(FORTUNE Magazine) - It is the instinctive wish of most American businesspeople, even those unlikely to be directly affected, that General Motors not go bankrupt. True, some people will say, "They had it coming to them." But the majority will be more practical, telling themselves that the company is so central to the economy, so sprawling in its commercial reach, that bankruptcy--"going into chapter," as restructuring folks say--is ominous almost beyond contemplation. And yet the evidence points, with increasing certitude, to bankruptcy. Rick Wagoner, GM's 53-year-old chairman and CEO, may say, as he did in a January interview with FORTUNE in his aerie of an office high above the Detroit River, "I know that things will turn around." But he cannot know that. He may not, deep down, even believe it himself.

Bankruptcy isn't going to occur next week. But down the road--say, past 2006 --its probability is high. That point of view seems supported by the opinions of the bond-rating agencies, which troubled companies must keep informed and which become virtual insiders in their understanding of a company's finances and operations. In recent months both Moody's and Standard & Poor's have made increasingly grim statements, bald in their talk of bankruptcy and laden with doubts that GM (Research) can turn around its reeling North American auto operations, now reduced to an embarrassing market share of 26%.

GM chairman and CEO Rick Wagoner

More from Fortune Will Mmmhops be a hit? NBA confirms L.A. Clippers sale to ex-Microsoft CEO Steve Ballmer FBI and SEC probe into Carl Icahn and golfer Phil Mickelson FORTUNE 500 Current Issue Subscribe to Fortune

In that percentage lies a harrowing, and maybe intractable, revenue problem. Says one GM executive: "There's no fix for us unless we get revenues stabilized."

Nonetheless, Wagoner and crew must also deal with the full range of GM's problems, and they add up to a Hummer-sized load. The company lost $8.6 billion last year, burning up billions of dollars in North America, earning too little back overseas. Its product mix in the U.S., heavily weighted toward trucks, pickups, and SUVs, is on the wrong side of gas prices. It has a finance subsidiary, GMAC, whose majority interest it needs to sell to keep that business healthy and itself in cash--and so far, no buyer has emerged. It is inextricably entangled in the bankruptcy of its biggest supplier, Delphi. In that imbroglio, as in countless others, it is up against a formidable and sometimes militant union whose ability to accept the full reality of GM's problems is not assured. The company is even under investigation by the SEC for accounting sins, as yet unrevealed.

And gravely, it is burdened by health costs, which it supplies for a population bigger than Detroit's--that is, for a total of 1.1 million employees, retirees, and dependents. Its thriving Japanese competitors, such as Toyota (Research), pay health benefits for their U.S. active employees and dependents too. But Toyota does not have GM's retiree health burden, a mountain that at year-end totaled an unfunded $64 billion and that, in annual effect on the bottom line, adds about $1,300 to the cost of every car and truck GM makes in the U.S.

Wagoner is exultant that he and the UAW gruelingly managed last year to make a deal that, if blessed by a federal judge, will cut GM's unfunded liability by around $15 billion and pare cash outlays as well. But that will still leave Wagoner facing a colossal competitive disadvantage. The cost is not his fault. Rather, it is a legacy dumped on him by CEOs of decades ago who gained a certain amount of wage restraint from the union--and labor peace for their own terms of office--by granting retiree health benefits that had neither large, immediate cash costs nor, under the accounting rules then applying, much effect on the bottom line. Today, with health-care costs exploding and the accounting rules stiffened, this mess has come home to roost. It is the problem, says Wagoner (almost certainly giving too little weight to his shortage of revenues), that more than anything else "affects the future viability of GM."

In character, today's GM is a weird and painfully scarred combination of businesses. It is a car company doing poorly, and it is an insurance company engulfed by obligations way beyond its ability to pay. Such an enterprise probably cannot escape bankruptcy. The securities markets flash their warnings with regularity. The prices of GM's bonds have fallen severely, and its stock plunged in December to below $19, the lowest price since 1982. In early February the stock was $23. Were it not for GM's dividend, $2 annually, the price would surely be lower than it is.

