This post was originally published on this site
I am an engineer by education, business manager by profession, and stock investor by passion. As an investor, I share the concern about the historically high market valuations for stocks.
An investment into ‘safe’ bonds would struggle to keep up with inflation. So, where to put the money?
As a business manager, I understand that only a true value proposition for the customer can sustain a business.
As an engineer, I have passion for new technology and an ingrained belief that scientists and engineers can solve a great many of the world’s problems: Replacing fossil fuel being one of the largest ones facing the world. Solar technology is one solution touted and has been a pretty hot space the past decade. However, is solar a good investment?
I like technology companies and have invested in quite a few. When doing so, I stick with Thomas Alpha Edison:
As for solar cells or other clean technology, unless they sell widely, proving their utility, they won’t make an impact. Solar is somewhere between 1-2% of the global electricity market.
One can violently debate whether global warming is real, but it only takes a trip to China and breathing the persistent smog there to understand that coal burning isn’t healthy for the environment or for the population. So, let’s just agree that less pollution is desirable. To make a real impact with solar, it has to become economical enough to penetrate a large part of the market.
I have come to appreciate that it’s way expensive to buy into new technology (no matter which one) as a first adopter. Not only is the technology expensive, it also hasn’t matured yet via market feedback. Typically, version 3 of anything has worked out most of the bugs. Version 1 gets rushed to market as ‘good enough’. In version 2, they fix the known issues, and in version 3, they fix the unknown issues they discover through market feedback. Cars tend to be more reliable, software is more stable, cell phones have their early issues worked out etc.
Cost, particularly when it comes to electronics and semiconductors, tends to come down exponentially. So, if you wait for 3-5 years, you get a much better product for half the price.
Which brings me to solar.
Saving the world with renewable energy is a great idea. But what does it take to sell it? And, if you can sell it, is it a good investment?
Bloomberg published the following graph of the price history of solar cells:
While forecasting a selling price reduction of solar cells from $76.67/W in 1977 to under $10/W just a decade later might have been conceivable, as the technology was young, I don’t believe even the most aggressive companies would have thought $0.36/W was even remotely attainable, nor profitable. What you see happening is a bust cycle of over-investment into solar cell capacity, particularly in China, and a subsequent disruptive price drop starting in 2009. I will agree that $0.36 is attainable but not really profitable.
Would you have invested into solar companies, if they had shown you this chart? Should you now? I believe it is bad news for the makers and their investors but excellent news for the consumer and the planet.
Solar cells are a semiconductor industry
The semiconductor industry has lots of examples for these exponential price erosion curves and a quite volatile history of profit boom/bust cycles, as a result. It takes huge capital investments to build capacity against rosy projections. When the capacity comes online simultaneously, overcapacity materializes and prices immediately drop down near variable cost, leading to P&L losses. Investment stops, low prices drive market penetration and start filling up the lines. Some profit emerges and the investment cycle starts again. This has played out e.g. in the D-RAM market when computers started to take off, and more recently, you can see it in the so-called ‘crystal cycle’ for flat panel TVs. Prices for consumers come down dramatically but profits remain elusive for the manufacturers.
Which is why the stock price development of First Solar (NASDAQ:FSLR) looks eerily similar to Intel (NASDAQ:INTC) over the first 10 years of their life. History in the stock market might not repeat itself, but it for sure has some parallels.
Given this insight, how can one make a profitable investment into solar?
Solar as a solid investment
Solar has been tickling my interest for several years now, but I patiently waited for the industry to mature and the costs to come down. I wanted a payback of less than 10 years. Why 10 you ask? I can reasonably expect to stay in my current house that long, the system will still be under warranty, and I don’t expect any components to break, so it’s a reasonable but not aggressive bet.
This year, panel pricing crossed the threshold, and I decided to go ahead with a 12.9 DC kW installation.
Several factors drove the decision:
The expected payback with the latest panel pricing after tax credit and utility energy rewards program, assuming a 3.2% annual electricity cost increase calculated to 9.2 years. Even if energy costs weren’t going to inflate, it would pay off in about 10 years. 3% inflation has been the long-term average in the US. Energy costs will rise, not the least because of the mandate to put renewable into the mix. (More on that in another article).
The stock market has been on a tear for the past couple of years and is clearly at inflated levels, which would lead to the expectations of lower returns going forward for the next decade.
