The auto bailout: How we did it

(Fortune Magazine) -- Without any experience in automaking or government, Steven Rattner left his Wall Street perch to wade into the largest restructuring in American history. The scale and speed of the rescue raised many questions, inspiring Rattner to write this account of a defining moment in capitalism.

Rattner, who led the group that did the hands-on work, believes passionately that the decision to intervene represented not "creeping socialism," as some feared, but a critical part of the effort to prevent economic collapse. Here is the full Fortune exclusive, where Rattner gives a behind-the-scenes look at the President's decision to impose tough medicine, the ouster of GM's Rick Wagoner, and the revolt of Chrysler lenders that forced bankruptcy. Rattner's account:

Reporting for my first day of work on the auto task force, I lingered outside a guard shack on Pennsylvania Avenue. My clearance into the fortress-like Treasury building was adrift in a bureaucratic haze, a fitting example of the chaotic start of my government service.

The reaction to my likely appointment had been as chilly as that February morning in the weeks since it had leaked. Few agreed that I was the right guy for the job. Some Michigan legislators -- of my own party -- questioned my knowledge about manufacturing in general and autos in particular. They were understandably suspicious about a Wall Street guy who hadn't even been to Detroit in three decades. Before long, the press piled on (Fortune.com catalogued the cars I own and pointed out that only one was American-made).

Part of the disappointment was caused by expectations that had been raised of an all-powerful "car czar" to come to the rescue, based on legislation that had been crafted in the waning days of the Bush administration. But that was never the Obama administration's plan; its car team was slated from the start to report to incoming Treasury Secretary Tim Geithner and White House chief economic adviser Larry Summers.

The administration was surprised by the reaction from the heartland and paused to regroup. To help defuse the emotion, it created two auto task forces: a cabinet-level group that would receive high-level briefings on policy options and an assemblage of talented sub-cabinet economic thinkers. But the day-to-day work would be done by yet a third "task force," consisting mostly of Wall Street refugees like me, beavering away in the basement of the Treasury building.

Thoroughly unnerved by the task ahead, I had second thoughts. The tight deadline and problems of the industry appeared impossibly daunting: collapse of market share, overwhelming structural costs, profuse bleeding of cash.

I needed the gentle e-mail push that came from Larry Summers. "Life will work out for you either way, though America will be better off if you do this," wrote Larry, whom I had liked and respected for 15 years. He and Tim believed that the fact that I barely knew one car from another was less important than my 26 years advising, financing, and investing in companies. In their minds, this was a restructuring challenge, not a management job.

And so I closed my eyes and jumped. Among the surprises along the way: We were shocked, even beyond our low expectations, by the poor state of both GM and Chrysler. Looking just at the condition of GM's finances and Chrysler's new-car pipeline, the case for a bailout was weak.

But on the other hand, as we surveyed the interconnected web of finance companies, suppliers, and related businesses, the potential impact of the likely alternative -- liquidation -- stunned us. We imagined that the collapse of the automakers could devastate the Midwest beyond imagination. We were determined not to fail. But as we started down the road, we saw mainly obstacles.

When the Obama administration took office on Jan. 20, it inherited nothing in the auto area: no staff, no stacks of analyses, no plans of any kind. The Bush administration had decided in late December that GM and Chrysler were not going to go bankrupt on its watch and had shoveled $17.4 billion of TARP money into the companies to keep them afloat, but without any meaningful stab at restructuring them.

Because of my vacillating and the government vetting process, my own appointment was not announced until Feb. 23, a week after the unveiling of the task forces. But since early January I had been helping out, including on the key matter of hiring.

To add critical expertise and help buttress us against attacks from the left, I recruited as my partner Ron Bloom, a tireless former investment banker with whom I had overlapped briefly at Lazard Frères but who had dedicated the previous 12 years to working on behalf of steelworkers. Ron's hefty Main Street experience could help balance my Wall Street baggage.

From the first leaks about my potential appointment, I had received many unsolicited résumés. One in particular caught my eye. In a long cover note, blue-chip banker Harry Wilson explained that he had grown up in Johnstown, N.Y., where his mother had been laid off from three textile mills as they sequentially closed, that he was the first in his family to have gone to college (Harvard and Harvard Business School), and that he was a registered Republican to boot. The addition of Harry cemented the principle that our hiring would be done without regard to political views.

