Intersolar 2011: Notes From the Solar Underground

Module prices at $1.30 a watt? First Solar opening a plant in India? Component inventory updates? All right here, from the trade show floor.

Every year in early June, tens of thousands of fair-minded citizens converge upon the fair city of Munich to take part in the world’s (after SNEC, the second-largest) solar trade fair, Intersolar Europe. As usual, GTM Research analysts were around to measure the pulse of the industry and get the latest dirt on industry developments. Below is a summary of the most noteworthy items.

Inventory Levels: Through the Roof, but Signs of Shedding

Component inventory reached record levels across the value chain in late April, a direct consequence of hitherto weak demand in the two PV markets that matter the most: Italy and Germany. In the case of Italy, the culprit was subsidy uncertainty for the second half of the year, which led project finance (and consequently shipments) to all but dry up. Recognizing this, German developers adopted a 'wait-and-see' approach toward module purchases. The thinking was that suppliers would be forced to lower prices in a bid to find a home for the gigawatts of orphaned modules that were expected to be Italy-bound -- and, as is discussed below, the strategy appears to have worked. Recent signs, however, have indicated that demand in both these markets has picked up substantially. Italy’s feed-in tariff scheme for H2 2011 was finalized in early May, and the revised FIT levels will still drive attractive rates of return. German demand also seems to have made a comeback, as prices have eroded substantially over the past three months (see below), and developers need to connect systems to the grid before the upcoming variable-rate FIT cuts that will take place in July. As a result, all signs point to a rapid pick-up in demand and shedding of inventory levels through the end of June, although demand visibility in Q3 and Q4 (especially) remains murky.

Module Prices at $1.30/Wp? Yes and No

The most talked-about tidbit doing the rounds of the trade show floor centered on current module prices. It is well known that despite stable ASP results from the public PV manufacturers in Q1 2011, component prices fell a long way from March to May given the above-discussed buildup in inventory. However, the number thrown up again and again was shockingly low: Tier 1 Chinese crystalline silicon module prices in the region of 90 to 95 Eurocents per watt, which at current exchange rates, amounts to U.S. prices of $1.30 to $1.35/Wp. Given that Trina’s current blended cost structure, for example, is roughly $1.16/Wp, this implies a gross margin of 10% to 17%, and the numbers for smaller producers lacking vertical integration would be even worse. Were producers so hard up for sales that they were willing to give up this much long-term pricing power? Account for the fact that European demand has been picking up of late, and clearly something wasn’t right.

The answer lay in the nature of the contracts that were being signed at these prices; rather than long- or medium-term contracts, these were largely inventory flush points representing one- to six-week delivery periods, hardly representative of average sales price (ASP) trends; Trina itself revealed that its Q2 ASPs would come in at around $1.65/Wp (we think they’ll be closer to $1.55/Wp). Clearly, there is a danger in accepting anecdotal numbers at face value, and rumors can spread quickly on the dance… er, the trade floor. Meanwhile, c-Si cell ASPs are in the region of $1.10/Wp, while wafer ASPs, having held up for so long, have plunged to around $0.70/Wp. Again, spot pricing for both is considerably lower ($0.95/Wp and $0.60/Wp). While these numbers are not representative of current blended prices, they do show how low prices could go in an environment of structural oversupply.

First Solar’s India Plant: It’s Happening

First Solar has made no secret of its ambitions in the nascent Indian PV market: the company signed sales contracts with project developers Moser Baer Clean Energy and ACME Tele Power for delivery in 2011 worth 40 MW and expects an additional 60 MW to be deployed in the cricket-crazy nation this year as well. Besides being attractively priced, First Solar's CdTe modules can provide higher energy yield (and therefore drive lower LCOE) in the hot, dusty climate there compared to conventional crystalline silicon (little brother Abound Solar also announced a sales agreement with an Indian customer in March 2011, touting similar advantages).  However, numerous sources at Intersolar indicated that the firm is likely to go one step further by setting up a module manufacturing plant in India at some point in 2013. If you’re wondering why, it’s because current stipulations of the Indian National Solar Mission mandate that modules must be produced domestically, past the current pipeline of legacy projects. In our view, however, depending on India as a demand center of consequence any time soon carries significant risks: the bragawatt-to-megawatt ratio is likely to be high, cycle times for projects will be long, and there are no sustainable financing solutions yet present. At the same time, nothing would prevent First Solar from shipping its Indian-made modules to Germany, Italy, and the U.S., just as is the case with its existing facility in Malaysia and its planned plant in Vietnam.

