The Los Angeles Times has a really good look at the failure of Solyndra, the solar-power company that went bankrupt earlier this month despite a $528 million Department of Energy loan two years ago.

This isn’t a story about the machinations of the Obama administration or an investigation into why the FBI raided the company as it was failing. It’s a story about Solyndra’s business, why it was so promising, and why it ultimately didn’t work.

The main reason: Plunging commodity prices made it impossible for Solyndra’s newfangled solar panels to compete with traditional silicon ones made in China. Silicon prices had gone nuts until the economic crash, up ten times in a “few years,” according to the LAT. Back in 2008, when Solyndra was ramping up, silicon-panel prices had soared to $4.19 per watt. They’ve since plunged to $1.25 a watt as silicon prices have fallen roughly 90 percent.

That’s good news if you want cheaper solar energy (and don’t care about Americans manufacturing the tech). It’s terrible news for a startup trying to knock off an existing technology.

The LAT makes it clear that the private sector was dazzled by Solyndra:

“It was revolutionary,” said Walter Bailey, a former Macquarie Capital investment banker who specialized in green technology and visited Solyndra in 2008. “You had some of the smartest money in the world getting behind it. It was a real company with a huge factory and an extremely unique product…

One investor, British billionaire Richard Branson’s Virgin Green Fund, bragged that it had selected only Solyndra from a pool of 117 solar companies seeking backing. Other investors included billionaire Oklahoma oil baron George Kaiser, and a fund that manages the money of the family behind Wal-Mart Stores Inc. Wall Street heavyweight Goldman Sachs Group Inc. was its lead investment banker.

“Very high-profile money was all over that company,” said Bailey, the investment banker. “Nobody else had anything as strikingly different as Solyndra.”

Keep that in mind when you hear people blaming public sector’s poor capital allocation for the failure and lost taxpayer money. As I wrote earlier, most venture capital investments go bust. The point of investing in them is that you think you’ve found something that has a chance to go very big.

But that hardly lets the Obama administration off the hook. The LAT notes that silicon prices had already plummeted below $100 a pound from $1,000 a year earlier by the time the Energy Department made the loan in May 2009.

Solyndra sold its panels for about half what they cost to build, but by early 2009, the price of traditional silicon panels had fallen far below its $3 a watt. That implies that Solyndra’s cost to build a panel was roughly $6 a watt. Even at the peak of the silicon bubble, traditional panels were only $4.19 a watt, and you have to assume the latter included a profit margin. Solyndra’s technology had other advantages: It weighs a lot less, doesn’t get blown around by wind, and is easier to install. Those factors are worth some kind of premium, but 40-plus percent? And by the end of 2009, silicon panels had sunk below $2 a watt, making them two-thirds cheaper than Solyndra’s. With those headwinds, it’s surprising that the company was actually able to grow its revenue from $100 million in 2009 to $140 million in 2010.

But the most consequential line of the story isn’t in the story at all. It’s in the accompanying graphic.

Solyndra panels head steady at more than $3 per watt

That looked okay in 2008, when the DoE was looking at the Solyndra loan and Silicon prices were in a bubble. Even in May 2009, it would have been totally unreasonable to think the price crash was an overcorrection. But that brings up one unspoken lesson from the Solyndra story: Commodity-market bubbles distort markets (I owe somebody a hat tip for that thought. Shoot me a note if it’s you).

TAGS: , , , ,

Ryan Chittum , a former Wall Street Journal reporter, is deputy editor of The Audit, CJR's business section. If you see notable business journalism, give him a heads-up at rc2538@columbia.edu.