Turning his boundless energy to recruiting, Harry blasted out an e-mail to his network, and we were soon deluged with hundreds of résumés. Compressing Wall Street's lengthy hiring process into a couple of weeks, we briskly assembled a mini investment bank (and a mini law firm led by Matt Feldman, a superb restructuring expert).

We inherited one key team member from the transition: Brian Deese, who would later be profiled by the New York Times as "the 31-year-old in charge of dismantling GM." Radiating wisdom well beyond his years, Brian not only had brains and judgment, but as a member of Larry Summers' National Economic Council team, he also had an understanding of the policy process that bridged my glaring gaps.

Our timing had been set by an arbitrary March 31 deadline that the Bush administration had imposed on the companies to meet a hodgepodge of conditions. And by coincidence, both companies would probably run out of money around the same time. That gave us only about five weeks to learn the auto industry, study the two companies' restructuring plans, develop a plan of action, and sell it to our superiors, including the President.

For Team Auto, as we called ourselves to avoid czardom, it would be the most intense period of work and personal disruption of our careers. For my part, I lived in a sterile sublet condo in Washington and would have to grapple simultaneously with the New York attorney general's investigation of my former firm, Quadrangle Group, and me about our actions in connection with an investment from the state pension fund. But we soldiered on.

Moving simultaneously down multiple paths, we began meeting with all the interested parties: labor, lenders, legislators, and suppliers. We naively assumed that stakeholders eager to see a rescue of the two companies would come with a set of "gives"; I was startled that each stakeholder meeting invariably included a set of "asks" from the government.

Case in point: parts supplier Delphi, which hoped for government assistance even though nearly 90% of its workforce was outside the U.S. An important part of our job was going to be to convince the stakeholders that the government wasn't going to be everyone's piggy bank.

Our physical surroundings set the tone: While Treasury staff labored to find us office space, we all worked out of a yellow-walled conference room, our desks facing the room's perimeter. Our first meeting with General Motors took place in a dingy, windowless room at the Treasury Annex.

When I asked our energetic young chief of staff, Haley Stevens, what we were going to give our visitors for lunch, she replied, "Nothing. Treasury has no budget for even bottles of water." It seemed harsh to expect our guests to go many hours without eating, so I gave Haley $100 and told her to go to a sandwich shop. That became our hospitality protocol.

Even before we met the two management teams, it was clear to us from the "viability plans" that the companies had submitted on Feb. 17 that GM and Chrysler were in a state of denial.

Both companies needed gigantic reductions in their costs and liabilities. They had way too many plants and workers for expected car volumes. And their labor costs were out of line with those of their most direct competitors, the Japanese "transplants" manufacturing in the South. The administration was united: No more money except in the context of a shared sacrifice and a truly viable restructuring.

I found it frustrating that so many pundits were suggesting that the government stay on the sidelines and let the two companies fend for themselves. With financial markets still largely frozen and no private capital available, there was no question that both would have slid into bankruptcy, run out of cash, closed their doors, and liquidated.

As early as January, I had doubted whether fundamental restructurings could be accomplished outside of bankruptcy. While certain changes, like renegotiating labor agreements, could be done without it because only a single point of negotiation was required, other important steps, like trimming dealer networks and reducing debt, involved innumerable individual actors and would be very difficult to implement absent the cleansing nature of bankruptcy.

But we were terrified about the prospect of it. Most important, we feared that consumers would be unwilling to buy such a long-lived product as a car with the prospect of holding a warranty from a bankrupt company. And we were fearful of how long a traditional Chapter 11 proceeding would take. Delphi, which had been spun off by GM in 1999, had been in bankruptcy for more than three years.

As we pondered this dilemma, I found myself jumped one day by a reporter on the matter of bankruptcy. I responded with the best reply I could summon up: "Bankruptcy is not our focus. Bankruptcy is not our goal." That narrowly crafted dodge became our mantra.

In adopting that sound bite, we were trying to shorten as much as possible the period in which buyers would avoid the companies' products over bankruptcy fears. As we began to plan for possible bankruptcies, we were guided by Tim Geithner's notion of putting "foam on the runway," by which he meant to develop as many ways as possible to reassure consumers about the prospect of buying cars from a bankrupt company.