Is there anything optimistic to say? Well, it is important to remember that giant auto companies have been turned around before. In 1980, aided by $1.5 billion in loan guarantees from the U.S. government and his own pitchman routines on television, Lee Iacocca brought Chrysler back from the abyss. Nearly 20 years later Carlos Ghosn, an improbable mixture of Lebanese blood, Brazilian birth, French education, and American business experience, grabbed tight hold of Japan's sinking Nissan Motor and restored it to industry prominence.

Yet these rescue jobs surely pale in comparison to what it would take to turn around General Motors, this giant so large that in the FORTUNE 500's first half-century it ranked No. 1 on the list in 37 years. (In our last list it was No. 3.) One Wall Streeter deeply familiar with the company recently stated the challenge starkly: "I would say that turning GM around is a harder logistical and managerial task than the invasion of Iraq."

This same Wall Streeter is not kind to the GM generals charged with the rescue job. Describing the company as a "sclerotic bureaucracy," he says a good remedy might be firing the top five people and replacing them with outsiders. A less acid form of criticism has been laid on by the camp of Kirk Kerkorian, whose Tracinda Corp. owns just under 10% of GM's stock. In January, Kerkorian's advisor Jerry York, a turnaround veteran himself (at Iacocca's Chrysler and Lou Gerstner's IBM), gave a long luncheon speech at the Detroit auto show that accused GM's executives of lacking "urgency" and "sense of purpose." York's reason for growling: Kerkorian's losses, about $172 million of them realized at this point, with the rest--another $223 million--sitting as unrealized losses on his books. York and Wagoner have talked about York's going on the GM board, but--as of early February, at least--they had not had a meeting of the minds. Maybe, one might guess, the Kerkorian camp has wanted both board representation and complete freedom to sell its stock.

That could have been a problem because GM's general counsel sent a memo last May to a sizable layer of GM executives telling them--for reasons he left quite vague--that they should refrain throughout 2005 from either buying or selling the company's stock. The prohibition, which still hasn't been lifted, is highly unusual because insiders normally have "windows" of time in which they can legally trade. In this instance, perhaps the insiders' deep understanding of GM's problems simply makes it unfair, and therefore also legally perilous, for them to be trafficking in its stock.

Wagoner, in any case, hotly disputes anybody's notion that GM lacks a sense of urgency. There's a "boulder hanging over our heads," he says, and it's causing the place to accelerate product introductions (like new models of its star Silverado and Sierra pickups, racing into showrooms soon) and otherwise operate with "breakneck speed." No doubt thinking back on GM's 98 years of existence--and to the pantheon of Alfred P. Sloan, Charles Kettering, and lesser gods--he strikes a poignant personal note as well: "Nobody's got a bigger stake in this than I do. I have this little sort of burden of history. I'm not going to be the guy that doesn't get this company going in the right direction."

If energy alone could do it, Wagoner might pull it off. Once a basketball player at Duke, he is tall and broad-shouldered, strong enough in fact to be at least metaphorically matched to the massive corporate weight he carries. A GM-er for the 29 years of his working life and CEO since 2000, he was in shirtsleeves as he talked to FORTUNE in January. Surrounding him were the usual suspects, miniature models of cars. He was animated to a degree that amazed one of his public relations people, who wasn't sure she'd ever seen him quite so expansive. And on his mind was the whole of the GM scene, including the need to keep growing in China, to build great products, to keep that boulder off his head.

Naturally, boulder evasion means that GM is deeply and broadly into cost cutting: It is closing plants to kill its excess capacity, terminating many thousands of people, negotiating with the UAW to free itself at least partially from the nearly un-American JOBS bank, in which laid-off union members get paid for not working. How many people are in the JOBS bank? Analysts ask that repeatedly and are refused an answer, probably because GM thinks nothing can be gained by hanging a number out there that Wall Street and the press can beam their attention on. But Sean McAlinden, chief economist of the Center for Automotive Research, thinks there were 5,200 employees in the JOBS bank at the end of 2005. He figures the annual cost of each to GM is at least $100,000.

