I have long said that solar technology is where computing was in 1971. I stand by that. I see the same mix of skeptical enterprise buyers and hobbyists or enthusiasts, a growing market at the front of the demand curve. Scaling supply to meet demand seems to be the main challenge.
So why aren't companies like First Solar (FSLR), LDK Solar (LDK) and Suntech (STP) being treated like Intel (INTC) or TI (TXN) were a generation ago? Why are most large solar stocks rated as "sell?" It's a disconnect a lot of analysts in the space want answers for. So here is mine.
While there are certainly some ready with conspiracy theories, a better reason is that how we make solar systems, the materials and manufacturing processes, are not nearly as settled as they were for chips in 1971. Back then, we understood the planar process. We knew that silicon oxide semiconductors were the way to go. Accurately microscoping designs and burning them into substrate were the main challenges.
That's not true for solar today. In a recent interview on Reddit, Applied Materials' Charlie Gay insisted that, contrary to the bears' assumptions, costs can continue dropping roughly 18% for each doubling of installed capacity, as it has for decades.
But how we get there may indeed change in fundamental ways. "Materials like quantum dots, organic and polymer absorbing layers for solar generation are on the horizon," said Gay, and there are over 70,000 people working these problems, an investment of $3 billion.
But will tomorrow's solar plants look more like chip plants, or printing plants, or chemical plants? That is harder to know. If you take chip processing as a given, then perhaps Hypersolar (HYSR.OB) has the answer. They would replace the glass on current systems with a patented material which, they say, acts as a microconcentrator, routes light of different wavelengths to where it can do the most good, and provides thermal management.
When the company announced this breakthrough in February, the share price zoomed. To a quarter a share. HYSR is currently listed on the OTC Bulletin Board for closer to a dime.
Or take QSolar (CNSX-QSL), which claims it will achieve "grid parity" later this year with a system that grows crystals in furnaces, cuts them with wire saws, then sprays them with an insulator. Sounds more like a chemical plant than a chip plant, doesn't it? Oh, and their Kristal systems come in a variety of colors, so they can be used for walls or to replace windows – no more mounting on the roof.
Unfortunately QSolar, like Hypersolar, is a thinly traded penny stock. (The most recent price is 35 cents a share, in Canada.) Neither company is going to be investment grade until it either makes a profit or gets bought by a larger firm that can scale its production.
There are as many ways to make an advanced solar cell as there are designs. Quantum dots are made in micro furnaces. Kodak (EK) hopes its knowledge of printing paper can build plants for Natcore (OTC:NTCXF) of China that will produce cells at 50 cents per watt based on a liquid phase deposition technology perfected at Rice University in Houston. (I'm a Rice graduate – class of 1977.)
So we don't know how tomorrow's ultra-efficient, ultra-dependable and super-cheap solar cells are going to be made. We don't know who is going to make them. The bears might be right, that cost declines for current manufacturers are going to get harder to come by. But by 2016 most of those companies might be replaced by firms you have never heard of, making cells in ways today's players can't imagine.
And that's the disconnect, it seems to me. The technology is in flux. But sales are growing, profits are growing, energy from oil and coal is being replaced by energy from the Sun, the Sun can meet our current energy needs 11,000 times over, as Gay notes, and solar photovoltaics will be able to meet the needs of 99% of the world's people in 10 years.
I just can't tell you how to invest in it with confidence. The big profits will be made, for some time, by savvy venture capitalists like Al Gore.