First Solar FSLR S
December 23, 2007 - 4:36am EST by
2007 2008
Price: 266.40 EPS
Shares Out. (in M): 0 P/E
Market Cap (in $M): 20,770 P/FCF
Net Debt (in $M): 0 EBIT 0 0
Borrow Cost: NA

Sign up for free guest access to view investment idea with a 45 days delay.


Like a moth to flame - I am attracted to shorting stocks that are rising extremely rapidly and have unbelievable upward momentum supporting fancy valuations.  I say this because shorting in this manner can sometimes (often?) be a trying experience.  As proven by the dot com boom, there is no telling how long the music will play - before it ultimately stops.  First Solar's valuation is already disconnected with the reality of several aspects of its fundamental business prospects.  It reflects unmitigated optimism and a supreme confidence that First Solar will be the dominant and surviving force in a huge new growth industry - solar panels.  This despite the reality of many competitors in a market in its still early throws of innovation and trialing of new ideas.  As a short idea, it has the additional merit of being one of Jim Cramer's favorite long ideas (any chance it many be overhyped favorite among momentum investors right now?).

As discussed below, however, FSLR has a number of challenges to over come to justify a valuation like the one it is currently being awarded.  While FSLR has great sales and earnings momentum right now, and a certain amount of locked in revenue from a small number of customers - mainly in Germany - there are a number of significant issues that the company will have to deal with at some point, any one of these can derail investor's wildly optimistic projections and lead to a significant multiple contraction (from the current level of approximately 133x 2008 Calendar Year Estimates of around $1.99 per share. 

In typical fashion, investors are linearly extrapolating FSLR's current business growth.  However, for a number of reasons, growth in 2009/2010 could end up disappointing investors.  In addition, solar has become a buzz world that investors are chasing today - sort of like dot com in the day.  Bankers are out trolling for companies to change their names to X SOlar or Solar X for the advantages that that could represent on flogging additional IPO or secondary offerings on a public hungry for "clean energy" and "renewables" based technology companies.  We all know how this story goes.

The Business

First Solar Inc. is a Phoenix, Arizona based manufacturer of solar electric (photo voltaic) modules (aka Solar Panels - panels that convert sunlight into electric current).  There are a number of companies worldwide that manufacture solar modules that have been producing and selling these modules around the world.  Some of these companies are public, while many are currently private.  A partial list of comapanies in the same basic space would include:  JA Solar Holdings (JASO), Evergreen Solar (ESLR), Solarfun Power Holdings (SOLF), Subpower Corp (SPWR), and Suntech Power Holdings (STP).  Private companies (including small divisions of large public companies in this business include BP Solar, Sanyo Solar, Kyocera, Motech, Nanosolar, Shell Solar, Mitsubishi and others.

Essentially, First Solar is a manufacturing company.  Their main (only) product is a 2 foot by 4 foot solar module which can be linked together with many other modules to produce electric current when exposed to sunlight.  The basic module (of 8 square feet) produces an average of 64 watts of electricity when the sun is shining.  Since solar power is based on "free" sunlight, the costs of building a large (or small) solar power plant include the following: the cost of the solar modules (bought from companies like First Solar), the cost of additional equipment (like inverters) necessary to take the electric current (which is direct current) and convert it into alternating current (the type we use on the power grid), the cost of the space (if any) required to place the panels, and the cost of labor necessary for the installation and maintenance of the "power plant."

The economics of Solar Energy

The key to understanding what is going on with the growth of Solar is to understand the economics of solar energy.  As described above, the cost of producing energy from solar panels is based on the capital costs.  This is so because the unlike "conventional" coal burning or gas fired power plans, solar panels do not require any consumable to produce electric current.  However, the capital cost per watt is higher upfront as illustrated below (all costs per Watt)

The cost of the solar module to customer:       $2.48  per watt (wholesale price - retail price would be higher)

Cost of Inverter, Grid Tie, etc                        $2.72 per watt

Total Cost per Watt                                     $5.20 per watt

As an aside, the retail price of solar modules in December 2007 was closer to $3.80 per peak watt because their is a retail market up to inventory, distribute and sell panels to small customers - those not buying direct in large quantity from manufacturers.

Now unlike a coal plant, a Solar panel will only produce electricity when the sun is shining, especially when it is shining brightly and directly onto the panel.  As a result, amount of kilowatt hours produced by a 1 watt solar panel in a year is a lot less than that produced by an equivalent 1 watt coal power plant which can run 24/7.  In fact, the amount of electricity produced in a year depends on the location - less in the north, more in the south.

