(FORTUNE Magazine) – Five years ago Alex Mandl turned his back on the biggest job in telecom. As the newly minted president of AT&T, he was heir apparent to CEO Robert Allen and thus a few years from running the biggest, most powerful telecom company in America. Instead the silver-haired, Austrian-born Mandl, now 57, left to run a startup telco in Virginia with no revenues and no business model; when he started, he didn't have even a secretary or a desk.

Money had a lot to do with Mandl's move--Teligent paid him a $15 million signing bonus and would ultimately grant him options for six million shares of stock. But in just as big a way, the startup appealed to Mandl's ambition. Congress had just passed the 1996 Telecommunications Act, aimed in part at opening the now $200- billion-a-year local phone markets to competition; Teligent was one of dozens of carriers popping up to stake claims.

Mandl bet his career that Teligent would emerge a leader in the race to offer voice and high-speed Internet access to businesses. And for a brief, shining moment, he was right. By 2000, Teligent was operating in 42 markets across the U.S. and Europe. Its stock was trading at $60 a share, and its market cap had climbed to $3.2 billion. Mandl's options were worth $360 million.

Yet today his dream lies in ruins. When the telecom sector collapsed last year, Teligent's funding evaporated. This May, with $1.4 billion in long-term debt and just $1.2 billion in assets, the company filed for Chapter 11. Its shares, which last traded around 56 cents, were delisted. Mandl's stake is worthless. Most humiliatingly, a young John Malone wannabe named Howard Jonas, a guy who once sold hot dogs from a cart in the streets of New York City, took control of the company for which Mandl had given up the crown at AT&T and pushed him out. "For the first four years, this thing was a terrific success," sighs a shell-shocked Mandl as he pushes aside a salad at a Manhattan restaurant in early June. "The last year was more difficult, to say the least."

In his misfortune, Mandl is not unlike scores of executives who jumped from well-heeled FORTUNE 500 companies to startups with thrilling ambitions and supernova burn rates. Teligent's crash is fairly typical too--after five years of deregulation, the Baby Bells still control 94% of local access, and dozens of startups have failed. But Mandl and Teligent were among telecom's brightest lights, and the saga of their rise and fall is far more interesting than most.

Many people thought Teligent had an especially good shot at taking on the Baby Bells, partly because its unique technology bypassed their networks. Teligent's system worked by putting antennas the size of large dinner plates on the rooftops of office buildings. Broadband access was bounced from these antennas to base stations on nearby buildings.

Its technology helped Teligent attract smart-money investors, including AT&T's Liberty Media Group, the Dallas buyout firm Hicks Muse Tate & Furst, and Japanese telecom giant NTT DoCoMo. The company recruited high-powered executives from AT&T, WorldCom, and MFS. "Teligent had a lot of things that, in Wall Street's mind, differentiated it," recalls analyst Riyad Said at Friedman Billings Ramsey.

Yet Teligent was hampered almost from the start by internal strife. Mandl's ambition put him in conflict with Teligent's board, which was controlled by the company's founding family. More so than many high-profile CEOs--more so than he admits even today--Alex Mandl was just a hired gun. He'd been lured to Teligent by the Associated Group, a firm owned by the Berkman family, a Pittsburgh clan that had made it rich in steel. In the mid-1980s, the Berkmans had shrewdly acquired radio spectrum in the nation's 31 largest markets. Companies would pay billions of dollars for such licenses in federal auctions a decade later. But because deregulation had not yet happened and the spectrum in question was an ultra-high-frequency band of little apparent commercial value (the military used it to control "smart bombs"), the Berkmans got it virtually for free. When deregulation came and the Internet emerged, they formed Teligent to cash in on this windfall. Their investment in Teligent totaled about $50 million.

Mandl faced unexpected obstacles from the start. The licenses came under fire: Cellular mogul Craig McCaw and Microsoft's Bill Gates challenged them before the FCC, complaining that Teligent's signals would interfere with the Teledesic global satellite network they were building. Teligent spent six months arguing its case; the FCC finally upheld the licenses but shifted them to a different frequency band.

Technology problems hit next. Teligent's software was plagued by glitches, and its hardware was much more costly and difficult to install than anyone had expected. Result: another six-month delay in the initial rollout.

