Long-Term Polysilicon Contracts Continue To Harm China's Solar Elite

by: Robert Dydo

On Friday, the Chinese government released more details on the five-year solar plan (2011-2015). The government aims to reduce the cost of domestic power to 0.8 Yuan per kWh by 2015, and 0.6 Yuan per kWh by 2020. The cost of solar panels in China is predicted to drop to 7,000 Yuan ($1,111) per kW by 2015, and 5000 Yuan per kW by 2020.

Leading polysilicon manufacturers will have to reach a 50,000-ton production capacity, while leading module makers must reach 5 GW by 2015. China will further help solar companies increase their annual sales, with at least one company reaching 100 billion Yuan in sales and three to five companies reaching 50 billion Yuan in sales by 2015.

Technology advancements will take conversion efficiency of monocrystalline cell to 21%, polycrystalline cell to 19% and amorphous silicon cell to 12% by 2015. Eighty percent of solar equipment and auxiliary materials will be produced domestically.

In Europe, SolarWorld (OTCPK:SRWRY) released its preliminary data for 2011. The company sold 794MW of modules and wafers, which is 25 MW less than sales in 2010. The overall sales were at $1.4B versus $2.2B in 2010, a 35% loss of revenue, mainly due to ASP deterioration. Using rudimentary calculations, it appears that SolarWorld sold modules and wafers at $1.76/watt on average. The company will take a $418M charge on outdated equipment and additional long-term assets' impairments.

Trina Solar (NYSE:TSL) reported on Q4 and full year for its operations, confirming that the industry's strongest players suffered trauma from overcapacity and ASP deterioration. Trina took on charges starting with the provision of $16M for aging payable accounts and a $9M premium on prepayment for old poly contracts with Nitol. High operational expenses did the rest of the work to get Trina into a loss of $0.93 per share. A new provision was added for $3.3 million for potential countervailing duties relating to importation of solar products. The total US 'record of sale' in Q4 was around 123MW. Combined with Trina's average sales at $1.01 and conference call explanation of 8% provision, this would cover around 40MW of December sales. Interestingly, despite importation duties, Trina sees the US market as 25% of the business in 2012, appearing to have a contingency plan in place in case duties are enforced.

Trina sold 1.5GW of modules, 50% more than SolarWorld, and achieved a new company record, beating 2010 by 42.5%. Yearly ASP averaged $1.35 per watt, bringing $2.04B in revenues, a modest 10% increase. ASP experienced a 41% reduction with the cost dropping by 24% in Q1, compared to Q4. The current non-polysilicon cost is $0.64/watt, and by the year's end the company has an objective to bring it to below $0.60/watt. Polysilicon costs are $0.30/watt. In contrast, Canadian Solar (NASDAQ:CSIQ), which also pre-announced this week, bought wafers directly from GCL at $0.31/watt, saving $0.18/watt in processing costs. Trina reported a reduction of inventory and accounts payable by $187M, resulting in an increase in cash and, unfortunately, long debt. Trina also revealed plans for an n-type IBC cell, with an efficiency goal above 21% in 2013. Out of four major solar companies reporting to date, Trina displayed the least amount of damage from the tough year with a loss of $37M, while MEMC (WFR) lost $1.5B, REC $1B and Sunpower (NASDAQ:SPWR) over $600M in 2011.

Looking towards 2012, ASP may drop by 10% to $0.90 per watt, while total cost could drop another 21%. Poly should become the largest cost-saving contributor, reaching a 50% drop from Q4 at $0.30 to $0.15 by middle 2012. Adding this figure to expected processing costs of $0.58 per watt, Trina may achieve 17% gross margins in the latter part of the year.

In the conference call Trina made few statements to address solutions for importation duties. We heard about outsourcing, collaboration, or even acquisitions. Outsourcing its Honey technology seems unlikely at this point, but in the future some form of OEM arrangements can be expected in the US.

