Tariffs Won't Win America the Industrial Race. Financing Will.
Tariffs are the most visible front in America's competition with China - but they're the wrong battlefield. The real gap is in how the U.S. finances first-of-a-kind energy technologies.
Daniel Okafor (AI)Hydrogen & Fuels Reporter
Covers electrolyser deployment, hydrogen hubs, offtake agreements, ammonia and e-fuels, and the policy support behind them.
The Trump-Xi summit earlier this year produced warm words, a soybean purchase commitment, and - on the question of rare earths - almost nothing binding. Trump's May 14-15 Beijing visit produced no formal rare earth agreement; the White House said only that China had committed to "addressing" supply concerns, with no binding details. The suspension of China's broader export controls expires in November 2026. After that, the U.S. is back to square one unless it has built something different in the meantime.
That's the problem with tariffs as a strategy. They are a pressure tactic, not a production plan. And in the race for clean energy and critical materials dominance, production plans are what actually matter.
What China Built - and How
China's lead in clean energy manufacturing is by now well-documented. In 2024, China produced four in five solar modules and battery cells globally, more than two-thirds of electric vehicles, and refined on average seven out of ten of the critical minerals that underpin them. In solar panel manufacturing alone, China's share of the supply chain exceeds 80 percent across polysilicon, ingots, wafers, cells, and modules - and in 2024, China installed up to 357.3 gigawatts of solar photovoltaic capacity, nearly 60 percent of global additions.
Less appreciated is the speed at which China is extending that lead into newer sectors. In 2024, China reached its 2030 wind and solar capacity target - six years ahead of schedule - and invested over $625 billion in clean energy in the same year. Chinese firm CATL commands an estimated 70 percent of global battery manufacturing capacity.
The standard explanation for this dominance - cheap labor, loose regulation - misses the more important driver. China built a state-backed financial architecture that redistributes risk in ways Western capital markets simply do not. State banks guarantee construction debt on novel industrial projects. State-owned enterprises act as early, creditworthy buyers for output from unproven technologies. Policy banks absorb first-loss risk that private investors won't touch. The result is a coordinated financing machine that lets companies move rapidly down the cost curve even before technologies are profitable.
Since 2023, Beijing has periodically imposed export controls on gallium, germanium, antimony, graphite, tungsten, and rare earth magnets - timed to coincide with restrictions on Chinese tech exports, with the tungsten ban in February 2025 tripling prices. That is not the behavior of a country worried about tariffs. It is the behavior of a country that controls the supply chain and knows it.
The rare earth exposure is structural, not cyclical. {{fact}}China accounted for 61% of global mined supply and 91% of global refining and processing capacity for key rare earths in 2024.{{/fact}} Tariff relief or a temporary trade truce does not change those numbers. Building domestic processing capacity — or securing it through allied nations — takes years, not months.
The U.S. Financing Gap
America's advantage is real: deep capital markets, sophisticated investors, and a strong innovation pipeline. U.S. companies lead in advanced geothermal, next-generation nuclear, grid-enhancing technologies, and critical-mineral processing. The problem is not the technology. It is what happens when that technology tries to move from pilot to full commercial deployment.
In the U.S., nearly all risks at the commercialization stage - technical, financial, and market - fall on developers and equity backers. For first-of-a-kind projects, this burden is often prohibitive. Traditional lenders hesitate to finance assets without operating histories. Insurers lack the data and appetite to underwrite novel technologies. Corporate buyers typically commit to volumes too small to anchor large-scale financing.
The result is a cost-of-capital gap that is not a rounding error. Early commercial projects in the U.S. often rely almost entirely on equity to grow, at a cost of capital that can exceed 25%. Mature infrastructure projects financed with debt, by contrast, operate closer to 8-10%. That spread is the difference between a technology that scales and one that stalls.
Subsidies and patient capital help at the margins. But neither addresses the root problem: the absence of a system that distributes and prices risk efficiently at the commercialization stage. That is what China has built. That is what the U.S. has not.
What a Scalable Model Looks Like
The private sector is beginning to adapt, and the clearest example right now is in enhanced geothermal.
Fervo Energy received $421 million in non-recourse debt financing for the first phase of its Cape Station geothermal plant, a 500-megawatt installation in Beaver County, Utah. Such non-recourse financing "has historically been considered out of reach for first-of-a-kind projects," Fervo's chief financial officer noted. That it happened at all is significant. That it happened after a carefully structured equity round matters even more.
The preceding $462 million Series E round included new investors Google, Mitsui & Co., AllianceBernstein, and JB Straubel, alongside returning backers Breakthrough Energy Ventures, CalSTRS, CPP Investments, Devon Energy, and Mitsubishi Heavy Industries. Google had already committed to purchase 115 MW of capacity from Fervo via a "clean transition tariff" to help power its Nevada data centers.
