The Unglamorous Approach to Impact Investing in Energy

In the fashionable world of “impact” investing, often the least sexy investment makes the most impact; nowhere is this truer than in the energy sector

Executive Summary:  The rapid growth of impact-oriented (“ESG”) investments in the energy sector prompts this article’s discussion of 1. Where that capital is going, 2. Where it should go, and 3. Why it isn’t going there. Market observers and participants see the current yield compression and “dry powder” ratios in the renewable energy infrastructure sector and wrongly declare this an indication of capital sufficiency – it is a sign of too few projects, not too much capital. The biggest impact therefore comes from addressing the supply of construction-ready projects, which is accomplished by i) injecting capital into the project development phase, or ii) offering more flexible long-term capital which changes the definition of what is “financeable”. Both approaches are hard and unglamorous. Meanwhile, “cleantech” investments – while necessary at some level – play the role of “shiny object”, drawing attention and capital toward SPACs and venture capital funds which, though well-intentioned, rob the infrastructure sector of the opportunity to displace fossil generation at the required pace. 

----------------------------------------------

Amidst the current barrage of demoralizing, near-apocalyptic modern phenomena, the decade-long emergence of “impact investment” funds, strategies, and mandates – and the resulting robust, diverse capital pool – stand as a refreshing island of benevolence and decency. Even the most cynical among us (rest assured I will have my moments herein) must be heartened by the near $1 Trillion USD[1] dedicated across the globe to “double-bottom line” investments, especially considering that ten years ago, using the same methodology, the pool of capital was $50B USD.  

Yet, as I stare at a press release for the 12th ESG focused “SPAC”, I wonder: is the impact investment community directing its energy sector allocation optimally?

The coming pages will touch upon 1. Where the money is going now, 2. Where it should go, and 3. Why it isn’t going there.

Backdrop

Briefly, my own history on the matter is peripherally relevant: I spent a few years in venture capital during the second half of the “cleantech” phase (06’-09’). It was fascinating, inspiring, exciting, accessible, and a great way to light a bunch of money on fire. Which I did[2]. As the CalPERS CIO famously said in 2013, cleantech was “a noble way to lose money.” In the energy sector, cleantech VC funds were mostly investing in shiny new objects that were meant to revolutionize the way we generated energy, like thin-film PV, fuel-cells, concentrated solar, O&M robots, tidal energy, biofuels, algae fuels, residential solar companies with an iPhone app[3], etc. Somewhere amidst all those pitchbooks and failed investments it dawned on me that perfectly good technologies for generating clean, renewable energy had existed for decades[4]. 

I became obsessed with the idea that I was wasting my time chasing unicorns with a blindfold on while the tools to address the problem – today and now – were available, but massively under-utilized. Suddenly the Vinod Khosla’s of the world seemed all form and no substance: spinners of fun, futuristic tales dumbed down for gullible investors like me. That these cool investments rarely seemed to graduate from “compelling science project” to “practical, useful, instrument of progress” only exacerbated the obsession.  Around the same time I met a fellow[5] who, despite once being part of an investment team that had the jaw-dropping track record of investing in the seed(!) rounds of Tesla, SunEdison, and Solar City (and almost nothing else), spent his days closing financing for actual solar power plants.  It was the antithesis of everything I’d seen and done: all substance, no fanfare, no chest-pounding, just head-down grinding on comparably boring deals which resulted in the existence of tangible operating power plants (like…next week, not in eight years). So I changed careers and went to work for him. The subsequent decade certainly hasn’t been a glamorous one for me, but it has been rich with substance, purpose, and tangible clean power plants getting built.

I share that history because I fear the impact investment community remains stuck where I was stuck in 2009, and we have some mission critical areas of need in the renewable infrastructure sector which that community is wonderfully suited to address, if only they woke to the true opportunities for “additionality”. 

A few definitions:

• “Impact Investments”: Investments made with the intention to generate positive, measurable social and environmental impact alongside a financial return.
• “Infrastructure Investments”: Investments in generation, transmission, distribution, or storage projects (assets), ranging from development through operations, which use existing technology.
• “Technology Investments”: Investments in new technology research, development, and commercialization; what is referred to herein and elsewhere as “cleantech”.