If a solar panel with a peak capacity of 1 watt were to run at 100% efficiency, i.e. would produce electric current (365 days * 24 hours), i.e. 8,760 hours per year.  I.e. at 100% capacity, the 1 watt panel (measuring around 4 inches by 4.5 inches at approximate current density) would produce 8,760 Watts per year (or 8.7 kWh - kilowatts per year).  In contrast, typical solar panels produce around 1.75 Kwh per year in California and 1.2 Kwh per year in Maine.  So lets assume 1.5 Kwh as the average per year.  this means that our average solar panel has about 17% efficiency - i.e. it is producing electricity 17% of the time.

Now if you spend $5.20 for a Watt of generation capacity, it will last for around 40 years (thorugh 25 is guaranteed).  Lets assume you can finance the entire capital cost at around 8% interest rate.  That means that you will be paying around  0.41 cents per year to get 1.5 Kwh of electricity - or a cost per Kwh of 27.3 cents with today's technology and costs.  This number has come down a lot over the last few decades (cost around $27.00 per peak watt for solar modules in 1982 - so 10x improvement in 25 years) and is expected to continue to decline further. 

The reason solar panels can get cheaper is through a combination of volume gains and better technology and manufacturing techniques.  Everyone realized that there is a huge market for solar once the cost drops below parity with coal (or at least the retail cost of electricity).   

However, it is a long way from the cost of fossil fuel generation.  A typical utility with a large base load coal plant can produce electricity around 4 cents per kWh - and they turn around and sell this to retail customers for around 11 to 12 cents - higher in some areas like Hawaii and lower in others.  Now this does not count the cost associated with carbon emissions from coal which could be worth another approx 2 cents per kWh added to the coal cost.  Of course these assumptions are sensitive to the price of coal, etc.

The above calculations are the reason why we dont all have solar panels on our roofs, factories, and office buildings.  The dream of solar pioneers has always been to drive its costs down to "parity" with coal.  Many companies and experts are predicting that we are going to be able to achieve that result in the 2010 and 2011 time frame if we extrapolate current trends forward.  Recently, due to demand growth (caused mostly by artificial European subsidies) demand has exploded much more rapidly than it would have without government intervention. 

Notice a few things.  While the cost of the solar module  the part made by First Solar can (and has) gone down rapidly, the cost of the other compenents (represented by the $2.60 per Kilowatt due to inverter, labor, etc. dont decline as rapidly - labor even can go up).  So more than half of the capital cost is not subject to the rapidly declining semiconductor economics - this represents an obstacle to getting the price down futher. 

The evolution of Solar Panel production technologies 

According to a private company called NanoSolar (more bout them later), there have been three waves in the development of Solar Technology as further descibed below (right from their website):

"The First Wave started with the introduction of silicon-wafer based solar cells over three decades ago.  While ground-breaking, it is visible until today that this technology came out of a market environment with little concern for cost, capital efficiency, and the product cost / performance ratio.  Despite continued incremental improvements, silicon-wafer cells have a built-in disadvantage of fundamentally high materials cost and poor capital efficiency. Because silicon does not absorb light very strongly, silicon wafer cells have to be very thick. And because wafers are fragile, their intricate handling complicates processing all the way up to the panel product.

The Second Wave came about a decade ago with the arrival of the first commercial "thin-film" solar cells.  This established that new solar cells based on a stack of layers 100 times thinner than silicon wafers can make a solar cell that is just as good. However, the first thin-film approaches were handicapped by two issues:

1. The cell's semiconductor was deposited using slow and expensive high-vacuum based processes because it was not known how to employ much simpler and higher-yield printing processes (and how to develop the required semiconductor ink). 

2. The thin films were deposited directly onto glass as a substrate, eliminating the opportunity of:

The Third Wave of solar power consists of companies addressing the above shortcomings and opportunities. Most every of the new companies address one or the other of the above aspects. One company -- Nanosolar -- brings together the entire conjunction of all seven areas of innovation, each break-through in their own right, to deliver a dramatic improvement in the cost-efficiency, yield, and throughput of the production of much thinner solar cells.

Now NanoSolar which is a private venture backed company claims to have combined a number of innovations that will drop their cost of production far below that of any other existing company (including First Solar - whose technology falls squarely into the 2nd wave of technology described above).  First solar is the leader in the 2nd wave technology and has a significant cost advantage over the 1st wave producers (many of the people on the list).  First solar's plans are currently in the US and Germany with additional capacity coming on in Malaysia this year.