I wouldn’t be surprised if the current administration prematurely guts the federal solar tax credit of 30%, although it had been extended until 2019. Without that, despite the astonishing fall of panel pricing, the paybacks don’t hold, so the window of opportunity might close.
Financing rates are still low for those that would need it.
With the soft cost of installation now at about 60%, the potential for further dramatic price reductions seems reduced.
Plus… as an engineer I REALLY REALLY want it!!! … but as a business person, I can’t succumb to a ‘want’ at this size of investment. Rationality will have to prevail, and a reasonable return on investment has to be provided.
So, here is the hypothesis I will test out in real time:
In 2017, it’s a better idea to invest in solar panels on my roof, generating my own energy, than investing an equal amount into the stock market (SPY ETF as a proxy) and paying my energy bills out of the stock market returns.
Really? Solar panels will beat the stock market? Yes, they will, and with lower anxiety levels and volatility to boot!
The Solar ‘Utility’ SOL-1
So, meet my private ‘utility stock’ SOL-1 of which I am CEO, CTO, joint shareholder with my wife and kids, and major customer:
Our annual energy usage last year was 14.930 kWhrs for which we paid $1,833 at standard local utility rates (12.3 cts/kWh). Actually, we paid a bit more, as we are subscribed to wind source, but I will not include that into this analysis.
We are installing a 12.900 DC kW system, which is projected to produce 13.000 kWhrs and save $1,713 annually, essentially covering almost all of our needs, and with a bit of good weather, possibly all.
Solar panels will be installed on the east, south, and west side of our roof, as the south-facing side doesn’t have enough room. East and west panels will be only about 80% efficient as the south side, which has been accounted for in the projections. If I could have everything facing south, we’d be about a year faster in payback.
Before we get into the projections, some of my big insights:
The power of compounding inflation:
If energy prices are going up 3.2% per year as projected, the electrical bill in 10 years will be 33% higher than today. Generating energy ourselves, we have actually a built-in hedge against inflation.
I am essentially running a dodgy utility company with some admittedly ‘sexy’ technology. I expect steadily growing ‘utility’ type of returns.
Solar panels degrade over time.
These ones are guaranteed for no more than 0.6% deterioration per year for the next 25 years retaining 83% of their capacity. Of course, the solar panel company needs to be around to honor this, so put your investor hat on and do some due diligence on who might live this long in the current market environment. With China’s producers being pumped up by their government, it might be tough to bet on local manufacturers. I will be using Korean panels made by Hanwha (NASDAQ:HQCL). Hanwha recently encountered some tough times, and I expect that its panels will be a better investment than its stock, but it is large enough to weather this storm. Suniva customers are likely out of luck if they have a warranty issue, as Suniva filed for bankruptcy. Great technology, but it didn’t make it. Be careful who you choose.
Solar panels are long-term investment, I mean looooong. But these are the decision horizons for typical utility companies. A coal or gas-fired plant has 30+ years of life expectancy.
For that time frame, my inner CTO (chief technology officer) chooses proven i.e. reliable technology over ‘first adopter’ type of innovations. While there are new technologies out, reliability scores higher for me than unproven ‘frontier’ ingenuity. Think about it as an equivalent to the Mars mission. This system will have to run for 25 years, and you essentially do not want to worry every week whether things will work. Panels don’t work, you don’t save money. And, having an installer on your roof every other month would be a major hassle too. This should be installed and churn out energy reliably without any intervention. Solar panels in space do, but they don’t face the earthly weather.
Inverters are the expected weak link. Warranty is 12 years for SolarEdge (NASDAQ:SEDG), which is the leading provider in the industry. Extended warranties for 25 years can be purchased for $750. Warranty for the solar panels is also 12 years on workmanship and 25 years of performance. The performance guarantee only applies to panels that haven’t failed completely. Bit of a catch there. If the panel underperforms, it will be replaced. If it dies after 12 years, you have to buy a new one. SolarEdge provides a free monitoring app that monitors energy production and alerts me and my installer to any issues. That app is also tracking panel degradation for the panel warranty.
I am not a fan of extended warranties, as they are designed to make money for the provider, and chances are if they sell you one, it will last exactly as long as the projected failure rate is low and expire before age-induced failure rates go up. It is wasted money. Try to negotiate an extension for low/no extra cost. The models assume a 25-year warranty on inverters. Anything goes wrong, they will be replaced.
