16.38%.  Just kidding, it’s obviously way more complicated than that — by David Riester  


I have good news and bad news for you.  The good news is that developers of renewable energy and energy storage projects have never had more – or better – development capital options.  The bad news is that sifting through the options and understanding the cost of different flavors of development capital is always going to be a journey requiring nuance and patience.  Your introductory call “so what is your cost of capital?” questions are still going to get “it depends” for an answer…for the foreseeable future.  For those just trying to find the cheapest capital to support development efforts – especially those who have not spent much time in the finance/investment corners of this, or any, industry – “it depends” can be extremely frustrating.  Just tell me what your bloody interest rate is!!!!!  We at Segue experience this literally every day, and it’s obviously a valid and important question to answer (eventually). But it’s safe to assume anyone who answers that question before you’ve told them anything about your needs is either i) misleading, ii) likely to reneg , iii) trying to move the conversation along, or iv) going to be out of business imminently.

Why?  Because: a) your “cost-of-capital” is your capital provider’s “return”, b) return must always be understood in the context of risk, c) risk is a function of what you need money for, and d) “development capital” needs vary wildly.

Developers understandably approach us assuming we have an off-the-shelf product that always has the same costs associated with it, and the biggest thing to sort out is if the projects/needs “qualify” (meet our minimum conditions precedent).  If yes, then you get the off-the-shelf “product”, if no then we go our separate ways.  For better or worse, that’s not what we – or most of our fellow development stage investors – do.  Such an approach would be too rigid for a very heterogenous and constantly changing sub-market.  In reality, we get excited about a lot of different flavors of development investment.  We do have off-the-shelf structures and documents, but they are meant to be malleable when it comes to a few things, including economics.  This should be good news; if this weren't the case, then a lot of opportunities/projects worthy of capital support would get quick “no”.  But with that flexibility comes a longer path to understanding economics.  It’s difficult to see how this would be an unwelcome trade-off for developers.

The “Efficient Frontier”

To begin, let’s start with the basics:

  1. Expected return and risk are inextricably linked.
  2. The riskier an investment, the higher the expected return ought to be.
  3. Risk is subjective; one cannot quantitatively determine risk, and reasonable people can disagree.
  4. The more “efficient” a market is, the more likely it is that people assess risk (or the appropriate return for that risk) similarly.

The institutional investment community use something called “the efficient frontier” to illustrate these concepts. The renewable energy industry would be well served to familiarize itself with the concept. 

Figure 1:  The Efficient Frontier – Traditional “Asset Classes”

The illustration above shows traditional “asset classes” and where they fall.  The blue line represents the best you can do in an efficient[1] market– anything below and to the right is achievable, anything above or to the left is not.  Usually, this tool is used to illustrate the benefits of diversification (you blend a bunch of those dots together and, in the aggregate, you get closer to the blue efficient frontier).  We’re focused on a different takeaway - a set of very large, liquid, data-rich, and comparably efficient markets can be plotted out in a way that paints a picture of the relationship between expected risk and expected return.  Because the pattern is established by global capital markets in the broadest possible sense of the word, this isn’t theoretical - the money has spoken.   

So what do hedge funds and commodities have to do with renewable energy?  Well, for starters the chasm of differences between i) an investment in treasury bonds and ii) an investment in emerging markets equity, is no wider than the same gap between, say, i) long term project recourse debt, and ii) greenfield project development capital.   The same framework can be useful to understand capital types and costs in the renewable energy sector.

Plot out all types of renewable energy investments on the same frontier, and it might look something like this:

Figure 2:  The Efficient Frontier - Renewable Energy Sector Investments

Now…in reality the renewable energy market is quite inefficient, so those dots don’t really press up against the efficient frontier in a clean pattern like the illustration above.  For example, tax equity should be off the page in the upper left waaaaaaay into the “not achievable” (theoretically) realm, and corporate equity should be much lower on the Y axis because historically venture capital style investments in corporate developer equity have performed miserably in the aggregate.  But the above is what investors believe they are signing up for, and that’s arguably more important than historical data in an inefficient market.   

Expected Return is generally measured using internal rate of return (IRR). If you conceptualize IRR as your average annual return, you’re close enough[2].  However, don’t make the mistake of conflating interest rate for an investor/lender’s return.  Not in our industry, at least.  Those upfront fees that didn’t enter the conversation until the draft term sheet came across?  That’s part of their return.  Commitment Fees? ”Administrative Fees”?  Yeah, that’s just obfuscated return too.

Risk is traditionally measured using standard deviation, but really that’s just a technical way of measuring the range of potential outcomes.  Maybe a better way to think about it in renewable energy is 1) what is the probability this investment underperforms, and 2) what is the potential magnitude of the underperformance? This is how risk is best assessed – probability and magnitude.  So, lets pick a handful of those renewable energy investment classes and discuss the probability and magnitude of downside scenarios:

Magnitude & Probability

When your probability and/or magnitude of risk is greater, you need a commensurately higher expected return to justify taking the risk.  This is one of the most fundamental principles of investing and capital market functionality.  Most renewable project developers don’t have investment and/or capital markets backgrounds, so understandably under-appreciate how important this is.

To help this land, a couple thought experiments:

Thought Experiment #1:

Imagine you’re a developer.  You just planted three new 100 MW project “seeds” in the form of site lease options and interconnection application submissions.  But that’s it – that’s as far as you’ve taken your idea so far.  You started work on the projects three months ago and spent a total of $75k of your own money.  NTP is 3-5 years out, CODs even more.  You get a phone call from an IPP offering to pay you $0.04/W ($12M) all cash up front.  What say you?   

