Since the beginning of the year, despite high profile
bankruptcies in the solar sectors, such as China’s Suntech or Germany’s
Conergy, solar indices have climbed markedly, on anticipation that the long
awaited consolidation was coming. For instance, the Guggenheim Solar ETF (TAN
US) is up 72% YTD while the Market Vectors ETF (KWT US) is up 46%. Those rises
barely cushion 5-year falls of 32% and 35% respectively. This morning, rumours that Foxconn, the supplier of Apple, would enter the solar industry also added
to the feeling that we reached a low point for this industry and that the
future would be bright – pun intended.
Should we join the solar party?
Why solar power?
The Sun is pouring a deluge of energy on Earth. We
receive around 120 petawatts (i.e. not reflected by the upper atmosphere) of
energy from the sun. A petawatt is one million gigawatts. In comparison, the
new EPR nuclear reactor from Areva has an electrical power of 1.6GW, which is
therefore 75 million times lower. The earth nuclear reactor fleet comprises
currently 432 units with 68 under construction. Assuming a 1GW output per
reactor, we humans generate 500GW a year with nuclear reactors, which is still
240,000 times lower than what we receive from the sun. Since nuclear is 14% of
the electrical power output globally, we can calculate that the sun delivers
32,000 times more energy than what humans produce! Harnessing this deluge of
energy makes therefore sense.
The average power delivered per square meter on earth is
calculated at 1366W/sqm.
Currently, solar panel technologies can allow for up to
20% of this power to be captured. But this implies peak insolation time, no
clouds and being close to the equator for the best sunlight angle. Therefore, normal
operating conditions yield an average of 100W/sqm with the current technologies.
Subsidies in the
Western world have allowed the solar sector to develop fast.
In Germany, initial subsidies were as high as €0.1kWh
(while they are currently stand at €0.16kWh) against a power price around
€0.05kWh. This has led to a massive build-up of capacity which has been very
profitable for the solar panel chain.
For instance, poly-silicon prices (a raw material for solar
panels, once a tight market) were priced at up to $400/kg in 2007 while they
are now below $20/kg.
A full panel cost €3.5/W in 2007. It is interesting to
notice that the largest solar panel companies experienced operating margins
well above 20% despite spending less than 1% of their turnover in R&D
(pharma companies spend more than 20% in R&D to get to 20%+ margins).
With the financial crisis and because of the massive
installation rates, subsidies were cut and the chain had to adjust its pricing.
This summer of 2013, as I write, panel prices have plummeted to €0.55/W (polycrystalline
panel technology) or €550/kW.
However, panel costs are just a portion of the equation.
Soft costs (permitting, red tape, grid connection fees…) are representing often
more than half the total bill. Germany is predictably the place where soft
costs are on average the lowest, at around €1000/kW. Large project should
benefit from economies of scale and make this fall towards €700/kW. In
California, soft cost can be as high as €4,000/kW. Politics determine a large
portion of soft costs, therefore, the progressive mindset we are seeing in the
US and China should serve the solar installers well.
Economics of solar
today
While prices have plummeted, efficiency of the
polycrystalline panels has barely improved in the last 5 years and they
represent the main technology on offer at an industrial scale.
I believe solar needs to be cost competitive without
subsidies in an environment where power prices look set to fall from current
levels. In the US, power prices for industrials have halved since 2008. Shale
gas and massive new coal supplies will cap long term power prices. Large solar
installation should also put a cap. Overall, energy will be cheaper in 10 years
than it is today.
Let’s calculate the NPV of a non-subsidized 1kW project
with the following hypotheses:
-
A long term price could be around €0.035kWh
-
Hardware costs around €500/kW
-
Soft costs at €700/kW
-
6 hours of effective sunlight per day
-
A 15-year financing with a WACC of 8%.
The first cash flow would be a negative €1200 followed by
15 annual cash-flows of €77.
This results in a negative NPV of €-540 or an IRR of
-0.5%.
Flexing the hypotheses, we find:
-
If the WACC falls to 6%, the NPV becomes €-452
and the IRR is also -0.5%
-
If the financing goes to 20years, NPV is €-444
and IRR is 2.5%
-
If the hardware costs halve, we find NPV at
€-290 and IRR at 2.55%
-
If the hardware costs go to 0, we find a NPV of
€-40 and an IRR of 7%
-
If we do the same over 20 years, we get an NPV
of €56 and an IRR of 9%.