Incentives paid out by the utility must be reported as taxable income, so your tax bracket matters. Your upfront investment is net after taxes, the utility incentives are gross before taxes.
If your system produces more energy than you use, the same applies. You generate taxable income with every check the utility company sends you. Unlike dividends, it’ll be at your full marginal tax rate.
When there is a power outage, your solar panels will automatically shut off to protect utility workers. I am all for protecting those hardworking individuals that braved many snow storms to restore power. But in the event of a power-out, the sun might shine, alas you are still out of luck, out of power and out of revenue. The only way to cure that is to install batteries and go ‘off-grid’. Unfortunately, batteries don’t follow a semiconductor cost curve, as they are mostly chemicals, so that will still be a while (Think a decade rather than years).
WAF: Wife Acceptance Factor. Key critical that your significant other agrees with the looks. The gamut in my family runs from ‘ugly’ to ‘cool’. Fortunately, my wife is on the ‘cool’ side. Don’t plan your system until you have agreement on the looks.
House owner associations are reportedly even less enamored with the looks of solar panel, so if you are in one, make sure you secure approval early.
Ground-based installations could require neighbor approvals, rooftop installations typically don’t.
Solar panels are coming down in price rapidly, but the rest of the installation costs remain fairly steady.
The NREL (National Renewable Energy Laboratory) keeps track of cost and, as of Q1 2016, has reported a $2.98/WDC on an average 5.6 KW residential system with the module cost being ~ $0.60/WDC or about 20%. Labor cost for project planning, installation costs, and the inverter pricing have barely moved over the last 2-3 years. Installation hardware and inverters are now almost 50% of the material cost.
If that trend holds, every 10% drop in module pricing only affects installed cost by about 2%.
We see an asymptotic curve here, which makes me comfortable that there isn’t as much to gain by waiting as there was in 2009. A 40% price drop for modules over that period still seems remarkable.
Where the stock market and installation cost were 2009, I am certain I could not have justified going solar.
The cost for our system calculates to an equivalent $2.12 /WDC after Federal Tax Incentive. Partially, this is driven by a larger system, which spreads out the fixed costs but also by the recent glut of solar panels, which has drastically reduced panel pricing. Quite a few suppliers are producing at cash cost, so the room for further dramatic price decreases has shrunk quite a bit. However, there is no question that the system will further come down in price. The counter trend to that is that incentives might get curtailed going forward.
All important Cash Flow for 9% IRR
Here is the forecasted cash flow for my ‘power plant’ as provided by my contractor:
It assumes 3.2% annual energy cost inflation, a 28% Federal and 7% State Income Tax Bracket (applied to the utility incentives), and a 30% Federal tax credit in year 1. Solar panels deteriorate at 0.5% per year. The system will offset 91% of the annual power consumption, which per my contractor, is conservative and includes a 3% adjustment for shading as well as adjustments for less favorable East-West facing panels.
Our utility will provide an additional 8cts per kWhr produced for 10 years incentive, which calculates to about $10.620. This is paid whether I use the energy myself or give it as surplus to the net. Energy consumption is net metered at consumer retail pricing of 12.3 cts/kWhr. So, I am selling my energy to the utility at what they charge me, utilizing their infrastructure as my ‘battery’ powering my household at night and in bad weather, and I get 10 years of incentives on top. In our state, there is no sales tax, no increase in property tax. This might be different, where you live.
The Internal Rate of Return (IRR) for 25 years calculates to 9%. If we assume no electricity cost inflation, it ends up being around 8%. This is NET, as I won’t pay taxes on this, assuming I don’t overproduce.
Not too shabby compared to a long-term expected stock market return of 6-7%, which comes with a 15% tax on dividends and long-term capital gains of either 15% or up to 23.8% for high income earners. For all calculations, I’ll be assuming a 15% tax.
The SPY ETF as a proxy for the S&P 500 currently yields about 2%. For the dividend investor that is looking to get electricity ‘free for life’, she would have to invest about $100.000 or so into the SPY to cover annual energy costs, assuming, that dividend increases keep up with the general inflation. SPY’s 20-year ROR was 6.5%, including dividends according to Fast Graphs.
