(FORTUNE Magazine) – On the face of it, the management team at General Motors has good reason to be declaring victory these days. To the amazement of countless skeptics, it has guided the world's largest automaker to impressive market share gains. Sales are booming, and the company is widely expected to hit its target of almost doubling earnings per share this year. And while the automotive sector has been pounded in the market of late, GM's stock (down 36% since January) hasn't fallen nearly as far as that of archrival Ford (down 55%) and it's about even with the S&P 500 index. Yes, management has done a wonderful job, it would seem, laying the foundation for a financial turnaround. Nevertheless, company shareholders should beware: They may soon find themselves stranded by the roadside with an empty gas can.

The reason is that GM is facing a vise-like financial squeeze in the coming years--with pressure coming from both ends of its balance sheet: assets and liabilities. And unless something radical happens, the carmaker is not likely to wriggle out.

Before we go any further, let's put GM's cost problem in perspective. The issue here isn't on the assembly line. Give CEO Rick Wagoner and his management team credit for trimming production costs where they could: The company is on track to cut its annual $71 billion materials bill, for instance, by some 3.5%--a sizable amount in an industry measured in gross tons. It has also managed to shave the number of hours it takes to build a vehicle by almost 20% over the past four years.

No, the real problem lies in what Wagoner can't control. Start with the massive pension and health-care bills related to its 460,000-strong base of U.S. retirees (for the record, that roughly equals the size of the U.S. Army). These are GM's "legacy costs," stemming from labor contracts the company signed with autoworkers in decades past. And they could very well soak up every dime of free cash flow the company generates over the next few years--yes, every dime.

The costs have also become a huge competitive disadvantage for GM: It now counts about 2.5 retirees for every active worker, compared with a one-to-one ratio for Ford and Chrysler. The Japanese and Korean carmakers, meanwhile, have virtually no pension obligations to their U.S. employees. GM's legacy costs soak up almost $900 for every vehicle it will sell this year, according to a recent Lehman Brothers report--a figure that could rise to $1,300 by 2005.

Part of that is the $1 billion GM is paying out this year for medical and life insurance for its retirees. That, too, will grow: Morgan Stanley estimates that the tab for future health-care expenses is a staggering $38 billion in today's dollars over the next 13 years, based on actuarial assumptions like health-care inflation rates.

Even more worrisome from a cash flow perspective is GM's pension plan. The company's obligations now stand at about $76 billion. At the end of 2001, the latest data available, the pension plan's assets were short by about $9 billion. GM concedes that the deficit has swelled significantly since then, thanks to the ongoing bear market. If the plan's assets slide another 10% by the end of 2002, says company treasurer Eric Feldstein, the unfunded obligation will top $20 billion--about the same as the company's current market cap. What's more, a further decline in interest rates will make that hole even bigger. In the simplest terms, that's because lower rates would increase the present value of the pension liabilities.

Don't be fooled into thinking this is a faraway problem--this is a now problem. Under current pension regulations, GM must pour more cash into the pension plan if the shortfall exceeds a certain level or face costly penalties. It has already plowed $2.2 billion into the fund in 2002, and most analysts believe the company will have to make additional contributions starting by 2004, though it may provide them as early as later this year.

UBS Warburg auto analyst Saul Rubin estimates that GM will have to infuse the plan with at least $13 billion more through 2007--about $1 billion shy of the cash flow he expects the company to generate (after capital expenditures and shareholder dividends) over that period. And that assumes that the pension plan experiences overall returns of 8% a year starting in 2003--at a time when ten-year Treasury notes are yielding about 3.5%, and equities, ahem, are returning less than that. And if we remain in a persistently low-interest-rate environment, Rubin figures the required payments could balloon to as much as $17 billion. Or as the analyst summed up in a September report: "While we would not want to suggest a risk of insolvency, it is fairly clear that once the pension fund is taken care of, there will not be much cash left over for shareholders."

But as we said, such unyielding liabilities are only half the problem. Indeed, the pension obligations might even be manageable were it not for the fact that GM is in Act 1, Scene 1, of a cash squeeze play. And this play doesn't end well.

For over a year, as the economy has wheezed and sputtered, GM's strategy has been to grab as much market share and to blow as many vehicles out of the showroom as it can. "This is a scale industry," says Chris Struve, an analyst at credit-rating agency Fitch. "These guys just have to sell a certain number of vehicles to cover their fixed costs or they're going to lose their shirts."

And in the battle to get that additional showroom share, GM has been using price cutting as its primary weapon. In August, GM gave away almost $2,600 per vehicle in customer incentives like 0% financing, nearly 25% more than the industry average. While it has partially offset those incentives by raising sticker prices, GM's net price per vehicle has actually fallen by about 1.5% from 2001 levels. It may not appear like much until you consider that GM sold about 4.5 million cars and trucks in the U.S. last year, at an average price of $21,000. That squeezes $1.4 billion of revenue out of the company's coffers.

