Cost competition between FSLR's CdTe module and c-Si modules from China

On Monday before market open, BofA Merrill Lynch cut estimates on FSLR to $19. Merrill cited margin pressures and concerns regarding project pipeline growth. The firm says its latest round of checks suggest a cost per watt for leading silicon vendors below $0.65 by the middle of next year, and it doesn't see how the company can make money at level.

Regarding "c-Si module cost going below $0.65/w", one needs to understand a few things before accepting the the statement at its face value. First, what is the necessary ASP for China's c-Si module makers? There are quite a few discussions on various message boards indicating 12-15 cents need to be added to the cost for China's c-Si makers to break even, given an array of other costs. An additional 5-10 cents profit need to be out there to sustain the business. Therefore, even if the c-Si module cost reaches $0.65/w, the ASP may need to stay at $0.8 to $0.85. Secondly, one needs to understand that the cost decline in c-Si module is largely the result of declining component and consumable costs. For example, poly is mid 20s instead of $80/kg. Now there are several components are selling at a loss, such as the front glass, the aluminum frame. The junction box and the back panel has small profit margin. The silver paste cost is more related to the cost silver and one should not count it as cost saver (it may actually get more expensive). It is safe to say that the huge stride to lower c-Si module cost is in its final phase and we should see small, incremental cost decline in the next few years. Thirdly, one need to understand that the c-Si side is not a monolithic player and there is a cost spectrum among the 20+ makers. Moving to high-efficiency created more disparity among the players. Therefore, not everyone can make modules at $0.65/w, or maybe only a small percentage of players can make it next year.

On FSLR's side, its pace to lower its module cost is like a tortoise in the past two years - if one can label c-Si side as a hare. However, its roadmap shows it is likely to achieve faster cost savings down the road. We already see CEO Ahearn took a few initiatives on the direction by terminating the CIGS development and moving more production to Malaysia. There is already announcement that FSLR made substantial progress in raising module efficiency to above 12%. With only one focus, I expect FSLR to gain momentum on its target of reducing cost to $0.55/w in the next two years.

Here I left out a few items such as the high efficiency game, the panel efficiency degradation, the FSLR 2008-2009 quality issue etc. I may address these issues later.