From that exhortation came the idea of the government guaranteeing warranties for GM and Chrysler cars. But our best idea was to employ a little-used part of the bankruptcy code, Section 363, which greatly streamlines the process by allowing a newly formed company to buy the operating assets from the bankrupt entity.

We recognized the importance of a trip to Detroit, so in March, several of us -- including Ron, Brian, and Diana Farrell (a star deputy of Larry's) -- made the journey.

Our planners made sure our schedule was carefully balanced among the old (a tour of Chrysler's Warren truck plant) and the new (test-driving the extended-range electric Chevy Volt and other cars of the future at the GM tech center).

What we didn't prepare for was the intense public interest in our visit to these hard-hit communities. Throngs of reporters awaited us at every stop while a news helicopter buzzed overhead. More peculiarly, the ensuing press coverage seemed wildly over-focused on our test drive of the Chevy Volt, as if the company's salvation rested on this one vehicle.

We applauded GM's increased emphasis on new technologies, which seemed to be driven by both commercial considerations and public relations, but we recognized that they couldn't possibly have any meaningful impact on GM's finances for at least five years.

The Volt, for example, could initially cost GM about $40,000 to produce while competing with conventional cars that sell in the $20,000 range. Even with the new $7,500 tax credit for plug-in hybrids, it will take years of improved efficiencies to make it profitable for GM.

Everyone knew Detroit's reputation for insular, slow-moving cultures. Even by that low standard, I was shocked by the stunningly poor management that we found, particularly at GM, where we encountered, among other things, perhaps the weakest finance operation any of us had ever seen in a major company.

For example, under the previous administration's loan agreements, Treasury was to approve every GM transaction of more than $100 million that was outside of the normal course. From my first day at Treasury, PowerPoint decks would arrive from GM (we quickly concluded that no decision seemed to be made at GM without one) requesting approvals. We were appalled by the absence of sound analysis provided to justify these expenditures.

The cultural deficiencies were equally stunning. At GM's Renaissance Center headquarters, the top brass were sequestered on the uppermost floor, behind locked and guarded glass doors. Executives housed on that floor had elevator cards that allowed them to descend to their private garage without stopping at any of the intervening floors (no mixing with the drones).

In my relatively few interactions with chairman and CEO Rick Wagoner, I found him to be likable, dedicated, and generally knowledgeable. But Rick set a tone of "friendly arrogance" that seemed to permeate the organization.

Certainly Rick and his team seemed to believe that virtually all of their problems could be laid at the feet of some combination of the financial crisis, oil prices, the yen-dollar exchange rate, and the UAW.

It seemed completely obvious to us that any management team that had burned through $21 billion of cash in a year and another $13 billion in the first quarter of 2009 could not be allowed to continue. Equally important, GM's February viability plan was more "business as usual" and not the aggressive new approach that we felt was essential.

In a mid-March meeting with Rick, I explored his feelings about his team and then tried to work the conversation around to himself. "I'm not planning to stay until I'm 65 but I think I've got at least a few years left in me," said Wagoner, 56. "But I told the last administration that if my leaving would be helpful to saving General Motors, I'm prepared to do it." He added that neither he nor his board thought that was a good idea.

I left my conversation with Rick at that. The next day, I met with Rick's deputy, Fritz Henderson. Another GM lifer (and son of a GM lifer), Fritz conveyed more energy and openness to change. The question for us was whether GM would be better off with Fritz or with an outsider, as Ford (F, Fortune 500) had done in bringing in Boeing executive Alan Mulally. While nervous about whether Fritz could bring the change GM desperately needed, I was considerably more nervous about the likelihood of recruiting a thoroughbred CEO in the midst of the turmoil.

Meanwhile, if ever a board of directors needed shuffling, it was GM's, which had been utterly docile in the face of mounting evidence of looming disaster. We decided to recommend to Tim, Larry, and ultimately the President a package that would include replacing Rick with Fritz as interim CEO, changing at least half of the board, and making an outside director chairman (which should be universal).

While we had separate teams working on Chrysler and GM from the start, the first and toughest decision that we faced was what to do with Chrysler.

Badly run after Daimler bought it in 1998, Chrysler had been sold nine years later at the peak of private equity mania to Cerberus Capital Management. Larded up with debt, hollowed out by years of mismanagement, Chrysler under Cerberus never had a chance. We marveled, for example, that Chrysler did not have a single car that was recommended by Consumer Reports.