The Firm to Fear: GCL-Poly

We all know that when it comes to competitive positioning in PV manufacturing, the top-tier Chinese players are the ones to beat. But who is it that's keeping those guys up at night? While the obvious answer would be Yingli, LDK, or Suntech, the name that popped up repeatedly in our discussions was GCL-Poly Energy Holdings. The company has scaled up the integrated polysilicon-wafer manufacturing model like no other: by the end of this year, it will reach polysilicon capacity of 46,000 metric tons and wafering capacity of 6,500 MW, up from almost nothing in 2008. Additionally, the company reported an obscenely low polysilicon manufacturing cost of $22.90/kg for Q4 2010, which would put it head and shoulders below the rest of the Siemens pack. While it’s hard to validate these numbers given the lack of disclosure, the firm’s progress is a testament to the speed with which Chinese capital can be deployed and mobilized (the firm’s financing partners include the China Development Bank and the Bank of China). Even in the industry's far-from-level playing field, some companies have it better than others, and it looks like GCL-Poly is at the top of that pack.

Western and Japanese Firms: Selling Energy Yield, but Who’s Buying?

The consistent marketing message emanating from the booths of the more established crystalline silicon producers from Europe and Japan (Q-Cells, Mitsubishi, and Kyocera, for example) was that of high module quality and, consequently, high energy yield in terms of kilowatt-hours per kilowatt. Given that there is no way these firms can compete on cost/price, and that top-tier Chinese manufacturers (Suntech, Trina, Yingli, SolarOne) have emerged as bankable suppliers, this is hardly surprising: as pure-play producers, there’s not much else they can tout. However, while it’s true that modules made by these developed-world firms have a longer operating history in the field and are a surer bet in terms of output predictability, there has been scant evidence so far of regional trends when it comes to yield characteristics for c-Si modules; from all accounts, top-tier Chinese modules perform as well (in some cases, better) than those from Western and Japanese competitors (see selected results from Photon’s 2010 yield measurement test here, which shows Taiwanese Win Win Precision and Trina Solar at the #2 and #3 spot, respectively). In our view, these firms are hanging on to scraps: as indicated in last week’s article on PV competitive positioning, in the absence of low manufacturing cost, some form of technology (related to efficiency, performance, or product) or business model (polysilicon, system sales, project development) innovation is required to compete with the Chinese; if nothing else, they need to create sales channels in emerging markets to buy them some time.  Without even that, their fate lies squarely at the mercy of still-lucrative feed-in tariff markets, and as recent developments have shown, the purse strings are getting tighter every day.

The U.S. Market: Promising, but Not Well Understood

With 887 MW installed in 2010 (5% of the global market), the U.S. holds a much smaller portion of Intersolar Europe's collective consciousness than Italy and Germany. However, it is clear that the U.S. is becoming the perennial "next" major PV market. In times of difficulty in European markets (see Q1/early Q2), suppliers point to the U.S. as the future destination of their newly expanded production capacity. (For recent examples, see Q1 earnings calls from Suntech, Trina, and Solarworld.) With the U.S. market anticipated to double again this year, expectations remain high -- but active U.S. market participants report cloudy skies ahead. Historically strong markets (California commercial, Arizona residential, New Jersey commercial) are slowing down, and plummeting module prices have yet to open up substantial near-term demand. While the U.S. market will undoubtedly grow this year, it is time to reset expectations and plan for the long-term, steady growth that the U.S. market is more likely to experience in reality.