NEW YORK ( TheStreet) -- Most solar panel manufacturers rely on polysilicon as their primary raw material to capture electricity from the sun. However, the price of the commodity has fluctuated sharply over the past two years as polysilicon producers ramped up production capacity to match rising demand for solar modules.

In 2007, undercapacity and short supply led the price to rise to $400 per kilogram, equivalent to about $2.80 per watt. This spurred manufacturers to lock in long-term contracts and invest in their own polysilicon production.

LDK Solar ( LDK), Yingli Green Energy ( YGE) and ReneSola ( SOL - Get Report) all had plans to begin polysilicon production for themselves, but some may now regret their expensive ambitions as polysilicon spot prices plummet, hitting $60 in the third quarter of 2009.

LDK Solar, a solar wafer producer, broke ground on its polysilicon production plant in August 2007, at a point when polysilicon was around $350 per kilogram on the spot market. It was to build a small plant and three larger plants, with an original plan to produce 150-350 metric tons of the material in 2009 at its smaller facility and a further 5,000-7,000 metric tons at its larger facilities in 2010.

LDK has now delayed its polysilicon production facilities several times due to issues in plant commissioning and poor market demand for wafers. The company lowered its 2009 target to 15-25 metric tons for the year. The company's larger project -- a 15,000-MT capacity plant -- was originally slotted to reach full capacity in 2009, but 10,000 MT of capacity have been pushed back until further notice. Production targets for the year have also been cut by 66% from the original estimate of 5,000-7,000 MT over 2009.

It will be several months -- assuming no more construction or processing delays -- before LDK starts making significant quantities of silicon. We continue to have concerns about the quality of the polysilicon produced since, as a new entrant, the company will not likely achieve anywhere near the efficiency or quality of the more experienced silicon players.

Also, the company expects its initial batches of polysilicon to cost the company around $80 per kg, with the cost falling to as low as $35 per kg over a few years, as the company benefits from economies of scale.

Intriguingly, just this past week, LDK sold a 15% stake in its polysilicon plant to a Chinese trust and investment company for $219 million. Although this will relieve some liquidity concerns surrounding the company's high level of debt, it may also be an acknowledgement that the project was ill-timed and proved to be more costly than anticipated.

On Jan. 8, 2009, Yingli Green Energy completed the purchase of Cyber Power for $77 million, a development-stage company, with plans to begin production of polysilicon in the second half of 2009. This acquisition will make the company fully vertically integrated, since it will produce everything from raw polysilicon to complete solar panels.

Yingli expects that it will be able to produce silicon for $60 per kg during top production periods (likely six months out), and then after stabilizing it, expects to reach costs of $30-$40 per kg, followed by costs as low as $25-$30 per kg after a year.

ReneSola, a solar wafer producer dabbling in polysilicon and modules, completed the first phase of its 3,000 MT polysilicon plant and expects to produce 200-250 MT before the end of 2009. The company declared that its current production cost is below $55 per kg, and it expects costs to decline to $50 per kg by the end of 2009 and below $40 per kg by the end of 2010.

While investments in polysilicon plants looked extremely profitable in 2007 and 2008, when the raw material cost more than $300 per kg, the projects have proven to be much riskier than anticipated. The spot price of polysilicon has fallen to below $60 per kg, making it much more difficult to produce at a profit without economies of scale.

All three companies will be producing batches at costs higher than market prices, so it will take time to ramp capacity to bring costs down. The market spot price is also expected to stay low, at least until solar demand shows more signs of stabilization. It is also evident that commissioning and ramping a poly plant is not easy.

LDK suffered numerous setbacks in its plant construction, causing it to push back production targets and delay mass production until further notice. These companies have invested significant capital in these projects, so it may be difficult to back out of completing them.

Until these solar players produce meaningful quantities of the raw material at low cost and sufficient quality, we are taking an "I'll believe it when I see it" approach.

-- Written by Battle Road Research in Waltham, Mass.

Battle Road Research (www.battleroad.com) an equity research firm, serves fund managers, analysts and financial advisers with an independent voice on technology, health care, solar power and education stocks. Battle Road analysts place an equal weight on industry and securities analysis in an effort to seek out stocks to buy and stocks to avoid. As an integral part of our research process, we tap into a network of industry sources who provide insight into the companies we cover. We present our conclusions in a straightforward buy, hold, sell format. As a matter of principle, we refrain from investment banking, company-paid reports, and personal investment in the stocks we research. Visit us on the Web at www.battleroad.com and www.battleroadblog.com.