Decarbonisation under
capital scarcity.

The energy transition was priced in dollars that no longer exist. Net zero pathways, the IEA’s stately roadmaps, the McKinsey decks, the Mark Carney speeches all ran on a single unstated assumption:

Money would be there when you needed it, and the price you paid for it would round to zero.

That assumption held for about a decade.

It’s no longer holding.

We’ve moved from a world where the ten-year US Treasury paid you nothing to a world where it pays you four percent and might pay you five. Central bank balance sheets are contracting, government debt loads are at peacetime records and defense budgets are climbing in every NATO capital. Demographic transition is moving savings into consumption and globalisation is being unwound on purpose, after decades as the great deflationary force of our lifetime.

Capital has become scarce again.

And few of the players in the climate conversation have updated their priors.

The transition is still discussed as if the ten-year yield were one and a half percent, and the marginal carbon-conscious investor had infinite duration in their portfolio. Neither remains true.

Renewables are almost pure capex. You build a wind farm, a solar array, a transmission line, and once it’s up, the marginal cost of producing electricity is approximately zero. Sun and wind are, famously, free. The whole thing’s a balance sheet bet on a thirty-year revenue stream you’ve already paid for.

A combined-cycle gas plant has modest capex and high opex; you’re paying for fuel every minute it runs. Natural gas prices set the economics.

When capital is cheap, renewables win. Their low operating costs dominate their high capital costs, once you discount the future at near-zero rates.

When capital is expensive, the math reverses.

Every percentage point on your discount rate shaves another percentage point off the present value of those however many decades of free electrons. Renewables become marginal, then unbankable, then irrelevant. The energy density advantage of fossil fuels reasserts itself in spreadsheets, where it had been waiting all along.

We’ve already seen this play out. Offshore wind in the UK and the US Northeast got destroyed in 2023 and 2024. Projects that penciled at three percent rates didn’t pencil at six. Avangrid paid the State of Massachusetts $48 million in cancellation fees rather than build the wind farm it had won the contract to build. Ørsted took a $4 billion impairment on its US portfolio. The UK held a contracts-for-difference auction and got zero offshore wind bids because the strike price was set in a different rate regime.

Rates rose into historical territory, and the math died.

Politics responded with industrial policy; the Inflation Reduction Act in the United States, the Green Deal Industrial Plan in Europe, the national champions strategies across Asia etc. All admissions that decarbonisation can’t fund itself in a high-rate world.

The state has to carry the cost of capital that the market won’t bear.

Subsidies, loan guarantees, contracts for difference, production tax credits and capacity payments; the acronyms differ by jurisdiction but the function is near-identical. They put the public balance sheet behind the private, because the private alone cannot make the numbers work.

And the state’s balance sheet isn’t infinite.

The forces driving rates higher are reducing the headroom for industrial policy. The IRA was passed in 2022 with an estimated ten-year cost of around $370 billion, but by 2024, Goldman Sachs was modelling its actual cost at over a trillion — and the Penn Wharton Budget Model came in even higher. Subsidies that scale with deployment scale with the success of the policy. The better the policy works, the more it costs. The more it costs, the harder it gets to fund in a world where every government department is competing for the same shrinking pool.

Economists call this fiscal dominance; the central bank can’t raise rates without bankrupting the sovereign, so monetary policy stops mattering.

We’re not there yet. But the trajectory is in that direction, and it’s accelerating.

China maintains integrated state capacity at scale. They can direct capital, mandate production, subsidise across entire supply chains, and write off the losses on a state-bank ledger that nobody audits.

They build more solar in a quarter than Europe builds in a year. They control roughly 80 percent of global solar manufacturing and a similar share of EV battery production. Their domestic EV market is now larger than the rest of the world’s combined. CATL alone produces more battery cells than every Western manufacturer put together.

And green virtue has nothing to do with it.

China is a different kind of economy. Capital allocation flows through political channels, and markets don’t get a vote. The cost of capital inside a state-owned enterprise is whatever the politburo says it is. They don’t have a discount-rate problem because they don’t really have a discount rate.

Western policy is facing four scenarios:

1. We can match that with our own state capacity, which we don’t have.
2. We can tariff our way out, which raises consumer prices and slows deployment.
3. We can accept that the energy transition will be largely Made in China.
4. We can pretend, for a while, that none of these are the choices on offer.

Most Western governments are taking the fourth option. The European Union is arguing about whether to put 100 percent tariffs on Chinese EVs, and the German auto industry is openly warning that it would be a disaster for Germany. The greens want the cars, the centrists want the jobs and the Treasury wants the tax revenue. No policy delivers all three.