Globally GM is once again striving to take advantage of its huge scale to reduce engineering and parts costs (it has paid lip service to this goal before). In the U.S. it can't easily cut brands----dealer franchise laws make that almost impossible--and that's a vise, because the eight brands it sells (Buick, Cadillac, Chevrolet, GMC, Hummer, Pontiac, Saab, and Saturn) are a big crowd for a 26% market share. At the least, the company is working to wedge Pontiac, Buick, and GMC into the same dealerships----under the same "rooftops"--and to sell within them, for instance, only one minivan model rather than two. GM's sales and marketing head for North America, Mark LaNeve, calls the cutting of each independent dealer a "little soap opera," in which entrepreneurs, in some cases with kids they expected to inherit their business, give up their turf by inches.

Perhaps most important for cost cutting at GM, Wagoner has just put a renowned chopper, Frederick "Fritz" Henderson, 47, into the job of chief financial officer. Henderson, a mustachioed and candid speed-talker, came from GM's European operations, where he reduced (though didn't eliminate) losses and got to be known as Chainsaw Fritz. He considers cost cutting an unending battle and approvingly cites the "continuous improvement" that is an integral part of the Toyota culture. In another example of non-arrogance, Henderson listened at a luncheon table, taking notes, as Jerry York gave his January speech. "He's a major shareholder," said Henderson later. "It's important that we listen to him." But he scoffed at York's claim that GM is blind to the depth of its troubles: "I'm in crisis mode and have been for years."

In all that GM is doing, there is a bleak awareness that no companies have ever turned around because of cost cutting alone. The essential partner is revenue growth--and as those losses in market share show, that has been the crucible for GM. In product design, it lost the magic long ago. "They need irresistibility and head-turners," says one car buff, "and they haven't had them." The man now on that case is product-development boss Bob Lutz, 74, who, after retiring from Chrysler, was hired by Wagoner in 2001. Tall, elegantly dressed, and outspoken, he is treated like a rock star at auto shows, often attracting more attention than his cars. At the Detroit show in January, touring GM's space with reporters, he was pleased to point out classy-looking car interiors--"some of GM's used to be grotesque," he said--and a level of fit and finishes that he judged superb. A reporter needled him: "Bob, I miss those bad fits, those gaps, that you had a while back. I used to store my quarters for tolls in those."

Lutz--and all at GM--are plainly battling the past, when many buyers gave up on its vehicles and turned to foreign cars. Today, GM has an enormous perception problem: a belief by too many U.S. consumers--particularly in the East, West, and some of the South, which pretty much leaves GM hugging the Midwest--that it doesn't make cars as reliable as those of foreign producers. That was indisputably true once. The current evidence, though, is mixed: Consumer Reports, a bible for many carbuyers, rates GM's improvements as "inconsistent" and ranks most of its cars as also-rans; J.D. Power, however, a leading arbiter of quality, gives many of its cars top grades. Meanwhile, GM people haul out comparison charts showing, for example, that a Chevrolet Malibu outdoes Toyota's Camry in just about every performance rating going, yet costs $2,640 less. Customers shrug their shoulders and keep on buying Camrys--their memories are long, and their motivation for returning to GM small.

The gist of GM's sales problem is summed up by Don Freda, a suburban New Yorker who has run an independent auto-repair shop for 52 years. What, he is asked, do you think about the quality of GM's cars these days? "They're very good," he answers. "They don't break like they used to." Then, immediately, "But nobody will buy them."

So it's no surprise that GM has been the impresario of incentives since 2001, when it immediately followed up 9/11 by launching the incentive program called Keep America Rolling. After that, the come-ons never ceased, so buyers quickly realized it would be idiotic to pay anything close to MSRP (manufacturer's suggested retail price). Last spring, when Wagoner personally took over the running of the North American business, he said that GM would reduce the use of incentives. But that pullback wasn't immediate: GM needed revenues in 2005, no matter their quality, and it kept on dishing the incentives out.