The question for us -- and ultimately, the President -- was whether any restructuring could save Chrysler.

Back and forth the debate went in Larry's small office in the West Wing. Our working team was joined by Diana and other administration economists. Gene Sperling, a Michigan native, argued eloquently and passionately for standing by America's heartland.

Austan Goolsbee, a fearless former University of Chicago economist who had been with Obama since the beginning of his campaign, led the charge against Chrysler, marshaling strong factual arguments. One was that letting Chrysler go would give a needed boost to GM (and also to Ford), since most buyers of Chrysler's strongest products -- trucks, minivans, and Jeeps -- probably would turn instead to the other Detroit automakers.

Harry maintained that a Chrysler liquidation could potentially add billions of dollars a year to GM's operating income in a normal sales environment, vastly increasing the value of the company.

The group was torn (at one point the vote was four to four) and so were Tim, Larry, and I. We intuited that from a theoretical point of view, the correct decision could well be to let Chrysler go. But this was not an academic exercise.

Repeatedly, Larry asked us for probabilities -- what did we think the chances would be of Chrysler making it for two years? For five years? Indefinitely? Pressed by Larry, I came down 51-49 in favor of helping. Comfortable that Chrysler could survive, Larry asked us to recast Austan's analysis into an assessment of the employment effects of a Chrysler liquidation.

We were all shocked to realize that when the collateral damage of a Chrysler shutdown was factored in (like lost jobs at dealers and suppliers), the short-term effect of a Chrysler shutdown could be 300,000 more unemployed, similar to what was lost across the entire economy in the month of July. And with the memory of Lehman's collapse still fresh, we imagined the potential for other systemic risk. That cemented matters for Larry and me.

Yet Chrysler's cupboard was bare. We did not believe we could underwrite its viability without a strong corporate partner, so we turned our attention to that single possibility, an overture from Fiat.

The Italian carmaker had been brought back from near disaster a few years earlier by its own new management team, led by Sergio Marchionne. Raised in Canada, rarely seen in anything but a black sweater, and with an insatiable appetite for the media, Sergio was relatively new to the auto industry but possessed a drive to win that was alien to the traditional Detroit culture. Fiat also brought its advanced products to the table -- small, stylish cars and fuel-sipping engines.

As our March 31 deadline approached, we got ready to brief the President. For the prior six weeks, Tim and Larry had been using their daily briefings of the President to keep him posted on our progress. And we had sent in a couple of lengthy memos for his evening reading stack, which indicated that the only decision over which his advisers were divided was whether to save Chrysler.

In our meeting in the Oval Office on March 26, Larry began to lay out the issues, only to be interrupted after a few minutes by the President, who said "Larry, I've read the memo," signaling that he wanted to dive into the decision items. Quickly, the question of whether to save Chrysler dominated the discussion.

As we went back and forth over Chrysler, the President himself seemed as torn as Larry and I had been. Suddenly, he realized that Austan Goolsbee, unofficial spokesman for the opposition, was not in the room. "Where's Goolsbee?" the President asked. A few minutes later, Austan joined the conversation. But the President had only about 20 minutes available before his assistant, Katie Johnson, came in with a note urging him to move on to his next appointment. "This is too important to try to decide in a rush," the President told us. "We need to get together again later."

For that early-evening session, we convened in the windowless Roosevelt Room, the only real conference room in the West Wing. Larry asked me to begin by summarizing the consequences of refusing aid to Chrysler, and from there the conversation took off. The group was sobered by my assessment that given the early stage of the Fiat discussions, there was only a fifty-fifty chance of reaching an acceptable agreement.

The President's political advisers were as torn as his task force: Polls universally showed the public strongly opposed to the auto bailouts. At the same time, the advisers recognized the severe economic and political consequences of a Chrysler shutdown across broad swaths of the industrial Midwest. We were dazzled that chief of staff Rahm Emanuel -- a former congressional leader -- could identify from memory the representatives in whose districts the large Chrysler facilities lay.

After about an hour, the President asked for any final comments and then said, "I've decided. I'm prepared to support Chrysler if we can get the Fiat alliance done on terms that make sense to us." And we were thrilled when the President said, "I want you to be tough, and I want you to be commercial."

