A Crack in The System

=20

A Crack in = The=20 System
By 1998, AIG Financial Products had made hundreds = of=20 millions of dollars and had captured Wall Street's attention with its = precise,=20 finely balanced system for managing risk. Then it subtly turned in a = dangerous=20 direction.

By Brady Dennis and Robert O'Harrow Jr.
Washington = Post=20 Staff Writers
Tuesday, December 30, 2008; A01

Second of three parts

For months, several executives at AIG Financial Products had pulled apart the data, = looking for=20 flaws in the logic. In phone calls and e-mails, at meetings and on their = trading=20 floor, they kept asking themselves in early 1998: Could this be right? = What are=20 we missing?

Their debate centered on a consultant's computer model and a new kind = of=20 contract known as a credit-default swap. For a fee, the firm essentially = would=20 insure a company's corporate debt in case of default. The model showed = that=20 these swaps could be a moneymaker for the decade-old firm and its = parent,=20 insurance giant AIG, with a 99.85 percent chance of never having to pay = out.

The computer model was based on years of historical data about the = ups and=20 downs of corporate debt, essentially the bonds that corporations sell to = finance=20 their operations. As AIG's top executives and Tom Savage, the = 48-year-old=20 Financial Products president, understood the model's projections, the = U.S.=20 economy would have to disintegrate into a full-blown depression to = trigger the=20 succession of events that would require Financial Products to cover=20 defaults.

If that happened, the holders of swaps would almost certainly be = wiped out,=20 so how could they even collect? Financial Products would receive = millions of=20 dollars in fees for taking on infinitesimal risk.

The firm's chief operating officer, Joseph Cassano, had studied the = model and=20 urged Savage to give the swaps a green light.

"The models suggested that the risk was so remote that the fees were = almost=20 free money," Savage said in a recent interview. "Just put it on your = books and=20 enjoy the money."

Initially, the credit-default swaps business would amount to a = fraction of=20 the half-billion dollars in Financial Products' revenue that year. It = didn't=20 seem to them like a major decision and certainly not a turning = point.

They were wrong. The firm's entry into credit-default swaps would = evolve into=20 insuring more volatile forms of debt, including the mortgage-backed = securities=20 that helped fuel the real estate boom now gone bust. It would expose AIG = to more=20 than $500 billion in liabilities and entangle dozens of financial = institutions=20 on Wall Street and around the world.

When the housing market tanked, a statistically improbable chain of = events=20 began to unfold. Provisions in the contracts kicked in, spurring = collateral=20 calls on swaps linked to $80 billion in questionable assets, requiring = the firm=20 and AIG to come up with billions of dollars in cash. They scrambled for = almost a=20 year to stave off the calls, but there were too many deals with too many = counterparties.

In September, the Bush administration concluded that AIG's position = at the=20 nexus of the deals meant that it could not be allowed to fail, = triggering the=20 most expensive rescue of a private company in U.S. history. So far, the=20 government has invested $152 billion in its efforts to save AIG. Federal = investigators are sifting the carnage.

Credit-default swaps exemplify the contradictions of modern finance. = At a=20 basic level, they serve as insurance, but they aren't regulated as such. = They=20 have allowed companies to free up untold amounts of capital that = otherwise would=20 be tied up as collateral for loans. They were sold both to reduce risk = and, in=20 some cases, to give clients room to take on more risk -- a key component = to=20 making money on Wall Street.

But in the end, neither the buyers nor sellers truly understood the = enormous=20 risks they were creating. Anyone could sell such a swap, and anyone = could buy=20 one, even if he had no stake in the transaction. Some buyers used them = to bet=20 against failing companies, prompting a debate among state regulators = about=20 whether this type of swap was a form of gambling.

The very nature of credit-default swaps put Financial Products at = odds with=20 itself, requiring it to deviate from the disciplined system that had = made it a=20 pathbreaker. Everything about the company -- its technology, its people, = its=20 rigorous culture of transparency and caution -- was designed to minimize = the=20 various risks that it shouldered while solving problems for clients.

