The State of the Thin-Film Photovoltaic Industry 54 comments
-
Font Size:
-
Print
- TweetThis
There are currently three major thin film players (defined as the three companies with over $1 billion in thin-film assets): First Solar (FSLR), Sharp (SHCAY.PK) and Applied Materials (AMAT). There are numerous other thin-film startups that are in such early phases of development that it's difficult to analyze potential production and capex costs if their technology ever matures. These are mostly CIGS companies and include Nanosolar, Miasole, Heliovolt, Ascent Solar (ASTI), DayStar Technologies (DSTI) and countless others. Most have struggled to get manufacturing efficiencies over 6% for industrial usage. This doesn't include the aerospace panels which can achieve 20% but are incredibly expensive.
FSLR has been the industry leader for years, but can they maintain their advantages in the face of such fierce competition? All signs indicate that they indeed can. It has been common knowledge that FSLR's technology is about two years ahead of all competitors. Many of the CIGS startups are at the same stage that FSLR was in 2006. They're all operating in the test-plant phase before scaling up mass production.
Cramer has been bullish on FSLR since the summer of last year when it was trading at $50. I've been buying aggresively since it was at $33 and it now accounts for the majority of my invesment holdings. Cramer recently became so emphatic about buying FSLR that he went bearish on all other solar stocks, with the exception of AMAT,, on the belief that FSLR would eat everyone's lunch.
So, according to Cramer, AMAT is positioned #2 in the solar industry. As a supplier of production machinery, they are not in direct competitors with numerous thin-film producers and don't have to worry about raising the heavy capital required for production. AMAT plans on selling $1.7 billion in thin-film equipment by 2009. This is a staggering amout which surpasses even FSLR. At first, this seems like very bad news for FSLR. When digging through the numbers, however, it may turn out to be quite the opposite. AMAT's capex/watt is stated at $3/watt as opposed to FSLR's $1/watt. Sharp's capex has been reported at approximately $2 billion/GW, or $2/watt. This may or may not be accurate, since Sharp is also considering purchasing AMAT product.
Basically, $1.5 billion in AMAT equipment will have the annual production of $500 million in FSLR equipment. The fact that so many competitors are willing to pay triple for AMAT's equipment shows the incredible lead FSLR has on the rest of the field. Secondly, AMAT has stated a goal of getting manufacturing costs of the panels down to $1/watt by 2010. FSLR, by comparison, has a goal of approximately $.60/watt by 2010 according to their grid parity roadmap. As of the end of 2007, FSLR's manufacturing cost/watt was already at $1.12. Unless the dollar crashes further, which would slightly increase costs but greatly increase profits, they should break below $1/watt this year. Another indication of their two-year lead: reports are that 40% of the cost in manufacturing conventional solar panels comes from the silicon. At $1/watt, FSLR's manufacturing costs would rival conventional PV even if their silicon was free. Recent forecasts indicate further polysilicon shortages well into 2009.
With $700 million in cash and the ability to raise billions more, it's hard to imagine any of the smaller competitors chipping into FSLR's two-year lead. The only thing that concerns me is whether bigger players with tens of billions in capital to invest, such as GE (GE) or BP (BP), get into the thin-film game. At their current pace of growth, FSLR's lead may be insurmoutable in a few years when grid parity is acheived. With a three year backlog in product and a 30% operating margin, I feel FSLR has been too conservative in its growth plans. During their secondary offering last July, they diluted shares outstanding by 10% in order to essentially triple output by 2009. I had expected them to announce another stock offering and at least four new 160 MW plants in the first half of this year. This is in line with analyst expectations, judging by 2010 earnings estimates which top $8/share. There's still a chance this might be announced upon earnings next month. The first to reach grid parity will see such explosive growth that their competitors will probably be relugated to "also-ran" status. In the solar arena, grid parity will be the same sort of monopolistic turning point which allowed Intel (INTC) and Microsoft (MSFT) to crush its competition in the late 80s/early 90s. Energy, a $5 trillion annual industry worldwide, is potentially even more rewarding.
Disclosure: I'm a full time investor with over 50% of my holdings in a large long position in FSLR.
Related Articles
|
-
The Seeking Alpha 100
See allThe top 100 stock
market authors
selected for publication
Seeking Alpha 100 » -
Fastest Climbers
Authors with the largestSee all
increase in followers
in the last 7 days
Fastest Climbers » -
Top Commenters
Top commenters,See all
as ranked by their peers
Top Commenters » -
Top Instabloggers
The 100 most active
Instabloggers.-
1
TraderMark
-
2
Don Dion
-
3
Cliff Wachtel
-
4
Mike Havrilla
-
5
James Quinn
Top Instabloggers » -
1
-
Top StockTalkers
TopSee all
StockTalkers.