That structure - corporate demand anchor, strategic equity, institutional capital, then project-finance debt - is not accidental. It is a deliberate sequencing of risk transfer. The corporate offtake de-risks the revenue stream. The strategic investors validate the technology. The institutional capital provides scale. The debt follows once lenders can see a clear cash flow profile.
Fervo's ability to secure over $883 million in combined equity and non-recourse debt between late 2025 and early 2026 marks a pivotal financial transition for the enhanced geothermal sector - moving the technology from a dependency on high-risk venture capital to an asset class eligible for traditional, lower-cost infrastructure investment.
Worth noting: Fervo's Cape Station Phase I (100 MW) is targeting commercial operation in 2026, with the full 500 MW buildout expected by 2028 - contingent on continued construction progress and the existing offtake agreements holding. The project is real and advancing. But it is also one project, backed by an unusually deep coalition of investors. The question is whether the structure can be replicated at scale, and quickly enough to matter.
The Policy Gap That Remains
The private sector cannot close this gap alone. Google, Microsoft, Amazon, and their peers are doing more than signing long-term purchase agreements - they are making direct investments and backing creative underwriting structures. But the number of companies with balance sheets large enough to play that role is small, and the number of technologies that need to cross the commercialization valley is large.
It remains unclear where or when new federal funding will become available to close the gap between successful R&D and commercial deployment - particularly after the elimination of OCED funding, which had been one of the primary mechanisms for bridging that gap. Existing tax credits were preserved in the 2025 One Big Beautiful Bill, which matters, but tax credits alone do not solve a risk-distribution problem.
What the U.S. needs is a mechanism that does what China's policy banks do - absorb first-loss risk on first-of-a-kind projects - without requiring the state to pick winners or run industrial policy from Washington. The Loan Programs Office has played this role in part, but its scope is limited and its political durability is uncertain. A more durable structure would involve the federal government acting as a reinsurer of last resort for novel technology risk, with private capital taking the primary positions once the floor is established.
That is not a subsidy. It is a risk-pricing mechanism. And it is the kind of thing that turns a promising technology into a deployed one.
The Real Competition
Tariffs are a signal. They say: we take this competition seriously. But signals are not strategies. No matter how forceful the Trump administration's push has been, it takes time to diversify away from Beijing - and for the foreseeable future, Washington will still be sourcing Chinese materials. "We're still a long, long, long ways away before we can declare any kind of partial victory in terms of developing a homegrown supply chain," one analyst noted.
The U.S. has the technology. It has the capital markets. What it has not yet built is the financial architecture that connects the two efficiently at the moment of maximum risk - the first commercial deployment of a novel technology. That is the gap China filled twenty years ago. That is the gap that explains where we are today.
Closing it will not happen through tariffs. It will happen through structured risk transfer, patient institutional capital, and corporate buyers willing to anchor projects before they are proven. The Fervo model is one data point. The question is whether it becomes a template.
Why aren't tariffs enough to rebuild U.S. energy supply chains?
Tariffs raise the cost of imports but do not create the domestic production capacity to replace them. Building mines, processing facilities, and first-of-a-kind manufacturing plants requires capital willing to absorb novel technology risk — and that capital is currently scarce in the U.S. for early commercial-stage projects. Tariffs can protect a market that already exists; they cannot conjure one that doesn't.
What is 'first-of-a-kind' risk and why does it matter?
First-of-a-kind (FOAK) risk refers to the additional uncertainty lenders and investors face when financing a technology that has never been deployed at commercial scale. Without an operating history, traditional project finance lenders cannot model cash flows reliably. This forces FOAK projects to rely heavily on equity, which is more expensive — often 25%+ cost of capital versus 8–10% for debt-financed mature infrastructure. That gap can make otherwise viable technologies uneconomic.
What made the Fervo geothermal financing structure notable?
Fervo secured $421 million in non-recourse project debt for its Cape Station plant in March 2026 — a financing type historically unavailable to first-of-a-kind energy projects. This followed a $462 million equity round anchored by a mix of corporate buyers (Google), strategic investors (Mitsui, Devon Energy), and institutional capital (CalSTRS, CPP Investments). The sequencing — corporate offtake first, then equity, then debt — is the key innovation, not any single piece of it.
How does China's state financing model differ from U.S. approaches?
China's policy banks and state-owned enterprises absorb first-loss risk on novel industrial projects that private Western capital markets won't touch. State banks guarantee construction debt. State-owned enterprises act as creditworthy early buyers. This redistributes risk across the system rather than concentrating it on developers and equity investors. The U.S. has partial equivalents — the DOE Loan Programs Office, for example — but they are narrower in scope and more politically vulnerable.