That meant hedging whenever possible, a Wall Street term for making=20 offsetting trades to balance risk. For transactions involving credit and = loans,=20 it also meant building an escape route so that the firm could get out = early if=20 it saw a deal going bad.

With credit-default swaps, there was no way out, and the risk was so = minute=20 that hedging was considered unnecessary, as well as problematic. Savage=20 remembers discussions about whether the firm's vaunted computer system = could=20 even come up with the proper values needed for the trades that hedging = relied=20 on.

All of that made Savage and the others wary. Skepticism was = hard-wired into=20 the company's culture, part of its mantra: Hedge if you can. Don't make=20 speculative trades. Above all, protect AIG's reputation and its = top-drawer=20 Triple A credit rating, which gave Financial Products credibility and = the=20 ability to borrow money at the cheapest rates. The rating was the fuel = for=20 Financial Products' innovation and success.

AIG's chairman, Maurice "Hank" Greenberg, had once warned Savage that = he=20 would come after him "with a pitchfork" if Financial Products did = anything to=20 harm AIG's AAA rating. No one saw credit-default swaps as anything on = that=20 scale. After conversations that included AIG executives, Greenberg = blessed the=20 new line of business. "There was a long discussion about it," Savage = recalled=20 recently, "and he said it was fine."

Greenberg said recently, "I don't think going into it in '98 was = wrong."=20 During his tenure, he said, he and his risk managers kept close watch on = the=20 swaps and the exposure they created.

Savage retired from Financial Products in 2001. When he left, = credit-default=20 swaps were still a small portion of the firm's business. Not long ago, = in the=20 dining room of his golf club in Florida, he reflected on the = significance of the=20 decision that he and his colleagues made in 1998.

Like his bosses at AIG, he still thinks it made perfect sense to give = swaps a=20 try. "The credit derivative business had just begun and because of our = role in=20 the derivatives business, it was very natural for us to have some = minimal=20 participation," he said.

Savage says he now sees that the decision sent Financial Products = down a path=20 at odds with its guiding principles. The firm's success had been built = on=20 assessing data daily, recalibrating assumptions constantly, = counterbalancing one=20 risk against another and making the hedges. The credit-default swaps = didn't=20 require that sort of attention.

"The different nature of those trades from any other trades that FP = had=20 done," Savage said, "opened the door to all the problems that came = about."

He added later: "In retrospect, perhaps those deals should never have = been=20 done."

2: 'A Watershed Event'

One of the firm's biggest advocates for credit-default swaps was = Joseph=20 Cassano.

Cassano, the feisty, hardworking son of a Brooklyn cop, did not have = the=20 pedigree of Financial Products' three founders, who hailed from places = such as=20 Bell Labs and the Wharton School. Cassano had worked with the trio at the = junk-bond=20 firm of Drexel Burnham Lambert, and had been one of 10 original recruits = who=20 left Drexel to start Financial Products.

A Brooklyn College graduate, the 42-year-old Cassano was not one of = the=20 "quants" who had mastered the quantitative analysis and risk assessment = on which=20 the firm had been built. He had no expertise in the art of hedging. But = he had=20 excelled in the world of accounting and credit -- the "back office," as = it is=20 known on Wall Street.

The founders of Financial Products made him the firm's chief = financial=20 officer. From the start, Cassano gained respect, in part because he and = his team=20 rarely made mistakes processing trades. He was smart and aggressive -- = sometimes=20 too aggressive, some executives thought. He had a mercurial temper, = occasionally=20 screaming at an underling. He swore, berated and moved on, sometimes = leaving=20 hard feelings in his wake.

"He was very, very good," recalled Edward Matthews, AIG vice = chairman. "But=20 he was arrogant."

He also was ambitious. He made plain to his bosses that he wanted = more than=20 the back office.

In 1994, Cassano got a chance.

The firm's founders had left in a bitter dispute with Greenberg, and = Savage=20 had taken the reins. He put Cassano in charge of the Transaction = Development=20 Group, a new unit hunting for business involving energy products and tax = credits=20 in the United States and abroad. He was also made chief operating = officer.

