During the early production scale-up of a solar cell / module manufacturing business, capital expenditures dwarf cash flows. In the case of successful companies such as First Solar (Nasdaq: FSLR), "steady state" quarters eventually shine through, but another capacity ramp is always just around the corner.
Couple these heavy and routine reinvestment requirements with the rapidly evolving nature of the business, and you have a recipe for disaster. Not only do you put your balance sheet at risk by borrowing heavily to build your production facilities, but you also have no assurance that your technology will be competitive long enough to make the investment worthwhile.
Solar companies such as Yingli Green Energy (NYSE: YGE) compound these issues by pursuing a model of vertical integration, where they involve themselves in every step of the supply chain, from manufacture to system installation. This approach will work for some companies, but it will break the backs of many others.
For me, the case of Evergreen Solar (Nasdaq: ESLR) both illuminates the pitfalls of the conventional model and points to a different way forward. The company got in a real jam as it tried to scale up its string-ribbon technology, and I lost faith early last year. At the time, it was evident that Evergreen's best way forward was to shift to a contract manufacturing model. That's the path the company has chosen for its panels, while keeping its proprietary wafer fabrication in-house.
Today I'm wondering whether a new generation of solar companies won't push that model to the logical extreme. In the semiconductor space, giant foundries run by the likes of Taiwan Semiconductor (NYSE: TSM), and UMC crank out chips for "fabless" companies such as NVIDIA (Nasdaq: NVDA) and Qualcomm (Nasdaq: QCOM). The latter companies are thus freed to focus on things such as design, marketing, and customer support. It's not so hard to envision a day when solar fabs will be ready to mass-produce whatever the next-generation solar visionaries can come up with.
Both TSM and UMC have dipped a toe into the solar space, so perhaps I'm not the only one drawing this analogy. I'll grant that it may not hold up, though, on account of a different intellectual-property landscape.
Disclosure: Author doesn't have a position in any company mentioned.