Five years ago, enterprise-software maker J.D. Edwards ( JDEC) celebrated an impressive IPO, its shares skyrocketing 55% on their first day of trading.

But one year ago, the Denver company's fortunes were much different in a tight IT spending environment, particularly in its sweet spot -- manufacturing.

Ed McVaney, co-founder, CEO and chairman at the time, was busy quashing rumors that J.D. Edwards was on the selling block while steering the company through its second restructuring in as many years.

Times have changed again. With IBM veteran Bob Dutkowsky at the helm of the company since the beginning of the year and new products in place, J.D. Edwards has become a turnaround story. The 25-year-old company has beaten estimates the past three quarters, as rivals such as Siebel Systems ( SEBL), Oracle ( ORCL) and SAP ( SAP) have missed numbers or have lowered outlooks. Because J.D. Edwards has fallen -- it's down 24% on the year -- it may now offer a better opportunity to investors than other software stocks.

"A year ago it became obvious that if the ship was even remotely pointed in the right direction, you could see some pretty good operating-margin expansion," said Josh Freedman, an equity analyst with Berger Financial Group, which owns J.D. Edwards shares in its Growth and Small Company Growth funds.

To be sure, J.D. Edwards, which reports third-quarter earnings Wednesday, still has a way to go. Third-quarter license revenue is expected to ring in at less than half of its peak in the fourth quarter of 2001, when it totaled $137.4 million and overall revenue came in at $282.7 million.

For the third quarter, the company projected license revenue would be $58 million -- up 7.2% sequentially and 16% from a year ago -- and earnings would come in at 6 cents a share -- flat sequentially but a reversal from a net loss of 3 cents a share in the year-ago period.

Wall Street analysts polled by First Call/Thomson Financial are expecting third-quarter earnings in line with company guidance on total revenue of $222.2 million -- virtually flat sequentially and up 6.1% from a year ago.

Buying in advance of the earnings report may not be the most prudent move. Credit Suisse First Boston analyst Brent Thill noted that the company lowered guidance below Street estimates in the past two quarters. Thill, who has a buy rating on J.D. Edwards, believes there is a small chance that could reoccur. (The CSFB technology group, while working for a different company, took J.D. Edwards public.)

JMP Securities analyst Pat Walravens, who has a strong buy rating on J.D. Edwards, also believes the risk is there. Current Wall Street estimates peg estimates for the seasonally strong fiscal fourth quarter, which ends in October, at 11 cents a share and revenue at $252.1 million.

That represents a decline in earnings from 18 cents a share a year ago and a 4.6% year-over-year increase in revenue. Walravens' firm hasn't done any banking with J.D. Edwards.

Opportunities in the Margin

Still, Walravens is bullish. One major reason: Walravens sees plenty of opportunity for J.D. Edwards to improve its operating margins.

That shouldn't be a huge challenge, considering that in fiscal year 2001, the company's pro forma operating margins were 0%. At 5% in the second quarter, J.D. Edwards' operating margin was the lowest among rivals PeopleSoft ( PSFT), (12%), Siebel (19%) and Oracle (35%), Walravens said in his note initiating coverage of J.D. Edwards at the end of May.

J.D. Edwards already has taken several steps to improve margins, including cutting about 1,200 jobs.

Because margins are so low, analysts have shied away from using a price-to-earnings ratio to value J.D. Edwards. Based on Tuesday's close of $12.43, shares are trading at about 35 times 2003 earnings, hardly a significant discount for a software stock.

Freedman prefers to use a price-to-sales ratio, with sales defined as trailing revenue plus cash. The company recently has been trading at between 1 and 1.5 times sales, just a little higher than the historical bottom of 1 for software stocks, he said.

However, J.D. Edwards does not look like a good buy to everyone. Morningstar analyst Mike Trigg agrees that the company has "definitely" turned the corner, has a strong base of customers and a good opportunity to expand margins. But he believes the stock is overpriced. Trigg figures the stock's intrinsic value, based on discounted cash flow, is only $10, so he's not recommending buying. His firm doesn't do investment banking business.

"I would just say that the optimism around the turnaround is already reflected in the stock," he said. "People are assuming IT spending will recover and these guys are going to do pretty well."

J.D. Edwards initially stumbled with supply chain management software and was late to the game with customer relationship management software, said Tom Topolinski, a Gartner vice president of worldwide software applications. In a note a year ago titled, "Too little, too late for J.D. Edwards," he questioned the firm's viability as a major software company.

Topolinski now says he doesn't believe it is too late for J.D. Edwards. "I think they certainly have made a lot of progress from last year to this year," he said. "It appears they do have potential to be a strong contender in the enterprise application market." He cited the successful launch about nine months ago of the company's CRM product.

A survey by J.D. Edwards of its 6,400 customers found that less than 9% of them have a CRM solution in place. That's "a huge greenfield opportunity from people customers who like you a lot," Freedman said.

And while SAP, Oracle and PeopleSoft also are touting a suite of products, J.D. Edwards has carved out a significant niche in the middle market, which it defines as companies ranging from $200 million to $5 billion. The middle market requires more hand-holding and may not want products as complex as those for sale by the likes of SAP, analysts say. And that, of course, is more good news for J.D. Edwards.