Cassano's portfolio included deals involving credit, so he played a = key role=20 in the credit-default swap debate going on inside the company. In 1998, = when the=20 investment bank J.P. Morgan came to Financial Products, seeking a = credit-default=20 swap arrangement, Cassano was among the most interested. After studying = the=20 proposals, he passed on the first deal. But he soon became a leading=20 proponent.

J.P. Morgan wanted to package a variety of debt on its books and = resell it.=20 The debt would be turned into bond-like securities, and layered like a = wedding=20 cake so that investors in the top tiers were first to get their money = back in=20 case of default. Investors in lower tiers earned a higher interest rate = for=20 taking greater risk.

The "structured" deal had an unwieldy name, the Broad Index Secured = Trust=20 Offering, so it was called "Bistro" for short. Because the debt in = Bistro was=20 diverse, the investment was considered exceedingly safe; if one kind of = debt=20 went into default, it was unlikely other kinds would go under at the = same time.=20 As an extra measure of safety, the Bistro organizers wanted Financial = Products=20 to write credit-default swaps on the top tiers to further reassure = skittish=20 investors.

As private contracts, deals like Bistro could be financed with = greater=20 amounts of borrowed money than regulators would allow if the deals were = publicly=20 traded. This high degree of leveraging would come back to haunt the = industry=20 later.

The structure was an early form of collateralized debt obligations. = CDOs were=20 a hit almost from the start. It would take several years and a housing = bubble=20 for CDOs backed by mortgages to catch on. At Financial Products, the=20 credit-default swap was only one of many innovations in play, but = Cassano was=20 passionate about how it could help the firm.

"It was a watershed event in 1998 when J.P. Morgan came to us, who = were=20 somebody we worked with a great deal, and asked us to participate," = Cassano told=20 an investment banking conference in 2007. "These trades were the = precursors to=20 what's become the CDO market today."

Even as Cassano spoke, the housing market was collapsing, the lack of = diversity of the CDO debt was being exposed, and the risk for Financial = Products=20 was rising.

3: 'It's the Hardest Thing'

By summer 1998, after four years as president, Savage found himself = thinking=20 even harder about risk, particularly credit risk. It was often difficult = to=20 quantify the likelihood that someone would pay back a loan.

Savage kept his distance from developing trades, with the idea that = he could=20 better maintain his objectivity about potential pitfalls. He sometimes = wondered=20 whether Cassano's enthusiasm for the credit deals colored his ability to = assess=20 them. Cassano's lawyer, F. Joseph Warin, said in a recent interview that = Cassano=20 took care to follow procedures that minimized risk.

Greenberg, too, kept at Savage about the risk, even while keeping on = the=20 pressure for greater profits. On Wall Street, investment banks and other = financial institutions were mad for private contracts called = derivatives, Wall=20 Street's jargon for a contract based on something trading now, but = settling in=20 the future. (A credit-default swap is a kind of derivative in which one = company=20 takes on the future credit risk of another.)

Derivative contracts accounted for more of the world's financial = activity by=20 the day. Some in Washington had taken notice, and thought investors and=20 regulators needed to know more about these privately arranged deals that = were=20 cloaked from outside scrutiny and clouded by complexity.

Brooksley Born, the 57-year-old head of the Commodity Futures Trading Commission, argued forcefully = for a=20 public debate about whether derivatives posed an unknown and growing = risk to the=20 world's financial system. She testified at least 17 times before = Congress on the=20 subject.

Her campaign gained no traction. More powerful regulators, including = Federal=20 Reserve Chairman Alan Greenspan, Treasury Secretary Robert E. Rubin and Securities and = Exchange=20 Commission Chairman Arthur Levitt, opposed Born. They and others said her = agency had=20 no authority over derivatives and that her call for action was casting a = "shadow=20 of regulatory uncertainty over an otherwise thriving market."