Figure 1: Renewable Energy Supply Chain

To digest information about where capital has been flowing, one must appreciate the capital intensity of the different segments of that supply chain:

Infrastructure Capital Intensity

A quick arithmetic exercise (and humor me here… an airtight methodology is not necessary to illustrate the point): 

The world electricity production in 2019, according to Enerdata, was around 27,000 TWh. 

The average annual growth of that number over the last 30 years is 2.75%. Let’s be energy efficiency optimists and say that slows to 2% going forward.

That takes the global production to ~49,000 TWh in 30 years. 

Subtract the ~7,000 TWh[6] of electricity produced by renewables (including hydro) in 2019, and we’re at 42,000,000,000,000 kWh needed to get to 100% renewable in 30 years. 

Assume that is, in fact, the goal and lets also assume (for the sake of limiting variables and staying in my comfort zone) we try to get there with solar alone. 

Assume the average solar irradiance over the next couple decades is 2,500 kWh/kW/year[7]. 

Divide 42,000,000,000,000 by 2,500 and you get 16,800,000,000 kW or 16,800 GW of new capacity needed. 

Assume we deem it prudent to get there linearly (divide by 30), and you get ~560 GW needed annually[8]. 

Leveraging Berkeley Labs’ work[9], we will apply an average cost of $0.75/w for 2020 installations to arrive at a capital need of ~$420B USD for renewable infrastructure in 2020.

As Figure 2 shows, invested capital is likely to be just over half that. Moreover, consider we are still picking low hanging fruit, so to speak, and the integration of more intermittent renewable sources is going to get harder and harder without a tremendous investment in storage and transmission. Some reports[10] project that in the next five years average global annual spend on storage will breach $100B, and by 2035 it will exceed $500B. 

Cleantech Capital Intensity

Investments in the research, development, and commercialization of new technology, however, are not terribly capital intensive. From 2004 to 2019 the average capital invested into companies pursuing new clean energy technologies was $9.2 million[11]. The reason is straightforward – these companies don’t have to buy as much stuff. The main capital cost is human capital. 

I. Where The Capital is Going

View from 100,000 ft.

Those familiar with the market and the data might be wondering how anyone could possibly deem there to be a shortage – on a relative or absolute basis – of investment funds focused on renewable energy infrastructure. The “capital overhang” in the equity markets for operating renewable energy projects is particularly noticeable – there is more capital than qualified projects, which has resulted in yield compression and dead-in-the-water capital seeking opportunities[12]. The statistics match that experience up here at 100,000 ft:

Figure 2: BNEF Investment by Clean Energy Asset Class

The exact nature of the investments in each bar is hard to unwind with precision, but for our purposes, it’s safe to treat the red bars as mostly “cleantech”, blue as entirely “infrastructure”, and yellow as a combination of the two.

The Bloomberg NEF data seems to strongly suggest that infrastructure is getting virtually all the money, and this chimes with the yield compression in the markets. Yet even the ~$250 B invested into infrastructure over the past four quarters looks well behind the pace. So, what the hell is going on? The supply of projects is too small to service the capital demand (and what is needed to displace fossil generation).  The knee-jerk conclusion that too much money was raised is myopic and short-sighted – the supply of capital is not the problem; the supply of projects is. If we zoom-in we can investigate if the impact investment community is focused on addressing that.

View from 30,000 ft.

The 100,000 ft level includes the entire capital market – governments, pension funds, sovereign wealth funds, NGOs, multinational financial institutions (e.g. World Bank, NADB, OPEC), state-owned enterprises, banks, SPACs, insurance companies, hedge funds, private equity funds, venture capital funds, corporations, family offices, individuals, crowdfunding pools, and probably a couple Ponzi schemes. The impact investment community is but a speck of dust on that picture. One way to get a higher resolution view is to look at the community through which a significant amount of impact investment flows: the venture capital and private equity community. 