It was not until this January that GM officially announced a new pricing program, built on the thought of "selling the product, not the deal." The program cuts the MSRP on most of GM's cars, a change aimed at still giving the buyer an attractive price, but not by way of ballyhooed incentives. The price of a Tahoe SUV, for example, is dropping from a 2005 level of $36,790 to $32,990 (an amount that Wagoner says could still be reduced by ad hoc incentives). A big reason for the change is that about two-thirds of carbuyers these days do comparison shopping on the Internet, where GM feels it must show a "real" price, as opposed to the fictional prices--before incentives--that it was presenting before. As still another part of its new marketing program, GM is planning to cut back on its large-scale sales to rental-car companies, which not only buy at a sharp discount but also quickly flip their vehicles into the resale market and thereby hurt the residual values of GM cars. In all this, it is important to keep remembering that GM desperately needs to at least stabilize its market share and simultaneously to extract profits from what it sells. Whether this plan will do any better than the others now discarded is deeply uncertain.

Acknowledging the risk--"The jury's out," he says--Wagoner nonetheless expresses confidence because he believes there is "inherent goodness" in GM's products that the market will begin to recognize. But he also knows that every car manufacturer has a provincial view of its own prospects: "We're all guilty," he says. "We go through our design studio and go, 'Wow, we've got great products. They're so much better than what we had. This is going to turn things.' What you forget is that the same discussion is going on in every design studio around the world." That doesn't necessarily make you wrong in your expectations, he says. But in the end, it's "a bet." And you don't know--can't know--whether this time it's going to bring in the revenue.

That's a sweat-out matter for GM in 2006. As the year begins, the world is also focused on the company's "liquidity"--its store of ready assets that would allow it to withstand further operating blows, should they materialize. On the balance sheet of its auto business at year-end, GM had $20.5 billion in liquid assets, made up of cash, marketable securities, and short-term assets in a VEBA--a "voluntary employees beneficiary association" that holds money dedicated to the payment of health costs. GM calls this amount of liquidity "strong." But this is a company whose auto operations had a negative cash flow last year of nearly $6 billion. Suppose that repeats this year? Or suppose--Wagoner himself volunteers this--that the price of gas goes to $3 a gallon? Or that industry sales of autos drop by, say, 5%, or there's an outright recession? Besides, GM can't possibly spend itself down to its last dollar. Jerry York figures that GM needs at least $5 billion at any given moment just to operate, and others say the amount might be $10 billion.

Success by GM in selling just over half of its finance subsidiary, GMAC, would help GM's liquidity--though that really wasn't the impetus for putting it on the market. Here are GMAC's finer qualities: It is a well-run company; a good earner, with profits of $2.83 billion in 2005 (before a goodwill write-off of $440 million) on about $22 billion in book value; and a dutiful corporate child that paid $2.5 billion in 2005 dividends to its parent. So why would GM be unloading this treasure? Because GMAC's raw material is money, and--thanks to its scruffy parent--it is losing access to its raw material. The issue here is that GMAC's credit ratings are linked to GM's and therefore have been repeatedly lowered. That means GMAC is no longer welcome--as it devoutly wishes to be--in the commercial-paper market, which is in effect a deep-pocketed bank with good interest rates. So GMAC has been funding itself more expensively, by selling off its loans or borrowing against them.

A sale of, say, 51% of GMAC to a financially strong buyer would presumably raise its ratings and put it right back in the commercial-paper market. As to what the sale might deliver to GM, that's a mystery. A deal has dragged, partly because prospective buyers are leery of the financial consequences should they make a purchase and then see their co-owner, GM, go bankrupt. There's also a downbeat qualification about the money that GM might get: The company carries out intracompany transactions with GMAC that ordinarily leave GM a net debtor. It is very likely that any buyer of GMAC, not wanting to be owed by GM, would insist that those debts be paid off as part of any transaction. That would reduce GM's take.

Moody's has said unequivocally that "the sale proceeds are critical to GM maintaining adequate liquidity." Standard & Poor's wants this deal done too. But it questions just how much GM would benefit. Said S&P's Robert Schulz in January: "GM will be giving up half of an asset that's provided a lot of earnings. At the end of the day, it's hard for us to get excited about that."