With that, the meeting dispersed. The recommendation to ask GM's Rick Wagoner to step aside had received barely a mention. New to business meetings with Presidents, I found Obama's style consistent with his "No drama Obama" image and on a par with the best CEOs I had spent time with. He was cordial without being effusive and decisive when his advisers were divided.

Rick Wagoner and his team were scheduled to be in Washington the following day, March 27, to discuss the progress of GM's restructuring. I was also aware that the GM board was meeting by phone every Friday during this crisis and asked Rick's assistant for a few private minutes with him, early enough to be sure it preceded that board call.

I don't know whether Rick had any inkling of why I had wanted to see him alone. His face was impassive as I said, "In our last meeting, you very graciously offered to step aside if it would be helpful, and unfortunately, our conclusion is that it would be best if you did that."

I told him of our intention to make Fritz acting CEO and he supported that idea, cautioning me against bringing in an outsider to run the company. "Alan Mulally called me with questions every day for two weeks after he got to Ford," he said.

As we continued our rather awkward conversation, Rick suddenly asked, "Are you going to fire Ron Gettelfinger too?" Startled by the reference to the UAW head, I replied, "I'm not in charge of firing Ron Gettelfinger," and Rick soon left to brief his board on our decision.

Later, I spoke to Fritz, who expressed enthusiasm for his proposed promotion but asked that he not be called "interim" CEO. "You can fire me anytime you want, but at least give me a better chance to succeed," he said. We agreed.

When I spoke with the GM board by conference call that Friday night, the reaction was violent, including veiled suggestions of mass resignations. The directors felt that one of their most important responsibilities -- hiring and firing of the CEO -- had been usurped by the government without any warning or consultation.

I explained our decision, including how we had chosen to implement it, which was partly to give Rick the dignity of putting his departure in whatever context he chose to with the board. After about an hour, I left the board to its deliberations. In the end, all the directors agreed to stay on through the restructuring.

In advance of the President's national address on Monday, a half dozen calls had been scheduled for him at short intervals with key Michigan legislators and Gov. Jennifer Granholm. The President sat at his bare desk in the Oval Office as the rest of us huddled around a badly functioning speakerphone ("probably from a well-connected government contractor," the President joked). For the Michigan delegation, the news that the President was prepared to put both companies into bankruptcy was unwelcome, to say the least.

When we later regrouped in Larry's office to compare notes, he was worried. For all of Larry's reputation as an academic, his years in public service had honed sharp political instincts. The pushback on his calls had convinced him that as financial guys, we had focused too much on the restructuring and not enough on supporting auto sales and a broader industry recovery.

During the transition, the Obama team had considered an idea being used successfully in Germany to induce car buyers by offering rebates on trade-ins of gas guzzlers. In response to Larry's concern, Brian reminded Larry of the idea and worked through the night to insert new language into the President's Monday speech, thereby giving birth to "cash for clunkers."

The news about Rick's departure had leaked out on Sunday before we started briefing the press and consequently ended up being a bigger part of the story than it should have been. I was stunned by the suggestion that the government, GM's only source of fresh capital, was somehow out of bounds for asking for the resignation of a CEO who had lost $13 billion of taxpayer money in three months and was now asking for more. But rightly or wrongly, the concept of Washington extending its iron fist to an industrial icon proved unnerving to more than just the Wall Street Journal editorial page.

The President's announcement that he was prepared for bankruptcy dramatically changed the nature of the discussions we were having with the stakeholders, none so much as those with Chrysler's 45 senior lenders.

Led by Jimmy Lee, J.P. Morgan's top banker and a long-standing business friend of mine, the banks had been insisting that, as secured lenders, they were entitled to repayment of their entire $6.9 billion. "And not a penny less," Jimmy said to me on more than one occasion.

From the outset that had struck us as ridiculous. The debt was trading at about 15¢ on the dollar, and according to Chrysler's analysis, the liquidation value of the company was around $1 billion. Clearly the banks didn't believe that the government would push back and let the lenders take over the company, if necessary. Until they heard the President speak on Monday morning. Almost immediately afterward, Jimmy called with a new message: "We need to talk."

The characters who assembled were straight out of central casting. Jimmy, who favors white-collared shirts and suspenders, arrived with an entourage of equally nattily clad colleagues to face our more rumpled team, including Sergio in his black sweater.

