by Stephen O’Rourke, Managing Director, Deutsche Bank Securities

How many capacity expansion announcements in solar photovoltaics did we hear of last month? If you guessed it was several dozen, you may have fallen short. In the past year, more than 75 solar companies were funded and billions of dollars in public money raised — largely for the purpose of building capacity. That creates a daunting specter of possible oversupply within the next couple of years. Considering cell capacity expansion plans in both crystalline silicon and thin films, as well as polysilicon production ramp plans over the next 18 months, it is quite likely that the supply of modules will exceed demand sometime in 2009.

The argument that the industry enters a sustained oversupply period and shakes out quickly is too simplistic. A more likely scenario would be the emergence of an oversupply situation of polysilicon modules which will drive meaningful module price declines. This will spark demand, potentially consuming available supply and rebalancing supply and demand at a much lower price point. This dynamic could occur iteratively as additional production floods the market, creating a situation in which the industry flirts with oversupply for a prolonged period, “rescued” by elasticity of demand. Such a dynamic could drive demand much higher, effectively consuming supply.

However, regardless of any new equilibrium, the damage will have been done. Should module ASPs decline faster than costs — which is highly likely over short periods of time — company margins will compress, stocks will decline, balance sheets may be challenged, and some companies might face extinction in the process. The market should act as a clearing mechanism for companies in distress. (Some governments, intent on establishing domestic manufacturing bases, could artificially prolong an industry rationalization, but this is almost impossible to predict.)

Considering how the industry is evolving, I expect an oversupply to manifest at the module level, driven by an adequate supply of polysilicon. I see a lagging impact up the value chain (i.e., polysilicon contract pricing will not collapse), and what could be a much more delayed impact down the value chain (i.e., system integrators may well enjoy a period of sustained margins as system ASPs decline at the expense of module price declines). Within the supply chain, I expect the non-differentiated, pure-play cell and module makers to suffer the most — and to suffer first.

While this may be a bad thing for many non-differentiated companies, it bodes well for the industry as a whole. Over several years the present unbridled build-out of capacity would be rationalized, industrywide manufacturing assets would be better utilized, significant consolidation would proceed — and most importantly, the cost/kWh of solar PV generated electricity would decline faster than present expectations.

In light of what could well be a few turbulent years around the corner, keep your eye on three types of companies:

– Those with long-term technology-based competitive advantages (there simply aren’t that many)
– Those with business models that hedge against collapsing module ASPs (those that integrate downstream to the customer); and
– Those positioned to sell energy rather than simply modules or systems.