Another boost to GM's liquidity could be gained by the company's embracing what Jerry York has labeled "equality of sacrifice"--that is, compensation cuts for most of its non-union constituencies. Such moves, for example, might cut the pay of GM's directors, who include such corporate folk as lead director George Fisher, retired CEO of Kodak, and Stanley O'Neal, CEO of Merrill Lynch. Base pay for a board member annually is $200,000, though each must put $140,000 of that into GM stock--an investment plan that hasn't worked out too well lately. Among the people escaping the entire GM flameout, it should be noted, is A.G. Lafley, CEO of Procter & Gamble, who exited as a director last spring. Given that a GM bankruptcy would no doubt put egg on the face of every board member, Lafley's departure possibly qualifies as Shrewdest Move by a Director in 2005.

York also wants GM's executives to cut their pay. Around headquarters in Detroit, there is muttering about this, since it would be a second blow: No bonuses were paid for 2005. In addition, the executives at the very top have definitely bled with the stock, because they are required by the board to hold multiples of their base salary in GM shares. The requirement for Wagoner works off his 2004 base salary, $2.2 million, and stipulates that he should own seven times that amount in stock, which is $15.4 million. Wagoner may have met that goal at one time; a precise answer about that has disappeared into proxy-statement fog. He for sure was still an optimist in March 2005, when he paid $1.5 million to buy 50,000 GM shares at about $30. Today, though, with the stock down closer to $20, he is way shy of the $15.4 million target.

What York sees as truly essential is a 50% cut in the dividend--pain for every shareholder, including himself and Kerkorian. Such a move may be symbolically important. It is not, though, an economic panacea. The dividend is only $1.1 billion annually--against a GM market value, in early February, of about $13 billion. Totally eliminating the dividend would, for example, not even cover one-fifth of GM's annual spending for health care (about $5.7 billion last year). There is another, quite unintuitive, point to be made about the dividend: With bankruptcy certainly a possibility, you could make a case that GM's directors might be doing their dead-level best for the shareholders by continuing to pay the dividend. That is, in a bankruptcy the shareholders are apt to reap nothing; for now, the dividend is something. A corollary to that thought is that any unsecured creditor of GM's who thinks bankruptcy will come should logically be protesting every penny paid to the shareholders, since any outflow of cash is money the bankruptcy estate won't be getting.

For GM, the problem of whether to cut the dividend is huge--in scale, way beyond the $1.1 billion. Were a reduction to be made, there would surely be national, and even global, headlines. That would agitate buyers who are already nervous about GM's viability--who worry, for example, about the company's ability to make good on its warranties. That's an unnecessary worry, but it still exists. It is probably not an overstatement to say that cutting the dividend would be a public relations disaster. On the other hand, not cutting the dividend gives the finger to the UAW, which has already agreed to a "giveback" of health-care benefits and from which GM needs many more concessions. "Why," the UAW is asking, "are we making sacrifices when the shareholders aren't?"

The GM board was scheduled to meet shortly after this issue went to press, and the betting here is that the dividend will be cut. That's because in these intense times, and in anticipation of a contract up for renewal in September 2007, GM desperately needs decent relations with the UAW. Keeping peace with the union right now almost has to outweigh a public relations problem.

The truth is that GM is essentially indentured to the UAW because of the union's power to strike. To that sign of bondage, add another: GM's hourly and salaried employees, present and past, essentially own this company, a fact we will prove by describing some bank accounts. At the end of 2004, the latest date for which figures are available, GM's pension funds (both inside the U.S. and out) had $100 billion in assets--which is wealth belonging to GM's employees, retirees, and dependents. To that you can add $19 billion that GM has put in a dedicated account for retiree health benefits. That makes $119 billion that GM has banked for its employees. In contrast, the shareholders of GM recently owned their grubby $13 billion in market value. That is a bizarre, Alice-in-Autoland result from 98 years in which capitalism supposedly reigned.

The union's leverage over GM affects everything that the company tries to do in cost cutting. The burning example is retiree health benefits, surely a competitive cost disadvantage if there ever was one. At various Berkshire Hathaway meetings, chairman Warren Buffett has envisioned what GM would do if it had contracted many years ago to buy steel at a premium price and had arrived at 2005 needing to get that cost back in line. "It would simply get out of the contract," Buffett has said. GM's retiree health benefits, arrayed against the benefits that the Japanese companies don't provide, are like paying extra for steel. But the odds against GM's breaking this contract are monumental.