Greenspan, in particular, argued a free-market view. He saw = derivatives as a=20 mechanism that unlocked efficiency, allowing dormant capital to flow = into the=20 system, greasing the gears of the world's economy. The Clinton = administration=20 and many congressional Democrats endorsed the notion that too much = regulation=20 stymied growth.

Greenspan pushed the idea that the marketplace was self-correcting, a = view=20 that he often espoused in speeches at economic conferences around the = world. He=20 invited Greenberg to attend one such meeting in Basel, Switzerland. = Greenberg=20 couldn't go, so he arranged for Savage to go. Chief executives of banks, = investment firms and insurance companies, as well as U.S. and German = regulatory=20 officials, filled the room.

Greenspan, already celebrated as an economic guru, commanded = attention every=20 time he spoke. The question he posed that day resonated with Savage for = a long=20 time.

"Do you folks find that you have enough information to make credit = decisions=20 in your businesses?" Greenspan asked.

Mathis Cabiallavetta, chairman of the board for the giant Swiss bank = UBS,=20 responded that his company knew well what it was up against.

Not well enough, as soon became clear.

In September 1998, Long Term Capital Management, a heavily leveraged = hedge=20 fund with mountains of derivatives, told Federal Reserve officials that it could not cover $4 = billion in=20 losses. Russia, swept up in an Asian economic crisis, had defaulted on = its debt,=20 and Long Term was besieged with calls to put up more collateral for its=20 investments. The collapse threatened the fortunes of investors from = tycoons to=20 pension funds.

UBS lost hundreds of millions of dollars. Cabiallavetta lost his = job.

The exchange in Switzerland, and the Long Term debacle, fueled = Savage's=20 unease. His mind kept turning over the problem of how to calculate the = risks of=20 credit.

"I've always thought about that," he said. "At the highest level of = finance,=20 this is a question of interest. Are you getting enough information about = the=20 loans that you're making to corporations? It's the hardest thing. . . . = You have=20 to look beyond the credit-rating agencies and make your own = decisions."

Savage recalled something that Greenberg had once told him.

"He said to me, 'I want you to understand that no matter what the = credit=20 rating is, no matter what other things you might understand, when a CEO = owes you=20 $100 million and is supposed to pay you on Friday, sometimes he just = doesn't do=20 it.' "

4: Exploiting A Seam

Financial Products' drive to keep ahead of its competitors took the = firm in=20 unexpected directions. It developed a reputation as an innovator with = one of the=20 most diverse toolboxes in the derivatives business.

That's how Cassano and his Transaction Development Group found = coal.

For a group of financial wizards, the coal business seemed an odd = turn. But=20 it was a logical extension of what the firm had been doing all along:=20 discovering gaps in regulations and markets.

A 1980 law, generated by the Carter administration, offered tax = credits to=20 companies as incentives to design and use synthetic fuel systems. The = aim was to=20 reduce U.S. dependence on foreign oil.

Associates at the Transaction Development Group had discovered that = many=20 energy companies were not making enough money to benefit from the tax = breaks.=20 But Financial Products' profitable parent, AIG, could use those credits = to=20 reduce its tax bill.

"One thing AIG had was ample income," Savage said. "So what we did = is, we=20 went out and we bought synthetic coal facilities."

The firm had no intention of becoming coal processors. Instead, it = arranged=20 to install the equipment -- bought for more than $225 million, as Savage = recalls=20 -- at coal facilities and power plants. The facilities leased and = operated the=20 machines at a discount, while AIG got millions in tax credits.

Financial Products hedged aspects of the deals and checked with = government=20 officials to make sure the arrangements qualified for the breaks. Savage = said=20 the idea was bold as well as clever. "We had the gumption to go out and = take=20 seven of these plants that were sitting around doing nothing," he said. = "We=20 carted [the machines] off to where they could be used, and it went = on."

Greenberg, too, was taken with the gambit. "It was opportunistic," he = said=20 recently. He once joked that he wanted to ride shotgun in the truck = carting the=20 machines around, Savage said.

