The solar bears are sure to be out in force today, with Solyndra following Evergreen Solar into bankruptcy. Bloomberg notes that privately-held SpectraWatt has also gone under recently.
Republicans are already calling for an end to government investment in the space, while at the same time praising moves to run a pipeline to Alberta in order to harvest tar sands. It's clear their priority remains fossil fuels.
Rather than getting into an environmental argument, I want to make an economic one, starting with a premise I feel absolutely certain of. I call it Moore's Law of Energy, after Gordon Moore's 1965 paper predicting silicon chip density would double every 12-18 months for the foreseeable future. But if you, and he, prefer, I'll call it Dana's Law of Solar Power.
The amount of solar energy in use will double every 12-18 months for as far out as I can see.
It's an easy prediction to make, because if this year's production is just a little ahead of last year's production, in terms of wattage, you get twice as much solar power in the next year that you got this year. In other words 1+1=2. And in this case not only is production of cells continuing to increase, but efficiencies are rising and costs are falling.
But here is where solar is different from silicon. Solar doesn't have to be silicon. The largest U.S. solar producer, First Solar (NASDAQ:FSLR), uses cadmium telluride, not silicon at all. The technology platform of solar, in other words, is subject to change without notice, meaning today's leadership may be worthless tomorrow.
Take Stion, for instance. This California-based company is using Mississippi government money to build a new plant in Hattiesburg, Mississippi. But they're not making solar the same old way. They're making thin film, not panels, and they've achieved breakthroughs in multi-layer CIGS technology as well as thermal management. In other words if you've been making polysilicon panels all your life, this is completely different, and it could put you out of business in terms of cost and efficiency.
Meanwhile, labs around the country continue to turn out breakthroughs. (Each time I start a paragraph like this I check Google News and find stories I'd never known of before.) Kentucky scientists write that adding antimony to gallium nitride can result in a simple system for splitting water into hydrogen and oxygen resulting in very cheap fuel for fuel cells. Stanford is among many colleges advancing the science of quantum dots, and Berkeley Lab scientists are working on cells that use the full spectrum of light, dramatically increasing efficiency.
Some of these breakthroughs will make it into the world as start-ups, and some of those start-ups will go into production, producing solar systems with technologies even Stion doesn't understand. What's certain is that costs will keep declining, efficiency will keep rising, and eventually the cost of producing solar power will fall below the cost of producing it with coal or oil, even when the device and infrastructure costs are layered-in. I'm guessing five years at most.
What does this mean for investors? To me it means avoid the technology and, for now, avoid the companies producing the cells. All that is subject to change without notice.
Focus instead on the after-market, on selling systems and on selling the power. Power is power, power has a price, and the numbers can be made to work there for some time. Companies like SunPower (SPWRA) have a much better outlook in the near-to-medium term and my guess is that when venture capitalists look to cash out of firms like SunRun and SolarCity you will see solid opportunities.
Sometimes, in technology, investment failure is a sign of industry success.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.