permit holder, the only private response is to filter the air after it has already entered buildings. Demand for air filtration products (furnace filters and portable HEPA units) is visible in retail spikes during forest fire episodes. However, back of the envelope calculations suggest the total addressable market is $0.3–0.6 billion of forest fire-attributable North American revenue in a particularly smoky year. That is paltry compared with California wildfire-mitigation and undergrounding budgets, which are already in the billions per year and are mandated. For investors, gaining exposure to the indoor air filtration market is not straightforward. Category leaders that tend to capture the spike in sales are mostly private, while air purifying products make up only a small portion of revenues of publicly listed companies that sell them.
A core principle of economics is that scarcity drives price. But several environmental assets are currently running counter to this principle. As we wrote in Environmental Assets in Short Supply (October 9, 2025), “many environmental assets have not re-priced in times of scarcity or are simply not priced at all.” We have tracked environmental scarcity across seven distinct resources since 2018:
- Critical minerals: What began in 2018 as a governance and supply chain concern about cobalt sourcing from the Democratic Republic of Congo has evolved into a geopolitical flashpoint across several materials that are essential for national defense, artificial intelligence and green technology. In our most recent note on the topic, we outline why inputs that were thought to be in relative abundance are now turning into an outright materials arms race (From Transition Play to Arms Race, May 28, 2026).
- Water: Water scarcity arises when reliable supplies of clean freshwater are insufficient to meet the needs of people, ecosystems, agriculture, and industry. It is driven not only by limited physical availability, but also by overuse, pollution, uneven distribution, and climate change. In recent months, investor attention has zeroed in on water needs derived from AI data centers (an issue we view as solvable and highly investable) especially in times of drought. See Water Woes: Ag, Not AI, April 16, 2026.
- Marine protein: Fish is the dominant protein source for many parts of the world and oceans have finite productive capacity; fish stocks are being depleted faster than ecosystems can replenish them. A 2025 FAO assessment of 2,570 marine fish stocks found that over one third are overfished, i.e. are harvested beyond biologically sustainable levels.
- Phosphorus: Investor attention to fertilizer scarcity was prompted by Russia’s invasion of Ukraine and the resulting sanctions on Belarusian potash, though underlying vulnerabilities have been present for over a decade. Indeed, phosphorus is a non-renewable, non-substitutable material with no synthetic alternative. Although the world has over 300 years of reserves, China is currently the largest producer, but Morocco and Western Sahara hold roughly 70% of global reserves, meaning long-run supply security rests on a single-country dependency rather than absolute physical scarcity.
- Soil: We have been tracking soil as an undervalued ecosystem since 2022 whose productive capacity is being eroded by short-term farming economics. Erosion, salinization, compaction, nutrient depletion, and land conversion all degrade soil far faster than it can be naturally replenished, making fertile soil a finite environmental asset. There is evidence that the premium paid for land with more productive soil is growing, suggesting that scarcity is beginning to be factored into price (please see our note Environmental Assets in Short Supply, October 9, 2025).
- Sand: Sand is a resource that is assumed to be infinite. However, in its usable construction-grade form, it is being depleted faster than natural replenishment. Prices are beginning to reflect this reality: in geographies that rely on natural sand for construction, prices reflect structural shortages, with inflation for construction-grade sand running far hotter than general price inflation.
- Timber: Primary forest loss runs at about 10 million hectares per year, mostly due to agricultural conversion. Old-growth carbon stocks and hydrological function cannot be replicated by plantation forestry on any relevant timescale. Of course, wildfires like this summer only serve to amplify the problem.
Three Investment Strategy Pillars
For each of the resources described above, our research has implicitly relied on the same three pillars to find investable equity opportunities within environmental assets in short supply.
- Beneficiaries are the owners of the scarce asset or the infrastructure controlling access to it: water utilities, phosphate miners, timber REITs, aquaculture operators, etc. are all beneficiaries of natural resources that are in short supply since they benefit directly from any increase in supply-constrained pricing of the underlying good. Companies in this category are most often materials, utilities and REITs sectors.
- Solvers are companies that reduce consumption, improve efficiency, or substitute away from the constrained input: desalination technology, manufactured sand, precision agriculture, battery chemistry that eliminates cobalt and nickel demand entirely are examples of industries that are solving for constrained natural resources. Companies in this category tend to be heavily weighted to the industrials and technology sectors.
- Protectors are companies that enable monitoring, certification, or active restoration of the resource base itself. When resources are constrained, these companies often benefit from regulation that is meant to ensure a more sustainable usage of resources in short supply.
We mapped our equity coverage against this framework of beneficiary, solver and protector. The table below highlights sample companies with high revenue alignment to each of the scarce environmental assets and pillars.