Figure 3: VC/PE Investment Distribution Across Supply Chain

That is based on a data set of all clean energy investments in the Cambridge Associates database from 2004 – 2019[13]. It’s not intended to capture every single VC/PE transaction done during that window, so the gross numbers are not the point, but at ~$50B, 1,620 companies, and 16 years the sample size is certainly significant when it comes to observing the capital-destination patterns. It’s a “barbell” distribution with over 75% of the capital going to technology development or operating projects. In some respects, this makes sense – venture capital is highly relied upon to fund technology development and commercialization; nevertheless, the dearth of capital in the middle part of the spectrum starts to stand out, given the shortage of construction-ready projects.

View from 1,000 ft.

If we zoom all the way in and isolate the impact investment community, the distribution becomes even less congruent with the broader capitalization patterns[14]:

Figure 4: Impact Investors Investment Distribution Across Supply Chain

That is quite a dramatic lean toward cleantech, and a shocking rebuff of the project development phase.

The current allocation of capital in energy is both insufficient and inefficient in driving our society towards a zero-carbon grid. While most of the money in clean energy goes to infrastructure assets, rough math suggests we are still underinvesting in renewable energy projects. Meanwhile catalytic, impact-driven capital is disproportionately weighted to cleantech, seemingly comfortable deferring thermal displacement until…later.  

II. Where Should the Capital Go?

One wades into dangerously subjective and passionate waters here. How do we define “impact”; and – maybe more importantly – how do we measure it.  Gosh, I don’t know; and there’s not really a right or wrong answer anyway. If you must go down this rabbit hole, allow me to direct you to the (far-more-qualified) work of GIIN[15], HBS[16], or Cambridge Associates[17].

Certainly we know that both technology and infrastructure investments are needed. The right answer is not at either extreme. Evangelists of aggressive investment in cleantech like Bill Gates rightly point out that the world cannot fully solve our climate challenges with existing wind, solar, storage, and fission technologies. But making progress on the technology front doesn’t keep me up at night – moonshot technology hunts are well-covered. Go take a spin through Breakthrough Energy Coalition’s website if you don’t believe me[18]. It’s just one highly visible representative from a huge sandbox. Organizations like that may never generate a positive return, but they will do well as a conduit for funding bold technology innovation. A look at the names behind their efforts leaves no doubt powerful people and institutions are focused on cleantech.

Chipping away at the generation mix at a steady enough pace to avoid passing the point-of-no-return[19]…now that keeps me up. We can’t just focus all our resources on discovering the silver bullet solution and assume that when we find it - if we do - we will be able to turn on a dime and implement it at scale. We must aggressively implement the best current technologies (read: build renewable power/storage plants) as we move up the technology curve, because there are no guarantees about the length and slope of technological progress, regardless of how much money we pour into speeding it up. 

The implementation of current technologies tends to get “stuck” in the same spot:

Figure 5: The Big Renewable Energy Bottleneck

The conveyor belt, of sorts, represented in the graphic above will only move as fast as the stickiest/slowest gear in the machine. Over the years, different gears have been problems at different times. We’ve had periods when panel supply was the constraint, others when the demand for clean electricity governed installation pace, others when prices-plus-subsidies left projects in the red, others still where infrastructure capital was insufficient or too expensive. 

I suppose this is where calling out some good news is in order: none of those things are big problems right now. Offtaker demand for clean electrons is decent and increasingly due to price instead of altruism or greenwashing; global module supply is fickle (and severely constrained in the post-pandemic restart), but can generally support the called-for pace with continued incremental efficiency improvements; human capital in the renewable sector continues to grow in quantity and quality; installation costs are coming down more or less in line with expectations; and…..the capital supply for construction/operations is there, it’s pretty cheap, and it’s ready for some hockey-stick growth.

