When I first started learning about climate change, I assumed that we needed to invent new technologies to solve the problem. I no longer believe this. By and large, we already have the technologies we need to decarbonize our economy and achieve the goals of the Paris Climate Accord. What’s holding us back is the lack of markets, regulations, and capital needed to promote the mass deployment of those technologies. In this post, I’m going to focus on the capital constraint.
Capital isn’t being allocated to where it’s most needed and impactful for climate change.
At first, it might seem strange to suggest there isn’t enough capital flowing into climate-positive investments. After all, estimates of climate-related investing activity are in the range of $500 billion per year, and asset managers representing over $40 trillion of capital have signaled their desire to allocate capital toward combating climate change. The problem is that this capital is unevenly distributed and far removed from the “dirty work” of real-life project development, meaning that it isn’t going to where it’s needed the most.
Using venture capital and private equity data as a proxy, David Riester at Lacuna found that investment dollars in this space follow a barbell distribution. At one end we have “moonshot” R&D companies. These companies are focused on the development of breakthrough new technologies, have a high chance of failure, and even the successful ones expect to take many years before completing their first real-world project. This sector is well funded by both classic venture capital funds and newer ones like Breakthrough Energy Ventures, a $1BN fund anchored by Bill Gates. At the other end of the barbell are the owners and operators of mature projects. This is the domain of large-scale asset managers like Brookfield Renewable Partners, a $30B fund that buys and holds assets like solar and wind farms that have already been developed and have long-term cash flows under contract. These two ends of the barbell consumed nearly 80% of all private capital allocated to renewable energy last year.

Looking at that distribution, it’s clear where the bottleneck is: project development. Project developers are the mitochondria of our fight against climate change. They do the unglamorous work of putting steel in the ground to turn existing technology into new infrastructure that we all benefit from. I call these companies the “new industrialists” because I believe they are carrying the mantle of the leaders of the first industrial revolution. If solar panels are today’s oil wells, then Sunrun and First Solar are today’s Standard Oil. But the state of venture capital in Silicon Valley in 2020 is such that I would find it easier to raise $10M for an idea-stage moonshot than $1M for a new industrial startup, even if the latter offered a higher risk-adjusted return both financially and in terms of climate impact.

Why is this the case? I have observed three myths that new industrial startups face when raising venture capital:
- Myth #1: New industrial startups often lack proprietary technology, and therefore cannot differentiate their product.
- Myth #2: New industrial startups lack a compounding advantage to scale unlike, e.g., software businesses that have network effects.
- Myth #3: New industrial startups just aren’t a good fit for venture capital because it’s hard to see how a new industrial startup becomes a unicorn.
Let’s go through each in turn.
Myth #1: New industrial startups often lack proprietary technology, and therefore cannot differentiate their product.
I think it’s important to distinguish between having proprietary technology and having a proprietary product. The latter is extremely important; the former, not so much. One secret of Silicon Valley is that hardly any of its software startups have truly proprietary technology. What they have is a proprietary product, built on top of free and open-source software. They take off- the- shelf technology and arrange it into a proprietary product for an end customer. This is precisely the recipe that new industrial companies follow.
For example, Sunrun ($7B+ market cap) is in the business of installing solar panels, yet they did not invent their own solar panel technology. Instead, Sunrun bundled third party solar panels with the convenience of online purchasing and an innovative financing model, which made it much more convenient and affordable for people to get solar. It’s this unique combination that makes Sunrun’s product valuable to the end customer, who doesn’t know or care which company invented the original technology.
Myth #2: New industrial startups lack a compounding advantage to scale unlike, e.g., software businesses that have network effects.
While we’d all like to be Facebook, most businesses do not require network effects to grow exponentially or to reach significant scale. For example, most SaaS and e-commerce companies have zero network effects. This has not prevented SaaS and e-commerce companies from raising billions of dollars each year in both public and private capital at sky-high valuations (at the time of writing, the average company in the BVP Cloud Index trades at 17-20X ARR, and Etsy trades at around 40X EBITDA). Instead, what makes these companies valuable is that they have become the automatic choice for customers in their respective markets. Need a CRM? Salesforce. Need a book? Amazon. This level of incumbency is extremely difficult to dislodge.
The incumbent new industrial company is NextEra. NextEra is the world’s largest producer of wind and solar energy. It didn’t get there through network effects but rather through superior execution and capital allocation. As its scale and reputation grew, it became the developer and asset owner of choice for utility scale renewable energy projects, and its sheer size and financial strength now enable it to take on larger and more complex projects than its competitors, feeding a flywheel of growth. At the time of writing, its market cap is $147 billion, larger than that of Exxon Mobil.
Myth #3: New industrial startups just aren’t a good fit for venture capital because it’s hard to see how a new industrial startup becomes a unicorn.