Jimmy made a virtuoso pitch about the importance of treating senior secured lenders properly, but the President had taken away his only real bargaining chip. The lenders knew that they would fare poorly in liquidation. Expecting to haggle, I had decided to put forward the smallest number I could imagine ($1 billion, equal to both the current market value of the loan and the liquidation value of the company).

Jimmy was apoplectic. As the negotiations proceeded, the lenders were particularly aggrieved that the UAW's health-care trust, which ranked below the secured creditors, was slated to exchange an $8 billion existing claim for $4.6 billion in notes and 55% of the equity in the reorganized company. While arguably close to a 50% haircut, it was a higher-percentage recovery than we were offering the banks.

The lenders felt that this represented an ideological decision by the Obama administration to tilt in favor of labor and against capital. That was simply not the case. At no time during our months of work did the White House ever ask us to favor or punish any stakeholder.

Many other unsecured creditors -- notably, suppliers and consumers holding warranties -- actually received 100¢ on the dollar. The fact was, Chrysler had to have workers, suppliers, and customers to succeed and therefore needed to give them more than called for by their rank in the capital structure.

Equally important, the White House never tried to use the auto restructurings to achieve any other policy goals. While we were at work, a new set of fuel-efficiency standards was negotiated by the White House with all the automakers -- without any involvement on our part. Nor did we impose any new technology mandates. Most important, we were under strict instructions not to get involved in day-to-day management decisions.

Pulling together the rest of the package had other dramatic moments, but by late April we were cautiously optimistic. In particular, Ron Bloom had expertly refereed UAW negotiations that included major concessions on many fringe benefits (including overtime and holiday pay), as well as the important recasting of the company's obligations for health care. We believed that Chrysler's labor costs would finally be competitive with the transplants.

By April 28 the time had come to give the lenders the bottom line that Tim and Larry had approved: $2 billion, which we felt was more than fair. "Take it or leave it" was my message to Jimmy. He took it, successfully bargaining for cash.

As the President's deadline drew nearer, we recognized -- somewhat to our surprise -- that the only reason bankruptcy might be necessary would be to address the problem of the banks. Under the law, the lenders' entitlement to full principal and interest could not be reduced without the consent of each of them.

The next day, just 24 hours before the President was scheduled to speak again to the nation on the status of the auto bailout, Jimmy came back to me with a startling new proposal: If we were willing to increase the $2 billion, he thought he might be able to win approval from all of the lenders, thereby avoiding bankruptcy. After some haggling, we agreed on a boost of $250 million and a 6 p.m. deadline.

As it approached, Jimmy asked for more time, which I doled out in short increments. Finally, by 7:30, we concluded that we were going to be unable to make a deal. I thanked Jimmy for his hard work, but I was quite angry that a handful of investors had pushed Chrysler into bankruptcy.

In his speech the next day, the President attacked the holdout lenders who sought "an unjustified taxpayer-funded bailout." I welcomed that language, not sufficiently anticipating the extent to which those words, together with the perceived disparate treatment of the various stakeholders, would be misinterpreted by a Wall Street community already shaken by repeated attacks from Washington.

We would spend many hours over the ensuing months emphasizing our deep conviction that the outcome of the Chrysler restructuring had virtually nothing to do with the heavy hand of government and everything to do with the fact that Treasury was the reluctant investor of last resort.

Every stakeholder did better under our plan than they would have in the alternative: a liquidation, in which the lenders would have gotten far less than the $2 billion they wound up with.

We reminded the critics that we pursued our bankruptcies under existing law through the traditional court process (all the way to the Supreme Court in the case of Chrysler). Finally, we noted the golden rule of Wall Street: He who has the gold makes the rules. Or as my father used to say to his unruly children, "He who eats my bread sings my song."

Our next patient was GM. A portion of our team, led by Harry, had been hard at work with GM's management trying to inject realism into its disappointing Feb. 17 plan. We believed GM needed to assume that U.S. car sales, which had peaked at 17 million in 2005, might not get much above 10 million for the next several years.

Through dozens of meetings and discussions, with Harry and his team asking tough questions, GM management came forward with a plan that accelerated the plant closings, eliminated the Pontiac brand, increased the job and dealer reductions, and added white-collar job cuts. (Buick and GMC were saved by Fritz Henderson's passionate belief in Buick's China appeal and GMC's reputation for ruggedness.)