Phew! To be competitive, we therefore need the panels to
be free or soft costs to disappear in a non-subsidized low cost of electricity
world!
What can we make of that?
This is a rough exaggeration but it has the merit to give
us some valuable conclusions:
a) Solar
needs another technology break
With current effective yields below 10%, the technology
is still immature.
Research is helping though. Concentrated solar panel is
being tested on a large scale (mostly in Spain by Abengoa, Acciona, etc. and in
the US by Mid-American Energy, the utility owned by Warren Buffett) and
suggests yields at least twice as high. However, the panel costs are more than
twice the price of current technologies. We therefore need to wait for some
economies of scale.
b) Subsidies
must continue somewhat
Germany poured more than €100bn in the solar sector and
this funding will stop in 2016-17. Japan is starting a subsidy plan for solar
which will require massive support as well. In the US, the subsidies are often
replaced by large tax credits, accelerated depreciation regimes or loan
guarantees for projects.
I would imagine that those subsidies are a bottomless pit
since they support technologies which cannot be economical on their own. I
would be more bullish when states start subsidizing new advanced technologies
such as HPCV.
c) Soft
costs need to be acted upon
Germany leads the world on soft costs thanks to political
will. Soft costs are often half the total system costs and hinder the spread of
solar.
One recurrent criticism of solar is the volume of land it
occupies.
Various researchers have calculated that covering 1/3 of
the earth power needs with solar will require less than 2% of emerged land
against 13% currently used for agriculture. This looks to be a low enough
number to dissuade NIMBY-ism.
d) Investment
conclusions: buy the projects and not the manufacturers so far.
At this point, it looks to me unclear whether the solar
hardware manufacturers will ever be a good investment proposition unless a
large subsidy environment returns. Solar needs a lot of “ifs” to be economical
at the moment as a stand-alone technology and there is a large manufacturing overcapacity
right now.
In terms of investments, I would focus on the following:
-
Equities:
I would buy two types of equities. The low cost poly-silicon
manufacturers (Wacker – WCH GY, REC – REC NO) and the solar plant owners /
installers. However, for the latter, I would differentiate between legacy
plants and new plants. Legacy plants, built before 2011, have high costs and
can be profitable only if the subsidy regime has been maintained. If you look
at Acciona in Spain – ANA SM, you will see how retro-active subsidy cuts have
damaged the equity story. The ideal equity is an IPO of a solar plant about to
be built with the IPO proceeds or built recently. It ensures that subsidies are
at a low enough level to be maintained.
If you know listed manufacturers
of concentrated solar panels, please let me know. It seems that Boeing has developed a technology yielding efficiency rates twice higher than the best panels from the current mainstream technology. Interesting to follow-up...
-
Fixed income: to continue on the solar plant
owners theme, some of those companies are looking for debt funding for new
projects. This is an attractive route to get a decent coupon (between 4 and 5%)
for long maturities (15-20 years) and with collateral. I remember a small
company called Etrion (ETX SS) which is looking for debt funding for a Chile
solar plant. The solar plant has a PPA with a local mining company and is
functioning without subsidies. In this case, the mining company gets the whole
power output so the soft costs are low (no need to link to a main grid, panels
installed on the land concession in the middle of nowhere), there is large irradiation
and Chile is pushing the industry with a good tax regime => lots of “ifs”
which do not reflect normal operating conditions but which are making sense in
this little case.
-
Unlisted: power plant projects will often be
looking for equity investors. Projects with some local subsidies can often
offer IRRs of 12-15% if you assume the subsidies hold over the loan payback
period. These IRRs are typically calculated over 20years. However, panels have a 25y guaranteed life but will probably resist 40 to 50 years, if not 100 years. The value of five extra years of operations is actually significant: since the marginal cost is nearly 0, the cash-flow going to the equity holder can be simplified as [revenue x (1-tax rate)]. Pension funds or endowments should be looking at such projects.
As always, looking for feedback and ideas to explore.
Biography: excellent article here.
No comments:
Post a Comment