This is a very inspiring and fascinating story about our former colleague Arash Bayatmakou. He suffered a severe spinal cord injury from an accident about a year ago. He's been in a wheelchair ever since without knowing what his recovery might be. This video shows both Arash's incredible spirit and the power of technological innovation.
SolarCity is at a $2bn valuation! What a success story for IPO investors and for the solar industry as a whole. The effects are already felt throughout the industry. Given the way SolarCity’s CEO pitches the company (We are a next generation energy company.), the interest in financing and owning the long-term contract is increasing. This should open financing options to consumers and is a welcomed relief after the expiration of the cash grant created a tax equity bottleneck.
Now, let’s look at this as a public market investor. Is $2bn a fair valuation for SCTY? Given that GAAP makes a mess of SolarCity’s financials and really doesn't help much in understanding its business, Wall Street seems to completely ignore traditional financial metrics. An interesting metric that Lyndon Rive used in the last earnings call was net present value per watt installed, which is currently at $1.12/W. So a rough, back-of-the-envelope calculation would be to estimate the expected growth that Wall Street prices into SCTY given this NPV/kW number: $2bn / $1.12 per watt = 1,785MW which is equivalent to 300,000 residential systems. Since they are currently at 35k-40k residential lease customers, that’s a lot of (undiluted) growth already priced in. Clearly this simple calculation does not take into account all the complexities of valuing a business like SolarCity but it’s at least a starting point. Let’s see how it all plays out.
I highly recommend this article to anyone who is interested in understanding tax equity for the residential solar industry. It's a complicated issue but Josh Lutton at Woodlawn Associates found a way to summarize the key issues in an easy-to-follow fashion.
This Huffington Post article nicely highlights the leapfrogging potential of solar as households go from no access to grid electricity to distributed solar. One million home solar systems is a remarkable achievement which will hopefully lead to more economic opportunity for many.
I am hopeful that more people in the world will benefit from access to affordable clean electricity in 2013.
We are near the finish line for SolarCity's IPO. Since my last post, the company released their price guidance of $13-15 per share. At roughly 10M shares offered this means a decrease of the target capital raise from $200M to $150M. At a $1BN market cap with stable post-IPO gains, this could still be a successful solar IPO. Within that target range, SolarCity would be the second most valuable solar stock (after First Solar).
So how should a public market investor think about valuing SolarCity? The drop in hardware prices has been helping the downstream solar market. However, that doesn’t mean that SolarCity is sheltered from intense competition we’ve seen in other parts of the value chain. (There are plenty of companies competing with SolarCity in each of its verticals.) I think there are three relevant components to hone in on to arrive at a reasonable valuation: (1) the NPV of the future cash-flows of their lease contracts (net of selling, building, financing and O&M costs), (2) expected growth rates of new lease contracts, and (3) regulatory risk factors.
Before dissecting each of these, a few quick thoughts on why other things don’t matter. Firstly, being “fully integrated” is not worth much in itself except maybe the brand value of having big green trucks drive around which could help with customer acquisition and therefore positively impact (1) and (2). However, the fixed cost of carrying that sort of infrastructure is scary. Any significant decrease in sales would mean idle resources with adverse financial effects. Secondly, parallels to Tesla (just because Elon Musk is involved in both) are overemphasized in the press. Maybe you can infer a certain aggressiveness on how much risk the company is willing to take. Thirdly, using Real Goods Solar as a comp (the other publicly traded residential solar company) is pointless. RSOL does not offer their own financing which is where all the juice is for SolarCity. There are no solar industry comps for SolarCity. If you are looking for comps, looking at consumer finance companies would be more helpful.
(1) NPV of Lease Contracts
Without the long-term cash-flows of its leases, SolarCity would probably be valued at a fraction of $1BN, especially in the light of the struggling solar stocks everywhere. But investors love the certainty of long-term contracts. During their road show, SolarCity stated that the NPV of an average lease contract is $15k, which struck me as high when I first saw it. But digging into the assumptions, it became clearer how they have arrived at that number.