That wasn’t a trick question.  You sell your projects.  Obviously.  If you hesitated there, it’s safe to say your intuition around risk and return might not be properly calibrated.

Now imagine you have three 100MW projects with site control, all discretionary permits, a signed LGIA, a tax abatement, and PPAs ready to sign once the project owner posts $10M of PPA security.  The projects are pretty thin, but isn’t every project these days?  It looks to pencil with a few cents of margin. NTP is about 6 months out.  You haven’t finalized an EPC agreement, or financing, but folks are taking your calls and expressing genuine interest.  You get a phone call from an IPP offering to buy them for $0.04/W ($12M) all cash up front.  What say you?  

That’s a much tougher call, no? It doesn’t seem crazy to wonder if someone might pay more if you ran a sale process or carried the project a tad closer to NTP.  The point is the answer isn’t objectively obvious, whereas in the first scenario it absolutely is.

The intuition applied to arrive at those two different conclusions for the “same” project (at different stages) is the same intuition capital providers apply to assess risk (and, therefore, return).  In the first scenario, the probability of the investment going poorly (e.g., some/all of the projects dying somewhere along the journey) is obviously dramatically higher than the same probability in the second example.  Deep down you know those lease options and IX applications are more likely to lead nowhere than they are to turn into profitable operating power plants for which you’ve been paid handsomely. Money talks (i.e. reveals your true assessment of probability/risk). 

So, by extension, we should not expect the cost of capital for early stage development capital to be the same as for late stage development capital.  The needs are completely different, therefore the risks are completely different, therefore the “return” ought to be very different, therefore your cost-of-capital expectations ought to be very different too.  That anecdote out there about a developer getting an 8% “line-of-credit” for “development expenses” in the form of posting a fully refundable interconnection deposit a few months before NTP? That’s not a very relevant comp for someone looking to capitalize an early stage project/portfolio.  Our nascent, inefficient development capital market hasn’t yet established the nomenclature to quickly differentiate between the different flavors of development capital needs, but that doesn’t mean all development capital is the same. Market anecdotes (especially outliers) combine with insufficiently nuanced language/descriptors, causing folks to conflate information, oversimplify, and talk past each other.  Ships pass in the night. 

Thought Experiment #2:

Pop quiz: you think the 3 projects in the prior example will have a 67% attrition rate (1 out of 3 become operating power plants), and for the projects that die you will invest (lose) half as much capital as you will invest in the project that “makes it”.  You want to get a 2x return in the aggregate, what multiple of invested capital (MOIC) do you need on the project that makes it to get a 2x in the aggregate?[3] 

Follow up question…  if it took you 3 years to get there, what would the IRR (“cost of capital”, if you insist) be?[4] 

Next follow up…using those basic assumptions, what are the odds of you getting a big fat donut (full loss of capital, all 3 projects die)?[5]

That number tells you the probability of your downside, and we already established the magnitude is severe, with a full-loss downside.  “Asks” with variations on that risk profile are put in front of us at Segue every week, and in many cases we’re into it.  But make no mistake about it, that’s a very high-risk investment. You face a 34.3% chance of losing your entire investment - is a 26% base case expected return really too rich?  Be brutally honest with yourself: would you invest a meaningful amount of your own money into that?  Your money is out for a few years and, if it goes well, you double your money, but you have a very real chance of losing every cent.  Reasonable people could disagree on the answer there – that’s the nature of subjective risk assessment – but at a minimum you’d have to think about it a bit, right?

An Approach to Consider

“Great speech Dave – I’m sure we’re all impressed. But how the hell do I know what I should expect the cost of capital to be for my portfolio, then?”

Glad you asked.  I’d start by focusing on it less.  Our little corner of the market is plenty competitive these days, so none of us “dev cap” providers are putting out greedy offers.  The belts are pretty tight, you’re very unlikely to get swindled.  Just make sure you talk to a couple different folks when your needs come into focus, compare the apples and oranges as best you can, but then focus on the stuff that really matters. 

It is very unlikely that your development capital provider’s “cost of capital” is going to be the difference between you retiring early or grinding it out for another couple of decades.  The quality of the assets you develop, the capital support they have, and the project monetization execution are the variables which will write your fate.  It follows that a developer’s capital raising priorities should probably be to pick an investor that will:

  1. Be economically aligned with your interests
  2. Actually close when you need them to; you have more important things to focus on and projects to move forward
  3. Understand the ride they are signing up for and are prepared to nimbly roll with the punches
  4. Help you cross the finish line (monetize your matured assets) when it’s time

Be humble, realistic, and pragmatic.  Humble in knowing that, while your shop might have a good track record and an experienced team, you’re still playing a tough game in which you don’t fully control the outcome…no matter how good you are.  Realistic in setting expectations of attrition and margin, appreciating the difference between business plans and real-world execution.  Pragmatic in not letting perfection be the enemy of the good when it comes to closing development capital, accepting that an inefficient market built on subjective assessments of risk is going to unearth some differing views.  In the end we all pretty much have the same goals – get some profitable renewable energy/storage projects built and make some money in the process.  Usually when the first of those goals is accomplished, the second one takes care of itself.  Obsessing over an unanswerable question (the title of this paper) is not the highest and best use of a developer’s resources.    


[1] If the concept of market efficiency is new to you, just think about a perfectly efficient market as one that functions very well where everyone pretty much always agrees on price, risk, and value.  So…the market for Microsoft stock is very, very efficient, while the market for a Banksy painting is very inefficient.

[2] IRR is actually somewhat complicated from a mathematical perspective, but for our purposes the marginal value of understanding the nuance is pretty minimal.

[3] Answer: 4x

[4] Answer: 26% in the aggregate

[5] Answer: 34.3%