You might reasonably wonder how it can be that the capital supply for construction/operations is ample if invested capital last year is only ~50% of what it needed to be. The answer lies in the difference between invested and available capital. Of the unlisted (private equity) funds pursuing the renewable energy sector and included in Capital IQ’s database, the split between invested and uninvested capital is 28%/72%. For reference, those same ratios are 54%/46% and 52%/48% in telecom and healthcare, respectively[20]. Anecdotally, I can offer that it has been eight or nine years since I’ve spoken with a lender or cash equity investor satisfied with how fast they were deploying capital. Not to mention the massive pools of money – Canada’s CPPIB and Norway’s GPFG come to mind - that are desperately trying to find a way to invest tens of billions of dollars into the sector over the next few years, but are not showing up on the radar because they are still sorting out their strategy, and need more places for their money to go. To twist the Field of Dreams tagline: They will build it…if the projects come. Which brings us to the critical point:

Today’s “sticky gear” is the supply of construction-ready projects.

What makes a “construction-ready” power plant?

1. All development items are completed (site control, entitlements, interconnection, etc.)
2. The project can be constructed profitably
3. The project and its characteristics are “financeable” – can garner the capital required to achieve the profit

At last we arrive at the crux of it: where the impact-oriented capital should go. There are two opportunities here for impact investors to grease up the sticky gear, if you will. Project “financeability” usually comes down to i) the revenue source and ii) the technology/components used in the system. Lenders and investors in the renewable energy industry were initially drawn to the sector because it offered long-term, low-risk, fixed-price, predictable revenues from investment-grade buyers. They have understandably become quite addicted to this revenue profile, which is problematic in that it severely limits the universe of potential power plant installations which meet the three criteria above. Capital providers will price-in the risk up to a point, but there are hard out-of-bounds lines all over the capital markets, and too many worthy projects sit on the wrong side of those lines.

Figure 6: The Bottleneck Within Development

If, say, an impact investor wanted to own operating power plants but wanted that capital to be truly accretive, they could either:

1. Price-in risk of less proven tech – Help bypass the chicken-egg problem (need operating projects to get financing, need financing to get operating projects). Where everyone else sees black and white, find a shade of grey by doing a deal with some technology risk and marginally improved economics.
2. Relax offtaker credit requirements – Lend to or purchase projects with electricity buyers which don’t check all the boxes for the big banks and private equity funds. That leaves plenty of room to make smart investments with good risk-adjusted returns.
3. Not require a power purchase agreement at all – I’ll wait for the gasps from the pearl-clutching crowd to fade. How dare I suggest something that the fossil generation world has been doing for a century? Which is to say…be prepared to[21] sell power/electricity/credits into the merchant power markets. Change is hard, and with increased risk must come increased returns; but the coal/NG generation industry seems to have made some money without long term contracts in place at the start of construction. This evolution is inevitable and healthy – we need catalyst capital providers.

Sardonicism aside, those are very powerful and defensible approaches which would get projects built which would otherwise wither.[22]

The best way to boost the supply of construction-ready projects, however, is to provide capital to projects still under development.

Figure 7: Types, Timing, Amounts of Infrastructure Capital

Figure 7 depicts the general capital inflows and outflows associated with bringing a renewable energy asset online. Most components of the capital stack are well-covered by institutional investors: banks or private equity credit funds supply the construction financing and permanent capital generally comes from those same lenders plus the equity infrastructure community (and tax investors). As already noted, the market for equity investments into these infrastructure assets is deep and if anything, oversupplied.

The keystone capital allocation that could unleash a greater flow of renewable energy infrastructure projects onto the grid is the first piece: development. Development capital comes from a few different sources (covered thoroughly in a separately published article to which my old colleagues at Lacuna also contributed). It is a small and inefficient market with a high-risk[24], high-return profile. Swaths of development projects that probably stand to pass construction-ready criterium #2 and #3 fail to make it to that stage because they are starved of capital and die on the vine. There are too few investors willing to take on development risk, despite the upside potential: this is market inefficiency – and a missed opportunity for “additionality” – at its most basic. 

If impact investors want to direct their capital in the energy sector to the best possible point of leverage for combatting climate change while earning good risk-adjusted returns, it is here: in the development space.

III. Why Isn’t the Capital Going Where It Can Make the Most Impact?

Why are impact investors by-and-large missing this opportunity? Because investing in the development stage of renewable energy projects is hard. Cleantech, in comparison, seems simpler and more familiar.