NanoSolar is an impressive company backed by some impressive people (including the Google founders).  Their claims may (or perhaps may not) be true - though I think they are credible (personally).  NanoSolar is going to come into the market in a big way in 2008 and then ramp up production quite rapidly.  While FirstSolar projecting getting to utility retail rate parity in 2010, NanoSolar claims it will be there in 2008.  Whether they do achieve that or not, the broader point is that First Solar is the current king of the mountain, but there is no MOAT to their business and there are barbarian hordes approaching from all sides.  Venture Capitalists are throwing tons of money at this problem and as are all the 1st generation companies and a number of Chinese companies.  FirstSolar's entire current business model rests of maintaining its lead as the low cost producer - however, there is a very high chance that it will give up this status in the very near (2008/2009) time frame under attack from the Chinese producers and the new venture backed companies with new ideas.  From a recent article, VC's have invested more than $340 million into 5 startups using competing (lower cost) technologies.

Customer Base Concentration and Pricing Model

Because FirstSolar is the current cost leader with increasing production volumes, revenues and profits are expanding rapidly.  More specifically, Revenues in the most recent quarter ending Sept 29 2007 expanded to $159 million (almost 4x the revenue of Q3 2006 of $40.8 million).  The revenue growth is not constrained by demand, but rather by supply - the company's ability to produce enough solar panels - which in turn is dependent on their manufacturing capacity.  As a result of the rapidly rising sales, gross margins and profits are also rising rapidly.  Gross margins were more than 51% in Q3 2007 and net income was 28.9% of sales in Q3 2007.  However, as mentioned, the main constraint is manufacturing capacity.  Currently, the company has 2 manufacturing sites, one in Perrysburg Ohio (75 MW Capacity)  and the other in Frankfurt Germany (125 megawatt capacity). 

The customer base is almost entirely in Europe, especially in Germany.  The reason for this is that Germany decided to provide tremendous subsidies for small scale renewable electricity producers to sell power to the grid.  This situation was created by the passage in 2000 of the EEG law (renewable energy sources act) in April 2000.  Germany provided an incredible incentive for developers of renewables generation, specifically, renewable generators would get a contract at a fixed price for 20 years.  The fixed price is based on the price of the starting year and goes down each year.  For example, ground mounted PV (solar) would be purchased for 0.3796 Euro per kWh.  I.e. approx $0.506 per Kwh.  This would be for developments signed in 2007 (building mounted subsidies are greater ranging from $61 to 65 cents per Kwh).  While this amount goes down each year by 6.5% for ground mounted sustems- presumably reflecting efficiencies and reductions in price of modules.  These are the tariffs that utilities are required to pay to buy solar power at these high rates and the cost is passed on to electricity consumers in their electicity tariffs.  As a result of these subsidies, PV demand has exploded.  Therefore, these subsidies have been characterized as Effective and Expensive.  These subsidies are the main reason that in 2006, PV demand from Germany was 750 MW peak or 52% of the global total of 1450 MW peak in 2006.   It is expected that there will be 1200 to 1500 MwH of new installed capacity in German Alone in 2007.  But after that, the amountof PV generation capacity growth may begin to slow substantially.  The Germany economic ministry recently prevailed upon the environmental ministry and a recent draft proposes a tariff drop by 9% in 2009, additional 7% in 2010 and 8% in 2011.  This proposal would increase the price drop from 5% in the existing law.  Of course, since this would happen starting in 2009, this actually increases the demand in 2008 as people try to lock in the existing subsidies for 20 years.

The level of investments is driven by the return that people earn locked in for 20 years.  This in turn is also a function of the cost of money since buyers of these small solar projects finance the cost and pay for the project over 20 years as they recieve the revenue from the subsidies.  If interest rates go up, the demand for these panels will also decrease fairly quickly.  The tariff reductions in the draft would also have a fairly meaningful impact on the economics of these projects.  In France and Spain, the subsidies are lower.  In 2006, the Spain PV market was only 60 Mwp - less than 10x the size of the German market.

In the case of First Solar, they sell almost all of their solar panels to developers and system integrators in Germany, France and Spain (all subsidy countries) which then resell the panels to end-user recieveing the goverment subsidies in their respective countries.  First solar is quite vulnerable to the tariff reductions that are signaled by the recent german moves.  However, for now, the Company has Long Term Sales Contracts signed by 12 key customers (with biggest one accounting for 21% of their sales).  These contracts call for deliveries from 2007 to 2012.  Based on the way these are described in the Company's filings, it seems as if the customers do not have the option to cancel their contracts in the event of market or tariff changes - however, contracts can still be abrogated through bankruptcy or voluntary default - a situation that can be caused by customers overzelously committing to long-term purchases of this kind.

