This article builds on that work, as well as recent reports from Morgan Stanley and a recent LCOE analysis from Lazard. We believe all of these articles are critical reading, as they provide a framework to analyze the industry. Our research reaches different conclusions because it is designed to serve our investor and developer partners, and because we disagree with previous assumptions about SREC prices, build costs and the cost of capital.

Investor returns will tighten

Not surprisingly, the reduction in the ITC means an overall smaller “economic pie” that can be split between solar developers, the EPC, the financier and the host customer. Assuming turnkey costs do not change, the ITC step-down correlates to a 3 percent to 5 percent unlevered post-tax return for the investor that is purchasing a project. This means that an investor that used to be securing a 9 percent return on their investment in a commercial project would receive a 4 percent to 6 percent IRR, if turnkey prices did not change.  

Structured transactions like a flip or lease-pass-through will either scale in size (portfolio sizes will need to increase to support the same investment) or disappear. Further, structured transactions will have a reduced impact on overall implicit IRR (these structures can effectively reduce IRR by 1 to 300 basis points in certain markets currently). Given this combination, we believe that there will be more tax-advantaged capital (like utility affiliate funds) buying solar projects.

This analysis can be illustrated through a “heat graph” comparing costs of capital, PPA rates and build costs. We originally provided this graph in 2014 and have updated it below. The reader can use his or her own assumptions to arrive at a conclusion. (Click to enlarge.)

The industry has matured to a point where investors are bidding on commercial projects within a 150-basis-point differential, generally between 8 percent and 9.5 percent. If build costs are not reduced from an all-in price of $2.09 per watt, as SolarCity recently accounted, investors will need to be comfortable with a 5 percent to 6 percent return for many commercial projects. That will not happen in the next 18 months. Instead, our industry needs to focus its energies on increasing build and development efficiency.

Build costs will come down

Sol Systems’ research team has run what we term our “Sol Map Analysis” to look at required build costs on a state-by-state basis, and to provide a national snapshot that summarizes the planned 2017 step-down.

This analysis runs a specific project through each state model simultaneously, calculating differences in taxes, average commercial retail electricity, SREC monetization, production tax credits, etc., to determine the break-even build costs for that state. In each iteration, we assume that the PPA is 90 percent of the expected retail electricity rate, a savings of 10 percent for the customer. Our model utilizes proprietary forward SREC curves based on those we see in the market.

We contemplate a tax-efficient buyer acquiring these projects. Structured transactions are slightly more efficient, generally lowering the overall effective cost of capital 100 to 300 basis points for projects depending on state incentives and electricity prices.

We ran four different scenarios in our model, with the worst-case scenario representing an 8 percent cost of capital, and the best case 6 percent. We do not make assumptions about build costs, but instead offer a state-by-state break-even build cost based on a stated investor hurdle rate and the average commercial retail electricity rate for the state.

We use EIA retail electricity rates, which, critically, do not include a consideration of demand / energy split in any given state. We have excluded Alaska, Hawaii and the District of Columbia from these charts, but those regions are on the extremes, one would expect, in all scenarios.

Similarly, we have excluded states where power-purchase agreements (PPAs) are either illegal or unproven, according to DSIRE. Finally, we measure addressable market by load, and not by available space or other technical limitations. These market sizes are most helpful for comparison purposes. 

Mapping the future: 2015 commercial market

Build costs: $2.10

Cost of capital: 8 percent

ITC: 30 percent

Approximate addressable market: 258 gigawatts

Utilizing relatively conservative cost of capital estimates of 8 percent, the United States looks like a relatively attractive place for commercial solar. Developers can build solar at a realistic price and commercial customers can save.

It is a challenging market, but one in which a properly aligned developer can succeed. We utilize best-in-class build costs of $2.10, which is aggressive but realistic for larger commercial systems. We believe that an 8 percent hurdle rate is realistic for larger systems. With these assumptions, the addressable commercial market in the United States is 258 gigawatts in our state-by-state analysis. 

2017 Aggressive scenario: 30% ITC with declining costs of capital and build costs

Build costs: $1.80

Cost of capital: 6 percent

ITC: 30 percent

Approximate addressable market: 437 gigawatts

In the best-case scenario, we assume that the commercial segment will secure acquisition capital at a 6 percent IRR for the investor, and that the 30 percent ITC will not change. We also assume that build costs are reduced dramatically in the next 18 months to $1.80, from $2.10.

This could happen as the industry expands and investors become increasingly comfortable with the asset class, but this is a full 200 basis points below where investors are buying large commercial projects today. Structured portfolios would be most likely to achieve this hurdle for investors.

If the industry can adjust this quickly, reducing both the build costs and also the cost of capital for commercial projects -- and the ITC does not step down -- the addressable market explodes in 2017 to 437 gigawatts, or almost a doubling of market size. Texas, Arizona and New Mexico, all relatively modest markets at the moment, become critical new commercial solar markets.  