The challenges of investing in development-stage renewable energy assets:

• Size. Development is too capital intensive to be bootstrapped (for most), but the check sizes are generally below threshold limits for traditional institutional infrastructure investors ($1-$10 M).
• Harder.  Development assets require very careful, specialized underwriting which usually calls for educating oneself on local laws, grid considerations, permitting, regulatory environment, subsidies, labor markets, etc.
• Heterogenous. Cookie-cutter approaches usually fail as each situation tends to be different than the last.
• Risk/Return.  Many such investments have the risk of significant capital loss; and, in any event, the range of potential outcomes is wider. Such risks are compensated for with higher potential returns.
• Shorter time horizons. The turnover of a development investment is ~1-4 years. That requires constant cycling of deal origination, execution, and monetization. 

It’s no big mystery as to why this space has been undercapitalized for a decade. It’s messy, tough to scale, riskier, and does not provide current yield in a market which badly wants it. 

But can’t you say all those same things about cleantech? Why do impact investors flock to cleantech if it poses many of the same challenges as development-stage clean energy infrastructure assets? My theories, based on 15 years of conversations, observations, and a very small amount of survey response data:

The allure of cleantech to traditional impact investors:

• Impact community needs to take the “big swings” other capital pools can’t. A reasonable thought and true, to some extent. Banks, infrastructure funds, pension funds, and other big pools of capital don’t have the mandate or structure to do cleantech. But don’t forget that the government, non-profit, and education sectors spend a lot of time and money on R&D too. Is it possible the desire to take the “big swing” is more about the allure of being heroic?
• More “bang-for-the-buck” in cleantech.  There is this idea out there that a couple million dollars injected into the perfect idea matched with the perfect team could change the world. And it could, sure. But there’s a certain arrogant naivete embedded in that motivation – namely the implied presumption that the investment team is going to hit the bullseye with one throw. If it were that easy, we wouldn’t face the problems we do. 
• The cleantech deals feel more familiar. A lot of impact investors have been (or are) active investors in the technology sector, often through venture capital funds. Cleantech investment pitches can feel very familiar to such investors. Here’s one: a cloud-based SaaS company which applies proprietary algorithms[25] to satellite imagery to routinize/scale the process of sourcing residential solar customers and automates the sales pitch (cost, system size, PPA rate), etc. That probably feels very “adjacent” to a bunch of other investments an investor has seen perform. Which is all well and good, but what if I told you that’s a worn-out idea which has been chased about 20 times in the last decade and never worked? Every month I receive a “what do you think about this company?” email from friends in the impact space staring at a pitchbook for that idea or the 2-3 others that pop up incessantly[26]. Meanwhile a wind project slated to displace a coal plant can’t find a few million for an interconnection deposit.
• Technology is sexier than infrastructure.  Once, at a conference dinner, I attempted to make my work sound interesting while sitting next to a woman doing white whale cleantech deals. I told what I thought was a rather wonderful story about finding financing for a quirky solar project. In response, she regaled the table with a tale about a recent breakthrough in cold fusion research which sounded like a scene from The Saint, and a new genetic modification for biofuel algae which could make it powerful enough to fuel space rockets. Cool. Websites with slick logos and stories about new companies pursuing ideas out of the science-fiction books of our childhood are just more fun. Alas, perhaps the time for such indulgences is passed.
• Job preservation and delayed failure. A quick refresher on some behavioral economics: when a decision maker (professional, politician, investment committee member, etc.) faces asymmetric penalties for failure vs. rewards for success, the individual naturally prioritizes avoidance of failure over pursuit of success. Put more simply: keeping one’s job becomes goal number one. With that comes a tendency to make choices i) the result of which is more delayed, and ii) where the negative outcome is understandable (“we always knew it was a longshot, but just imagine if it worked!”). So, with that context, compare i) a fusion energy moonshot investment that funds five years of research, with ii) a development-stage investment in some California energy storage projects which could fail, but are not expected to, and the point of potential failure would arrive in the next year or so.