Over the next several years, AIG reaped $875 million in benefits from = the=20 deals. It was a coup for Cassano and his group. Although it wasn't = Cassano's=20 idea, Savage said, he guided it from concept to reality.

"He says he thought about it for six months," said Savage, who came = to=20 appreciate Cassano's single-minded focus. "He made a lot of money for = the=20 company."

5: 'It Would Be Joe'

In fall 2001, Savage decided to call it quits. He had moved his = family to=20 Florida and briefly considered whether he could manage the commute. The = Sept. 11=20 attacks made that sort of arrangement seem impossible. He told Greenberg = of his=20 plan to leave.

Cassano emerged as Greenberg's candidate to take over. Some = colleagues=20 questioned his qualifications to manage a team that was heavily = dependent on=20 quantitative skills. Though he was the firm's chief operating officer, = some=20 colleagues thought he wasn't as conversant with the complex calculations = of risk=20 that remained at the heart of its business. Beyond that, few liked his=20 chip-on-the-shoulder demeanor.

Greenberg had come to know Cassano through board meetings over the = years.=20 Cassano had won Greenberg's confidence. The two shared a number of = qualities.=20 Both were strong-willed, and both disliked criticism. Greenberg knew = that, like=20 him, Cassano had made AIG the center of his life. He knew about = Cassano's=20 temper, but he appreciated his grit and drive to make money in the = derivatives=20 field, which was becoming more crowded with competition.

Cassano had one other virtue that helped him land the top job: He = followed=20 directions from Greenberg and Matthews, the parent company's = leaders.

"He told us that in no uncertain terms, that he was -- that all of = his people=20 up there were -- smarter than anybody we had at AIG," Matthews said. = "And he=20 made it clear that he listened only to two people: He listened to Hank Greenberg and he listened to me."

Cassano would need all the smarts he could muster. He was taking the = reins at=20 a challenging juncture. Financial Products was now a $1 billion = operation with=20 225 employees working on a multitude of derivatives deals for clients, = involving=20 hundreds of billons of dollars in obligations. But in early 2002, when = he=20 replaced Savage, the derivatives industry was coming under a shadow.

A high-flying financial company called Enron was just starting to melt down. Because Enron had=20 systematically abused derivatives as part of its fraudulent corporate=20 accounting, some kinds of derivatives became the focus of regulatory = scrutiny=20 and fell out of favor. Structured deals for corporations were a large = part of=20 Financial Products' business.

The firm would need to make up lost revenue. "The response to Enron = really=20 reduced the toolbox for Financial Products," Savage said. "It wasn't at = all=20 clear to me where the profits were going to come from."

Under Cassano, Financial Products would grow, take on more risk and = become=20 more top-down than before. The culture that had characterized the firm = from the=20 outset -- one that relied on informed skepticism in which just about = anyone=20 could question dubious aspects of a trade -- would change, according to = people=20 who worked at the firm.

Cassano disputes the notion that the culture had changed, according = to Warin,=20 his lawyer. "FP worked closely and had healthy discussions with its = internal=20 auditors so they would fully understand the business and investments," = Warin=20 said. "Mr. Cassano encouraged this oversight, review and open=20 communication."

6: Clearing The Way

In 2002, the regulatory debate over one of those lines of business,=20 credit-default swaps, was going nowhere. The swaps had fierce critics. = Some saw=20 them as insurance deals that ought to be subject to the same regulation = that=20 governed the writing of homeowners' policies or car insurance. Others = saw=20 certain swaps as gambling: Because anyone could buy a swap, even someone = who had=20 no stake in a particular asset, some critics thought those swaps were = like a=20 poker game in which spectators placed bets among themselves on who would = win the=20 hand.

Some regulators had a hard time seeing the financial value in certain = swaps=20 -- especially in deals used to remove debts from a corporation's = books.

But those regulators were fighting a lost cause. In the waning days = of the=20 Clinton administration, Congress had passed the Commodity Futures = Modernization=20 Act, which preempted derivatives from oversight under state gaming laws = and=20 excluded certain swaps from being considered a "security" under SEC rules.

