Solyndra Solar: Unemployment Checks Make a Better Investment Than DOE's Millions

Solyndra is closing a solar-panel factory, delaying previously announced expansion plans. Contrary to management's spin, even a cost-cutting do-over by the photovoltaic (PV) module maker won't brighten its prospects for survival.

Headquartered in Fremont, Calif., the privately-held company manufactures cylindrical modules, incorporating copper indium gallium diselenide (CIGS) thin-film technology, for the commercial rooftop market.

Just weeks after opening the first phase of a new fabrication plant, called Fab-2, Solyndra will close an older and less efficient facility, Fab-1. Phase I of the new plant, mostly built with a $535 million loan-guarantee from the Department of Energy, was originally intended to add an additional 250 megawatts (MW) in production capacity to Fab 1's annualized run-rate of 110 MW.

As for Fab-2's production scale-up, company spokesman Dave Miller pegs run-rate for 2011 at about the 140 MW, ramping up to 300 MW by year ending 2013.

Chief executive Brian Harrison told the NYT in an interview that Fab-2 was much more efficient and cost-effective than its existing facility. "We're adjusting our plans to be more in line with where the market is and where our business is at the moment," said Harrison.
Solyndra once planned an initial public offering (here's its S-1 registration statement with the SEC). Suffice to say, when market sentiment for high-tech offerings soured, Solyndra management shelved the IPO, opting instead to raise needed R&D monies via venture capital channels instead.

Unfortunately, even a Houdini incarnate couldn't do much more than conjure up a few financial tricks to distract from unpleasant tasks that lay ahead -- like expanding capacity with a dearth of free cash flow:

  • The company will need to raise an additional $642 million in capital -- required for land, buildings, improvements, manufacturing equipment and other start-up costs for Phase II of Fab-2;
  • During the last two fiscal years ending January 2, Solyndra installed 31.6 MW of rooftop PV systems -- but in doing so, accumulated a deficit of $557.7 million in "start-up" costs;
  • At present, the company holds about $64 million in a fund called "contingency reserves," payable to DOE should the company implode. Look for this liability to treble come 2011, due to tangible net equity -- well, let's just say below zero;
  • Revolving credit facilities that are payable at 15 percent. and, the company owes almost $961 million to venture capitalists, currently noted as "preferred" equity on the balance sheet.
By the end of next year, chief executive Harrison confidently predicted to Greentech Solar's Michael Kanellos in an interview last week, early forecasts envision Fab-2 "all-in" prices for a Solyndra rooftop system, including modules, racking and installation of around $3.50 per watt -- or lower -- mostly though throughput efficiencies.
Based on the average sales prices of solar panels sold in the fourth fiscal quarter of 2009, our average sales price was $3.24 per watt, which was $1.29 per watt, or approximately 66%, higher than the $1.95 average sales price per watt of leading crystalline silicon (c-Si) photovoltaic manufacturers during the same period. As a result, certain system owners who focus more on the up-front price of solar panels than on achieving the lowest levelized cost of electricity per kilowatt hour, or LCOE, may choose the product offerings of those competitors that have a lower initial panel purchase price.
Troubling -- flaws in the soundness of Harrison's business model immediately jump off the page:
  • Although Fab-2 is highly automated, what is the actual "learning curve" required to absorb savings moving forward? Unknown.
  • It's going to take significantly more than a backlog of 800 MW, in my opinion, to lower fixed production costs from throughput quantity gains.
  • The company has sparse experience in third-party purchaser debt financing.
  • Given few signs of production rate constraints from c-Si challengers on the Chinese mainland, can Solyndra thwart -- or afford -- the market-share battle that's shaping up?
  • Payback period -- duration of time for system to pay for itself -- is estimated in the neighborhood of seven to 10 years. Unless erstwhile VP Al Gore is on the Board of Directors, try getting any director to think beyond the next Wall Street cocktail party.
As most observers know by now, financial incentives set by governments -- such as the "feed-in-tariffs" that make utilities pay above-market rates to renewable-power producers, or investment tax credits that lower total construction costs -- play an integral role in boosting demand. Unfortunately for Solyndra, the solar landscape looking out beyond mid-2011 is frightening, like a scene from The Walking Dead.

For the fiscal year that ended January 2010, the company derived approximately 49.6 percent and 15.3 percent of aggregate revenue from rooftop projects installed in Germany and Belgium. In January, solar subsidies are set to decline by double-digit percentages in most European Union nations. Ergo, demand will slow -- or solar panel makers will need to slash prices (absorb margin losses) to hold onto their market share.

"The sun'll come out
Bet your bottom dollar
That tomorrow
There'll be sun!
Just thinkin' about
Tomorrow...." ~ Broadway musical Annie
I applaud Harrison for holding on with such vigorous and unbridled optimism. But let's face the facts:
  • Solyndra has no vertical integration strategy to capture additional cost-savings, up or down the nodes of its supply chain;
  • Thin film technology uses about one-hundredth the amount of silicon found in crystalline silicon PV. Solyndra might want to rethink its business plan, as its core advantage has gone "poof!" For example, ReneSola (SOL), a leading global manufacturer of solar wafers, expects to produce 3,000 to 5,000 metric tons of PV-grade silicon at a production cost below $35 per kilogram by June 2011, according to recently filed regulatory reports;
  • More proof -- vertically-integrated challenger LDK Solar (LDK) is targeting manufacturing costs nearer to $30 per kilogram by 2012; and,
  • Access to cheap Chinese debt financing, cheaper local labor, and the cheapest electricity costs means that even with silicon priced between $50 to $60 per kilogram, traditional c-Si Chinese manufacturers like Suntech (STP) and Yingli (YGE) can still eat Solyndra's lunch -- even with average selling prices as "high" as $2.00 per watt.
"Our costs [for manufacturing the module] will be below $2 by the end of 2011," Harrison told Greentech Solar. At the time, Solyndra will sell the modules, potentially, for around $2 to $2.35 per watt, while the balance of system might run $1.10 to $1.35 per watt.
"At those prices, Solyndra will be a cash-positive company," Harrison added.
Solyndra's ability to achieve or sustain positive gross margin and profitability will ultimately depend on its ability to simultaneously increase its production capacity and reduce manufacturing cost per watt faster than its average selling prices decrease. Increasing ramp-up efficiencies will require construction and build-out of Phase II (a key component of margin growth moving forward) for Fab-2.

Unfortunately, as mentioned, the company's finances are an auditor's nightmare. Will DOE save the day again? In September 2009, management applied for a second round of financing from the DOE, in the amount of approximately $469 million -- funding necessary to expand Fab-2 capacity from 250 MW to about 500 MW.

With declining solar panel prices back in play, President Obama might just want to reconsider shutting off this spigot. Or, maybe Obama ought consider that sage advice by (soon-to-be) ex-House Speaker Nancy Pelosi -- something she once said about how unemployment checks were a cheap way to stimulate the economy! To look at Solyndra's flow of red ink, maybe Pelosi had a good idea.