Conclusion

The situation calls for a sense of urgency and renewed investment in accelerating the pace of fossil generation displacement - even if that means pulling some capital away from the hunt for the One Right Tool Which Could Save The World If We Find It. Most of the capital we need to accelerate deployment is out there right now, but our capital “Swiss army knife” is missing a couple implements – development capital, and permanent capital with a novel approach to underwriting revenue and technology. Capital surgically directed toward those places of need will act as the engine grease which loosens up the sticky gears in our renewable power plant deployment machine. 

Ideally, we can bring forth this capital without cannibalizing the search for the better energy technologies of the future, but if some of the capital in that community must be redirected to serve the infrastructure needs, impact investors serious about their goals should pivot accordingly. 

Footnotes/references:

[1] $715 B at end of 2019; from, Annual Impact Investor Survey 2020, Global Impact Investor Network Research Team

[2] As a 26-year-old Senior Associate I couldn’t really “do” anything, so that’s romanticizing a bit. But I was certainly a willing, table-pounding voice in the room. Let’s just say I was a big Nanosolar fan.

[3] This was a big deal back then

[4] The solar cell was invented at Bell Labs in 1954. Nearly all solar installed today uses the same technology they stumbled upon that day in 1956. The Idea Factory, by Jon Gertner, tells the discovery tale well

[5] Tomakin Archambault

[6] Ourworlddata.org

[7] Obviously assumes quite a bit of efficiency improvement; again, let’s not split hairs that’s not the point of the article

[8] For those keeping score at home, 178 GW of new renewable capacity was installed in 2019 according to Irena

[9] Berkeley Laps produces a very nice annual review of utility scale solar costs. https://emp.lbl.gov/sites/default/files/lbnl_utility_scale_solar_2019_edition_final.pdf

[10] Lux Research “Global Energy Storage Market Forecast 2019”, accessed 9/4/20

[11] Data from pitchbook.com database, universe of all cleantech transactions into solar, wind, hydro, other renewable generation “seed”, series A, series B, series C rounds of financing; certain rounds the proceeds of which are known to have been used for asset financing were excluded (e.g. late stage Bloom Energy rounds)

[12] Remember that thought.

[13] Methodology: start with Cambridge Associates’ spectacular database (https://www.cambridgeassociates.com/private-investment-benchmarks/), which does most of the work but does not breakout Project Development/Construction/Operations, so we used a combination of Crunchbase and Pitchbook data to apply the % breakdown between those categories to the broader universe size of the CA data set. The methodology is not perfect, but I believe directionally useful.

[14] Data from the Global Impact Investing Network “Annual Impact Investors Survey: 2020”, Goldman Sachs’ Clean Energy Impact Report, and Cambridge Associates’ private investments database; “construction” and “operations” not delineated so combined; the universe are those who self-identify as impact investors, and the data is self-reported and therefore subject to differences in interpretation and application of definitions.

[15] https://thegiin.org/imm/

[16] https://www.hbs.edu/socialenterprise/Documents/MeasuringImpact.pdf (some nice work from a couple MBA candidates)

[17] https://www.cambridgeassociates.com/wp-content/uploads/2016/10/Understanding-The-World-of-Impact-Investing.pdf

[18] Breakthrough already gets so much attention I hesitate to mention them; there are a handful of other fabulous institutions with brilliant people, money, and the latitude to shift the technology frontier. Ecosystem Integrity Solutions, Prime Impact Fund, come to mind

[19] Bill Mckibben - founder of 350.org and considered, by many, to be the preeminent expert on this point - thinks we may have passed it. Which is unfortunate.

[20] CapitalIQ database of all active private equity funds which self-report mandate in clean energy infrastructure (or telecommunications, healthcare); accessed early September 2020

[21] Just because you don’t require a revenue contract on day one doesn’t mean you’ll never have one 

[22] There isn’t room in this piece to launch into the deeper discussion these concepts warrant; watch this space

[24] Though considerably lower risk than cleantech, generally.

[25] If I had a nickel…

[26] Crowdfunding investment pools, app-based community solar customer procurement, drone-based imaging/analytics/services