(FORTUNE Magazine) – Rick Priory was on top of the world. It was early 2002, and as CEO of Duke Energy he had taken a conservative electric utility and plunged it headlong into the newly deregulated power market. During his five years as chief executive he had acquired everything from gas reserves to pipelines to power plants, turning Duke into one of the most diversified energy players around. In his eyes, he was fulfilling his long-held goal of turning Duke into a "master architect" of the "new energy economy." And now that Enron--the one company that had overshadowed Duke--had crashed, Priory could hardly contain himself. "You can only do smoke and mirrors so long," he told FORTUNE at the time. We were certainly impressed. With $60 billion in sales in 2001, Duke had just shot up to No. 14 on the FORTUNE 500 list; the company's share price had recently doubled, to a high of $47. From our perspective, Duke's impressive numbers were a strong endorsement of Priory's longtime claim that Duke was the "un-Enron"--so much so that we titled our April 2002 story just that.

Well, Priory was wrong, and so were we. Duke turned out to have a lot more in common with Enron than anyone thought. By chasing after what proved to be short-term trends (sky-high electricity prices, the prospect of fully deregulated power markets), Duke got caught up in a depressingly familiar series of problems--government investigations, indictments, lawsuits, and, yes, accounting scandals. Priory built or bought dozens of new power plants just as the electricity market was collapsing, leaving Duke with tons of worthless assets and billions of dollars of debt. Duke shares have dropped to $22, the level at which they traded a decade ago, and Priory is out of a job.

Now his replacement, CEO Paul Anderson, is desperately trying to reverse the damage. Anderson has cut spending, sold assets, and scaled back trading. But Wall Street hates the stock--virtually all the analysts covering it rate it sell or hold--and short-sellers are swarming. The problem, as Anderson freely admits, is that he has no idea how to fix Duke's troubled merchant-energy unit, which generates and sells power to utilities and other wholesale customers, and is projected to lose $300 million this year.

Amid the debacles at Enron, Tyco, and the rest, Duke's fall from grace has largely been overlooked. It shouldn't be. For years Duke was the envy of the industry, winning awards for innovation and customer service. It was the ultimate widows-and-orphans stock--safe, consistent, reliable. The utility's tumble shows how the lure of easy money can upend even the most conservative and well-regarded company. And it underscores the dangers of basing corporate strategies on the assumption that current trends will last forever.

Astructural engineer by training and a 20-year Duke veteran, Rick Priory became CEO in 1997 at what he later called the most exciting time in the industry's history: regulatory barriers falling in the U.S., new markets opening up overseas, and energy demand projected to soar. Priory attacked Duke's sleepy culture, replacing many top executives with outsiders from banking, telecommunications, and other recently deregulated industries. In shareholder letters and conference calls with analysts, he tossed around new-economy jargon like "first-mover advantage." (Priory declined FORTUNE's request for an interview.)

It's hard to overstate how big a change this was for Duke. The Charlotte company started out a century ago as a single electric-power station on the banks of South Carolina's Catawba River. By harnessing the power of the river, founder James "Buck" Duke sold electricity to a sole customer, a textile mill in nearby Rock Hill. Over the years Duke evolved to meet the region's growing appetite for power by embracing new sources for energy--coal in the 1930s, nuclear power in the 1950s, and natural gas in the 1990s. It even helped shape the landscape of the region by recruiting textile mills from the North.

By the time Priory took over, electricity consumption in the U.S. had exploded as new appliances, tech goods, and the Internet swallowed more and more juice. And because of strict environmental regulations, among other factors, the industry had failed to build enough new power plants to meet the demand. As a result, prices were spiking--as much as 200% in some regions between 1997 and 2000--and capacity shortage led to periodic blackouts.

Moreover, industry analysts were churning out increasingly bullish forecasts. The Energy Information Administration, a division of the Department of Energy, proclaimed that U.S. industry would have to build between 1,300 and 1,900 new power plants over the next two decades to meet long-term demand. Doing some quick math, Vice President Dick Cheney, in a speech at the annual meeting of the Associated Press in April 2001, put it this way: "That averages out to more than one new power plant per week every week for the next 20 years." For good measure, the Veep added, "It's time to get moving."

