Climate Venture Capital Is Quietly Becoming Project Finance
The power law doesn't reward de-risked early-stage bets.
Today's piece is more extreme than it needs to be. But the shift is real. If early-stage VCs keep chasing safer bets, how do they return the capital they promised their LPs?
Sit in enough climate diligence calls and you hear the same questions: What are your unit economics at scale? Have you secured feedstock? Where is the offtake agreement? Who is your EPC partner?
These are good questions. They are also the wrong questions to ask a company raising a Seed or Series A.
Those questions come from project finance. They are what an infrastructure investor asks before writing a check against a physical asset with predictable cashflows. The point of that diligence is to confirm the risk is already gone, when binary risk is off the table.
Venture capital exists to do the opposite. It funds the part of a company’s life when nothing is proven and most of it might not work. The returns were always meant to come from taking that risk.
Something changed. Climate investors started importing the infrastructure checklist into the earliest rounds. They want to see unit economics and commercial traction. They are asking founders to look de-risked at the exact moment risk is the whole point.
Where the Capital Went
The numbers show it. US climate tech VC reached about $29 billion in 2025, the third-highest year on record, but ten late-stage deals captured 28% of all investment. Underneath that headline, deal count fell to 1,545, down 18% and the lowest since 2020, with Seed down 19% and Series A down 22% as investors backed away from new entrants. At the same time, Series C hit an all-time low while Series D and later rose 41%.
The money did not shrink. It walked up the stage ladder to where the risk was already gone. The industry called this a flight to quality. It looks more like a flight from the job.
Venture Capital Was Built on the Power Law
Venture returns follow a power law. A tiny number of investments carry the entire fund. Horsley Bridge found that 6% of US venture deals between 1985 and 2014 generated 60% of the asset class’s returns. You do not get those outcomes from companies that already locked their offtake and their unit economics. You get them from the bets that looked too early to make.
So if you wait for the de-risking, the math stops working. You pay a near growth-equity entry price for a venture-stage company, you inherit a ten-year lockup, and you give up the one thing that justified the illiquidity to your LPs: the shot at an outlier. A fund full of safe entries is just a worse-priced infrastructure fund. It cannot return the multiple it promised.
Which leaves every climate GP with one question. If you are no longer taking venture risk, how do you plan to deliver venture returns?
The Noise is Real
There is a defense, and half of it is true. The market got loud. It is easier to start a company than it has ever been. AI collapsed the cost of building, with developers moving 55% faster on work they used to do by hand. National labs will license IP and partner with startups that could never have funded the research themselves. And talent is everywhere.
When everything looks like a company, investors reach for the cleanest filter they can find, as milestones cut through the noise. A signed offtake is easy to see. Proven unit economics are easy to verify.
But milestones are the laziest filter on the table. Everyone can see the same signed contract. There is no edge in spotting a company that already de-risked itself, and no premium left to capture by the time it has. The real edge in a crowded market is the oldest one in venture: knowing the people. Conviction built before the proof exists, from time spent with founders, is the only thing that never shows up on someone else’s term sheet.
What Founders Should Do About It
If investors are going to underwrite like bankers, founders should stop performing for them.
The milestones VCs now demand are the exact milestones non-dilutive capital was built to fund. Grants, catalytic capital, and project debt all exist to pay for the de-risking that equity should never have funded. A development loan can finance the work that produces a signed offtake. That offtake is the same one the VC wanted to see. The difference is you got there without selling a third of your company.
Flip the order. Use the cheapest capital to clear the milestones, then walk into the equity conversation already de-risked, holding the stronger hand instead of asking for one. The investor who wanted proof now has to compete for a company that no longer needs them as badly.
If the business can reach cashflow, the exit problem that haunts climate venture stops being your problem. You do not need a fund-returning IPO to justify your existence. You need customers. A company that pays for itself owes its future to no one’s ten-year clock.
That is the part worth holding onto. The investors who win the next decade will be the ones willing to look stupid early, to back a founder before the contracts exist, because that was always the assignment. The founders who win will be the ones who stopped waiting to be picked.