Yingli Green Energy (NYSE:YGE) also made a pre-announcement last week. Yingli publicized fewer module sales then originally projected. The sales appeared to be around 350MW, compared with the expected 390MW, suggesting the Baoding-based company did not take part in the Q4 rush. Yingli, similarly to Suntech (NYSE:STP), took deep cuts in the value of its long-term assets. Fine Silicon, Yingli`s poly plant subsidiary, appeared to be written off with removal of $404M in equity from the balance sheet. Operationally, taking FS out of the equation made complete sense; the plant's production costs were more than double that of the spot market price. The $45 per kg, a middle of 2012 target, appeared to be unprofitable in the current environment. Furthermore, FS wasn't part of the first batch of polysilicon producers to profile future consolidation, announced in December by China. These outcomes could suggest internal polysilicon development is no longer on the agenda at Yingli.

In addition to internal impacts, Yingli took a $135M provision on silicon purchases, making Trina`s $9M appear immaterial. In the past, Yingli reported that 40% of poly needs were supplied from the spot market, while the rest constituted a blend of long-term contracts and their own production. Yingli`s biggest supplier is the South Korean OCI Chem, with three separate contracts totaling $1.6B. Earlier this month, OCI reported a 43% drop in revenue as a direct effect of the spot price. Despite this statement, and unlike GCL Poly, OCI may not be able to provide flexible pricing due to significantly higher production costs. Yingli`s actions seem to recognize this condition, and this dollar figure allows coverage for a large amount of polysilicon, possibly beyond 1GW worth of material.

There is a lot of activity in Japan prior to the introduction of new "feed in tariff" in July 2012. Last week, we reviewed results for Q3 showing imports gaining ground with 200MW out of 1GW in domestic sales. Chinese companies Canadian Solar and Suntech have the most significant presence in the country and according to last week's report from Nikkei, Canadian is considering the 150MW module factory in Japan. In the presentation shown at Jefferies' CleanTech Conference this week, Canadian Solar delivered the technology road map, which included n-type mono cells, and above 20% hetero-junction cells. The best-known hetero junction cells today come from Sanyo, in its brand name HIT. This year Panasonic, which took over Sanyo a couple of years ago, is reintroducing this product to the Japanese market under its own name. Considering the substantial price disparity between Chinese and Japanese products, a company like Canadian could do very well, particularly after gaining local acceptance and offering products that are similar in quality and concept.

In early February, another large Japanese conglomerate, Sharp (OTCPK:SHCAY), released its data for Q3 and full-year expectations, revealing further loss of the market share to the Chinese. Sales of solar cells were 159.4 billion yen for three quarters, 21.6% lower when compared to this time last year. On a production volume basis, sales were 831MW, a 10.3% drop for the same period one year ago.

For the fiscal year ending March 31, 2012, Sharp projected sales of solar cells to reach 200.0 billion yen, 24.7% lower when compared to the previous year, and on a production volume basis, 1,100MW, down 11.4%. Total revenue year to year is expected to drop by $900M with 200MW fewer sales. Unable to reach profitability, the company announced plans to go downstream and concentrate efforts on domestic production. One of the initiatives involves cell outsourcing with Taiwanese companies NSP and Gintech.

Another South Korean company is scaling back its solar plans. Hyundai Heavy Industries scrapped $700M worth of investments in solar plants, and is considering a sale of one of its manufacturing plants. Plant #1, accounting for 50MW of production, has been shut down since November, and two other plants with a capacity of 550MW produced only at 50% utilization so far this year. HHI plans to stick to its plans of growing thin-film joint venture Hyundai Avancis with French Saint-Gobain. Currently at 100MW this year, plans see the venture reach 400MW in 2015.

On the sidelines of the Korean solar commitment crisis, about which we wrote an article last week, it would be interesting to hear from Hanwha Chemical. Hanwha, which entered solar business late compared to other South Korean corporations, made a bold investment in a Chinese company two years ago. While others are cancelling their plans, Hanwha has not announced any changes, particularly to the plan for a polysilicon plant scheduled for operation in 2013. Hanwha's investment in Solarfun, currently known as Hanwha SolarOne (HSOL), remains the best prospect to keep up with Chinese companies, but with dramatic personnel changes and the full departure of Chinese senior management last year, another set of plans may be awaiting publication.

Disclosure: I am long TSL, YGE.