Priory certainly did. During his tenure as CEO he spent billions building and acquiring dozens of new power plants all over the country. On his watch Duke rolled out more than 15,000 megawatts of generating capacity in North America, enough to power 15 million homes and more than any other company except onetime giant Calpine. Almost overnight the merchant-energy unit Priory created, Duke Energy North America (DENA), became the star of the company. In 2001 it earned more than $1 billion, or about a quarter of Duke's total profits.

Despite Duke's rocketing share price, Priory was disappointed that his efforts didn't generate more enthusiasm among Wall Street analysts. The problem was Enron. It wasn't just that the Houston-based energy giant, unlike Duke, was stocked with young hotshots. The company had a far sexier business model too. Enron, as you may recall, pursued its now infamous "asset light" strategy, ditching power plants and pipelines to focus on trading.

Duke, by contrast, was a hybrid. The main purpose of trading at Duke--and the reason energy trading came about in the first place--was to prevent volatile energy costs from playing havoc with the budget. Trading allowed Duke to sign long-term contracts that locked in prices for a plant's biggest expense, natural gas. But you couldn't build a company on trading, Priory used to tell analysts on the Street. "You have to be able to produce and deliver energy as well as trade it." When Enron crashed in late 2001, Priory was triumphant. "Our values are rooted in North Carolina history," Priory bragged to FORTUNE in early 2002. "We're grounded, so we didn't fall victim to short-term fads or accounting tricks."

The victory dance didn't last long. Following revelations of sham trading among power companies, in May 2002 the Securities and Exchange Commission launched a widespread industry investigation. At the SEC's request, Duke reviewed its trading activities. It discovered that some traders had been conducting "roundtrip" or "wash" trades. Essentially, the traders were swapping power with counterparts at other firms merely to boost revenues.

Duke eventually admitted that employees had executed 89 roundtrip trades over 32 years, totaling about $217 million in revenues. In April of this year a Houston grand jury handed down indictments against two former DENA executives and a former trader, charging them with fraudulently booking $50 million in profits from more than 400 roundtrip trades. All three pleaded not guilty to all charges; a trial is expected early next year.

Duke Power, the company's regulated electric-utility business, also ran into trouble. In October 2002 an independent audit ordered by regulators in North and South Carolina found that top executives had underreported regulated profits by $123 million between 1998 and 2000. (Duke, remember, is prohibited from earning more than a certain return; if profits are too high, state regulators can cut the rates Duke may charge customers.) The regulators produced an embarrassing body of evidence--depositions, spreadsheets, e-mail--to support the charges. (Read one e-mail: "We are currently looking at reclassification entries as discussed ... to help with our 'allowed return' problem.") Duke settled the charges by paying $25 million to its customers, admitting no wrongdoing.

Duke was also one of dozens of companies accused of manipulating the California energy market in 2000 and 2001. It has agreed to pay more than $210 million to settle claims that it overcharged customers in California, Oregon, and Washington. In still another matter Duke paid $28 million in 2003 to settle charges that it manipulated the natural-gas market. Duke denied any wrongdoing in all cases.

Yet those seemed like minor annoyances compared with Duke's real problem: too much generating capacity. Priory's building binge couldn't have been more poorly timed. From 1999 through the end of 2003, the industry cranked out some 200,000 megawatts in North America, boosting total supply by nearly 25%. But demand slowed drastically. The stock market collapse and the 9/11 attacks sent the economy into a tailspin, while the California energy crisis and Enron's bankruptcy led many state legislatures to abandon plans for electricity deregulation. Also, a steep rise in the price of natural gas took a big bite out of profit margins. As a result, DENA saw earnings tumble 89% in 2002 and in 2003 posted a $3.3 billion loss. With his strategy in ruins, Priory stepped down in October 2003.