While some regulators had expressed concerns about the act, President Clinton's economic team had agreed that = derivatives=20 should not be regulated. Clinton signed the measure, which was part of a = larger=20 bill.

"By ruling that credit-default swaps were not gaming and not a = security, the=20 way was cleared for the growth of the market," Eric Dinallo, the superintendent of New York State's = insurance=20 department, told a Senate committee during recent hearings on the role = of=20 derivatives in triggering the financial crisis. "None of this was a = problem as=20 long as the value of everything was going up and defaults were rare. But = the=20 problem with this sort of unregulated protection scheme is that when = everyone=20 needs to be paid at once, the market is not strong enough to provide the = protection everyone suddenly needs."

7: 'We Made Some = Mistakes'

In August 2002, one Financial Products' innovation caught the = attention of=20 federal investigators. The year before, Financial Products had been = pitching a=20 new way for companies to shed bad debts, and it had found a customer in = PNC Financial Services Group, which had $762 million in=20 underperforming assets it wanted to unload.

Ordinarily, the bank would need to account for the falling value of = those=20 assets, which would mean a hit to its profits. Associates at Financial = Products,=20 working with accountants, thought they had found a way to solve PNC's = problem:=20 Create "special-purpose entities" to take on the unwanted assets.

Federal investigators alleged, however, that the deals were a sham. = To make=20 the transactions look legitimate, Financial Products had set up a = company to=20 "invest" in the entities, while receiving an equivalent amount in the = form of=20 fees, according to the investigators. Structuring the deal this way = violated=20 securities laws, FBI agent Randy Tice asserted in an affidavit filed in = federal=20 court as part of the simultaneous settlement of a criminal case and an = SEC civil=20 complaint.

AIG and two Financial Products subsidiaries agreed to pay an $80 = million fine=20 and give back $39.8 million in the fees that it had earned, plus $6.5 = million in=20 interest. PNC paid a $115 million fine.

The government announced the settlement on Nov. 30, 2004. In the wake = of=20 Enron, the investigators were sending a message. "We are pleased that = AIG has=20 accepted responsibility," said Christopher Wray, an assistant U.S. = attorney=20 general. "There is no place in our markets for financial transactions = that lack=20 economic substance."

But authorities demanded more. The settlement also required AIG "to = implement=20 a series of reforms addressing the integrity of client and third-party=20 transactions." A group of senior AIG executives would review complex=20 transactions from the previous few years, working with an independent = monitor=20 chosen by the Justice Department, the SEC and the company.

In other words, the government had concluded that Financial Products' = internal controls -- the disciplined system that had once made the = company=20 different from its competitors -- had faltered.

Cassano, who had not arranged the transactions but signed the = settlement for=20 Financial Products, later described the PNC deals as an anomaly. "We = made some=20 mistakes in those transactions, and we suffered dearly for that," he = said in=20 2007 at an investors conference. "And we've gone to great lengths to = correct the=20 things that allowed the transactions to occur."

Greenberg said recently that Financial Products had consulted its = legal and=20 accounting experts before going forward with the special entities. The = board of=20 directors also had looked it over, Greenberg said. "We thought it was = proper,"=20 he said.

The settlement is still a source of grief for the former AIG chief = executive,=20 who had to swallow the costly settlement and the independent monitor. "I = took a=20 bullet for them," he said. "I went out in front. I didn't have to do = that. It=20 was their deal."

But the case had another consequence for Greenberg. It brought AIG = into the=20 sights of another skeptical investigator: New York Attorney General = Eliot L.=20 Spitzer.

8: Foot Faults

After the PNC case became public, a tipster approached Spitzer's = office.=20 Insurance companies, the tipster said, were selling policies known as = "finite=20 insurance." The tipster thought the policies were a fraud.

Done right, finite insurance expressly limits the losses an insurer = can=20 suffer. Done wrong, it isn't insurance at all because neither side takes = any=20 risk. Instead, it's an accounting trick that can help both parties = improve the=20 appearance of their balance sheets.