Whether you agree with this one or not depends on your definition of “venture capital”. If you define “venture capital” as “Silicon Valley-style venture capital”, then I agree. But this is a function of the specific investment strategy employed by Silicon Valley-style venture firms and not venture capital as a whole.
Silicon Valley venture capital funds expect to lose money on most of their investments and earn all of their returns from their top 20%. This investment strategy requires unicorn-like outcomes just to survive. As a result, Silicon Valley-style venture capitalists need to be unicorn hunters, and to them, new industrial startups don’t look like baby unicorns.
Why not? I think one reason could be that Silicon Valley venture investors just don’t know how to pick winners in the space. Project finance, the electricity markets, carbon pricing schemes, and tax credits are as crucial to new industrial startups as they are alien to most VCs, which makes it impossible for them to responsibly invest in the space. Also, new industrial startups break most VC pattern matching, and their founders float in different circles. The founders of a new industrial startup are more likely to have studied geology at Penn State than computer science at Stanford, and worked at Generate Capital instead of Kleiner Perkins.
Another reason might be that the new industrial startups offer a different set of investment opportunities than a typical Silicon Valley startup. There’s an opportunity to invest in new industrial startups on three levels: parent-level equity, project-level equity, and project-level debt. The way most venture funds are organized means they can only invest in the first, even though all three offer compelling risk-adjusted returns, especially if an investor fully recycled the project-level cashflows into new investments.
New industrial startups require a new category of venture capital.
What’s clear to me is that a few of today’s new industrial startups will be tomorrow’s industrial giants, but they need a new category of venture investor to help them along the way, especially at the earliest stages of their lifecycle. In particular, I think new industrial startups need an investor who can take early stage risk and both the parent and project levels.
The fund I imagine would be a hybrid of Silicon Valley-style venture capital and classic project finance. Like a Silicon Valley venture fund, we would provide “first check” equity capital to help founders at the parent-company level to bridge them from idea to completion of their first commercial project. But like a project finance fund, we would also invest at the project level, and earn our return through a steady stream of cashflow distributed from the parent and projects over time. To my knowledge, such a hybrid type fund strategy does not exist yet.
The opportunity to define this category reminds me of how accelerator and seed investing in Silicon Valley were overlooked categories before Y Combinator and First Round Capital filled the gap. Or how SaaS, e-commerce, and technology private equity were all misunderstood in the investment community before Emergence, Forerunner, and Vista came along. The goal would be to build the lead investor of choice for the best new industrial startups, no different to how each of the aforementioned funds now top the list of venture funds active in categories that they pioneered.
There are some obvious questions with this strategy (hit rate? returns? don’t you still need unicorns?) that I’ll explore in a follow-up post, but my intuition is that I think you can earn outsized returns with this strategy if you’re creative on structure and aggressive on recycling of capital. In particular, I think indie.vc’s approach might be a great start on how to think about structuring parent-level investments in new industrial startups. I also think that having some form of permanent capital is a necessity for such a strategy, because I suspect most new industrial startups do not want to be under any time pressure to sell their parent-level entities in order to provide liquidity to their venture investors.
My favorite thing about this idea is how you can grow with your best companies over time. The challenge I have with Valley-style venture investing is that the best companies in your portfolio invariably outgrow you, and thus it’s hard to maintain your ownership levels in your best companies over time, let alone double down. Indeed, almost every Silicon Valley VC I know would love to multiply their exposure to their top performing company but can’t, either because their LPA limits total exposure to a single name or, more commonly, they can’t get their full allocation in subsequent financings and instead get diluted over time. Perversely for the investor, this gets worse the better each company does, as allocations in subsequent financings become more competitive in proportion to a company’s performance.
By contrast, in the hybrid strategy I propose, the investor can grow with their best companies and increase their exposure to them over time. This is because an investor’s best companies will, by definition, develop the most and best performing projects, and the checks an investor writes at the project level are much larger than the check written initially at the parent level. For this reason, getting pro rata rights to participate in project-level financings proportionate to our parent-level ownership will be key.
In effect, we are making a highly asymmetric bet with each new investment. If a new industrial startup fails to execute on its project pipeline, we lose only the relatively small amount invested at the parent level. However, if that startup succeeds, then we can multiply our exposure to that startup many times over by investing at the project level. In this way, we can tolerate a high loss rate while also earning above-market returns, because on a dollars-invested basis most of our money is invested in our best performing companies.
Thanks to my OnDeck writers fellowship, Rishi Taparia, Janice Tran, and Nancy Yu for reading drafts of this, and credit to Dave Riester at Lacuna for his writing on current state of play for early stage project finance in the energy sector which helped frame some of what I’d been seeing in the market. His writing is here.