At least one of Mandl's duties should have been easy in the growing Internet boom--raising money. Abundant capital was available, and Mandl was happy with a deal Teligent made in 1997, in which NTT DoCoMo put up $100 million for a 10% stake. But it wasn't enough. Teligent, like many emerging carriers, consumed cash like fire burning through dry brush. Mandl wanted to stick with private capital until Teligent was up and running in several major markets. But board members were unwilling to either put more money into the startup or dilute control by taking on more private investors. Instead, Teligent was forced to go public sooner than Mandl had planned--its system was barely operating, in a few cities in Texas.

The IPO, in November 1997, was a hit despite the CEO's misgivings. The stock surged 20% on its first day, giving Teligent a market cap of $1.3 billion. Yet Mandl knew the game had changed. From then on Teligent would depend primarily on the public markets, something it could do successfully only by hitting Wall Street's targets. In 1998 analysts wanted to see two things: rapid expansion into many markets and access to thousands of office buildings. These were strange new metrics for Mandl; Teligent's plan had been to sign up customers and expand its networks in tandem so that growing sales would help defray build-out costs. But Mandl says analysts reminded him constantly that Teligent was behind fiber-optic access providers like XO Communications in the number of markets served.

So Mandl played to win on Wall Street's terms. He pushed hard to expand Teligent's service; starting in mid-1998, the company built systems in 42 markets in 18 months--"A record!" Mandl says proudly. Pleasing Wall Street came at a heavy price, however. Since negotiating leases for roof rights and installing equipment can take months, Mandl and his team decided they couldn't wait until critical masses of customers had signed up. "Clearly the more conservative cash approach is to get the customer first," says Mandl. "But the customer may not have wanted to wait, and with Wall Street pressuring you to get roof rights, you err on the side of putting in networks."

Increasingly Teligent's biggest battlefront was not the marketplace but the boardroom. Mandl learned the hard way that while he was the face of the company, the Berkman family was its backbone. "I felt sorry for Alex," says an ex-insider. "He was supposed to be the strategic-activity guy, raising money and doing mergers and acquisitions. But the board never allowed him to do any of that."

By 1999, a year of telecom consolidation, Teligent had grown substantial enough for management to begin hearing overtures from long-distance giants looking for local connections, including WorldCom and Sprint. Instead the Berkmans crafted a different deal. In June the family sold Associated Group--including its stake in Teligent--for $3 billion to AT&T's Liberty Media Group, led by John Malone, former head of cable giant TCI and a major AT&T shareholder. The sale was a coup for the Berkmans--Teligent accounted for about a third of Associated Group's value, so their $50 million investment had grown to $1 billion. For Mandl, the news was less great: He found himself working again, if indirectly, for AT&T.

Teligent's acquisition by Liberty wasn't set to close until the end of the year. But the looming change did nothing to ease Mandl's struggles with the board. By the end of 1999, Teligent's capital expenditures had surged 43%, to $262 million, as it raced to plant antennas across America. Sales, meanwhile, were barely detectable: The company racked up a net loss of $529 million on revenues of $31 million. To keep growing, Teligent needed a lot more cash.

Investors were more eager than ever to pump in money--Teligent was now one of Wall Street's darlings. But the board, says Mandl, wanted to put the company on a capital diet. In December, Teligent sold $500 million of preferred stock to Microsoft and Hicks Muse--but turned away an additional $500 million, Mandl says. In April 2000, with Teligent's market cap near its all-time peak of $6 billion, the board rejected a $200 million convertible preferred stock deal--too expensive, Mandl says he was told. "I objected strongly to turning down those investments," he says. "With that $700 million, we would have had a fully funded plan. We'd have a company today." Retorts a former director: "Teligent had plenty of money. We wanted Mandl to slow down the build-out."

Wall Street, meanwhile, twittered with speculation about what John Malone saw in this rooftop antenna business. In early 2000 his intent became clear. Liberty and Hicks Muse each purchased stakes in several of the same telecom startups. Next they proposed to merge two of the startups with Teligent and create a nationwide carrier providing multiple kinds of broadband access: wireless, DSL, and fiber-optic.

If the dot-com craze had continued--it was peaking just then--the deal could have made billions for Malone and his allies. Instead it fell prey to the dot-com crash. Talks dragged on until one of the startups went bankrupt in the fall and the deal collapsed. Meanwhile Teligent's stock cratered, falling from a March peak of $97 a share to $2 by year-end. Mandl found himself eking out revenues by reselling other companies' phone services--and Malone began looking to cut Liberty's losses.