Still, that has been a tradeoff GM has been willing to make. Its market share now stands at 28.2%, up three-tenths of a percentage point from a year ago. And the pricing sacrifices have been more than offset by volume gains, most of which have come from trucks and SUVs, like the Chevy Suburban and the GMC Envoy, which generate far more profit for the company than cars.

Lehman Brothers analysts estimate that GM will sell an additional 395,000 trucks and SUVs and an extra 75,000 cars in 2002. The trucks, however, are the company's golden goose, hauling in an average variable profit (before fixed costs) of about $7,000, compared with just $4,000 for the cars. All told, the volume gains could bring in an additional $3 billion.

But that drive for size could soon hit a brick wall. Many veteran Motor City watchers insist that the red-hot pace of car sales over the past four years won't continue. Industry sales are expected to reach 16.8 million units this year, not far from the record set in 2000. But many forecasters believe the rate is unsustainable and that demand could plummet to 16 million units--or even fewer--in 2003. Car buyers view the current 0%-financing plans as the deal of a lifetime. As a result they're buying cars sooner than they would normally, thereby stealing demand from future years.

There are plenty of other factors that could dampen buyers' willingness to shell out big bucks for a new car. The average age of vehicles sitting in consumers' driveways has declined significantly--that is, there are far fewer late-model clunkers waiting to be traded in. Used-car prices are tanking, making them an increasingly attractive alternative to new vehicles. And the looming specter of a conflict with Iraq, economists warn, could undermine consumer confidence and slow big-ticket spending.

GM, for its part, says the fears of a significant decline in auto sales are overblown. "Just about every analyst in the world has systematically underestimated U.S. vehicle volume since 1999," scoffs GM treasurer Feldstein. The company's own market research gurus confidently predict industrywide sales to be in the "mid-16 million" range next year.

Even if GM is right, industry analysts question the company's ability to defend its share of the lucrative truck and SUV segment from hungry foreign manufacturers. After all, it took overseas carmakers less than a decade to boost their share of the car market from 35% to more than 50% today. And while Asian manufacturers like Nissan and Honda currently hold just 24% of the light-truck and SUV market, it's doubtful they will be content to leave it at that. "High returns attract more investment. And guess what? Everybody and his brother is entering the SUV segment--some are even entering the hallowed pickup-truck segment," says Van Bussman, senior vice president of global forecasting for J.D. Power & Associates and a former chief economist at Chrysler. "I anticipate that the Big Three are going to continue to lose share for the next several years."

That would put even more pressure on GM--and its Detroit rivals--to raise the stakes in the incentives war. Over the long haul that's a no-win situation for GM. While it can't endure for long the lower margins that will come from aggressive price cutting, it can't halt them either. For evidence, look no further back than September: GM's unit sales dropped by 13% that month largely because it suspended its 0%-financing program for the first 18 days. To regain the lost ground, GM escalated the incentive war, offering 0% five-year financing on its newest-model cars and minivans. Not that GM's price advantage lasted long: Ford and Chrysler quickly responded with similar offers.

"The fact that the next model year is already extremely discounted is unprecedented," says Fitch's Struve. When GM began offering 0% financing in the aftermath of the terrorist attacks, the deals on next-year models were limited to three-year terms. Consumers, of course, are the beneficiaries of that largesse--a five-year, no-interest loan on a $25,000 car, for instance, lets buyers pocket an extra $275 a month, compared with a three-year deal--and it's unlikely now they'll ever settle for less. "It's becoming increasingly clear that the auto industry is getting the consumer to the point where unless they're offering a 0% five-year deal, they're not going to sell anything," says Gary Lapidus, an auto analyst at Goldman Sachs. And if that's the case, Lapidus warns, GM may be unable to sustain its market share and volume growth--even if demand for cars remains robust.

That brings us to another factor in the company's financial squeeze. The escalation in 0% financing deals, Lapidus says, is stretching the balance sheet of GM's financing arm, General Motors Acceptance Corp., close to the breaking point.

With about $200 billion in assets, GMAC is one of the largest financial services companies in the world. Its primary mission is to assist GM's automotive arm in selling cars by offering loans and leases to consumers; in 2001, GMAC financed almost two million new-car purchases. But GMAC's lending activities extend far beyond automobiles--it's also the nation's seventh-largest originator of residential mortgages. Even that has proved to be beneficial to GM's auto operations: Through subsidiaries like Ditech.com (yes, the company familiar to CNBC viewers for its ubiquitous commercials), GMAC has played a direct role in fueling the home-mortgage-refinancing boom and in turn propping up demand for GM cars. (As Ditech suggests on its website, one worthwhile use for all that extra cash might very well be a new car!)

But remember, GM loses money on every 0% car loan it offers. From an accounting perspective, it treats that cost as a marketing expense, just as it would any other type of sales incentive, like cash rebates. Here's how it works: GMAC offers the consumer a 0% loan. It then turns to GM to reimburse it for the cost of the money it had to borrow to make that loan, plus a built-in profit margin.