Top StockTalkers »
This article has 54 comments:
I've begun studying solar stocks and I like your investing style (similar to mine w/ Apple!). Pls recommend links to blogs and other information sources I should use.
Zawy, I doubt FSLR purchases any equipment from AMAT. FSLR has been in the business longer and is very secretive about their proprietary manufacturing techniques. It wouldn't make sense to allow AMAT to develop their equipment and risk them selling it to competitors.
i like FSLR but like other investors i have some issues with the telluium inventories.
one must be very careful when a stock is so richly valued and yet has an issue with it's main raw material' which it doesn't fully disclose in it's reports.
2. CIGS companies do not use Tellurium, which can be an advantage on the long run. (although is too early to invest CIGS, I agree)
3. The question is not that FSLR can achieve aggressiv growth, but how much of this already priced in.
nanosolar looks like a probable winner in the solar sector but it isn't public and not planning to become one at this stage.
WFR is a good way to play the solar game with relative safety.
at this stage the traditinal PV has lot of room to grow and big names like SPWR STP will probably make money for investors.
as the solar world is just starting to cranck one should make a mix of companies as it isn't clear which company will become a winner and what technology will prevail. one can use an ETF or choose few stocks.
don't put all the eggs in one basket since the risk is huge' unless you are a trader and as such use trading system rules.
Which publicly traded CIGS company can be interesting?
cheers,
Turpi
Steve, I guarantee you Nanosolar is unable to produce at $.60/watt today. They've spent hundreds of millions to produce maybe 20 MW so far. They're having all kinds of problems and missing all kinds of goals on their roadmap. I remember when I heard in Feb. 2007 how Nanosolar was going to produce 400 MW at $.60/watt by the end of 2007 I nearly passed on buying FSLR. At the same time FSLR was supposed to turn profitable in 2009. The track records speak for themselves. I also drove by their headquarters and I was not impressed. By 2010 the smaller companies that fail to achieve breakthroughs will find it difficult to obtain financing. For a newer technology to displace an older one it has to be far superior. With conversion efficiencies of 6% and countless manufacturing problems CIGS has hardly proven itself superior. The lower the efficiency the higher the installation costs will be.
It is flexible,sticks like a "band aid" to the roofing substrate,and has been around for a long time relative to competiton. You don't want to be spending a bunch of money on a new solar technology.
exactly what i meant when i said we need to buy more than one stock.
i am almost sure that solar will be huge in the years to come, but i can't tell with certainty who'll be the big winner so i prefer to buy a portfolio and trade part of it on regular basis
i have FSLR on my portfolio but i keep the position size close tp other solar stocks i hold.
for the more advanced traders/investors you can use selling puts on monthly basis till you reach the price that you're willing to pay for the stocks you want to own, if prices seem too high for you. the logic is that in bull markets like the solar there will be wild shakeouts and till they come we can use extreme volatility to "milk" some $ from the market
the last part is only intended to those that know options professionally
p.s. congrats on your conviction for holding FSLR through thick and thin :)
Spectrolab 3j cells are now up to 47% efficient. Distributed CPV systems are expected to amortise at around grid cost when they go into production, probably about the same time we can buy EV's & PHEV's in a few years time.
Together they will reduce our elctricity and transport costs substantially.
That's stunning at only 4 MW/year. The entry may be slightly outdated but it's still an eye opener. Like I said all the CIGS companies have major obstacles to overcome.
This year FSLR will cross the $1/watt threshold which will make their product cheaper than monocrystaline even if silicon were free. With by far the largest margins in the business it's no wonder FSLR is the first solar stock to touch new highs. People used to think STP was the "safe play" in solar because of their low PE. Obviously they were wrong as STP has been hammered this year. If there's oversupply in 2009 and beyond FSLR will be hurt, but not nearly as much as STP and the like. They will be struggling to stay in the black.
Stick a few in the sky over the desert, tether them to the earth, and have a side business selling "shade".
This is why I'm turned off by the CIGS companies. None of them have delivered on their timetables to date. Nanosolar said they would produce 400MW by the end of 2007. Yet with only a few hundred million dollars, the math didn't add up. Sure enough they produced perhaps 1MW. Perhaps the building is big enough to produce 400 MW some day but they forgot to mention it was 95% empty right now. It's the exact opposite of what you have with FSLR which has risen by 30%+ 3 of the 4 quarters they've announced earnings thats to underpromising and overdelivering.
Also, Norsk Hydro (huge Norwegian aluminum building materials producer), just upped their stake in the company to 35%, and they have recently announced some other promising strategic partnerships.
It seems likely to me that if this company can reach an annual capacity comparable to some of the bigger players like FSLR, they will blow away the competition as far as cost per watt. I mean, at the end of 2007 FSLR was at over 200MW annual capacity, and they could not produce modules for a dollar per watt yet...