They assume a low discount rate of 6%. Given the lack of default history, nobody truly knows what a reasonable discount rate is. And by "default" I don’t just mean customers defaulting on their payments. Investors have to take into account the impact of non-performing assets since that risk is largely on SolarCity (panels with worthless warranties from bankrupt solar companies, trees overgrowing that nobody wants to cut down, settlements for mis-selling, problems around transferability when a house is sold, etc.). Therefore, I would be more comfortable with a discount rate of 8-9%. This would have a very material impact on the NPV (as it reduces it to $9-11k).
The other aggressive assumption in the calculation is the terminal value of the asset after the 20 year contract runs out (they call it “Renewal Option”). In their road show slides, SolarCity assumes over $17k of future cash-flows (at “year 20” dollars) which discounted back to today is $5k. Since the customer can (among other options) ask the system to be removed at SolarCity’s cost, some systems might have negative value to Solarcity. And if there is perfect information, the customer knows that they have a very strong negotiation position to bargain for a lower rate. But again, nobody really knows what will happen after 20 years. If I had to pick a number, it would be closer to $1-2k in PV. Taking these two together, the NPV of an average customer would be about half of the $15k.
The other big uncertainty is the true cost of customer acquisition. Everything below the gross margin line in the P&L is a big mystery. They claim the majority of those costs are investments in future growth. It’s hard to tell with the disclosures given.
(2) Future Growth
Currently SolarCity is sitting on roughly 35,000 leases. At $15k NPV, that would equal $500M in enterprise value. At $9k, it’s $315M of EV. Factoring in future growth, you can see how SolarCity and its underwriters came up with $13-$15 per share (or was it the other way around?). I think it’s realistic that the company can add 2-3x of new contracts over the next few years. However, can they do so at a reasonable customer acquisition cost? And how much more capital will be required? The company is swallowing up a lot of capital and additional capital can lead to significant dilution to equity holders (or levering up of the balance sheet). SolarCity’s brand is the big x-factor to its growth strategy. On the one hand, brands can be very powerful to fuel growth and create shareholder value. However, brands are very expensive to build and continuous investment is necessary. It’s hard to think of a residential installed product with a branded construction experience. ADT, the home alarm business, might be an analog but customers value ADT’s brand because of trust that it will protect them. I don’t know of any big residential roofing or electrician brands.
(3) Regulatory Risk
Any company that heavily relies on government subsidies is exposed to regulatory risk. Additionally, any business in the energy industry has to live with changes in the regulatory environment. Below I’m highlighting the key risks to SolarCity:
As an entrepreneur, I am comfortable taking these risks. As a public market investor, they have to be taken into account when comparing SolarCity to other investment opportunities.
The macro story of the residential solar industry is strong and I am confident that overall industry growth rates will stay high for the next few years. Costs are still on a downward trajectory that will ultimately make solar economically viable to more and more Americans. SolarCity’s business model might have the winning formula if they can acquire positive NPV customers (with reasonable DCF assumptions). And do so at scale. From the information I have seen, it is difficult to tell what the fully loaded cost of a new customer is. As an investor I would also be concerned about SolarCity’s thirst for capital over the last years and its sustained losses. Our experience has been that SolarCity has always been pricing their product very aggressively in the marketplace. If that low pricing is not sustainable, increases will negatively impact growth rates and can put SolarCity in a tough spot given its high fixed cost structure. Overall, I think SCTY might be an interesting investment but not a $1bn valuation. Especially considering the negative market sentiment towards all things solar. I hope the above framework will help anyone who is interested arrive at a reasonable valuation.