Thanks to the Federal Reserve's interest rate cuts, GMAC's cost of borrowing has steadily declined--it now averages about 4.5%--and that has bolstered GM's ability to offer the 0% deals. As a result, GMAC has been growing too: It provided financing to almost 46% of retail customers who purchased GM cars during the second quarter of 2002, up from about 41% in 1999. And that percentage could rise even higher now that GM is offering no-rate five-year loans on 2003 models.

But while that may move cars for parent GM, consider the effects on subsidiary GMAC's balance sheet. Like any other financial institution, the more loans GMAC makes, the more debt it incurs to obtain the money to make those loans. And as debt balloons, the company must also see an increase in equity in order to keep its overall debt-to-equity ratio in line with the level demanded by credit-rating agencies and bond investors. For GMAC, that ratio currently stands at about ten to one. If GMAC allows that ratio to spike much higher, its credit rating could be downgraded, which would increase both its cost of borrowing and therefore the cost to GM of offering 0% financing programs. To raise its equity level, however, GMAC can't issue shares; instead, it has to generate more profits on its own or get additional money pumped in from parent GM.

Ratings agencies also pay close attention to the credit quality of the customers who qualify for GMAC car loans. Should the economy dip back into recession, GMAC would almost certainly experience greater defaults on the loans it makes. In that case, the rating agencies could demand that it hold even more equity against the debt.

Until as recently as 2000, GMAC was generating so much profit that it not only met the equity requirements of the credit-rating agencies but also had plenty left over to kick back to GM. Throughout most of the 1990s, in fact, GMAC was a veritable cash cow for the automaker, returning more than $8 billion in "dividends" to its parent. But in 2000 and 2001 the financing division ceased generating those dividends--and worse, became a cash drain on GM. As GMAC aggressively expanded its auto and residential-mortgage-lending businesses, the parent company was forced to pump a net total of $1.6 billion into its subsidiary.

The danger of the new crop of no-interest loans is that GMAC's debt burden will continue to grow--and suck up more of GM's cash. The question is, Will GM be able to come up with it? Lapidus says probably not. It has no capital to spare. Indeed, tangible equity (total assets less intangible assets and liabilities) in GM's automotive business is already negative.

Of course, if GM eases off on the 0%-financing deals, GMAC won't need the additional money. But then--as it saw in September--sales are likely to get hammered. "It's this catch-22 problem," says Lapidus. "They're damned if they do and damned if they don't."

GM's Feldstein insists that such fears are exaggerated. "We believe GMAC can sustain an adequate level of growth without meaningful equity injections on the part of GM," he says. But he acknowledges that the hefty cash GMAC had for so long doled out to GM's automotive arm could be a thing of the past. "What it may mean is that we can't turn on the spigot in a large way, as we might like to," he says.

That's not exactly reassuring for GM shareholders, given that the company may need that cash badly for its retiree obligations. At the very least, says UBS Warburg's Rubin, the company's stock dividend "could be in jeopardy long term." Ford has already had to bite the bullet on that front: In October 2001, it slashed its dividend in half and followed that up with another 33% chop in early 2002. (See related story in First.)

GM has also been down this road before. It last trimmed its dividend in late 1992, bringing the annual payout down by more than 70% from its 1990 level. (The dividend has never recovered.) At the time, the company was also facing a pension shortfall that ultimately ballooned to $22 billion by 1993. Most of that came about as a result of a low-interest-rate environment that both hit GM's return on bonds and increased the present value of its pension liabilities. The carmaker eventually dug itself out of that hole. But the circumstances that allowed it to do so simply aren't in place today.

For starters, the automotive industry back then was emerging from a period of weak demand brought on by the recession. The Gulf war had ended--and with it, the fear of sky-high gasoline prices. And, of course, GM's pension fund became a beneficiary of the longest and strongest bull market in history.

Today, by contrast, demand is poised to slump. There's a good chance that we may be heading into a war with Iraq, and GM can hardly count on a repeat of the stock market bubble. What's more, the current excess capacity in the auto industry is bound to keep the pressure on pricing.

In fact, the problems plaguing GM aren't a whole lot different from those experienced by high-cost players in other faltering industries. "It's conceivable that you could see the automakers looking a lot like the steel companies," says Raymond James analyst Greg Salchow. Ironically, the only thing that may rescue GM from years of serial restructuring and subpar returns is a collapse in car demand so severe that the automakers union would be forced to make significant concessions on worker benefits and plant closures. "The doomsday scenario," says Scott Hill, an auto analyst with research firm Sanford Bernstein, may actually be the one that saves the industry from long-term malaise. Until then, GM investors may find that the company's cars are a better deal than its stock.

FEEDBACK jrevell@fortunemail.com

RESEARCH ASSOCIATES Dana Vazquez and Doris Burke