2017 positive case: 10% ITC, aggressive drop in cost of capital and declining build costs

Build Costs: $1.90

Cost of Capital: 6 percent

ITC: 10 percent

Approximate Addressable Market: 239 gigawatts

A potential (but optimistic) scenario, would include the step-down to 10 percent in the ITC, and a less aggressive reduction in cost of capital and build costs. In that scenario, we see the commercial addressable market shrink from 258 gigawatts to 239 gigawatts, a small decrease of 7 percent.

We would note that this is an aggressive drop in the cost of capital of 200 basis points, but a fairly realistic build cost as developers scale and equipment costs come down. In this scenario, markets remain fairly stable, with a reduction in penetration. 

2017 base case: 10% ITC, declining cost of capital and declining build costs

Build costs: $1.90

Cost of capital: 7 percent

ITC: 10 percent

Approximate addressable market: 138 gigawatts

Unless there is a policy bridge to extend the ITC, we think this is the realistic scenario for commercial solar. In this scenario, the cost of capital naturally declines as project economics become less reliant on tax benefits. There is also slight decline of capital, and together they lead to a weighted average cost of capital for these systems of 7 percent. There is also a continued drop in solar build costs based on scale and technology.

As a result, we see a decline in the addressable market from 258 gigawatts to 138 gigawatts, a 47 percent reduction in the market. The commercial segment retreats to core markets, including California, the Northeast, and the Mid-Atlantic.

2017 worst case: 10% ITC, no decline in cost of capital and slight build-cost drop

Build Costs: $1.90

Cost of Capital: 7.5 percent

ITC: 10 percent

Approximate Addressable Market: 138 gigawatts

Finally, the worst case scenario is a market in which the ITC steps down to 10 percent and investment hurdles do not change. In that case, the commercial segment survives in a much smaller pool of states, primarily driven by high electricity prices and SRECs.

We should extend the 30% ITC

As we approach the expiration of the ITC, we are able to more accurately predict the impact that it will have on the solar industry, and where the industry needs to improve in order to survive. When asked, many solar executives maintain that they are not worried about the expiration of the ITC, and even go so far as to say that it would be good for the industry. We disagree.

The "good for the industry" hypothesis is premised upon the assumptions that 1) there will be more cash to lever, and 2) there will be lower transaction costs. We won’t argue either of those points.

However, a drop in the blended cost of capital for a project from 8 percent with a 30 percent ITC to even 5 percent with a 10 percent ITC does not yield a higher, or equal, takeout price for the developer, regardless of transaction costs. Nor does it produce a better structured return for a tax equity investor or sponsor. We encourage those of you who think otherwise to model an actual project with whatever aggressive debt terms you can imagine.

We point out that the original reasoning behind the 30 percent ITC also still holds true -- it offers solar operators a rough approximation of the tremendous tax benefit offered to fossil operators, who simply write off their fuel as an expense. 

Fewer markets, but large markets

The good news is that while only a limited number of states will be attractive for developers looking to do commercial solar, those states represent a disproportionate part of the addressable market. Based on current state electricity rates, and current estimated build costs, we estimate the current addressable U.S. market for commercial solar to be 200-300 gigawatts.

With the ITC step-down, and with decreasing build costs, that market is likely to shrink to between 150 and 250 gigawatts (which, it is worth noting, may not be a shrinkage at all). The heat chart below provides some useful parameters for that analysis. 

While there may be a reduction in the current market, we estimate that reduction to be around 20 percent to 35 percent, not a wholesale destruction. We make a number of conservative assumptions about state incentives and the viability of PPAs that will probably be revised in favor of solar over time.

Additionally, as overall construction and development costs come down because of scale and technological development, and as storage technologies enable solar to viably attack demand as opposed to merely energy charges, dormant state markets will re-emerge.  Finally, the addressable market will expand even further as industry participants like ourselves become better at evaluating "off-credit” hosts.

It is clear from this analysis and others that all market participants hold the future of the industry in their hands: smart investors will become more comfortable with the asset class, sophisticated financiers will lower transaction costs, developers and EPCs will streamline their processes, and suppliers will continue to drive down input costs.

We say "will" because the commercial market has such enormous untapped potential: the industry has installed fewer than 10 gigawatts of the hundreds that the market may be able to support. We fully anticipate that the commercial market will be a large part of the solar future.

***

Yuri Horwitz is chief executive officer of Sol Systems. Mr. Horwitz oversees strategic development and managing the firm’s relationships with institutional and private capital clients.

Eric Scheier oversees analytics at Sol Systems, including REC price forecasting, market research, and aspects of project acquisition that require a quantitative approach.