That's proved by the halting course of events since early last summer, when Wagoner let it be known that GM proposed to unilaterally change its health plan for retirees so as to cut $20 billion off its liabilities. Legally GM may have been within its rights to do this; at least, that's what some court precedents say. The UAW, though, predictably protested, claiming that the benefits were vested and citing precedents of its own. To have sorted out this argument in court would have taken years that GM didn't have. So a compromise was reached: The UAW would hire Lazard to come into GM (at GM's expense) to determine just how bad its financial condition was. Lazard, whose work was led by Jim Millstein, head of its restructuring team, sent in about 20 people and took months to do its work.

GM wasn't given Lazard's report to read. But one executive thinks he has a good idea what it said. He is Steve Girsky, who until last summer was an all-star auto analyst at Morgan Stanley (where he ridiculed GM's incentive programs), and who then made a high-wire leap into working for GM as Wagoner's roving aide-de-camp. Says Girsky: "Usually the union's perception is that management is making things look worse than they are. This time, when Lazard came back to the union, I think it may have said, 'Management is not exaggerating; in fact, things may be worse than it thinks.'"

So the union did a "giveback," agreeing after some deadline-packed weeks to an intricate change in the retiree health benefits that will hurt the retirees moderately but also require the actives (who ratified the deal) to chip in by giving up some cost-of-living raises they were due. GM gets relief of only about $15 billion in its liabilities, not $20 billion. And its expected cash savings of about $1 billion a year will arrive only slowly. A further stickiness is that the agreement is not yet final. Its fairness must be okayed by a judge, who at a hearing scheduled for March can expect to be asked by at least some retirees why in tarnation they didn't have a say in all this.

Beyond any savings that the new plan ultimately delivers, it provides GM with a psychological lift: A union that detests deadlines and mid-contract negotiations stepped up when pressured and gave something back. Doesn't GM's success here suggest that it could perhaps erase its whole retiree health problem by driving ahead with unilateral actions and dragging the union along? No, says Girsky, absolutely not: "If you have to push to that extreme, you'll be building cars with three wheels. At the end of the day, these guys you're dealing with are the ones who build your products."

Wagoner is impatient with people who want GM to blast ahead on the union front, pushing hard to gain cost savings. "Excuse me," he says he asks them, "were you around in '93, '94, '95, '96, '97, and '98 when we took 16 strikes?" (It sounds as if he might know the exact dates of those strikes as well as he knows the birthdays of his three sons.) He quotes the critics further: "Well, you dolts, don't you know how to deal with the union so it doesn't go out on strike?" And Wagoner grits his teeth and answers, "Gee, I thought the assignment was to get competitive." His bottom line about union negotiations: "There's an art to this thing."

Right now, he and GM are working their art at auto-parts maker Delphi (No. 63 on the last FORTUNE 500 list), in whose arcane crisis GM is inextricably--and dangerously--entangled. The distilled story here is that GM split off Delphi in 1999, retaining an incestuous relationship with it and sticking it with GM-sized wages and benefits that exceed those paid by Delphi's competitors. Delphi ultimately floundered under this weight and hired have-gun-will-travel Robert S. "Steve" Miller as CEO. Out of bullets, Miller put the company into bankruptcy last October and quickly started talking about ways to exit with competitive costs. He then called on the UAW to accept lower wages--for example, he wanted to cut pay for skilled workers by 54%, from $27 an hour to $12.50. (These figures do not include benefits.) Later he took that proposal off the table. But the UAW had by then become so infuriated with both the man and the plan that it refused to negotiate with him.

