A funny thing happened at the tail end of the 2006 IPO market: Two nearly identical Chinese solar-panel companies went public. One rose while the other dropped. A look at why suggests that investors are still doing their homework when it comes to evaluating new offerings.

The two companies are Trina Solar ( TSL) and Solarfun Power Holdings ( SOLF). Both were young with surging revenue. Both were profitable. Both saw revenue growing. Both were struggling with the shortage in silicon, which is expected to keep prices mounting for several more years.

Yet Trina has held close to its offering price while Solarfun dropped like a stone on its first day.

I looked at Trina Solar last week, but Solarfun contrasts in telling ways. Most notably, the stock -- which priced at $12.50, the midpoint of its range --dropped to $9.91 on its first day of trading. Although it closed at $12.06 on Friday, it has yet to trade back above its offering price.

Why? Solarfun was a little greedier than Trina, for one. There were simply more shares to go around. Whereas Trina initially filed to raise $66 million and ended up raising $98 million after it lifted its offering price, Solarfun filed to sell $186 million with 12 million American depositary shares (equal to 5 Solarfun shares) priced between $11.50 and $13.50 a share.

Like its competition, Solarfun appears to be on sound financial footing. The numbers it gives in its Securities and Exchange Commission filings show that the company has been profitable at the operating and net levels in 2005 and 2006, impressive given that its operations began bringing in revenue only in August 2004.

Solarfun's revenue grew to 386 million renmimbi ($49.6 million) in the first nine months of 2006 from $11 million in the same period the year before. The cost of revenue grew to $34.3 million in the first nine months of 2006 from $9.77 million in the year-ago period. That pushed Solarfun's gross margin up to 31% from 13%.

The company's operating margin grew even more quickly, rising to 20% in the first nine months of 2006 from 6% a year earlier. But there's a fly in this sweet-smelling ointment: General and administrative expenses were $4.06 million in the 2006 period, a huge jump from the prior year's $347,000. The bulk of those increases had to do with a "compensation charge" that Solarfun gave to key investors before its IPO in the form of discounted stock.

Here's how the company explained it in its prospectus: "We recorded a share compensation charge of $1.3 million, which related to a sale of our ordinary shares to Linyang Electronics, a company controlled by our chairman and chief executive officer, at less than fair market value by other shareholders of our company and a share compensation charge of $1.55 million as a result of the issuance of series A convertible preference shares to Good Energies Investments."

In other words, Solarfun allotted its pre-IPO stock to its CEO and an outside investor, U.K.-based green fund Good Energies Investments, at a discount to its fair market value.

Such unorthodox payments may not catch the attention of regulators, but they feel dodgy in an investment climate where dozens of U.S. executives have resigned in the wake of options-backdating scandals.

And if investors don't feel encouraged by the footnote disclosures in Solarfun's prospectus, they may also not like the things the company says when it discusses its handling of the protracted silicon shortage. Silicon is a key material not only in computer chips but also in solar modules and panels.

A key to appraising solar companies is looking at their risk sections. Everyone knows rising silicon prices is a risk; what's important are the tactics companies use to get around the problem. Solarfun is relying on costlier short-term contracts and prepaying silicon suppliers, which is risky if its suppliers can't get their hands on suppliers.

This already happened in 2006. "Some of our major silicon-wafer suppliers failed to fully perform during 2006 on their silicon wafer supply commitments to us, and we consequently did not receive all of the contractually agreed quantities of silicon wafers from these suppliers," Solarfun said.

While its rivals have taken resourceful approaches to the silicon shortage -- Trina is buying scrap silicon at bargain-basement prices, while Suntech Power ( STP) relies on long-term contracts -- Solarfun seems to lack a coherent plan. Like other China-based solar companies, it relies on lower labor costs (that is, when it's not jacking them up through discounted stock allotments).

So Solarfun sold more shares with a less-coherent strategy to manage silicon prices and some questionable disclosures in the footnotes of its income statement. If it had been the first Chinese solar play out of the gates, investors might have been more forgiving.

But it wasn't. Instead, it's vying for capital in a sector that's growing more crowded -- even as the industry's prospects remain challenging for a few years. That will keep investors cautious for the time being.