Amazingly Malone found a willing taker in Howard Jonas, the founder and chairman of a $1-billion-a-year long-distance and calling-card telco in Newark, N.J., called IDT. A telecom scavenger, IDT scarfs up pieces of telecoms at bargain prices and sells discount services to low-income customers. IDT also developed Net2Phone, a service that lets users make calls cheaply over the Internet. The polar opposite of the cosmopolitan Mandl, Jonas is a rumpled, Bronx-born, self-described libertarian and a father of eight who looks much younger than his 45 years. He has already written an autobiography, called On a Roll: From Hot Dog Buns to High-Tech Billions.

Jonas had been an avid John Malone fan for years. On the wall of his office hangs the handwritten draft of a worshipful letter he sent Malone in 1997, after an acquaintance introduced them. Malone took a shine to the young entrepreneur but soon became wrapped up in the sale of his cable-TV empire to AT&T and fell out of touch.

Jonas used actress Goldie Hawn to reestablish contact. In early 2000 he had Hawn to dinner at his house. IDT was now developing Internet content, and Hawn wanted to talk about starting a women's sports network. Jonas had his own agenda--he desperately wanted to sell part of Net2Phone to AT&T but couldn't get anyone's attention there. He called Malone on the pretext of discussing Hawn's project. "Malone thought a women's sports network was a terrible idea," says Jonas. "Then I asked what he thought about AT&T buying part of Net2Phone." Malone liked that better and quickly arranged a meeting. That March, an AT&T-led group bought 39% of Net2Phone for $75 a share in cash. The deal put $1.1 billion in the pocket of Malone's admirer.

When Malone called Jonas this year about helping Liberty get out of Teligent, Jonas was more than happy to oblige. In April, Malone traded his Teligent stake--worth $1 billion a year earlier--for $37 million in IDT stock. (Jonas also swapped IDT stock for Hicks Muse's Teligent holdings.) Jonas became Teligent's chairman and installed IDT executives in the ranks. For Mandl it was the end of the line. Jonas says that Mandl stepped down as CEO "by mutual agreement" and that he is being paid as a consultant. Mandl says he has no affiliation with Teligent.

Why would Jonas buy a stake in a nearly bankrupt company? He says it was a favor to Malone and Hicks Muse. "There's no question that this asset was a millstone for them, and we felt that we would pave the way for a better relationship," says Jonas. "In life, if you do someone a favor, sometimes they do you one."

Jonas also may have thought he smelled a bargain--which would explain why, one Friday morning in late May, he is visibly upset about Teligent. He walks in and out of a conference room mumbling and is short on pleasantries. IDT President Jim Courter tries laughing it off when asked what's wrong, saying Jonas is merely "cranky" today. When Jonas finally sits, he blames his mood on Teligent's creditors. He's offered to buy Teligent's assets for $250 million--"probably double the liquidation value!" he says. No way, said the banks, which are holding $750 million in Teligent notes and want him to raise the offer by 50%. Though Jonas makes noises about how he could turn Teligent around--he says he sees himself rallying the sales force and personally going office to office to sell Teligent's services--his heart isn't in it. Later that day Courter calls to say that Jonas has resigned as Teligent's chairman. "We wanted to signal the banks that this is a sincere offer and [that if they reject it] we're not going to stay engaged or invest more," Courter says.

Mandl, meanwhile, still believes in Teligent. He argues fervently that it could have--should have--survived the telecom bloodbath. It just needed a little more time, a little more money. "In summer 2000 the rules changed," says Mandl. "Even though we were meeting our targets, there was no more capital available." Unlike the founders, he didn't walk away rich. He sank nearly all of his ballyhooed signing bonus back into Teligent, and his six million options are worthless.

These days Mandl is fielding job offers. There were rumors he was in talks with struggling Pacific Century Cyberworks, but a move to Hong Kong seems unlikely. Over dinner he asks if the New York housing market is cooling at all. Still charming, Mandl is more subdued than in previous meetings. Asked if the Teligent experience has changed him, he rubs his hands and looks away. "Maybe I've matured," he says with a rueful laugh before turning somber. "I've clearly never experienced anything like this, and frankly, I don't want to experience anything like that again. We built a terrific base, and to watch that slipping away is a very painful and very stressful experience." It is also something Mandl can't quite walk away from yet. Industry insiders say he is trying to engineer a buyout of the wreckage. And that would be one last reason Teligent's tale is more interesting than those of other telecom bombs: It may have an epilogue.

FEEDBACK: jcreswell@fortunemail.com