GM, though, was always in this picture, because it and Delphi are virtually joined at the hip. The separation agreement back in 1999 included deals between GM and the union that made GM contingently liable for post-retirement benefits (mainly pensions and health) owed to certain employees if Delphi ever failed to provide them. With Delphi now in the soup, GM is on the hook--though for how many dollars is uncertain, because other Delphi/GM transactions figure in too. At the latest oracular word on Delphi, GM was estimating a total cost between $3.6 billion (an amount indeed taken as a special pretax charge in 2005) and $12 billion, with the most probable figure toward the low end of that range. Whatever the cost, most of the need for cash will hit GM slowly, because it will be liable for the retirement benefits only as they come due.

In the meantime, the bankruptcy court has told Delphi, GM, and the union to come up with a plan by August for getting Delphi out of bankruptcy. "This is three-hand poker," says Wagoner with a small grin, and the stakes for these struggling parties are way beyond Vegas. Delphi wants to emerge fit for competition; GM wants to hold its bailout costs to a minimum and also extract better prices from Delphi; and the union wants the highest pay possible. Talking about the cardholders at this table, Steve Miller allows that he, Wagoner, and the president of the UAW, Ron Gettelfinger, have the three toughest jobs in Detroit. His listener waits for the kicker, and it's surprising. "The job that's the toughest," Miller says, "is Gettelfinger's, because he has to get elected."

That's technically true--there's an election this June--but UAW incumbents seldom lose. Nevertheless, the whole Delphi affair has so angered the UAW that it could at some point haul off and order a strike. That is a true peril for GM. It cannot stand a strike. GM is hugely dependent on Delphi parts, buying well over $10 billion of them a year. Were there a strike of any duration, GM would not be able to move fast enough to line up alternative sources of supply. A strike would be a killer, enough to push it over the bankruptcy cliff.

That would not be good for the union, whose members would be threatened with cuts in wages, pensions, and above all, health benefits. But unions have called strikes before that didn't make sense, and it could happen again.

There are people who assert that GM should just get the suspense over and file for bankruptcy, thereby paring its liabilities to a manageable size and meanwhile continuing to sell its products. When he is asked about this scenario, Bruce Clark, a senior vice president of Moody's and its lead automobile analyst, turns very serious. "Bankruptcy would have significant costs, which any company would want to weigh," he warns. This is not the airlines, he says, where the dollar risk is a plane ticket, worth maybe a few hundred. With car companies, in sharp contrast, you have big-ticket items, $20,000 and more. "And you have buyers," says Clark, "to whom the warranty period is very important and who have a general expectation that service will be available." In other words, a buyer just might avoid any company in bankruptcy. And avoidance is hardly what GM needs; it's already had enough of that.

There is the big question, naturally, of whether some sort of extraordinary intervention might save GM from bankruptcy. Should it care to make the effort, the UAW would have the means: massive givebacks. Does that sound likely? No.

Then there's the U.S. government, whose bailout ability has been proved often. Clearly Washington could at some point decide that a GM bankruptcy was a nightmare it couldn't face and could step in with a massive infusion of money that would buy the company time--to shrink back, perhaps, to a viable size. Beyond that, what is Washington to do? Urge its citizens to pick up a Malibu with their tax refund? Pass national health insurance with the snap of a finger?

Certainly it won't be President Bush pushing the GM cause. He gave American auto companies the back of his hand in January, telling the Wall Street Journal that they needed to both deal with their costs and come up with "relevant" products. That no doubt mainly referred to hybrids, to which GM is a Johnny-come-lately. There's an implied message in what Bush said--"W to GM: Drop Dead." Which it probably will.

There's a note to be added about our old friend "urgency." It turns out that GM has a program, mentioned in last year's proxy statement and still extant, through which it will match up to $5,000 given by any of its 140,000 U.S. employees (and its directors as well) to a college or university. Matching-gift plans, especially big, splendid ones such as that, are admiringly welcomed by many people in the U.S. But when you are a company running on empty--and GM is--does not the continued existence of that program say that somebody never exactly caught on to what a sense of urgency is all about?

That raises a point made by a fellow--you will see why he doesn't want his name mentioned--in the GM dealer organization: "I can't really believe," he says, "that the people who got GM into this mess are going to be the people who can get GM out."

REPORTER ASSOCIATES Patricia Neering, Oliver Ryan

FEEDBACK cloomis@fortunemail.com