The tipster urged Spitzer's office to examine finite insurance and = suggested=20 several companies for scrutiny, including AIG and Gen Re, another large=20 insurance company. Spitzer's office sent subpoenas to companies, seeking = more=20 information. Not long after, a black binder from another tipster arrived = at=20 Spitzer's office in Lower Manhattan. Four inches thick, the binder held=20 confidential documents from Gen Re. The documents appeared to show that=20 Greenberg had arranged bogus transactions with Gen Re that made it look = as if=20 AIG had $500 million more in insurance revenue than it had actually = earned.

Spitzer and his people could not believe their luck. It was a case on = a=20 silver platter. They decided to question Greenberg right away, instead = of the=20 usual approach of working slowly toward such a big potential target.

On Feb. 9, 2005, Spitzer told his people to begin work on a Greenberg = subpoena.

That afternoon, coincidently, Greenberg announced AIG's latest = earnings in a=20 conference call with industry analysts and others. During the call, he=20 complained indirectly about Spitzer's investigation of the insurance = industry,=20 suggesting that the probe was overkill and Spitzer was wasting his = time.

"When you begin to look at foot faults and make them into a murder = charge,=20 then you have gone too far," Greenberg said.

Greenberg's remarks were reported online that afternoon and Spitzer = happened=20 to see them. Irked, he asked a deputy how soon the Greenberg subpoena = could go=20 out.

That evening, Spitzer was to speak at a dinner with senior executives = at Goldman Sachs, in an elegant conference room at the = investment=20 bank's headquarters. Among those in the audience: Henry Paulson, then Goldman's chairman and chief = executive. The=20 next year, he would become Treasury secretary and head to Washington, = where he=20 eventually assumed the central role in dealing with AIG's = near-collapse.

As Spitzer waited to deliver his remarks, a deputy came in and = whispered into=20 his ear: The Greenberg subpoena had been faxed to AIG. A few minutes = later,=20 Spitzer alluded to Greenberg's comments earlier in the day.

"These are not foot faults," Spitzer recalls saying. "But second, too = many=20 foot faults and you lose the match."

9: 'No Choice'

The end of Greenberg's reign at AIG came with a phone call March 13, = 2005. He=20 was in a private jet on his way back to New York from a visit to Key Largo, Fla. The AIG board of directors had called a = meeting=20 that Sunday to consider allegations from Spitzer that Greenberg had been = personally involved in the fraudulent deal with Gen Re.

The board had asked Greenberg to call. Frank Zarb, a veteran Wall = Street=20 executive and board member, told Greenberg that Spitzer had issued an = ultimatum:=20 Greenberg had to resign.

"I had no choice," Greenberg said recently. "No choice."

Earlier this year, four Gen Re executives and an AIG executive were = found=20 guilty on federal fraud charges. Later, AIG restated earnings from 2000 = to=20 2004.

Greenberg, referred to anonymously in federal documents as an = unindicted=20 co-conspirator, maintains that what "we did, from AIG's perspective, was = perfectly proper." In a recent interview, he tore into Spitzer: "He = destroyed a=20 company. And for what?"

Spitzer said recently that the activities at AIG were too important = to=20 ignore. Events have solidified his view. "AIG, as we have now all seen," = he=20 said, "was at the center of the web of the entire financial system."

Greenberg blames others for his company's downfall. He says his = forced=20 departure left AIG without the strong hand it needed to protect against = future=20 excesses. He said AIG and Financial Products were prepared to hedge any=20 transaction "if we thought there was going to be a potential = problem."

Matthews put it this way: "What bothers us about this is we had a = climate of=20 risk management which seems to have evaporated after we left."

By then, though, the company had already taken a deeper dive into=20 credit-default swaps, including an expansion into the subprime mortgage = market=20 that would eventually trigger the improbable.

The crack in the Financial Products system was about to get a lot = wider.

Wednesday: Downgrades and downfall.

Staff writer Bob Woodward contributed to this report.