I wouldn't discount < 200 micron Si as everyone else here has. I've switched from FSLR to CSIQ. CSIQ has been ramping production even faster than FSLR and their projected PE is five times better. They also expect to increase margin by a factor of 3 compared to FSLR just keeping their 45% margin. FSLR and CIGS will not produce enough to satisfy demand. Si has it's place. But if demand does wane, Si can be crushed. That "tripling" of margin i mentioned for CSIQ is going from 5% to 15%. If demand wanes away from the more expensive cells, FSLR will not suffer a bit but the Si players will go negative margin overnight.
It's a fascinating field with a lot of entertainment to come. But they won't be able to supply as much energy as nuclear which is why i'm also in PKN.
I have recently gone to several different Solar Conferences and have become more interested in Concentrated Solar or CSP. Do you forsee and of these companies like Ausra going public anytime soon? Also, a new company called Sopogy, who will be manufacturing Micro CSP panels plans to hit the distributed generation market pretty hard. Have you heard of this company and what kind of insight can you give me regading their strategy of focusing on the CSP market on a distributed generation level between 1-20mw? Interestd to hear your thoughts..
Colin, CSP may be in all the headlines right now with all the huge several hundred megawatt panels, but much like wind the technology has been around so long that the best way for them to reduce costs is simply by increasing system size. I don't think smaller distributed systems are the way to go. CSP may have low materials costs but there are also maintenance costs which would increase with smaller systems. Besides that, you lose power output/acre when building outside the Arizona and Nevada deserts. I've looked at the major alternatives to PV but they all cost at least $2.00-$2.50/watt, including nuclear. At that price there's presently no urgency for CSP as it's not quite at grid parity. However, I'm an adamant peak oil theorist and if Oil were to break above $200 in the next few years, which has been my #1 wildcard reason for investing in FSLR (best return on energy investment among the major players), wind, CSP, EVs, nuclear, and coal would be the only real ways to make up the slack. Attention would shift away from the "future" in PV towards today.
I'm not an expert in CSP but I can't imagine one of them coming through with a breakthrough proprietary technology. So if they did IPO I wouldn't be particularly interested. Colin, do you have any info on the installed costs for distributed CSP?
I've also owned STP, SPWR, CREE, CCJ, and ZOLT over the last two years but in the end I always sell them off when their stocks start behaving improperly e.g. weak relative strength. I would still consider buying SPWR if polysilicon prices drop drastically, CREE if LED prices drop drastically, and CCJ or FXI if I can't find anything more attractive to buy.
What I'd really like to invest in is an Electric Vehicle company like Tesla that can challenge the major automakers. I'm surprised there aren't major EV players out of China or India dominating the market already. Unfortunately they're losing their window of opportunity as the GM Volt and next-gen Prius will come out in a couple years.
www.uic.com.au/nip08.h...
I agree ASTI is really early in the game, especially in light of nanosolar apparently not living up to the hype.
The chart in the link above is interesting: the at $0.08/kWh and quickly rising for gas and oil, solar could NOW be used by power companies as peaking units to offset peaks in air conditioning units during sunny days. Those peaking units are a major reason gas has increased so much in recent years. 10 to 5 years ago, there was a flood of U.S. power companies installing gas turbines as peaking units. I went to one of them being installed and thought "this is insane". I knew even then gas didn't need to be used for that kind of thing. Now flywheels made by BCON could replace those turbines....as well as evening out the output of wind farms.
I agree about the possible gains, which is why I initially invested in ASTI and DSTI last October. Because of the inherent risk, these were small investments. I figured that if the time came when one of these companies became more distinguished as a winner, I would pull my money out of the other company and put it all with the winner (possibly along with more). I have long since identified DSTI as a looser, and pulled my money out accordingly. As for ASTI, I recently upped my stake in the wake of the Bear Sterns debacle. I am glad I did as the PPS has doubled since then.
While ASTI has yet to truly prove themselves, it is looking more and more to me like this is the CIGS company to invest in. They have big money on their side (35% owner Norsk Hydro), they have great management (which so far has not failed to deliver on a single stated goal), they recently announced IOC (initial operating capacity) for their 1.5MW pilot line, and they have amazing and unique technology.
Ascent is the only player who seems to have succeeded in rolling out monolithically integrated CIGS modules on a flexible plastic substrate. Their process does not require expensive vacuum deposition chambers, and unlike Nanosolar, Ascent's monolithic integration technique allows them to roll out CIGS PV at the module level, rather than the cell level. Here is some data from their recent 10K. It's pretty impressive:
google.brand.edgar-onl...
If they can pull off what they say they can, I believe Ascent Solar will revolutionize the PV industry.
1)They fail to make apples to apples comparisons. You have to look at FSLR's installed costs (what it will actually cost the end consumer) to get a true apples to apples. FSLR's equipment requires considerably more electrical components and taken into account its cost advantage is much less , say around 10%. Even if it is more, it is not fair to comapre it to its $2.20 competitor price.