Nuclear power is among the most capital-intensive ways to generate electricity ever invented. A single AP1000 reactor takes a decade to build and runs upwards of fifteen billion dollars before it produces a watt; by the logic of capital scarcity, it should be the worst-positioned technology going.

But nuclear sells dispatchable, baseload, zero-carbon, weather-independent electricity for sixty to eighty years from a single capital outlay, while solar farms last twenty-five years on a good day and wind turbines last twenty. Once you let the discount rate stretch the timeline, nuclear’s lifecycle economics look better than any renewable, even at five percent rates and accounting for cost overruns that would sink a normal hedge fund.

Which is why every government that takes both decarbonisation and capital scarcity seriously is now building nuclear. France, the UK, the UAE, Korea, Japan and the US under both the previous and current administrations. The exception is Germany, which closed its last reactor in 2023 and is now buying coal-fired electricity from Poland to keep the lights on.

Nuclear is the asset class that should win the decade. Whether the political coalition exists in any Western democracy to build it is a different question.

Oil and gas spent the back half of the 2010s on a capital strike. ESG pressure cut off financing, sovereign-wealth funds divested and banks ran from anything that drilled. The supermajors stopped exploring and started returning cash to shareholders.

That capital strike is over.

Aramco is back in Western capital markets. Exxon and Chevron have done the largest US oil deals of the decade. Norwegian and UK North Sea production is being extended past every previously announced sunset date. The conversation in energy investment has flipped from divestment to security-of-supply, and the financial weight has gone along with it.

Capital scarcity selects for short payback periods and existing infrastructure. Thirty-year capex bets get beaten every time the credit cycle turns.

Decarbonisation, in the form it was sold to the public, was a luxury belief.

The decarbonisation politics of the 2010s could only be held by societies confident in their own abundance, with the budgetary slack to absorb the transition costs and the geopolitical room to make idealistic choices about the energy mix.

Capital abundance funded a lot of luxury beliefs in the 2010s. ESG investing, diversity mandates, stakeholder capitalism, long-tail philanthropy etc. The whole apparatus of corporate progressivism was, in part, a financial product that worked when money was free and broadly amoral shareholders weren’t watching the bottom line.

But shareholders are watching now.

ESG funds are bleeding outflows; BlackRock has scrubbed the language from its marketing materials; and climate banking alliances are losing members on a rolling quarterly basis. The Net-Zero Banking Alliance lost six of the largest US banks in early 2025. The US has formally withdrawn from the Paris framework again. European governments are pushing their own targets back without admitting it.

This is what capital scarcity does to politics. It strips away whatever was being subsidised by the bull market and exposes what people will actually pay full price for.

What remains is the floor // the base of the curve. The median voter, the median investor, the median state, the median household, paying full freight for what they want without the asset-price tailwind that hid the cost. The answer turns out to be less than we thought.

Cheap capital made the transition look like it had no losers. Solar panels for everyone, EVs for everyone. Heat pumps for every home. Insulation grants for every retrofit. The bill would land somewhere abstract, on the future, on someone else’s quarter, on a balance sheet far enough away from the household to feel costless.

Expensive capital makes the bill concrete and visible.

Heat pumps are real money, and EVs still cost more than ICE cars. Insulation is real money. Higher electricity bills under capacity-payment regimes are real money. The cost falls on the people who can least afford it, and they vote.

This is the political mechanism behind the green backlash in Europe. The Dutch farmers, the German AfD vote, the French gilets jaunes, the rightward drift of the British countryside are all protests about who pays the bills.

A movement that built its political coalition around abundant cheap capital is poorly positioned for the politics of expensive capital.

The transition slows in the West. Nobody announces it as a slowdown, but the projects stop penciling and the subsidies stop scaling. We get a lumpier, geographically uneven, politically routed decarbonisation. Physical efficiency stops being the organising principle and fiscal energy takes over.

State capitalism wins where state capacity exists. China builds the panels and the cars, the Gulf states build the megaprojects on sovereign-wealth-fund balance sheets and the US carves out specific sectors for industrial policy and lets the rest fend for itself.

Adaptation budgets grow, and mitigation budgets shrink in real terms. We start spending more on sea walls and air conditioning and grid hardening and disaster relief, and less on the carbon ledger. This is rational under capital scarcity, but it’s also an abandonment of the 1.5-degree target, whether we admit to it openly or not.

The discourse changes. The moralism of the 2010s gives way to the cold pragmatism of the late 2020s and 2030s. Phrases like “energy security” and “supply chain resilience” do the work that “climate justice” used to do. They mean something similar but they sell better in a recession and they pass legislatures that wouldn’t pass the alternative.

The people who built their careers around the old framing have a choice: update, or keep performing the rituals of an era that’s already ended. Most will keep performing; that’s how priesthoods work.