On Friday, second-quarter results were released by ReneSola Ltd. (ADR)(NYSE: SOL). The company previously increased its quarterly guidance and yearly guidance in the area of module sales, which had been confirmed by the sales in the quarter. Module shipments of 159.7MW and 344MW of wafer shipments produced $233.0M in revenue. Module sales increased75% period over period from 90.9MW in Q1. The overall GM results were 0.7%, but the company had broken down the module GM out of the calculation and offered 12.2%.  Both revenue and COGS had been affected by unique items. Revenue received a credit benefit from the insurance cost due to the lower average selling prices. This benefit equalled to $7.9M. The costs were increased by $15.5 of provisions over the polysilicon pricing in the company’s inventory versus the market’s prevailing price. 

The ASP had been rounded to $0.75 per watt, and after receiving the insurance benefit costs have been lowered to $0.66 per watt. If the benefit was removed, module costs would go up to $0.70 per watt, leaving 7.1% gross margin.  Sales of modules included 76.1MW of the Virtus module, which is an impressive seven- fold increase from 10MW in Q1 and is a factor in higher costs.  Trina Solar Limited (ADR) (NYSE: TSL) had not disclosed the percentage of penetration of Honeycell-equipped modules in the Q2 shipments, but their processing cost was $0.66 in Q2.  Virtus module, due to wafer cost, wasalways considered to be higher than regular processing of polycrystalline modules.

This increase is already being addressed by the new wafer technology. ReneSola has been famous for its quality and R&D initiatives, which regularly deliver new products to the market. Quoting the news release, “ReneSola is now beginning production of its Virtus II modules, which will use the Company’s new Virtus A++ manufacturing technology, a unique technology developed by ReneSola. This technology does not require the use of a crystalline seed in the crucible to produce higher-efficiency solar wafers and results in multicrystalline ingots with a more uniform crystalline grain and distribution.” Not only sounding revolutionary, processing costs offered for this product are substantially lower than processing averages used in Q2. A++ wafer requires processing costs of $0.12 using the newest lines, which the company is expected to lower to $0.11 by the year’s end.  In addition, a new product line of microinverter has been introduced, to capitalize on the full solution offering. The company expects to bundle this product with high-conversion modules. 

Going back to results from Q2 and wafer costs, the company has taken a $15.5M inventory cost adjustment on poly, which was included in COGS.  When using the 344MW of shipments sold at $0.31 per watt, and allocating costs to wafer while eliminating provision, overall margin versus ASP was close to zero.  The processing wafer costs averaged at $0.17, balancing out poly costs per watt to $0.14 or $25.40 per kg.  

Regarding this dynamic, the company had offered costs improvements in the area of $0.63 per watt for all-in costs by Q3. The opportunities here are to reduce cell/module processing toor below $0.33 per watt from what we believe has been $0.39 in Q2.  There is also an opportunity to find a cent or two from overall wafer processing dropping to $0.15 by end of Q3.  Unfortunately, with expectations of ASP for modules already being quoted at $0.65 during the conference call, this will show – at best – flat gross margins in Q3. 

We have learned more about ReneSola’s project portfolio.  There is an existing 20MW facility in Qinghai, China. Another 9.7MW was built in Bulgaria. There are approximately 70MW of open projects for the remainder of the year.  ReneSola,publically, has been very conservative about engagement in EPC, yet their small but firm figures suggesttaking the path of other US-listed companies. Within last six months those pipelines had taken on interested volume, which range from 1GW for Hanwhato smaller ones like Canadian Solar with 590MW, Jinko with 235MW.  

A balance sheet review offers similarities to peers. The cash account increased by $5M, but so did the overall interest-bearing liability by around $20M. Another $41M was added to accounts receivables, suggesting sales on credit in the quarter. The practice of holding onto cash added $121M in accounts payables, while other accounts were reduced. Overall, $112M in liabilities wasadded in the quarter. The company has $0.42 of cash to a dollar of debt, 2.5 times more than JinkoSolar Holding Co., Ltd. (NYSE:JKS). Based on recent results on interest payments there is a separation of rates between tier-1 categories of companies versus the tier 2s.  Trina and Canadian Solar seem to have rates below 5%, while ReneSola and Jinko are above. The exception is GCL-Poly, which had a 5.89% interest rate based on the amount of payments made in first half of the year.  This largest manufacturer of poly and wafers added $1B in debt in six months, and for the first time in three years had a net income loss in the first half of 2012. 

ReneSola’s net income loss was $34M and $0.40 per ADS.  ReneSola used an income tax benefit of $16M to reduce $51M EBIT results, which included – small in comparison – a Forex loss of $4.5M,indicatingthe majority of the company’s business remains in China. In the matter of guidance for Q3, the company sees 530MW of sales, including 170MW of modules, to reach the range of $220M in revenue.  The full-year guidance was also upheld, which brings up a similar concern we expressed with Canadian Solar’sguidance with their ability to manufacture their guided shipments.  Based on 1.2GW of module capacity and to-date sales of 250MW, includingthe high end of the Q3 guidance, Q4 sales are to be at 179.4MW, which is easily obtainable against capacity for the rest of the yearto meet 600MW of module sales.  

The problem is the wafer situation.ReneSola repeatedly said that they do not count wafers sales toward their own module segment.  Therefore,by simply using the calculation of what has been shipped to date, including modules with guided volume for the Q3, counts are at 1500MW. The company runs its capacity at 100%, which has been changed down to 2.0GW in the last 20-F. This bringsfew concerns. First, 500MW per quarter of wafer production would be short of the 205MW of wafers needed to meet the lower end of the yearly guidance of 2.2GW.  Second, the high-end guidance would require 400MW more than the company can produce.  Furthermore, inventory turn is about 80 days, so some carried inventory would have to cover the gap between production and sales in Q4, but it can cover more than the company can produce. Unless the company had forgotten to update investors on the current wafer capacity, there seems to be not enough wafers available.  We already know that the current running cell capacity is 240MW, which obviously forces the company to use OEM services. 

Sincecalculation suggests that a portion of modules will not be made with the company’s wafers at all, Taiwanese cells most likely will be purchased, somewhat explaining how the company sees a big entry into the US market in Q4.