2)The promise of solar is that it would be distributive generation but FSLR's tech is not practical to sit upon a roof top as it needs to large a footprint. Such distribution costs can add 5cents kwh to get it into the grid.
COME ON - A MARKET CAP GREATER THAN 20BIL WHEN THEY ARE ALMOST ENTIRELY RELIANT ON GERMANY FOR SUBSIDIES AND HAVE SO MANY COMPETITORS BREATHING DOWN THEIR THROATS!
ASTI reports an expected Capex of $2/Wp capacity, or $0.30/Wp if they are able to churn CIGS out for 7 years. Converting from peak to actual is about $1.5/W, comparable to nuclear capital expenditure.
Where is ASTI going to get $220M plus $50M in R&D and G&A over the next 4 years?
www.nrel.gov/pv/thin_f...
I assume Ascent will use a similar coating. If anyone here has more info regarding Ascent's addressing of the encapsulation issue, please post a response. It would be greatly appreciated.
I would like to have a 25 kWh flywheel unit to get me through the night if i had a 10 kWp solar cell installation powering the house. Make that 50 kWh flywheel and 20 kWp solar if I need an additional 1 gallon of fuel per day for a car. Nuclear would be a lot cheaper at $300 a month at 15 cents/kWh. For 20 kWp future Si solar cells/thermal solar (you can forget thin film as explained above) might cost as little as $100,000 installed and $50,000 for the $1/kWh storage overnight. This is the nuclear and solar cell cost for the equivalent of 2 gallons of fuel a day for house electricity and avg use of one car: $150,000 loan ($1,200/month for 25 years) for solar or $300/month for nuclear. The average per capita of the U.S. is twice as much, 4 "gallons" a day of energy (including manufacturing, food, etc, everything).
Dr. D. Sarangi
(Email: d.sarangi@sify.com
Email: debajyoti.sarangi@gmai...
The most astonishing thing about FSLR to date has been their manufacturing prowess which often overlooked but cannot be overstated. Less than 1 year ago their lines were rated at 25MW/year. They are already up to 44MW/line which greatly reduces costs and increases revenues. As they were already running 24/7 on 4 shifts/day to achieve 25MW/year/line the gains were obviously major leaps forward. I suspect one reason they haven't announced 4 new plants yet is because they're still trying to tweak out more production from their old ones.
en.wikipedia.org/wiki/...
It's for standard latitude tilt, south facing solar systems. 6-7 kWh/d/m^2 for most of southwest, 5-5.5 for southeast, yearly average. Multiply by 13% efficiency x 365 x 25 yr to get the amount of 7117 kWh/25yr/m^2 for 13% eff modules. A m^2 of FSLR will someday be 13% producing which is 130 Wpeak/m^2 (1 kW/m^2 max Sun x 0.13) and at $3/Wpeak means $260/m^2 for FSLR. For 6 kWh/d/m^2 this yields $260/7117 = $0.0365/kWh/25yr
I used $1/Wp profit to help justify current stock price and as long as there is no competition, they don't need to lower it. We can reduce it to $0.50/Wp to get $0.03/kwh/25 yr. $1/Wp installation might also be possible, bringing it down to $0.024/kWh/25yr
These numbers are for large installations. Small installations might be twice as high, less than the $0.08/kWh you've calculated.
A 10kWp solar system would provide only 1.7 kW when averaged over 24 hour. A household of 2 requires about 6 kW averaged over the course of a day to power the house and car. This would require a 30 kWp solar system at the house. Someday, an installed system at the house might be only $4/Wp including energy storage and therefore cost only $0.05/kWh, but that's a $120K loan to be paid off, something like $900/month on a 25-year loan, and the system is used up by the end of the loan. Suddenly it doesn't look like a $0.05/kWh system (half price from the utlity) anymore. That's the importance of making a distinction between Wp and actual watts: the large initial outlay. So the $2/W capital cost of coal and nuclear can't be compared to the $1/Wp capital cost of solar cells, because that $2/W is 24 hr/day for 40 years but the $1/Wp for solar cells averages out to $5/W over 24 hours for 25 years.
If you have some numbers showing how solar distribution is a lot cheaper than nuclear distribution, I'd like to see them. Likewise, i'd like to see a source that says daytime power is 5 times more valuable.
30 nuke plants are being built right now. Construction itself takes 4 years (from WNAI web site). In the U.S., it's obviously a little more difficult than that.
solarbuzz.com/Consumer...
www.eex.com/en/
The conference brings together financiers, investors, and renewable energy project developers to network, strike deals, and to learn how money is made in the industry. In short, it is a "must attend" event for anyone involved in the renewable energy sector. For more information, visit www.REFFWallStreet.com.