The result is a diversified universe, albeit one that is weighted toward Beneficiaries. This is not a surprise: investors can easily buy a phosphate miner or a timber REIT when they expect the price of the underlying commodity to increase. However, it is more difficult to buy exposure to wetland restoration, aquifer recharge, or pollinator habitat protection, since financial vehicles to do so are not prevalent. Indeed, the Protector role has the lightest coverage of our three pillars: where restoration and conservation activity exists, it is overwhelmingly private, philanthropic or embedded as a minor revenue line inside diversified conglomerates (certification bodies, testing labs, satellite monitoring firms) rather than expressed through dedicated public companies.
Why This Matters Now
We believe that two developments in the past eighteen months – the buildout of AI infrastructure and the increase in global trade barriers – have made it more relevant for investors to consider the supply-side of physical inputs within equity portfolio construction.
Regarding artificial intelligence, AI data center buildout is simultaneously accelerating demand for water (cooling load), copper (grid and data center wiring) and critical material (rare earths used in photonic chips for AI data centers and 6G networks). The AI buildout comes on top of what were already separate scarcity stories with independent underlying drivers. Take water. Agriculture usage and more frequent droughts had already been pushing pressure on utilities and industrial companies to improve water efficiency as articulated clearly by executives at Schneider Electric. Water cooling demands from AI act to accentuate these forces and it’s happening across other environmental inputs like rare earths, copper, etc.
Meanwhile, supply side constraints are also occurring due to geopolitics. A less friendly global trade backdrop means that both raw inputs and finished goods flow less freely at a time when greater raw materials are needed to build domestic capacity (reshore). China’s April 2025 rare earth export restrictions provided a glimpse of the potential for supply chain disruption in strategic sectors like defense and technological infrastructure.
The Paradox at the Heart of the Grid
AI is both a source of the problem and the most credible solution to it. Global data center electricity consumption reached approximately 415 TWh in 2024, growing at 12% per year since 2017, and the IEA projects this figure could reach up to 1,200 TWh by 2030. A single AI-focused data center today consumes as much electricity as 100,000 households. At the same time, the electricity grid is increasingly unreliable: the North American Electric Reliability Corporation (NERC) flagged rising shortfall risk under both extreme and normal conditions in its latest Long-Term Reliability Assessment, and grid outages already cost U.S. businesses roughly $150 billion annually.
The irony is that the same computing paradigm generating this demand surge is also the most capable tool available to manage it. The question for investors is not whether AI will change how electricity is produced, distributed, and consumed. It will. The question is which companies in the stack are positioned to capture durable, recurring revenue from that shift.
Two Domains, Two Mechanisms
Grid Optimization
AI’s most direct contribution to grid efficiency operates at the level of real-time sensing, forecasting, and control. Smart meters and IoT sensors generate continuous operational data; AI algorithms process that data to build a live picture of grid state and issue control signals in milliseconds, far faster than human operators can respond.
The IEA’s 2025 “Energy and AI” report identifies AI-based fault detection as one of the highest-impact near-term applications, with documented reductions in outage durations of 30-50%. Given the $150 billion annual cost of outages, a 30% improvement represents tens of billions in recoverable economic value, even before accounting for carbon savings from avoided backup generation.
Virtual Power Plants (VPPs) represent the next frontier. These systems aggregate small-scale distributed resources, including behind-the-meter solar, batteries, demand-response loads, and EV chargers, and coordinate them through AI to respond to grid signals in real time. The U.S. Department of Energy estimates that tripling VPP scale in America alone could reduce grid costs by $10 billion per year, primarily by deferring expensive peak capacity additions. As the share of intermittent renewables grows, the economic case for AI-driven aggregation and dispatch becomes progressively stronger.
Industrial Process Optimization
Industry and commercial buildings together account for roughly 64% of U.S. electricity demand, making them the largest addressable market for AI efficiency tools. Here, the applications are more granular but no less material. We see 4 key industrial sectors where AI could markedly optimize production and reduce energy costs:
Steel: Steel is an instructive case. ArcelorMittal (MT-US) partnered with energy AI specialist Energiency to deploy machine learning models that predict and control energy consumption continuously across its smelting operations, achieving 2-5% annual energy savings on an energy bill that represents roughly one quarter of total production costs. A separate deployment using C3.ai’s energy management platform at a different major steel producer generated $14 million in annual energy cost savings at a single hot roll mill, alongside a 40 MW reduction in monthly utility demand charges.
Cement : Cement, which another thermodynamically constrained industry, can also achieve more efficiency using AI. A 2026 peer-reviewed analysis found 3-9% net energy and CO2 reductions from AI-optimized kiln controls, after explicitly subtracting the energy cost of running the AI itself. For an industry running near thermodynamic limits, 5% is financially significant and competes favorably with far more expensive physical decarbonization options.
Cooling : For cooling and facility management, the IEA independently cites a 23% cooling energy saving from AI-optimized controls as a representative industrial deployment benchmark. Across optimizable industrial subsystems more broadly, documented savings commonly reach 20-30% of the targeted system’s energy consumption, though total plant-level savings are typically 5-15% given that no single subsystem represents the entire load.