Now it's up to Priory's successor, Paul Anderson, to write the company's next chapter. On a recent afternoon, he is speaking to a group of employees in the company's bland auditorium. It's a beautiful summer day in Charlotte, but the optional "open forum" meeting is packed with people eager to hear how his turnaround plan is progressing. Duke's problems are legion, but Anderson, wearing a gray jacket and blue shirt unbuttoned at the collar, is characteristically upbeat. "We're a much stronger company than we were a year ago," he says, and rattles off a list of achievements, including paying down $2.2 billion of debt, shedding more than $1.5 billion of assets, and making "tremendous strides" settling regulatory and legal issues. "We've basically beat our goals for the entire year already," he says, "and it's only July."

Anderson, who became CEO last November, doesn't fit the mold of a typical utility boss. Despite his reserved manner, he drives a purple PT Cruiser with orange flames to the office and likes to ride Harleys. And he has developed a reputation as a turnaround artist. He cleaned up a struggling natural-gas pipeline company, PanEnergy, which he sold to Duke in 1997, becoming Duke's president and chief operating officer. In 1998, Anderson left Duke to take over as CEO of Australian mining and steel conglomerate BHP Ltd. after slumping commodity prices and disastrous investment decisions led to record losses. He restored it to profitability before merging it to create the natural-resources giant BHP Billiton.

In a recent shareholder letter, Anderson discussed the carnage in the electricity market, saying the industry resembles a "bombed-out village." But, he added, Duke is "one of the few recognizable structures remaining." He has a point. Unlike some pure-play merchant power companies such as Mirant, Duke has avoided bankruptcy, largely because of the strong cash flow from its core utility operations. Duke Power, the regulated electric utility, has more than two million customers in North and South Carolina. And with 17,500 miles of interstate pipelines (mainly the result of the PanEnergy acquisition), Duke's natural-gas unit boasts one of the largest networks in the country. Together those two businesses earned $2.7 billion last year.

That income will provide much-needed support while Anderson tries to solve the problems at DENA, the merchant-energy unit. He's already sold a third of its assets, mainly in the Southeast, an area plagued by surplus generating capacity. But the business unit is still projected to lose $300 million this year. Along with overcapacity, DENA is suffering from skimpy "spark spreads," or the difference between the cost of natural gas and the price of the electricity it generates, which have made it harder to generate power profitably. And a good chunk of Duke's earnings are still at the mercy of volatile commodity costs because of "mark to market" accounting, in which companies must use current prices to estimate the fair value of long-term energy contracts that may not be settled for years. Unpredictable mark-to-market swings accounted for $90 million of DENA's $521 million first-quarter loss.

How will Anderson fix DENA? That's a question he hears all the time these days--from analysts, investors, even employees. "It always comes back to DENA," he says. "Where is DENA going, and what's it doing?" While he has promised the Street that he will have DENA breaking even by 2006, he freely admits he doesn't know exactly how that will happen. "We don't have a road map that gets you from here to there," he says. Instead, he offers a telling analogy: "It's like eating an elephant. You just do it one bite at a time." So where is he now? "Way past the trunk and someplace into the middle. We're getting kind of bloated, though."

Anderson has considered merging DENA with another merchant player, but as he sees it, all the likely candidates are flawed. "They're either too small, they don't have enough fuel diversity, [or] they don't have enough geographic coverage," he says. "They all have problems." He could shut DENA down. But the costs--mainly to fulfill existing contracts, among other things--would be prohibitive, and Anderson believes the unit will be worth far more if Duke can hang on to it for a few more years. Wall Street analysts say it may take until the end of the decade to work off the excess generating capacity nationwide. But Anderson believes the markets where DENA is now concentrated, California and the Northeast, will recover sooner.

Anderson and his lieutenants have a lot riding on Duke's turnaround. Anderson gets paid only in stock, which he can't sell until 2007. And for Duke's top executives to receive full bonuses this year, earnings will have to hit $1.20 a share or more, but they get no extra pay if profits fall below the dividend level of $1.10 a share. (The current Wall Street earnings consensus is $1.20 a share.) The bonus plan may offer some reassurance for income investors lured by Duke's rich 5.1% yield, which Anderson has vowed to maintain. But aiming to return a whopping 92% of earnings to shareholders leaves little room for error. And even if Anderson can meet the dividend commitment, he has a long way to go before Duke is once again a suitable stock for widows and orphans.

FEEDBACK dstires@fortunemail.com