And under Ron and Harry's guidance, GM reached an agreement with the UAW that was at least as favorable to the company as the one that Chrysler had signed. All told, GM's debt and related liabilities were reduced from $120 billion to $55 billion, and $8 billion a year of North American structural costs were eliminated.

But like Chrysler, GM had a "brand equity" problem. Its cars often sold for several thousand dollars less than comparable models made by the Japanese and Korean "transplants," yet still weren't moving off the sales lots. This problem could only be solved with time and good products; an important part of our investment thesis was that GM's cars were better than the market gave it credit for.

Financially, however, GM was actually in worse shape than Chrysler. One simple indicator of that was the amount of capital the government ended up injecting: $12 billion into Chrysler and $50 billion into GM, a higher proportion for GM in comparison with its revenue.

So we were faced with a tough decision about how to contribute that capital. If we made our investment as a loan, GM would be saddled with unmanageably large obligations. The only viable alternative -- proposed by Harry -- was to inject most of our capital as equity. All of us -- especially Tim and Larry -- hated the idea of government ownership of private companies, but we simply had no responsible alternative.

We had not anticipated one benefit of putting Chrysler on a faster track than GM. The successful completion of Chrysler's restructuring established a clear precedent and signaled to GM's stakeholders that we had leverage. That was helpful because, with hindsight, some form of GM bankruptcy was inevitable from the start. Unlike Chrysler, with its 45 lenders, GM had thousands and thousands of bondholders, many of them individuals. There was no way that we were ever going to corral enough of them to avoid at least a quick-rinse bankruptcy.

On June 1 we once again gathered in the President's office so that he could call local politicians and Ron Gettelfinger. And we once again stood behind the President as he noted the progress Chrysler had made in bankruptcy and announced that GM would be following suit. Happily, just as the Chrysler precedent sped along our negotiations with GM's stakeholders, so too did Chrysler's successful navigation through bankruptcy help us with GM.

By now as well, both the White House and the public had gotten used to the idea of bankrupt automakers (and Chrysler's sales in bankruptcy had been better than any of us predicted), and so the anxiety that accompanied the Chrysler bankruptcy was much diminished as GM's case unfolded. When the GM restructuring was completed, the only surprise was that GM had beaten Chrysler's record by three days.

While the two automakers dominated news coverage of our efforts, we were simultaneously addressing myriad related problems, including decisions whether to bail out the auto-part suppliers too (generally, no) and whether to recapitalize the auto-finance companies (in the case of GMAC (GJM), yes).

All told, the intensity of fighting multi-front wars was almost overwhelming, even for my younger colleagues used to 24-hour workdays. One, David Markowitz, fainted from lack of sleep during the President's third address and was taken to the White House infirmary, where he had a chance encounter with a President in search of a couple of Tylenol.

Because we were operating under TARP, we encountered relatively little congressional intrusion -- until the two automakers virtually simultaneously announced their dealer-reduction plans. Our dentistry had finally struck a nationwide nerve.

Every congressional district had dealers, many of whom were leading figures in their communities. We were inundated with an avalanche of calls, letters, and demands. We patiently worked through each grievance and explained hundreds of times that the companies -- not the government -- made the decisions about which dealers to close.

But the episode left an indelible impression on me: If we hadn't had TARP money available and had had to seek congressional approval, I am convinced that one or both of our two automakers would have been forced to liquidate.

Fortunately, our restructurings survived, and the companies began to operate as private enterprises, just as the President had outlined and just as we had hoped. With that, the workload of the auto task force dropped precipitously, and many of us prepared to leave Treasury, satisfied that we had given these companies the best possible chance to succeed.

Like any patient that undergoes major surgery, a successful recovery is far from assured. For Chrysler, the biggest challenges are its need to regenerate its product line and manage a significantly leveraged balance sheet. In the case of GM, the overarching question is whether, without an infusion of new blood, its management team can implement the massive cultural change that is essential. But by dramatically lowering the break-even point for both companies, we believed we were creating a healthy margin for error.

On a muggy July evening, I left Treasury through the south entrance, the opposite end of the long building from where I had first tried to enter so many months earlier. A tourist who had stopped to take photos of the statue of Alexander Hamilton recognized me. "You guys did good work," she said, seeming to channel Hamilton's belief in industrial policy and the importance of manufacturing. "He would have been proud of you." I hope her confidence in us proves justified.

Reporter associates: Marilyn Adamo and Scott Cendrowski