Commencing operations in mid-2006, China-based LDK quickly established itself as one of leading, low-cost manufacturer of silicon-based solar wafers, the principal raw material used in the production of photovoltaic (PV) cells and modules. The company entered first-quarter 2010 with wafer capacity of 1.8 GW and expects to reach capacity of 2 GW in the second-quarter, up from 580 MW in March 2008, chairman and chief executive officer Xiaofeng Peng announced to listeners on the company's fourth-quarter 2009 earnings call.
Spot prices for usable, solar-grade silicon soared above $300 per kilogram in March 2007, from $40 per kilogram just a few years prior, due to chronic shortages of usable, solar-grade silicon. Polysilicon feedstock comprised 80.7 percent total cost of goods sold in 2007, up from 75 percent in 2006. Consequently, management predicted that "in-house" production of all elements of the solar supply chain -- polysilicon, ingots, and wafers -- would help the company to mitigate rising raw material costs through process cost improvements, such as silicon recovery gains (installation of closed-loop recycling systems). LDK broke ground on a polysilicon production plant in August 2007.
By controlling each node of the manufacturing value chain -- and locating the plant adjacent to its existing low-cost (wages) wafer facility in Xinyu City, China -- LDK was also confident it could reduce overall production costs, improve product quality (such as conversion efficiency rates), and enhance its core competitiveness in an increasingly crowded (and commodity-like) market for solar wafers.
Unfortunately for LDK, reality isn't determined by what its management says, but what the evidence shows us. Chairman Peng (who owns 67 percent of the voting stock) and his management team underestimated industry demand and the technological complexities involved in refining metallurgical silicon into the consistent, higher (purer) quality required for use in solar power, resulting in processing delays and cost overruns: in 2009, aggregate annualized production was about 6,000 metric tons -- significantly lower than nameplate capacity of 16,000 metric tons. Construction costs for the silicon plant are nearing $2 billion -- well above the forecasted budget of $1.2 billion outlined in the 2007 annual report.
Additionally, too much industry capacity coming on-line in recent years has collided with slowing demand for solar cells, culminating in dramatic falloffs in component prices across the value supply chain. Aside from a penny or two fall in wafer conversion costs (US$0.32 per watt in 4Q:09), LDK has yet to wring out any of the margin or competitive advantages it envisioned with its vertical supply integration strategy:
- In the fourth-quarter of 2009, LDK reported wafer average selling prices (ASP) of US$0.83 per watt, down from US$2.24 per watt in 2007. The significant decline in wafer prices forced the company to write down the carrying values of its inventories by almost US$500 million in the last two years.
- High-cost inventory continues to compress operating margins: the average cost of polysilicon consumed in the fourth quarter was US$70 per kilogram, above spot market prices of between US$50 and US$55 per kilogram. Gross margin dropped in two years from 32.5 percent to 9.9 percent. The company still has another 2,900 - metric tons of purchased feedstock (at the average cost of $US 70 per kg) to work through -- which will likely take another two quarters, according to chief financial officer Jack Lai.
Although success remains elusive, management did note on the earnings call that wafer ASPs had begun to stabilize in first-quarter 2010, and would likely increase through the third quarter due to global wafer capacity constraints and a rebound in demand from end-users.
Stability in wafer pricing has continued through the first week of April, confirmed Pvinsights; but, the photovoltaic (PV) research firm cautioned that most Asian solar cell makers were aggressively expanding their capacity. Running their factories at full utilization would once again end up with supply outstripping solar demand (possibly, as early as mid-2011), opined some analysts.
Despite its inability to capture wafer cost advantages through the ownership of more of its upstream supplier chain, management is nonetheless moving forward with its vertical integration strategy: on February 26, LDK announced an investment in the downstream PV module market, betting that excess production could now be sold through its own distribution chain. Seemingly ignoring a failed upstream strategy that high volume, in-house manufacturing was the key toward cost reduction, LDK looks to be headed down the same dead-end road: "We look forward to realizing cost efficiencies by bringing crystalline module manufacturing in-house," said CFO Lai.
As the proverb goes: "The devil is in the details. With its new in-house polysilicon factory already a debt-ridden millstone around the company's neck, any plans by management to expand module capcity will likely sink the company for good.