Based on data for polysilicon costs from Q2 results, two major module makers from China showed .43 cents cost for poly per watt, and depending on the grams per watt in a spread between 5.8 to 6 grams this converted to the average cost around $72 per kg. The spot price being below $40 per kg suggests savings of .19 per watt, however there will be no immediate impact to all for Q4, since many makers purchase their polysilicon on long-term contracts. Those are negotiated on a quarterly basis, thus current spot pricing may influence renegotiations for pricing in Q1 of 2012, but not sooner.
Couple companies to receive an immediate benefit would be Jinko, which buys almost all of poly from the spot market. Second best situated is Yingli with 30% purchases of poly made from the spot. In case of Yingli their Fine Silicon plant yields poly at $65 per kg. Consequently, procurement on spot acts as an offset to its expensive production, nonetheless, price of poly is expected to drop by 25% in Q3 and another 10% in Q4 for this producer.
Trina Solar has renegotiated its long-term contracts with GCL for the last quarter, and this trend is expected to continue. Trina is the largest customer of GCL with contracts estimated at 15.6GW of polysilicon and wafer products reaching out as far as 2020. Ten cents per watt are expected to be shed from poly costs for Trina in Q3 and perhaps another 4 cents for Q4. Other recipients of GCL reductions will include Canadian Solar, Hanwha, Suntech and Sunergy with 4 to 7GW in long term contracts covering next 5 years.
Another large provider, OCI Chem of South Korea, is expected to offer similar support and reduce long term pricing. YGE is the largest Chinese customer of OCI with $1.7B in long-term contracts, followed by Suntech with $1.6B, and Trina with $700M among fully integrated module makers.
Ja Solar should be seeing relief in their cost of silicon and also wafers purchases, having 10GW of products contracted from GCL and $1B from OCI. However Ja Solar, dominantly cell manufacturer, is expected to struggle with its bottom line due to dramatically low margin on cells and massive of fairly similar in cost and quality output from Taiwan.
Majority of the relief from OCI will be seen when deliveries start in beginning of 2012. As in the value chain downstream, upstream producers have variable costs and are not equal in costs. OCI production costs are higher than GCL. It will be interesting to see how the business relationships will develop from here. Some of the discounts may not be offered equally across the customer base and the size of the relationship financially may be the driving force behind it.
Wafer makers should find relief. Conversely, those with polysilicon production of their own may find themselves in a position which will threaten their bottom line if their production is above the average market price. Some may forgo plans of expansion if poly pricing does not warrant or provides the limited benefit only. Two major wafer makers Renesola and LDK should remain comfortable, where Renesola may abandon pursuit of new capacity in silicon beyond 2012.
Reduced price of silicon will have an adverse effect on margins of companies, which sell silicon as their core competency. Companies like Daqo and REC and mentioned already GCL and OCI will see loss to their margins. Less flexible producers like Hemlock and Wacker may be faced with cancellations of orders. Some contracts may not be executed due to industry exits and bankruptcies, putting even more silicon on spot market. New polysilicon manufacturing entrants may abandon the idea of expansion into the segment, unable to compete with the scale of large producers and future overcapacity.
The next beneficial development for the industry should be pricing reduction in installations and integration costs. However depending on fragmented demand, this may not be clear until first quarter of 2012.
