I just reviewed the registration statement for Silicon Valley-based start-up Solyndra, Inc. I blogged about Solyndra previously; its innovation is to use solar pv film in the form of racks of cylinders for installation on flat roofs. Cool stuff.
Reading the prospectus left me both impressed -- and puzzled.
First, "start up" may convey an image different than reality. The scale of this operation is indeed audacious. Since inception in 2006, Solyndra has raised $970 million from venture capitalists, landed a $535 million federal loan guarantee, and ramped up employment to 747 souls. These guys -- founders, investors -- evidently really believe in their concept. Solyndra filed to go public in late 2009.
On the other hand, the 150 pages of company description fails to provide the data necessary to grasp the fundamental economics of the business. The statement provides data on theoretical production capacities, and the projected costs of plants, but no information on production costs or management expectations for pricing levels or trends. Future cash cow, or ponzi scheme? No way to tell.
What we do learn is that solar fabrication involves ferociously high up front costs.
I'm old school: revenue minus costs equal profits; profits provide a return on investment.
Here's what the company provides on investment and production
- Fab 1, now built: capacity 110 megawatts annually
- Fab 2 Phase 1, under construction: capacity 250 MW
- Fab 2, Phase 2, construction to start soon: capacity 250 MW
- Total fab construction costs: $1,702,000,000 [I inferred Fab 1 costs from per-MW costs of Fab 2; these may be understated]
- Start up costs: $505,000,000 at Q3 2009. These are accumulated losses incurred to build the business, but not capitalized.
- Private financing: Umpteen rounds of VC investments totaling $970 million;
- Public financing: one load guarantee from the federal government (DoE) for $545 million for Phase I, and a second loan gaurantee applied for, totaling $469mm, for Phase II.
If all goes to plan, in a few years Solyndra will have 610 MW of production capacity, producing panels with ~200 watts of output, or about 3 million panels. It will have spent at least $2.207 billion to get to that point, a figure that will certainly go higher as the company incurs additional losses to ramp up capacity and production.
Let's assume that the company will achieve the capacity (and related process improvements) that it expects -- and that it sells out all its capacity. Will Solyndra make money?
At the end of 2009, PV panels were shipping from wholesalers for an average price of $4.30 per watt. I'll make a crude assumption of a 20% wholesaler margin, suggesting factory prices of about $3.60. At this price a sold out Solyndra would earn revenue annually $2.2 billion ($3.60/watt * 610 million watts).
However, solar panels are heavily subsidized, and are not economically viable (except in rare, off-grid remote applications) at this price level. Federal subsidies are scheduled to end in 2016 (as does Solyndra's first loan guarantee). Industry conventional wisdom is that panels need to wholesale for $1.50/watt or less to be economically viable. Let's assume a "future" factory price of $1.25/watt, which is $1.50 wholesale with the same 20% markup.
At $1.25/watt a sold out operation generates $762 million in annual revenue.
Return on Investment
As a potential investor in Solyndra, I'd like to see at least a 15% rate of return on our investment, in particular given the inherent risks and uncertainties of the venture. 15% on $2.207 billion investment is $330 million per year. In other words, Solyndra would need after tax earnings of about 15% of revenue at today's prices (since annual revenue and total investment are conveniently equal in this accounting). But at a future net price of $1.25/watt, the required net income of $330 million is actually 43% of sales, a level that would make Solyndra one of the most profitable companies in the country on a return on sales basis -- better than Google.
The company does not provide any useful data on what its future costs of production or operation are likely to be. Based on what we do know, management of those costs will be the determinant of success, ssuming the market likes its innovative product. If Solyndra can drive down its costs rapidly and substantially, it may be successful. If not, it will ultimately face problematic economics due to a reduction or termination of federal customer subsidies, and/or better production economics at competitors.
The founders and investors of Solyndra have real cajones to make a speculative bet this large. I will them well. After reviewing the prospectus, I remain at least skeptical of the fundamental economics here.
There is an old saying in the real estate business that hotels always make money eventually, but often not for their original owners. I have little doubt that Solyndra's fabs will make money for someone. But for whom?