One of the more intriguing sections in Solarcity’s S-1 are the details around the rather unusual investment structure of the Series G venture round. In Q1 2012, a group of investors (mostly Elon Musk, SilverLake Kraftwerk and Valor Solar Holding) invested $81M in Solarcity at a valuation of roughly $1.3BN. You might cringe a little at the valuation but when digging into the S-1 you'll find that the Series G group negotiated a very nice structure to protect their investment. Here is the fine print from the S-1:
“Each share of Series G preferred stock is initially convertible at the option of the holder into one share of our common stock. Pursuant to the terms of our amended and restated certificate of incorporation, upon the closing of this offering, each share of Series G preferred stock will automatically convert into a number of shares of common stock equal to the quotient obtained by dividing (A) the original issue price of $23.91 per share by (B) the product of (i) the public offering price in this offering (before deducting underwriting discounts and commissions), multiplied by (ii) 0.6. However, in no event will one share of Series G preferred stock convert into more than approximately 2.47 shares or less than one share of common stock as a result of this conversion adjustment mechanism. As a result, the outstanding shares of Series G preferred stock will convert into no more than 8,372,069 shares and no fewer than 3,386,986 shares of common stock.”
When I first read that my head was spinning. A glass of my favorite whisky (Caol Ila in case you were wondering) and a spreadsheet helped me figure it out. Here is an attempt to explain it in plain English:
The Series G group created a partial-hedge for their investment. If the IPO price is between $16 - $40 per share, their investment will be worth $135M at IPO no matter what. A very nice cash-on-cash return of 1.67x in under a year. If the issuance price is below $16, their investment will end up somewhere below $135M. (They would still get their money back if it’s as low as the Series F round of $9.68.) If the issuance price is above $40, their investment would be worth more than $135M.
A few implications:
In my next post I’ll share a few points on how to think about valuing Solarcity’s IPO.
I’ve often said Solarcity’s IPO will be a watershed moment for the residential solar industry. After all Solarcity is a market leader, has raised piles of money and will be the first residential financing company to go public. On Friday the big day had arrived and Solarcity published it’s S-1 (you can read it here).
From other companies in the industry we know that the installation of solar systems is brutally competitive. A quick look at RSOL’s financials gives you a good idea: they lost $4M on $21M of revenues in Q2 2012 and currently have a market cap of $15M which is well below book value. Basically, the market is assuming the company will stay unprofitable and any value left will go towards debt holders.
So how can Solarcity avoid getting thrown into the same pot? They will be pointing towards their fully integrated business model, their strong brand and the recurring cash-flows from their residential solar leases. In this post I will focus on what I think is the most interesting of the three: the profitability of their residential lease product. Especially when considering that Solarcity has often been selling cash-purchase solar systems at negative gross margins over the last few years. Great for customers, crummy for investors.
But since most of Solarcity’s business is coming from solar leasing, understanding the economics of those leases is crucial. Here is a key disclosure made in the “Management’s Discussion and Analysis of Financial Condition and Results of Operations”:
The gross margin on the leased systems looks very high (85% and 68% respectively) which intuitively does not make sense given the highly competitive nature of the industry. To better understand this disclosure, we have to dig deeper into the financials and the accounting treatment of operating leases.
Solarcity’s revenues are recognized over the term of the lease (usually 20 years) whereas the main capital costs are depreciated over the useful life of 30 years. Customers have different choices at the end of the 20 year lease agreement (including having the system removed from their homes) and there is no data on what will actually happen (the first leases will expire in 2027). Depending on which option the average customer chooses, GAAP might produce inflated gross profit margins on these leases. The effect could be significant: if all costs were to be recognized over 20 years, Solarcity's gross margins could be 30-50% lower.
Unfortunately, that still doesn’t paint the full picture of the profitability of Solarcity’s leases. The consolidated financial statements create a complex web that is very difficult to untangle. They leave us guessing how the economic rent is divided between the third-party fund investors and Solarcity. Without knowing the economics (NPV, IRR, etc) of one lease unit or one investment fund, it’s unclear whether Solarcity is currently widely unprofitable because it’s investing in future profitability or whether it’s just generating unprofitable business. It will be fascinating to see how public investors will receive the stock. It will largely depend on whether they will buy into the company's growth story and whether they can get comfortable with the accounting and profitability of Solarcity's leases.