‘But wait’ you say, the system loses value over time, the SPY goes up. So, my stocks will retain additional value. Plus, I can sell my ETF anytime and/or move. With a solar system, I am tied to the house. Very true indeed! This is strictly for a long-term investor, as you ‘prepay’ your electricity bill for 10 years before the benefits really kick in. So, this isn’t for the short-term investor.
Which raises the question, what the residual value of a solar system is, depending on its age?
Let’s, for kicks, assume that the system is worth NOTHING after 25 years (although, clearly, if it still produces 85% of its initial production in year 25, it is unlikely it would drop to zero the day after).
How does the net present value of the future cash flows compare? NPV essentially discounts future money streams to the present day.
Call me surprised, when I realized that even without any incentives, the investment breaks even over 25 years. Not that I would make that investment, but then, electricity cost can be as high as $0.33/kWh in Hawaii, so it is highly attractive there. One of the reasons their utility is in trouble.
SPY vs. Solar Utility
Now, let’s assume we invest the same amount of money into a SPY ETF as the initial cost of the solar system minus the Federal tax credit, so $25.375. The dividends would be $431 (($25.375×2%)-15%) per year after tax. Clearly, not enough to cover utility bills of $1,891 per year. So, we would need to sell off some shares to cover the difference. With 15% capital gains tax and not counting transaction fees, that amounts to $1,459 in the first year that I must sell. In future years, this number will be escalating with energy cost inflation and for the fact that I am generating less dividends on my now reduced principal.
Let’s look at the results here assuming yield stays constant at 2%, and ROR, excluding dividends, is 5.5%. The value of shares sold includes 15% capital gains tax to be deducted.
And, there you have it: By the year 20, the principal funds are exhausted. The residual value of the SPY portfolio is zero.
What is the likely value of the solar system after 20 years? Well, it should be the net present value (NPV) of future cash flows starting with year 20. Assuming again the solar system lasts exactly the warranty period of 25 years and then inactivates, the NPV assuming a risk-free return of 3% is $16.143. If you want to be picky and pay for taking it off the roof in year 25, maybe it’s $14.000.
I would argue that you should get at least some of that value reflected in a higher property selling price.
The same argument holds if you sell earlier.
In year 10, the NPV of future cash flows is $31.000. To break even with the SPY investment, you need someone to give you $19.900 for the ‘utility’, when you sell your house. Which is equivalent to about a 10% discount rate. Who wouldn’t love a guaranteed return of 10%?
So, at any given time, as long as you get a reasonable payback when selling the house, you should come out ahead. The sale of your house could also be tax-free, but you do pay brokerage commission.
Risk of Return Sequence: A 41% stock market drop?
Now, this is a pretty ideal scenario for the stock market. It keeps humming along, unfazed by its current high valuation. Not likely for the next 10 years.
The Shiller PE10 ratio of the S&P 500 stands at 28.7 vs. a historical mean of 16.7. In other words, the market looks overvalued right now. That, in general, tends to lead to lower returns going forward. As we are withdrawing principal, we are exposing the SPY portfolio to sequence of return risks.
Unlike utilities, or electrical bills, the stock market tends to be volatile, and when withdrawing principal, starting out with a bad return year has an outsized effect.
Let’s assume for a moment the following: The next three years, the market corrects part of its overvaluation. A full correction to historic valuation averages of ~ 16.7 P/E could result in an admittedly gut-wrenching 41% drop, which would feel like 2008 all over again.
Let’s take a more moderate scenario: Earnings keep growing at 5.5% per year. But multiples contract towards a more reasonable level. In year one and two, the stock market valuation drops by 10%, and in year three, by 5%. With a 5% earnings growth, it would bring the PE down to about 18.6.Then, the stock market starts growing again at its average of 5.5%. That isn’t doomsday scenario and is certainly in the cards, given the current high valuations. Bear markets tend to have drawdowns of 20-30%.
Here is what happens with the portfolio:
It depletes by year 12.
Solar looks to be a low-risk, guaranteed return alternative that will beat the stock market represented by the SPY ETF over the next 12-20 years.
In order to test this hypothesis out, I will track actual cash flows and earnings from my solar system vs. a SPY model portfolio and report back to the readers at regular intervals. Stay tuned for part 2, where I report on the actual installation and plan to further stress test the model by looking at my own ‘utility’ company vs. other investment opportunities.
Disclosure: I am/we are long INTC.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.