Chemicals: Chemical production offers one of the highest AI optimization ceilings in heavy industry. Energy costs represent 40-60% of variable production costs at specialty chemical and petrochemical facilities, and process complexity (multi-variable reactor management, distillation train sequencing, heat integration across unit operations) creates exactly the conditions where narrow ML models identify non-obvious savings that rule-based controls miss. AI-optimized process control at chemical plants has demonstrated sustained energy reductions of 3-8% on targeted units, with the strongest results on distillation columns where temperature and pressure gradients interact in ways that respond well to continuous model retraining. At industrial scale, a 5% energy reduction at a large petrochemical complex running 200-400 GWh annually translates to $3-8 million per year at current industrial power prices, with no capital expenditure on physical equipment.
A useful way to frame AI’s efficiency contribution is to ask what energy is consumed by the AI relative to what it saves. The answer varies dramatically depending on the type of AI deployed, and investors who treat all AI as equivalent in energy intensity will systematically misread this opportunity.
Industrial process optimization does not rely on large language models. It uses narrow, purpose-built machine learning models, often small enough to run on edge hardware or modest on-premise servers. A factory consuming 100 GWh per year that deploys a continuous ML optimization model on local hardware might consume 0.01-0.1 GWh annually on inference. If that model saves 5% of plant energy, or 5 GWh, the implied leverage ratio is 50:1 to 500:1 in energy saved per unit of AI energy consumed. This is not marginal: it is strongly net positive, and it is precisely why the IEA’s analysis of well-scoped industrial applications concludes that AI is energy-additive rather than energy-neutral.
The energy paradox narrative applies to large language models and generative AI, where a single query consumes 3-30 times the energy of a typical web search, and training runs consume electricity equivalent to small towns. But these architectures are not what drives a cement kiln or optimizes a steel furnace. The AI that matters most for industrial efficiency is lean, specialized, and deployed close to the process. The companies building and selling that software are not competing against data center scale-up; they are competing against the baseline of unoptimized industrial operations, and winning.
Where the Investment Opportunity Sits
The most defensible long-term positions across this theme sit with the large-cap infrastructure companies in the coverage universe. Their combination of certified field service networks, multi-decade customer relationships, and systems integration depth creates a moat that pure software entrants cannot replicate quickly.
Schneider Electric (SU-FR) is the most fully integrated play in the coverage universe. Its EcoStruxure platform spans building management, grid automation, and industrial control as a unified software stack, and recurring software and services revenue is growing as a share of EBITA. The company’s dominant position in data center power distribution gives it direct exposure to the AI infrastructure build-out, independent of any efficiency software thesis.
GE Vernova (GEV-US) sells grid software through a customer base it already serves with physical grid hardware and services contracts, which substantially lowers sales friction. GridOS, GEV’s advanced distribution management system, addresses dispatch optimization, fault detection, and VPP orchestration directly, covering the three highest-value AI grid applications identified by the IEA.
Eaton (ETN-US) is the most hardware-proximate name in the group, but its power management franchise is increasingly software-defined. Its fastest-growing segments (data center and industrial power quality) are exactly where AI-driven efficiency generates the clearest and most measurable ROI for end customers.
ABB (ABBN-CH) brings the deepest industrial automation capability in the group. Its ABB Ability digital suite is most concentrated in process industries (chemicals, mining, pulp and paper) where energy optimization ROI is highest and willingness to pay for proven software is well established. ABB’s multi-year portfolio restructuring has improved capital returns without eroding its technical position.
C3.ai (AI-US) is the primary enterprise software plays with revenue concentrated in industrial energy management. C3.ai’s documented results in steel energy forecasting are among the most quantified in the public domain.
Itron (ITRI-US) and Landis+Gyr (LAND-CH) offer small-cap exposure to the smart metering and grid sensing layer, which is foundational infrastructure without which AI optimization algorithms have no data to work with. Landis+Gyr is exclusively a smart grid company; Itron earns approximately 67% of revenues from smart grid products. Both carry higher execution risk than the large-cap names, but provide concentrated exposure to grid digitization spending.
Vertiv (VRT-US) manufactures the thermal management systems and power distribution infrastructure that sit inside data centers and determine how efficiently electricity is converted into compute. As AI chips generate five to ten times more heat per rack than conventional servers, the hardware required to cool them becomes more complex and more expensive, expanding Vertiv’s addressable market with every GPU cluster installed.
The YTD performance dispersion within the coverage universe is instructive. Vertiv’s 75% gain reflects the market’s pricing of several years of AI cooling demand; the thesis is sound but execution risk is now fully reflected in the price. The large-cap infrastructure names (Schneider, GE Vernova, Eaton, and ABB) have been relative laggards and offer better risk-adjusted entry given backlog-supported earnings visibility and growing recurring revenue. The pure software layer (AI-US, AZPN-US) requires higher conviction on near-term revenue inflection; C3.ai in particular has not yet demonstrated a durable path to profitability despite strong individual customer case studies.