Last year I wrote about the cost of residential solar electricity. It included a sensitivity analysis that links financing and installation costs to the levelized cost of distributed solar electricity. We always expected that both of those input costs would be declining over time through the virtuous circle of a growing and competitive solar market. As the market grows, costs will be driven down through economies of scale and competition. Lower costs means a larger addressable market as more and more customers can financially benefit from distributed solar electricity. And on it goes. However, as installation costs have fallen dramatically (by 30-50% in the last 3 years), financing costs have declined relatively modestly. As financing has a very significant impact on the future state of the industry, it’s worth exploring why that side of the cost structure has been so sticky.
There are two main drivers of financing costs in the solar industry: (1) how cheap the capital is that is used to finance the capital outlay to build the system and (2) how intense the margin pressure is on the financing providers:
(1) Cost of Capital
Cost of capital in the solar industry is a complicated beast. The financing mechanism is very complex and requires paying lawyers and accountants a lot of money to set up. (Those fees easily add up to over a $1m per fund.) The fund complexity and lack of long-term data is making the asset class less interesting for many conservative investor groups like pension funds, large companies or many banks. Additionally, as we are moving from a cash grant to a tax equity world, the supply of capital will continue to be restricted. As long supply is limited, costs will not decline as rapidly as install costs have.
This one is closely linked to the above: as long as it is difficult to put together the financing structures to finance distributed energy assets and as long as large financial institutions are waiting on the side line, it will be difficult for new solar financing companies to enter the market, innovate and drive down economic rents. Over the last couple of years we have seen a few companies like Sungevity, Suncap and a few others try hard to dethrone Solarcity and Sunrun, but they have not managed to hold on to temporary market share gains. Others like SunEdison have exited to market quickly.
So where are we heading from here?
Given the complexities around monetizing the tax credit, I don’t predict financing cost to decline significantly in the next 12-18 month. Over time, increased interest in this fast growing market, more operating history of the asset class and innovative financing approaches (including securitization) should drive down the cost of capital. After all the risk-adjusted return profile of this asset class is very attractive. Especially in this low-interest world we will be in for the foreseeable future. The result will be lower cost of distributed solar electricity to the end consumer and more markets reaching grid parity.
Last month Germany set a new world record. According to the IWR, the total solar energy production reached 22 gigawatts during the cloudless day of Friday, May 25th, 2012. No other country produces that much solar electricity. (Total US installed capacity was approximately 4 gigawatts at the end of 2011.)
Cleantechnica reported a more relevant number: on that day those 20+ gigawatt of solar installations produced 190 gigawatt hours of electricity which equals about 14% of total electricity consumption in Germany. Looking at the chart on the left, you can see that PV solar is making up a steady 8-12% of total weekly consumption in the month of May. Solar is no longer a niche electricity generation but is now in the same league as coal, gas and nuclear. The next challenge is how to incorporate this increasing generation capacity into the grid. That is a major problem that German policy makers and power companies are now forced to figure out. The rest of the world's energy industry is watching closely.
It's been an exciting week. We just announced that Gen110 received venture backing by Kleiner Perkins and got a ton of press about it. Here is an example: "Venture Backing for Distributed Generation Firm Marks a Shift".
So what has changed since the funding? Not much really. It's obviously an honor to be part of the Kleiner network. Plus the (self-inflicted) pressure has certainly increased. But success in this industry comes down to day-to-day execution, setting up the company for long-term success and withstanding short-term shocks. The residential downstream market is getting more and more competitive as installation and financing costs are coming down. As less reputable new entrants are trying make quick profits in the short-term, KPCB's backing allows us to think longer term and build a distributed energy powerhouse. It won't be easy but having KPCB behind us is a critical step to success.