When I started writing this newsletter, I wanted to follow the market, observe how emerging sustainability tech actually becomes financeable, and connect with others building in the space. Well, lately the headlines have been entirely dominated by power markets and AI data centers. I could talk about Crusoe’s $3.9B raise or Rune’s $40MM raise this week, both targeting factory built modular compute blocks, but no, I actually want to share a few new financing recipes I found so I don’t bore you!
Strip off the green premium and sell it separately, unbounded by location:
If you have a new technology to make industrial commodity in a more sustainable way, it likely costs more. The process is new and subscale; inputs and capital are more expensive. This is the green premium, and many companies die on this hill because the broader market doesn’t want to pay it. Take green steel, for example. A hydrogen direct reduced iron plant costs billions of dollars to build. You cannot finance that kind of capex without an absolute guarantee that someone will pay a premium for the decarbonized product. But the corporate buyers willing to pay that premium don’t actually need raw steel, and they are probably not located anywhere near a steel mill.
Google just bridged that gap by signing an agreement for certificates tied to up to 91,000 tonnes of near-zero-emissions steel from Stegra’s plant in Sweden. They are buying the environmental attribute, decoupled entirely from the physical delivery of the metal. We saw the solar industry do this exact same thing with renewable energy credits decades ago (although electrons move along the grid much easier than steel). Once that premium becomes a tradeable, verifiable instrument measured against a standard, in this case, the IEA’s near zero emissions definition, a corporate buyer can sign a long dated contract for it. That gives the project a contracted revenue stream it can actually take to a bank. The plant itself might be Swedish, but the mechanism and the buyer behavior are relevant for the US. I can’t wait for the day a sponsor pledges a ten-year strip of green steel certificates as loan collateral.
Get a price floor from your offtaker, and make them your shareholder. Get a contract, not tax credits.
Another roadblock is commodity price risk, incredibly common in the critical minerals and natural resources space. A minerals processing plant is normally unfinanceable with project debt because a lower cost competitor can flood the market and crush your margin any day. The good news is you have a deep enough market that you can always sell 100% of your output. The bad news is you are exposed to revenue you have no control over, carrying a fixed cost structure you may have very little control over. The DoD just structured a deal with MP Materials for their Texas magnet facility that includes a $400MM investment for a roughly 15% equity stake, a 10 year offtake agreement, and crucially, a price floor. It turns a highly speculative equity story into a stable project financing story, and the value created by enabling new domestic production and risk for providing a firm price floor is paid out via long term equity value accretion, at least that’s what they tell tax payers. Just as importantly, a binding contract survives a change in political administration infinitely better than a tax credit (you know what I am talking about). With the bipartisan Critical Materials Future Act moving through Congress right now, backed by two Republican and two Democratic sponsors to authorize another $750MM, every non-Chinese processing project in North America is going to try this new recipe. Markets may read these structures as bailouts. I think there is more. They are creating an entirely new contracted asset class with a sovereign counterparty’s credit sitting right behind it. Sounds familiar? ;)
Turn capex into a repeatable essential service and finance it with your customer’s credit, not yours
Finally, look at industrial water. Gradiant just locked down $300MM in new contracts since the start of 2026 to manage water across five semiconductor manufacturing fabs in New York, Virginia, Idaho, and Utah. Industrial water reuse used to be a bespoke, one-off capex headache sold plant by plant. Gradiant just proved it has become a contracted, repeatable service business with real backlog. Fabs are the ideal anchor credits for this kind of structure. They have investment-grade balance sheets, decades of site life, and water access is a fundamental license to operate input. They willingly sign long contracts because they have no other choice. The fab simply needs the water to exist.
When you look across green steel certificates, sovereign backed mineral floors, and fab water contracts, the pattern here is a new party steps in to take on a specific risk to solve the risk allocation problem. The risk isn’t disappearing, it’s just being repackaged and SOLD to the exact entities that can actually afford to hold it. And that makes these new financing recipes for getting hard assets built when the traditional markets say no.