Materials Issues

In addition to the customer concentration and tariff issues, there are significant issues related to the cost, availability, and safety of the materials used to produce the Company's thin film solar panels. 

Specifically, First Solar produces Cadmium Telluride based solar panels.  It turns out that Cadmium is extremely toxic substance.  Tellerium is also toxix though at a much lower level.  A recent article on the company ( points out that a new European environmental directive took effect in July 2006.  The company has indicated in its filings that it does not believe the these derectives apply to their panels, but that this could change.  As a result, the company recently adopted responsibility to reclaiming and recycling their panels at the end of their lives.  According to the company, it sets aside a certain amount (not disclosed how much) for this long-term liability (not clear where the amount being set aside is sufficient or not). 

The other (bigger) issue is that prices for both Cadmium and especially Tellurium are increasing rapidly.  In 2000, the price of Tellurium (mostly a byproduct of copper mining) was $3.82 per pound, in 2004 it was $10 per pound. In 2007, the price hit $100 per pound.  Currently, total world production of Tellurium is estimated to be around 200 tons.  Meanwhile, there are several application in Flash Memories and now Solar that consume Tellurium.  Mining supply is a significant constraint.  First solar is rapidly increasing its investments in Panel Manufacuring capacity. 

Between the Ohio Plant (75 MW peak per year capacity) and Germany (100 MW Per year in Germany), the company's total manufacturing capacity was only 175 MW peak.  With the addition of the 1st phase of the Malaysia plant (also 100Mw), the capacity will increase to 275 Mw in the 2nd half of 2008.  The land for the Malaysia facility can accomodate a second 100 MW plant and they also have options on land for another 2 plants at 100 MW each.  

The critical ingredient is Cadimium Telluride and the company only has 1 supplied for this critical ingredient.  The Company's own demand will come to represent a significant percentage of the gloabal market.  As the price of Telerium rises, the price paid by the company goes up each quarter.  Since prices are declining on the top line and potential increasing in input costs, this could put some pressure on margins. Moreover, given the inelestacity of supply of Tellerium, this material could ultimately represent a significant physical bottleneck to the comany's approach requiring them to go to someother material or technology - one in which they may not hold any edge.  This issue represents a significant risk factor for FSLR.  A similar issue already played out with the first generation panel manufacturers who were silicon based.  The price of silicon rose rapidly and affected the margins and competitiveness of the first generation manufacturures and something similar could easily happen to FSLR as well.

The Valuation

Currently, they are selling the capacity they produce at $2.48 per Watt, but this amount declines contractually around 5% per year.  So in 2008, they will generate revenue of around $530 million based on the capacity they have now (175 MW) plus 50 MW (100 MW for half year) assumed and price decline to $2.36 per Watt.  Assuming net margin goes up again to 30%, the company would generate net income of $159 million. 

Dividing by 78 million shares currently outstanding get to EPS of $2.03.  Now, it is highly likely given the price of the stock, that the company will do another secondary offering.  Given the stratospheric PE multiple, the interest on the cash would actually be accretive to these EPS estimates.  However, depending on the size of the secondary offering, it could put some pressure on the stock price.

If everything goes well, the company would generate around $1 billion of revenue in 2009 @ a 35% margin (assumed) is $350 million in net income.  Still 57x today's stock price.  Effectively, the company is selling at more than a 30x multiple of 2010 estimates which are highly suspect because of the many competitive and other issues cited above.

Tactics of Shorting

The biggest issue with this short will be timing it properly.  Given the current momentum, this is no immediate upper limit to the price action.  For technical reasons alone it is not inconcievable that FSLR could hit 400 before it ultimately collapses to where it should be (below $150 per share).  Therefore, I would suggest iniating a short position with only a third of the total amount you would ultimately like to be short.  Another option would be to Short but have Buy Stop Orders 10 to 20% above your short price as a protection against runaway losses and if it does keep rising consider reinitiating the short at higher levels. 

Another option would be to write really out of the money long-term calls or buy long-term puts (Jan 2009).  While it is difficult to predict the short-term price action (likely higher if anything), the factors above should play out somtime in 2008, and euphoria could be reduced substantially.  Given the momentum crowd that owns this stock, it could very rapidly decline to its $150 level or below. 


Germany tariff reductions, customer issues with pricing under LT contracts, increasing interest rates dampening demand in German, rapid increaseas in Tellurium prices, Signifcant capacity additions and increase of competition from many other new and existing solar panel manufacturers, a large secondary offering announced, Cramer gets hit by a bus :)
    show   sort by    
      Back to top